Six Sigma Handbook - 2007 PDF

VILLANOVA UNIVERSITY
LEAN SIX SIGMA

HANDBOOK
@ by Quality Council of Indiana - All rights reserved

VILLANOVA
UNIVERSITY
Professional Education Online

WITH A LITTLE HELP FROM MY FRIENDS.
JOHN LENNON/PAUL McCARTNEY
Acknowledgments
We thank our friends and professional associates for their assistance, particularly
Tim Brenton, Steve Adriano, and Vicki Shaw who helped with this text. We would
appreciate any comments regarding improvement and errata. It is our concern to be
accurate.
This Handbook is produced in collaboration between Villanova University and

Quality Council of Indiana. Any recommendations for improvement should be
directed to:
Bill Wortman
Quality Council of Indiana
602 West Paris Avenue
West Terre Haute, IN 47885
TEL: 800-660-4215
TEL: 812-533-4215
FAX: 812-533-4216
qci@qualitycouncil.com
http://www.qualitycouncil.com
First Edition - April 2, 2007
008
VILLANOVA UNIVERSITY INTRO-3 LEAN SIX SIGMA HANDBOOK

The Authors
Ron Crabtree
Ron Crabtree, CPIM, CIRM, is President of MetaOps, Inc., a consulting firm that
specializes in strategic business transformation. He is an internationally recognized
expert and author on business process improvement. Ron serves as adjunct faculty
for Villanova University developing and teaching Lean Six Sigma for the University
Alliance Online. Mr. Crabtree writes the "Lean Culture" Department in APICS
Magazine and is active in consulting and conducting seminars nation-wide.
In addition to a B.A. in Management and Organizational Development, Mr. Crabtree

is certified through APICS in Production and Inventory Management (CPIM) and in
Integrated Resource Management (CIRM). Currently, Ron serves as a Director-at-
Large for the Greater Detroit Chapter of APICS and is also a member of ASQ. Mr.
Crabtree is considered a leading master six sigma black belt by his peers.
Edwin Garro
Edwin Garro has spread quality and continuous improvement knowledge in Central
America for 20 years as college professor, quality manager, and general manager.
Edwin is a partner at Performance Excellence Solutions, an education, consulting,
and human capital organization. Mr Garro is also a partner and CEO of
Ludovico.Produccion Grafica, a printing shop, where he practices what he teaches.
Edwin is a senior member of ASQ and founding member of ASQ Section 6000, IMU
Costa Rica. He pioneered the first ASQ certifications in Central America. Edwin has
trained more than 400 professionals for ASQ certifications. He is an ASQ CQE,
CQM/OE, and CSSBB. Edwin has a B.S. in Industrial Engineering from the Instituto
Tecnologico de Costa Rica, and an M.S. in Manufacturing Engineering (Graduate
Student of the Year) from the University of Massachusetts.
Glenn Gee
Glenn Gee is a Senior Quality Engineer at Champion Laboratories. Glenn previously
worked as a Manufacturing Extension Director, advising over 300 companies on
such diverse issues as lean production, quality, economic issues, strategic planning,
marketing, finances, and product innovation. Glenn is an Adjunct Professor for
Eastern Illinois University and has taught ASQ section review classes for the CQE,
CCT, and CMQ/OE. He holds five ASQ certifications.
Glenn has a B.S. in Industrial Engineering from Purdue University and an M.S. in
Industrial Technology from Indiana State University. Glenn is a Registered
Professional Engineer and a Senior Member of ASQ. Glenn is a certified Project
Management Professional and a Certified Training Consultant. Mr. Gee is a member
of the Board of Governors for Quality Systems Registrar, Sterling, VA.

The Authors (Continued)
M. Dale Metcalf
M. Dale Metcalf holds an M.B.A. in Production Management and Organizational
Development from Indiana University. Mr. Metcalf's early professional career
spanned 25 years, including manufacturing floor supervision, middle management,
vice president of corporate training, and project leader with an international
consulting firm. He is a certified ISO-9000 lead auditor, member of ASQ, SME, and
is a Ford Global 8-D problem solving trainer. Dale's consulting and training firm,
Metcalf Training Group, Inc., began operation in 1995.
Metcalf Training specializes in all facets of lean (TPS) manufacturing, with an

emphasis on plant and divisional level implementation. Workshops include Ford
Global 8-D problem solving, root cause analysis, corrective actions, TPM (total
productive maintenance), 5-S implementation, PDCA, kaizen, gemba events,
statistical process control, quick setups, process mapping, project management,
JIT, kanban, and poka-yoke.
Omar Mora
Omar Mora is the founder and CEO of Blackberry and Cross a consulting firm
located in San Jose, Costa Rica. Omar received a B.S. in Industrial Engineering at
Universidad Internacional de las Americas and a M. S. in Industrial Engineering at
Universidad Interamericana de Costa Rica.
Mr. Mora is a Senior Member of ASQ and founding member of ASQ Section 6000, in
Costa Rica. Omar Mora is an ASQ Certified Quality Engineer and Six Sigma Black
Belt. He is also an APICS Certified Production and Inventory Manager. Mr. Mora
developed the first Lean Enterprise Certification in Latin America.
Terrill R. Paradise
Terrill Paradise holds a B.S. in Quality Management and an M.S. in Engineering
Management from Kennedy Western University. Terrill is a Senior Member of ASQ
and holds certifications in the CQT, CQI, CQA, CQE, and CSSBB areas.
Mr. Paradise's professional career includes 19 years experience in quality

engineering and management positions with major automotive manufacturers. He
has participated in design teams for a variety of Honda and Toyota products. Mr.
Paradise has been employed at Columbus Regional Hospital for five years. His
initial focus was process design, redesign, and improvement. Over the past two
years he has worked as a black belt in the deployment of the lean sigma program.

The Authors (Continued)
Wesley R. Richardson
Wesley R. Richardson is the Quality Knowledge Manager at Quality Council of
Indiana (QCI). In this capacity he writes, edits, and reviews materials created and
published by QCI. He has over 28 years of quality management experience,
including a commercial metallurgical testing laboratory, a medical device
manufacturer making MRI scanners, and a company manufacturing tungsten carbide
products for the coal mining and metal cutting industries. Wes has a B.S. in
Metallurgy from Massachusetts Institute of Technology, an M.S. in Metallurgy from
Case Western Reserve University, and an M.B.A. from the University of Kentucky.
Wes is a Senior Member of ASQ and currently holds twelve ASQ Certifications.
Bill Wortman
Bill Wortman is the CEO of Quality Council of Indiana - a quality publishing firm
located in Terre Haute, Indiana. He is a Senior Member of ASQ, former Chairman of
Section 0919, and Deputy Director of Region 9. Mr. Wortman currently holds eight
ASQ Certifications. Bill has instructed over 9,000 individuals in quality
fundamentals, including certification training for five ASQ Certifications. Mr.
Wortman has a B.S. in Metallurgical Engineering from N.C. State University. He
worked most of his professional life in the aluminum industry in a variety of
progressive technical and production management positions before starting Quality
Council of Indiana in 1988. Mr. Wortman has been author, co-author, or editor of
more than 30 quality related books and training CDs.

LSS Handbook Contents
I. LEAN SIX SIGMA OVERVIEW ••• •••••.••. .• •••••••• 1-1 VII. ANALYSIS TECHNIQUES . ....... ...... . .......... VII-1
LSS BODY OF KNOWLEDGE ••• •••.•.• .. . •• .••.•. 1-4 SEVEN CLASSICAL WASTES •• •••.••. . ••.••....• VII-2
GLOSSARY •••.••••••••••• •• . •••.•• • • .• • •• •••• I-g VARIABLE RELATIONSHiPS .. ........ . .... . ..... VII-6
MULTI-VARI ANALySiS . ..... ....... .. ......... VII-6
II. LEAN SIX SIGMA GOALS . . .. • .. • • .. .. . • . . .. • • • • •. 11-1 LINEAR CORRELATION .......... .......... .. VII-12
VALUE OF LEAN SIX SiGMA .... ...... .. ........ 11-2 HYPOTHESIS TESTING ............ .. . .... . .... VII-24
ORIGINS OF SIX SIGMA & LEAN . ............. ... 11-7 BASIC CONCEPTS ................. ......... VII-24
LEAN PIONEERS ............... .............. 11-12 POINT & INTERVAL ESTIMATES •.•••••••..• ••. VII-32
QUALITY & SIX SIGMA GURUS • . • . . • . • • • . • • • . •• 11-22 MEANS TESTS. .. .. .. . • . • . • . • • . • .. • . • • . .. ... VII-35
ORGANIZATIONAL LEADERSHIP ...• . ••.•.•.•.• U-39 Z TEST ...•.•.•••.•.•....•.•. ...••.••••. VII-35
METRICS AND GOALS ••..•.•.••. ... . • ••••. ..• 11-42 t TEST . . • • • . . . • . • • • . • . • . . • . • • • • . • . . . • • .. VII-37
REFERENCES ................ ..... . ......... 11-48 PAIRED t TEST .•••••••.••.•.• .•• .• •••••.. VII-39
P TEST .•••.•...• • .• •••• . • • .•••••...••.. VII-40
III. LSS PROJECT MANAGEMENT ..•••••••.••••• ... . . 111-1 VARIANCE TESTS ............ ..... . .... ... .. VII-41
LINKING PROJECTS TO GOALS ••.••••••. . • .. ••• 111-2 CHI-SQUARE TEST •••••• • . ....... . ••.••• . VII-41
PROBLEM SOLVING METHODOLOGIES .• .••. . •.• 111-5 F TEST .•.••••••• • • •• ••. .•••.••••••..••. VII-47
PROJECT SELECTION .••••• • ••• • .•• .. .• • • •.•• 111-13 ANOVA •.•.••••••••.•.•.•••.••• • .•.•...••. . VII-51
STAKEHOLDER ANALySiS .. ........•.•• ... • 111-15 ROOT CAUSE ANALYSIS . . • . • . . .. . . • • • • .. .. . ... VII-57
CUSTOMER DATA ..•••••. .. ••..••.•••. ••• •. 111-17 REFERENCES .••.••.••.. .. ... . •••.•••••••• • •. VII-60
QFD .•• .•••••.•••. ••..........••.••.•.•••. 111-24
BENCHMARKING •• . .••.•...••.••. ••••.•• .. 111-27 VIII. IMPROVEMENT TECHNIQUES •••.••. •.•••.•.••••. VIII-1
RISK ANALYSIS .•...• • •.•.•...•...• • • ••.•••.. 111-30 ELIMINATING WASTES ...... .... . ............... VIII-2
PROJECT MANAGEMENT TECHNIQUES . ••. • ••• . 111-33 5S .•••••••••.•.•.•.•••. .••.••. •.•.•.•.••••. VIII-2
PLAN ELEMENTS ...•••.•.••••.•..••• .•••• . 111-33 KANBAN (PULL) •.•.••.•. . •. .••• •.•.••••••.. VIII-16
PLANNING TOOLS ................ ... .... .. 111-36 POKA-YOKE ••..•.•..•... •••.•.• ..•.••••••.• VIII-18
PROJECT DOCUMENTATION ••.•••.• •••••••• 111-48 SETUP REDUCTION ....... . . ................ VIII-21
REFERENCES ......................... .. ... . III-53 FLOW IMPROVEMENT •.•. ... . .•.•. . ••• ..•••• VIII-25
QUICK RESPONSE MFG ....... . .............. VIII-27
IV. TEAM ORGANIZATION & DYNAMICS ••• .•• •••.. . .. IV-1 KAIZEN •••.•••.•. .•.•.....•••.••••.•••.••••• . VIII-28
INITIATING TEAMS ............ ..... .. ...... . .. IV-2 THEORY OF CONSTRAINTS ••••• • •.••••• •••••• • VIII-29
TEAM ROLES ••.•••.••.•••••••• .•• •.••.•. .•.. IV-15 DESIGN OF EXPERIMENTS ...... . ....... . . .. ... VIII-34
TEAM STAGES .••••••.•.•••••• . . •• . •••.• . .• . IV-24 IMPLEMENTING & VALIDATING SOLUTIONS . . •.• • VIII-58
TEAM DYNAMICS ••. •••.•••••.•...... . •.•• ... IV-28 MEASUREMENT SYSTEM REANALYSIS •.. ••. .•.• VIII-68
CONFLICT RESOLUTION •.•...• •.••••••••.• .•• IV-37 REFERENCES ........ . ................ ...... . VIII-70
TEAM TOOLS .•.•••••..•.••••• •.••• . •••.••••• IV-40
PERFORMANCE EVALUATION •••.... • .• . ..•.•. IV-43 IX. CONTROL CONCEPTS .. .. . . . .. ................... IX-1
REFERENCES .•.••...•••.•.•. ••.••.•••••• .•• IV-46 QUALITY CONTROLS •• •• . ••• •.. •••.•• . ••.•••• ••. IX-2
CONTROL PLANS •.••• .••.••••.••.....• . • ••••.•. IX-5
V. DEFINING OPPORTUNITIES . ..................... V-1 CONTROL CHARTS ••.•••••.•••• •••..••••.. . •..• IX-11
PROJECT CHARTER ..•• ••.••.• ••.•• • ••.. •••.•• V-2 TOTAL PRODUCTIVE MAINTENANCE .•••• .• •••••• IX-40
A3 REPORT ••.•.•••.•.• . ..... .• .•• .••. ..• ••.. V-7 VISUAL SYSTEMS .....••.•.•.••.•.. .•.•..•.••.• IX-47
DEFINITION TOOLS. . . • . • . • . . • . • • • . • . • • • . . • • .. V-11 STANDARD WORK ................ .......... . .. IX-50
AFFINITY DIAGRAMS .......... ....... . ... .. V-11 TRAINING REQUIREMENTS .. .. ... .. ... . ..... .. .. IX-54
CAUSE-AND-EFFECT DIAGRAMS •••• . ••••. .. • V-13 REFERENCES •••.•.••••...••. . ••.••..•.. . ••••• IX-61
PARETO DIAGRAMS ••••.•.••..•••.••••••..• V-15
CUSTOMER INPUTS ••••.•.•..•.••• •. ••••.••. . V-19 X. DESIGN IMPROVEMENT .••.•••.•. ••.•.•.••.••••.•• X-1
LEAN THINKING •.••• • •.• ••• . ••• •.••••• ..•••• V-24 DFSSJlDOVlDMADV .. .. • . . • . • . .. • .. • .. . • • .. • .. . •• X-2
CYCLE TIME REDUCTION •.••••. • .••.••. . . .•.. V-30 QFD-DESIGN ••••..• . ••••••••..•• •• ••••••••••.• X-11
VALUE STREAM MAPPiNG .... ... . ...... ... . .. V-36 ROBUST DESiGN ..... . ..... .. . ................ X-16
PROCESS MAPPING ••••.. .•••.•••.• . • •.•••••. V-43 FMECAlFMEA ••• ••• ..• •• •••• .••••.••.•••.••••• X-18
SPAGHETTI DIAGRAMS ... . ... . .......... ..... V-49 DFX ••••• •• . •.• ..•• •.•• ••••••.••••...••.. • ...• X-25
REFERENCES •.••.•.••. •..•• . • ••••••• •••••.. V-51 TRIZ • ••••••.••....•••••••••..• ••.••••• •••••.• X-30
SYSTEMATIC DESIGN .••.•.•.•.••.•••••.. ••• ... X-32
VI. MEASUREMENT TECHNIQUES •...•••...••..• ••.• VI-1 CREATIVE DESIGN .••• • ..•••.• •••..••. •. . .•• . .. X-33
PROCESS ANALYSIS ••••.• •••.•••.•.•••.. • .•.• VI-2 REFERENCES ..•••••. • •.•..••. ••••••.•••.•.... X-37
WORK INSTRUCTIONS ••.•. ••.•• •••• ••.•• •••• VI-3
TAKT TIME •.••••••••...•• ••••• .• • . ••.••••.• VI-5 XI. MANUFACTURING AND SERVICE CASE STUDIES .•.•. XI-1
LSS METRICS •••..••.•.•••.•• • • •••••.•• .•.• VI-9 QCI SPAGHETTI ...•.•••. ••••..••.•••.•• . .• ...•. XI-2
DATA COLLECTION .......... ... ...... ... .... VI-15 SOFTWARE LICENSE IMPROVEMENT ..••.•. ••• ..•. XI-6
COLLECTION METHODS .... . .. .......... ... VI-18 LSS AT LUDOVICO . .. ....... . ........ . ...... ... XI-11
MEASUREMENT SCALES ... ....... .... .... . . VI-22 LEAN IN SERVICE ••••.•••• ••• . ••.•••• . ••••••. .. XI-20
MEASUREMENT SYSTEMS ••.••• •.•• . •••.••••. VI-26 MEDICAL CENTER A3 ....... .... . ............. .. XI-27
MEASUREMENT TERMS •••••••..••. ••••••.. VI-29
REPEATABILITY & REPRODUCIBILITY •• . •.••. VI-32 XII. APPENDIX .•• •••••••••• . •• .• ••••••• .•••.. •• .. . XII-1
PROCESS CAPABILITY ANALYSIS ••• •.•• ••••• . • VI-39 TABLES .••. ••••••.•••• ••• • .•• •••••.•••••••••• XII-2
NORMAL DISTRIBUTION ............. . . .... . VI-41 INDEX ...••..• • .•• ••••.••.. .• .... ....•.•••... XII-13
PROCESS CAPABILITY .••.•• ••••...•.•.• ••• VI-45 ANSWERS •• •• • • ••••••••••.• • .•..•.. . • . .••.•. XII-25
CAPABILITY INDICES .••.•••• ••••• . •••. .•.•• VI-53
REFERENCES .••••.•.••. . . . • .. ••••.•••• •• ••• VI-59

LSS Handbook Question Contents
Handbook
Handbook Section %LSS Questions
I. LSS Overview
II. Lean Six Sigma Goals 10% 40
III. Lean Six Sigma 10% 40

Project Management
IV. Team Organization 10% 40
V. Defining Opportunities 11% 44
VI. Measurement Techniques 12% 48
VII. Analysis Techniques 12% 48
VIII. Improvement Techniques 15% 60
IX. Control Concepts 12% 48
X. Design Improvement 8% 32
XI. Manufacturing and

Service Case Studies
XII. Appendix
Totals 100% 400
The fully explained solutions to all 400 questions are available through QCI in the
LSS Solutions Text.
Quality Council of Indiana

TEL: 800-660-4215
hHp:llwww.qualitycouncil.com

I. INTRODUCTION
Professional Education Certifications Online
A JOURNEY OF A THOUSAND MILES MUST BEGIN WITH A

SINGLE STEP.
LAO-TZU
VILLANOVA UNIVERSITY 1-1 LEAN SIX SIGMA HANDBOOK

I. INTRODUCTION
LEAN SIX SIGMA BIBLIOGRAPHY
Preface
This text is designed to provide professionals with a review of fundamental
knowledge and skills. It is also intended to be a Handbook for those interested in
taking the examinations offered by the Society for Manufacturing Engineers and
Villanova University on-line. It is anticipated that the American Society for Quality
will adopt a lean enterprise or LSS certification BOK in the Ilear future. Test
questions have been fabricated by the authors in most cases. They are provided at
the end of each Section and are printed on blue paper for easy distinction and
removal if required during an examination.
5IX 5IGMA CON5ULTANT

WE DON'T HAVE I HOPE
THE FIR5T 5TEP ANY PROBLEM5. 50MEONE
15 TO IDENTIFY WHAT'5 THE GIVE5 ME
YOUR PROBLEM5. 5ECOND 5TEP? A BELT.
I
A little Dilbert® six sigma humor with permission of Scott Adams and United
Feature Syndicate, Inc.
Lean Six Sigma Bibliography

Some of the bibliography sources recommended by the authors include:
Ballis, J. (2001). Managing Flow. Dallas: Brown Books.
Breyfogle, F.W., III. (2003). Implementing Six Sigma: Smartel' Solutions Using
Statistical Methods, 2nd ed. New York: John Wiley & Sons.
Deming, W.E. (2000). Out of Crisis. Cambridge, MA: The MIT Pre~;s.
Dennis, P. (2002). Lean Production Simplified. New York: Produc:tivity Press.
George, M. L. (2001). Lean Six Sigma. New York: McGraw-Hili.
Goldratt, E. (1992). The Goal. Great Barrington, MA: North River Press.
VILLANOVA UNIVERSITY 1- 2 LEA,N SIX SIGMA HANDBOOK

I. INTRODUCTION
LEAN SIX SIGMA BIBLIOGRAPHY
Lean Six Sigma Bibliography (Continued)

Harry, M. & Schroeder, R. (2000). Six Sigma: The Breakthrough Management Strategy
Revolutionizing the World's Top Corporations. Doubleday.
Hirano, H. (1995). 5 Pillars of the Visual Work Place. New York: Productivity Press.
Imai, M. (1986). Kaizen: The Key to Japan's Competitive Success. New York:
McGraw-Hili.
Juran, J.M. (1999). Juran's Quality Handbook, 5th ed. New York: McGraw Hill.
Kaplan, R. (1996). The Balanced Scorecard. Cambridge, MA: Harvard Business

Review.
Kobayashi, I. (1995). 20 Keys to Workplace Improvement. New York: Productivity

Press.
Levinson, W.A. & Rerick, R.A. (2002). Lean Enterprise: A Synergistic Apprt;)ach to
Minimizing Waste. Milwaukee: ASQ Quality Press.
Naumann, E. & Hoisington, S. (2001). Customer Centered Six Sigma. Milwaukee:

ASQ Quality Press.
Pan de, P.S., Newman, P.R., & Cavanagh, R.R. (2000). The Six Sigma Way. New York:
McGraw-Hili.
Rother, M. & Shook, J. (2003). Learning to See. Cambridge, MA: Lean Enterprise
Institute.
Sharma, A. & Moody, P. E. (2001). The Perfect Engine. New York: Free Press.
Shingo, S. (1995). A Study of the Toyota Production System. New York: Productivity
Press.
Tague, N.R. (2004). The Quality Toolbox, 2nd ed. Milwaukee: ASQ Quality Press.
Womack, J. P., Jones, D. T., & Roos, D. (1991). The Machine that Changed the World:
The Story of Lean Production. New York: Harper Perennial.
Womack, J. & Jones, D. T. (2003). Lean Thinking: Banish Waste and Create Wealth
in Your Corporation. New York: Free Press.
VILLANOVA UNIVERSITY 1- 3 LEAN SIX SIGMA HANDBOOK

I. INTRODUCTION
LSS BODY OF KNOWLEDGE
Lean Six Sigma

Body of Knowledge
This Lean Six Sigma (LSS) BOK was developed principall)' by the authors.
Consideration was given to three other sources:
• ASQ/Navy LSS BOK

• SME Lean Certification BOK
• Villanova Lean Six Sigma Training
I. Introduction
A. Bibliography
B. Lean Six Sigma BOK
C. Lean Six Sigma Glossary
II. Lean Six Sigma Goals
A. Value of Lean Six Sigma

Recognize the unique features of lean and six sigma methodologies as well
as the synergies that organizations are achieving by combining them into lean
six sigma.
B. Origins of Lean Six Sigma

Describe the origins of both lean and six sigma principles.
C. Lean Pioneers
Recognize the origins of various lean enterprise techniques (Taylor, Ford,
Toyoda, Ohno, Shingo, etc.).
D. Quality and Six Sigma Gurus

Identify and distinguish those gurus having a contribution tC) the business and
technical foundations of six sigma (Deming, Juran, Shewhart, Ishikawa,
Taguchi, Harry, etc.).
E. Organizational Leadership
Recognize the key organization roles and responsibilities in support of lean
six sigma. Describe how process inputs, outputs, and feE!dback impact the
enterprise system as a whole. Recognize the benefits c)f using strategic
balanced scorecards, SWAT analysis and benchmarkiing to determine
improvement needs.
VILLANOVA UNIVERSITY 1- 4 LEA.N SIX SIGMA HANDBOOK

I. INTRODUCTION
Lean Six Sigma Body of Knowledge (Continued)

F. Business Metrics and Goals
Identify the key drivers for most businesses (profit, market share, customer
satisfaction, efficiency, etc.) and recognize how these key metrics and
scorecards are utilized for improvement goals.
III. Lean Six Sigma Project Management
A. Linking Projects to Goals

Describe how projects are linked to organizational goals such as minimization
of variation, meeting customer needs, controlling costs, eliminating time
traps, etc.
B. Problem Solving Methodologies

Identify and describe such problem solving methodologies as PDCA, PDSA,
Classical, DMAIC, IDEA, and Ford 80.
C. Project Selection
1. Process Elements
Describe the impact that people, materials, energy, equipment, and
information have on project selection.
2. Stakeholder Analysis
Recognize the importance of stakeholders (suppliers, stockholders,
management, employees, customers, and society) on the viability and
impact of projects.
3. Customer Data
Describe the importance of internal and external customer data in the
creation of improvement projects. Identify how surveys, focus groups,
complaints, etc. can be used to gather this data.
4. QFD
Identify how quality function deployment can be used to ensure that
customers wants and needs are adequately heard.
5. Benchmarking
Recognize how process, performance, and strategic benchmarking can be
used for project selection. Define the basic benchmarking sequence.

I. INTRODUCTION

D. Risk Analysis
Describe the various phases of risk management such as risk planning,
assessment, analysis, and handling. Identify various risk management
processes such as identifying, analyzing, planning, tra,cking, controlling,
mitigation, and communication.
E. Project Management Techniques
1. Plan Elements
Identify and describe the stages of project manauement: planning,
scheduling, and controlling. Distinguish major project elements such as
project scope, milestones, goal statements, and required resources.
2. Work Breakdown Structure

Describe how a work breakdown structure is used to develop a detailed
listing of activities required to complete a project.
3. Planning Tools
Describe and differentiate the features of major project planning tools such
as PERT, CPM, Gantt Charts, and AND diagrams.
4. Project Documentation
Recognize project documentation techniques such as status reports,
milestone reporting, lessons learned, and document archiving.
IV. Team Organization and Dynamics
A. Initiating Teams
Recognize the importance of improvement teams to both the company and
individuals. Describe basic team objectives and the need for management
support. Identify a variety of team arrangements for both lean six sigma and
other objectives.
B. Team Roles
Define and describe the roles and responsibilities of partici[pants on both lean
six sigma and other teams including team members, sponsors, process
owners, black belts, green belts, champions, etc.
c. Team Stages
Describe the main stages of team evolution including f4lrming, storming,
norming, performing, and adjourning.
VILLANOVA UNIVERSITY 1- 6 LE)I,N SIX SIGMA HANDBOOK

I. INTRODUCTION

D. Team Dynamics
Recognize how team behaviors such as groupthink, floundering, feuding,
dominate and reluctant members, attributions, digressions, etc. can impact
team performance.
E. Conflict Resolution
Describe how communications, conflict resolution, and negotiation
techniques are essential for effective team performance.
F. Team Tools
Define and apply basic team consensus techniques such as brainstorming,
nominal group technique, multi-voting, etc.
G. Performance Evaluation
Describe how team performance can be assessed both during and at the end
of a project.
V. Defining Opportunities
A. Project Charter
Describe the elements and importance of a project charter.
B. A3 Report
Describe the applications of A3 reports as project definition tools.
C. Definition Tools
Apply common problem definition tools such as affinity diagrams, cause-and-
effect diagrams, and Pareto diagrams.
D. Customer Inputs
Translate customer feedback such as CTQ trees, survey analysis, VOC
techniques, and Kano analysis into opportunities for improvement.
E. Lean Thinking
Understand key lean thinking concepts such as value, value stream, value
flow, pull value, and perfection.
F. Cycle Time Reduction

Describe key considerations in the reduction of process cycle time and their
importance in the elimination of waste and improvement of capacity.
VILLANOVA UNIVERSITY LEAN SIX SIGMA HANDBOOK

I. INTRODUCTION

G. Value Stream Mapping
Define the benefits of VSM in viewing an entire system for improvement
opportunities.
H. Process Mapping
Recognize the importance of process mapping and its appllication to process
improvement. Contrast process mapping with the VSM technique.
I. Spaghetti Diagrams
Understand the use of spaghetti diagrams in depicting emlployee movement,
information flow, and work flows.
VI. Measurement Techniques
A. Process Analysis
1. Procedures and work instructions

Recognize how procedures can be used to measure and control process
performance.
2. Takt Time
Define how takt time measurement can form a basis fClr an improvement
in work flow. Describe the benefit of small batch sizes.
3. Lean Six Sigma Metrics

Understand the major six sigma metrics such as DPU, IJPMO, Yield, RTY,
and six sigma quality levels. Recognize those determinations that are
commonly used in lean improvement such as throughput rate, lead time,
and process cycle efficiency.
B. Data Collection
1. Types of Data
Recognize and differentiate between variable, attribute, and locational
data. Describe how attribute data can be converted to variable data.
2. Data Collection Methods

Describe how data forms, data coding, and check she4:!ts are utilized for
data collection.

I. INTRODUCTION

3. Measurement Scales
Describe and apply the four major measurement scales: nominal, ordinal,
interval, and ratio.
4. Data Accuracy
Recognize basic data accuracy considerations and describe the
importance of random sampling.
C. Measurement Systems
Recognize the need for measurement system analysis and gage R & R.
Describe measurement error and common measurement terms.
D. Process Capability Analysis
1. Normal Distribution
Describe the application of histograms. Understand the use of z values in
determining normal distribution information.
2. Process Capability Studies

Understand how process capability studies are valuable tools in
measuring both current status and future performance.
3. Capability Indices
Recognize process capability and process performance indices. Describe
the difference between long-term and short-term capability.
VII. Analysis Techniques
A. Seven Classical Wastes

Recognize and analyze the familiar forms of muda in the workplace:
overproduction, excess inventory, rejects, wasted motion, additional
processing, unnecessary transport, and waiting.
B. Variable Relationships
1. Multi-Vari Analysis
Create multi-vari studies and interpret the difference between positional,
cyclical, and temporal variation.
2. Linear Correlation and Regression

Understand the application of regression models for estimation and
prediction. Recognize the significance of the correlation coefficient.

I. INTRODUCTION

C. Hypothesis Testing
1. Fundamental Concepts
Recognize fundamental hypothesis testing concepts such as the null
hypothesis, test statistic, types of errors, one-tail i:lnd two-tail tests,
practical versus statistical significance and adequate !;ample size.
2. Point and Interval Estimation

Define and determine point estimates and confidence inltervals for various
sample sizes.
3. Means Tests
Apply various average based tests such as z, t, paired t, and p.
4. Variance Tests
Apply and interpret the results of variance based tests such as F and chi-
square.
5. Analysis of Variance (AN OVA)

Define and apply the ANOVA technique and interpret the results.
D. Root Cause Analysis

Recognize the importance of root cause analysis. Differential between
subjective tools and analytical tools.
VIII. Improvement Techniques
A. Eliminating Wastes
1. 5S: Apply the 5S housekeeping and workplace organiz,ation technique to

improve performance.
2. Kanban (Pull): Recognize how the kanban system is w;ed to minimize in

process inventories and contribute to just-in-time manufacturing.
3. Poka-Yoke: Identify a variety of poka-yoke mechanism:!; to mistake proof

products and processes.
4. Setup Reduction: Describe how set-up reduction expands production

capacities, reduces inventories, and minimizes wastes"
VILLANOVA UNIVERSITY 1-10 LEJI.N SIX SIGMA HANDBOOK

I. INTRODUCTION

5. Flow Improvement: Understand the role of continuous flow manufacturing
and other work flow improvements.
6. Quick Response Manufacturing: Identify and understand the quick

response strategy.
B. Kaizen
Identify how Kaizen tools and techniques can be utilized to accomplish
process improvement.
C. Theory of Constraints
Understand the theory of constraints. Recognize the key TOC terms -
throughput, inventory, and operating expenses. Describe the drum-buffer-
rope strategy.
D. Design of Experiments
Define and describe common DOE terminology. Apply the basic elements of
experimental planning and execution. Construct randomized, Latin square,
full factorial, and fractional factorial designs.
E. Implementing and Validating Solutions

Use improvement methods such as main effects analysis, multi-vari studies,
and post improvement analysis to implement and validate solutions.
F. Measurement Systems Re-analysis

Understand the need to improve measurement capability as process capability
improves.
IX. Control Concepts
A. Quality Controls
Describe the function of quality controls such as written procedures and work
instructions in directing product and process performance.
B. Control Plans
Define how control plans are developed and understand how they help hold
the gains from improvement activities. Identify who creates these plans and
maintains their use and effectiveness.
C. Control Charts
Describe the benefits of control charts (SPC) in controlling process
performance. Identify special and common causes. Construct and interpret
various types of control charts. Understand the pre-control technique.

I. INTRODUCTION
Lean Six Sigma Body of Knowledge (Conltinued)

D. Total Productive Maintenance (TPM)
Describe how total productive maintenance can be utilizE!d to minimize the
major negative losses to equipment effectiveness. Us~~ TPM metrics to
evaluate performance efficiency, operating rates, and ()verall equipment
effectiveness (OEE).
E. Visual Systems
Distinguish how visual displays can be effectively used to make problems
apparent, clarify targets for future improvement, and influence and direct
employee behavior.
F. Standard Work
Identify how standards and standard work techniques can be used to
minimize wastes and ensure more consistent performanc~~.
G. Training
Recognize the importance of employee training as both a preventive and
control technique. Identify the importance of management support, training
needs assessments, necessary resources, and other training fundamentals.
X. Design Improvement
A. DFSS
Understand how the basics of DFSS are applied. Distinguish between
techniques such as IDOV and DMAC.
B. Quality Function Deployment (QFD)

Describe how QFD can be used as a design tool.
C. Robust Design Concepts

Understand the robust design process.
D. FMEAlFMECA
Define and distinguish between FMEA and FMECA. De~;cribe design and
process FMEAs (DFMEAs and PFMEAs).
E. Design for X (DFX)

Understand how design constraints dictate design for COStl:; (DFC), design for
maintainability (DFM), design for assembly (DFA), and othler considerations.
F. Special Design Tools

Understand how special design tools such as TRIZ, systematic design, and
creative ideas can benefit product and process design.
VILLANOVA UNIVERSITY 1- 12 LE.lI.N SIX SIGMA HANDBOOK

I. INTRODUCTION
LEAN SIX SIGMA GLOSSARY
Lean Six Sigma Glossary*

5Ss - Refers to the five Japanese words: seiri, seiton, seiso, seiketsu, and shitsuke
which are shorthand expressions for the principal techniques of maintaining an
effective, efficient, workplace. A brief explanation of these terms follows:
Seiri (Sort) - Eliminating everything not required for the work being performed.
Seiton (Straighten) - The efficient placement and arrangement of equipment and

material.
Seiso (Sanitize) - Maintaining tidiness and cleanliness in the workplace.
Seiketsu (Standarize) - An ongoing, standardized, improvement process.
Shitsuke (Sustain) - Discipline with leadership.
5 Whys - A simple technique used to reveal the root cause (as opposed to the
symptoms) of a problem. This approach asks the question "why" until the root cause
is finally discovered.
Andon Board - A visual control device in a production area. It is typically a lit

overhead display, giving the current status of the production system and alerting
employees to emerging problems. The number of lights and their possible colors
can vary. However, the traditional colors and their meanings are:
Green - No problems
Yellow - Situation requires attention
Red - Production stopped; attention urgently needed
ANOV A - An abbreviation of analysis of variance.
Autonomation - Automation with a human touch. A partial English translation of

jidoka.
Cellular Manufacturing - An approach in which manufacturing work centers (cells)

have the total capability necessary to produce an item or group of similar items.
Continuous Flow Manufacturing (CFM) - Descriptive of material moving one piece

at a time, at a rate determined by the needs of the customer, in a smooth and
uninterrupted sequence, without WIP.
CPM - An event oriented, project planning technique meaning critical path method.
* Adapted from Wortman (2000)2 and Northwest Lean Network (2006)1.

I. INTRODUCTION
Lean Six Sigma Glossary (Continued)

CTa Tree - A tool to translate initial customer requirements into numerical or
quantified needs for a product or service.
Cycle Time - The normal time to complete a product or service op.~ration. This is not
the same as takt time.
DFSS - An acronym meaning design for six sigma.
DFX - An acronym meaning design for X, where X represents an attribute such as

service, assembly, or manufacture.
DMAIC - The core problem solving methodology used by many lean six sigma
companies. The term refers to the steps: define, measure, anallyze, improve, and
control.
DPMO - An acronym meaning defects per million opportunities.
DPO - An acronym meaning defects per opportunity.
DPU - A term meaning defects per unit.
Enterprise Resource Planning (ERP) - ERP adapts the techniques of MRPII to all
areas of an organization (as opposed to the manufacturing arena). ERP is usually
implemented as a comprehensive business software solution.
Error-proofing - A technique for preventing production errors by designing the

manufacturing process, equipment, and tools so that an operation literally cannot
be performed incorrectly (see poka-yoke).
EVOP - Stands for evolutionary operations in experimental desi~ln.
Flexible Manufacturing System (FMS) - An integrated manufacturing capability to

produce small numbers of a great variety of items at low unit costs. A FMS is also
characterized by low changeover time and rapid response times.
FMEA - A design review process referring to failure mode effect analysis.
FMECA - A design review process referring to failure mode effect c:riticality analysis.
FPY - A product yield term meaning first pass yield.
Gantt Charts - A form of bar chart used to display project planning activities.
VILLANOVA UNIVERSITY 1- 14 LE,lI,N SIX SIGMA HANDBOOK

I. INTRODUCTION

Gemba - The shop floor or work site.
Heijunka - A production scheduling/leveling tool, essentially used to distribute

kanban cards in an efficient manner.
Hoshi" Planning - A management policy or strategy deployment approach. A

method for establishing goals (and supporting policies) and ensuring that they are
the primary focus of the organization.
Inventory Turns - The number of times inventory is consumed in a given period.
Jidoka - A form of automatic inspection of each item produced. Production is halted

and workers are notified, if a defect is detected. Toyota expands the meaning of
jidoka to include the responsibility of all workers to function in a similar manner.
Just-in-Time (JIT) - A production scheduling concept that calls for any item needed
at a production operation (whether raw material or finished item) to be produced and
available precisely when needed (not earlier or later).
Kaizen - The philosophy that every process can and should be continually evaluated
and improved in terms of the time required, resources used, resultant quality, etc.
Kanban - A card or sheet used to authorize production or movement of an item.

When fully implemented, kanban operates according to the following rules:
1. All production and movement of parts and material takes place only as
required by a downstream operation.
2. The specific tool which authorizes production or movement is called a kanban.

This word means card or sign, but can refer to containers or devices.
3. The quantity authorized per individual kanban is minimal, ideally one. The
number of circulating kanban for an item is determined by the demand rate for
the item and the time required to produce or acquire it.
Kano Model - A methodology to analyze customer needs by considering

dissatisfiers, satisfiers, and delighters.
KPIV - Stands for key process input variables.

I. INTRODUCTION

KPOV - Stands for key process output variables.
Lean Enterprise - The efficient performance of all aspects of an organization, from

the beginning of the supply chain, thru the production prOC4!SS, including the
customer base.
Lean Manufacturing - The philosophy of continually reducing waste in all areas and
in all forms. This English phrase often refers to the Toyota production system.
Level Loading - The smoothing or balancing of the work load in all steps of a
process.
Line Balancing - The equalization of the cycle times for units of the manufacturing
process, through the proper assignment of workers and machine~; to ensure smooth
production flow.
Mistake-proofing - A manufacturing technique for providing a si!~nal when an error

is about to be introduced into the production process. This can be as simple as the
use of a checklist.
MRPIMRPII- Material Requirements Planning - A technique (usuailly augmented with

software) for planning production material requirements, based on historic usage,
historic production, delivery lead times, and economic order size costing. MRPII
software programs have the added capability for capacity planning, scheduling, and
shop floor control. The scheduling and shop floor components of MRPII are often
unreliable (unless perfect forecasts are available) and tend to introduce huge
overhead costs in terms of inventory and production lead times"
Muda (waste) - A Japanese term meaning any activity that consumes resources but
creates no value. Those activities and results that should be eliminated. Many
references cite the following seven categories of waste:
1. Excessive or early production

2. Delays
3. Most movement and transport
4. Poor process design
5. Most inventory
6. Inefficient performance of a process
7. Making defective items

I. INTRODUCTION

Mura - Inconsistency or variation.
Nagara - A smooth production flow, ideally one piece at a time, characterized by

synchronization (balancing) of production processes and maximum utilization of
available time. This includes overlapping operations where practical.
Nemawashi - Lining up your ducks. (A little humor)
Non-Value-Added - Those actions that the customer is not willing to pay for. Any
activity that does not add value to the product or service.
NPV - An acronym representing net present value. This calculation considers cash
flow, time, and interest rates.
One Piece Flow - The concept of reducing production batch sizes to a minimal
amount, preferably a single unit. This can have dramatic effects on raw material,
WIP, finished goods inventories, production lead times, quality, and costs.
PDCA - A general problem solving methodology representing the steps: plan, do,
check, and act.
PDSA - Deming's modified problem solving methodology representing the steps:

plan, do, study, and act.
Perfection - The complete elimination of muda so that all activities, along a value
stream, create value.
PERT - An event oriented, project planning technique meaning program evaluation

and review technique.
Point of Use Inventory - Inventory that is delivered to the location where it will be
consumed.
Poka-Yoke (Mistake-proofing) - A means of providing a visual or audible signal to

indicate a characteristic state. Often referred to as "error-proofing", poka-yoke is
actually the first step in truly error-proofing a system.

I. INTRODUCTION

Pull System - A manufacturing planning system based on the communication of
actual real-time needs from downstream operations. A pull sy:stem is in contrast
with push systems, which schedule operations according to theolretical downstream
results, based on "best-guess" planning, MRP, or other methods.
Queue Time - The time a product spends awaiting the next processing step.
ROA - A financial and project analysis term meaning return on clssets.
ROI- A financial and project analysis term meaning return on investment.
RTY - A product yield term meaning rolled throughput.
Seiban - Seiban is the name of a Japanese management practice taken from the
Japanese words II sei II , which means manufacturing, and II bun II , which means
number. A Seiban number is assigned to all parts, materials, and purchase orders
associated with a particular customer's job or project. This enables a manufacturer
to track progress.
Sensei - One who provides information; a teacher or instructor.
Setup Time - The time required to change over a machine or pro(:ess from one item
or operation to the next item or operation. This time can be divided into two types:
1. Internal: Setup work that can be done only when the machine or process is
not actively engaged in production.
2. External: Setup work that can be done concurrent to normal operations.
Shojinka - Continually optimizing the number of workers in a work center to meet the
type and volume of demand imposed on the work center. Shojink,a requires workers
trained in multiple disciplines and a supportive work centel' layout (such as
U-shaped or circular).
Single Piece Flow - A situation in which one complete product proceeds through
various operations like design, order taking, and production, without interruptions,
back flows, or scrap. This is in contrast with batch-and-queue a,rrangements.
SIPOC - A term implying a high-level process map focusing On! suppliers, inputs
processes, outputs, and customers.
VILLANOVA UNIVERSITY 1-18 LEJ~N SIX SIGMA HANDBOOK

I. INTRODUCTION

Six Sigma - A concept that implies a highly disciplined approach to deliver near-
perfect products and services on a consistent basis. The value originates from a ±
4.5 sigma that accomplishes a ± 1.5 sigma shift over time.
Skills Matrix - A work cell visual control depicting all work activities. It provides
assistance in the cross-training of team members.
Small Lot Principle - Effectively reducing lot size until the optimum of one piece flow
is realized.
SMED (Single Minute Exchange of Die) - Literally means changing a die on a

machine in a minute or less. Often, the key to doing this is by converting internal
setup time to external setup time. One common variation of SMED is the single digit
setup, which requires performing a setup activity in a single digit number of minutes
(fewer than ten).
Standard Work - A precise description of each work activity, specifying cycle time,
takt time, the work sequence of specific tasks, and the minimum inventory of parts
needed to conduct the activity.
Takt Time - Takt time is the available production time divided by the rate of customer
demand. For example, if customers want 480 widgets per day and the factory
operates 960 minutes per day, the takt time is two minutes. Takt time becomes the
heartbeat of any lean organization. Takt is a German term for rhythm. Takt time is
the rate at which customers demand a product and is not the same as cycle time.
Triage - A system used to sort workloads into categories in a service environment.

This method facilitates prioritization of the workloads by urgency, level of difficulty,
or length of activity, to reduce delays in performing the service.
TRIZ - A Russian abbreviation for "the theory of inventive problem solving." The
term is pronounced "trees." It consists of 9 action steps and some 40 basic
principles.
Value - From the perspective of the customer, value represents those aspects or
features of products or services that they are willing to pay for.
Value-Added - Those steps that transform raw materials or activities directly into the
features for which the customer assigns value.
Value Stream - The specific activities required to design, and provide a specific
product, from concept to launch, from order to delivery.

I. INTRODUCTION

Visual Control - The placement, in plain view, of all the tools, parts, production
activities, and indicators of production system performance, such that the status of
the system can be easily and quickly understood.
VOC - A term meaning an activity to hear the voice of the customer.
Waste - All overproduction ahead of demand, waiting for the ne):t processing step,
unnecessary transport of materials, excessive inventories, unne!cessary employee
movements, and production of defective parts.
Water Spider (mizusumashi) - An individual who performs a wide range of tasks

which allow other workers to perform value-added tasks.
WBS - An acronym meaning work breakdown structure. This activity expands an

improvement project into a detailed listing of activities.
WCM (World Class Manufacturing) - The philosophy of being thle best, the fastest,
and the lowest cost producer of a product or service. It implies the constant
improvement of products, processes, and services in order to remain an industry
leader.
WIP (Work-in-Process) - Inventory that exists (in batches) between workstations.
Work Cell - The layout of machines or business processes Ilf different types,
performing different operations in a tight sequence, (typically a U-shape or L-shape),
to permit single piece flow and flexible deployment of human ef1~ort.
Work Center - One process station in a work cell.
References
1. The Northwest Lean Networks retrieved from http://www.nwlean.netl
November 15, 2006.
2. Wortman, B.L.,et. al. (2001). CSSBB Primer. Terre Haute, IN: Quality Council
of Indiana.
VILLANOVA UNIVERSITY 1- 20 LE.lIN SIX SIGMA HANDBOOK

I
II. LEAN SIX SIGMA GOALS
OUR PLANS MISCARRY BECAUSE THEY HAVE NO

AIM. WHEN A MAN DOES NOT KNOW WHAT
HAR_BOR HE IS MAKING FOR, NO WIND IS THE
RIGHT WIND.
SENECA (4 B.C. - 65 A.D.)

VALUE OF LEAN SIX SIGMA
Lean Six Sigma Goals

The subject of Lean Six Sigma Goals is presented in the followil1g topic areas:
• Value of lean six sigma • Quality and six sigma gurus

• Origins of six sigma and lean • Organizational leadership
• Lean pioneers • Metrics and goals
Value of Lean Six Sigma

There is an ongoing debate in some organizations regarding the difference between
lean and six sigma, and whether they are mutually exclusive. T1oyota in particular
is credited with making lean a well-known approach as embodied in the Toyota
Production System (TPS). Lean is about eliminating wastes, taking time out of
processes, and creating better flow. Asked about the essence of lean (TPS), Taiichi
Ohno summarized it as, "All we're trying to do is shorten the timE! line ... from order
receipt to collecting the cash for the goods or services provided."
Six sigma has been defined in a variety of ways. One definition states, "Six sigma
is ... a business strategy and philosophy built around the concept that companies
can gain a competitive edge by reducing defects in their industriial and commercial
processes." (Harry, 2000)24
A few key characteristics of lean and six sigma are discussed and compared below.
There are some explanations from the points of view of lean and six sigma purists.
Topic Six Sigma Lean

Improvement Reduce Variation Reduce Waste
Justification Six Sigma (3.4 DPMO) Speed (velocity)
Main Savings Cost of Poor Quality Operating Costs
Learning Curve Long Short
Project Selection Various Approaches Value Stlream Mapping
Project Length 2 - 6 Months 1 Week - 3 Months
Driver Data Demand
Complexity High Moderat.!
Table 2.1 Comparison of Lean and Six Sigma Charact.!ristics

Value of Lean Six Sigma (Continued)

Both six sigma and lean focus heavily on satisfying customers. Six sigma makes
customers the primary driver for action in a "war on variation" and identifies
opportunities that promise a large, fairly immediate, financial reward. Lean
considers customer inputs and conducts a "war on waste."
One of the selling points that some six sigma gurus tout is that six sigma zeroes in
better on "big bang" improvements. Black belts are expected to target and achieve
large bottom line savings in projects every year.
Both six sigma and lean empower people to create process stability and a culture
of continuous improvement. The cornerstones of a lean strategy are tools such as
value stream mapping (VSM), workplace organization (5S), total productive
maintenance (TPM), kanban/pull systems, kaizen, setup reduction, teamwork, error
proofing, problem solving, cellular manufacturing, and one-piece flow.
Many problem identification and problem solving techniques are commonly used
with both lean and six sigma methodologies. These include brainstorming, cause-
and-effect diagrams, 5 "whys", Pareto analysis, 8-0s, FMEAs, and others. Both six
sigma and lean methodologies have a heavy emphasis on careful problem definition.
Six sigma better promotes a rigorous, systematic process to find the true root
cause{s) of the problem.
Value stream mapping (VSM) is the principal lean diagnostic tool. It is credited to
Toyota, who called it material and information flow mapping. The methodology was
developed into a viable tool for the masses by Rother and Shook in 1998 in the text
Learning to See (2003)50. VSM creates a visual representation of what is happening
in a process to improve system performance. Process mapping is a tool favored by
the six sigma community and is best used to identify the inputs, outputs, and other
factors that can affect a process. {Crabtree, 2004)9
Should six sigma and lean coexist in any organization? Ron Crabtree feels the
answer to this question is self-evident: Yes. He feels that lean approaches should
precede and coexist with the application of six sigma methods. Why? Put simply,
lean provides stability and repeatability in many basic processes. Once stability has
taken hold, much of the variation due to human processes goes away. The data
collected to support six sigma activities thereby becomes much more reliable and
accurate.


In management presentations, Ron Crabtree (2006)10 depicts IE!an and six sigma
tools on a linear continuum with lean six sigma in the middle. Mr. Crabtree asks if
major business problems fall into the following categories:
• There seems to be a lot of waste

• There is a need to minimize inventories and redundancies
• There is a need to improve work flows
• There is a need to speed up processes
• There are human mistakes
If so, then lean tools should be utilized to:
• Eliminate wastes • Simplify processes

• Increase speeds • Improve flows
• Minimize inventories • Mistake proof processes
However, if organization challenges exhibit the following attributes:
• There are quality issues

• There is excessive variation
• There are complex problems
• There are challenging root cause identifications
• There are numerous technical considerations
In these cases, six sigma tools should be utilized to:
• Minimize variation
• Apply scientific problem solving
• Utilize robust project chartering
• Focus on quality issues
• Employ technical methodologies
Most executives recognize that they have a combination of both sets of issues.
Placing lean six sigma in the middle of this continuum reflects a more holistic and
synergistic approach. If a specific problem requires only lean clr six sigma tools,
then that is perfectly ok. Lean six sigma is a relatively new p~lradigm providing
broader selection approaches. If the only tool in a company's bag is a hammer, then
all problems start to look like a nail. It is best to have a tool kit with a broader set of
tools, principles, and ways of thinking. (Crabtree, 2006fo (Crabtree, 2004)9


What has been occurring for some time (at least the past several years) is a marriage
of lean and six sigma initiatives into a unified approach called lean six sigma or
some variant of this nomenclature. Presented graphically, if lean specific projects
represent a 6% corporate improvement over time, and six sigma initiatives represent
another 6% improvement, then a combination could potentially represent an
improvement of 12% (or more). Refer to Figure 2.2.
12 LEAN SIX SIGMA LEAN SIX SIGMA
!zw 10
:E 8
w
>
0
a: 6
a.
~ 4
~
0
2
TIME-t TIME ... TIME ...
Figure 2.2 The Graphical Marriage of Lean and Six Sigma
Various authorities tout improvements (margins, inventory reductions, waste

eliminations, etc.) ranging from 2% to 20%. These percentages depend upon the
industry and the initial measurement base.
Quality Digest (November, 2006}48 cites research from Avery Point Group (a search
firm specializing in lean and six sigma placement) indicating that lean and six sigma
are destined for eternal togetherness. According to Avery Point Group
approximately one-half of employers are looking for employees with both lean and
six sigma skill sets.
Six years ago, books published on the combined use of lean and six sigma were
virtually nonexistent. Today, they represent almost one-half of the lean books and
25% of the six sigma books published. Tim Noble, manager of the Avery Point
Group states, "Those companies that perpetuate the divide between six sigma and
lean are clearly missing the point. The two are clearly complementary tool sets, not
competing philosophies."


An increasing number of organizations (manufacturing, service, hospitals,
municipalities, military, insurance, etc.) have been unifying their efforts into a lean
six sigma approach. The mechanisms ofthese combinations vary widely. The most
effective approaches include management direction and involvement, a cadre of
trained specialists, the use of teamwork, the use of project management, team
member training, the humane treatment of people, an understandable problem
solving methodology, and some mechanism to apply the appropriate tool(s).
On the following pages are additional descriptions of six sigma and lean enterprise.
This book unifies the discussion of lean six sigma by use of the DMAIC problem
solving approach. Obviously, other systems would work equally well, as long as
they are communicated and known to the organization. Refer to ·rable 2.3 below for
some applications of the various lean six sigma tools at various problem solving
stages.
Define Measure Analyze Improve Control

Value Stream Prioritization Regression DOE SPC
Mapping Matrices Analysis
Charter - MSA Studies 5 -Whys Kaizen Visual
Problem Events Controls
Statement
Voice of the Capability Cause· Effect TOC Control Plans
Customer Studies Diagrams
Communication Videotaping Root Cause Pull Systems TPM
Plans Analysis
CTQ Issues Time Studies AN OVA SMED/SUR Standard Work
Business SIPOC Multi-Vari 5S or6S Procedures and
Results Analysis Work
Instructions
Benchmarking Collecting Hypothesis Work Flow Training
Data Testing Improvement Requirements
(Missouri Enterprise, 2007)40
Table 2.3 Lean Six Sigma Tools in a DMAIC Matrix.
The student should note that there are a multitude of effective t()ols in addition to
those listed above.
VILLANOVA UNIVERSITY 11·6 LEAN SIX SIGMA HANDBOOK

ORIGINS OF SIX SIGMA AND LEAN
Six Sigma Introduction

Six sigma is a highly disciplined process that focuses on developing and delivering
near-perfect products and services consistently. Six sigma is also a management
strategy to use statistical tools and project work to achieve breakthrough
profitability and quantum gains in quality. It has been stated that product
characteristics with six sigma process capabilities (Cpk > 1.5) are of world-class
performance. The average American company is at a four sigma level. (Harry,
1998)23. This would be the equivalent of 0.6% defective or 6,210 defects per million
opportunities. Snee (1999)59 describes six sigma as, "A business improvement
approach that seeks to find and eliminate causes of mistakes or defects in business
processes by focusing on outputs that are of critical importance to customers."
Motorola, under the direction of Chairman Bob Galvin, used statistical tools to
identify and eliminate variation. From Bill Smith's yield theory in 1984, Motorola
developed six sigma as a key business initiative in 1987. Many credit the resulting
improvements as a key factor in Motorola winning the Malcolm Baldrige Award for
Business Excellence in 1988. Dr. Mikel Harry, who had led the corporate effort,
subsequently left Motorola and later founded the Six Sigma Academy. The purpose
of the Six Sigma Academy is to accelerate the efforts of corporations to achieve
world-class standards. (Harry, 1998)23
Sigma is a statistical term that refers to the standard deviation of a process with
regard to it's mean. In a normally distributed process, 99.73% of measurements will
fall within ± 3.0 sigma and 99.99932% will fall within ± 4.5 sigma.
Motorola noted that many operations, such as complex assemblies, tended to shift
1.5 sigma over time. So a process, with a normal distribution and normal variation
of the mean, would need to have specification limits of ± 6 sigma in order to produce
less than 3.4 defects per million opportunities. This failure rate can be referred to
as defects per opportunity (OPO), or defects per million opportunities (OPMO).
Motorola® is a registered trademark of Motorola, Inc.

Six Sigma Introduction (Continued)1

Figure 2.4 illustrates the ±1.5 sigma shift and Table 2.5 provides some indications
of possible defect levels.
+-~~~~~~--~----~--~~~~~
-6 -4 -2 o 2 4 6
Figure 2.4 The ± 1.5 Sigma Shift
Sigma Level ppm

6 sigma 3.4 ppm
5 sigma 233 ppm
4 sigma 6,210 ppm
3 sigma 66,810 ppm
2 sigma 308,770 ppm
1 sigma 697,672 ppm
Table 2.5 Defect Levels
Note that Table II in the Appendix provides defect levels at other sigma values.
Various authors report slightly different failure rates based mainly upon rounding
effects and slight miscalculations. Most of the differences occur at levels less than
3 sigma. However, in looking at this situation objectively, companies with less than
3 sigma capability, and with ± 1.5 sigma shifts, probably won!'t be around long
enough to undertake a six sigma improvement effort anyway.
It should be noted that the term "six sigma" has been applied tOI many operations
including those with distributions that are not normal, for which a calculation of
sigma would be inappropriate. The principle remains the same, dl~liver near-perfect
products and services by improving the process and eliminating defects. The end
objective is to delight customers.

Six Sigma Introduction (Continued)

The six sigma steps for many organizations are described as DMAIC:
Define: Select the appropriate responses (the "Ys") to be improved.
Measure: Data must be gathered to measure the response variable.
Analyze: Identify the root causes of defects, defectives, or significant

measurement deviations whether in or out of specifications. These
are the "Xs" (independent variables).
Improve: Reduce variability or eliminate the cause.
Control: With the desired improvements in place, monitor the process to

sustain the improvements.
Modified from (Hahn, 1999)22
Harry (2000)24 proposes that the entire six sigma breakthrough strategy should
consist of the following eight elements:
R Recognize the true state of your business

o Define what plans must be in place to realize improvement of each state
M Measure the business systems that support the plans
A Analyze the gaps in system performance benchmarks
I Improve system elements to achieve performance goals
C Control system-level characteristics that are critical to value
S Standardize the systems that prove to be best in class
I Integrate best in class systems into the strategic planning framework
Because of the integration of a number of tools, such as lean manufacturing, DOE

(design of experiments), and DFSS (design for six sigma), six sigma has been
referred to as TQM (total quality management) on steroids.
The business successes that result from a six sigma initiative include:
• Cost reductions • Productivity improvements

• Market share growth • Customer relations improvements
• Defect reductions • Product and service improvements
• Culture changes • Cycle time reductions
(Pande, 2000)44

Six Sigma Results

Motorola credits the six sigma initiative for savings of $940 million over three years.
AlliedSignal (now Honeywell) reported an estimated $1.5 billion in savings in 1997.
GE has invested a billion dollars with a return of $1.75 billion in 1998 and an
accumulated savings of $2.5 billion for 1999. (Hahn, 1999)22
Harry (1998)23 reports that the average black belt (or green belt) project will save
about $175,000. There should be about 5 to 6 projects per year, pIer black belt. The
ratio of 1 black belt per 100 employees, can provide a 6% cost rt!duction per year.
For larger companies, there is usually 1 master black belt for every 100 black belts.
Snee (1999)59 provides some reasons why six sigma works:
• Bottom line results

• Senior management is involved
• A disciplined approach is used (DMAIC)
• Short project completion times (3 to 6 months)
• Clearly defined measures of success
• Infrastructure of trained individuals (black belts, green belts)
• Customers and processes are the focus
• A sound statistical approach is used
Organizations that follow a six sigma improvement process for :several years find
that some operations achieve greater than six sigma quality. When operations reach
six sigma quality, defects become so rare that when defects do occur, they receive
the full attention necessary to determine and correct the root caUSie. As a result, key
operations frequently end up realizing better than six sigma quallity.
Companies that have embraced six sigma include:
• Motorola • AlliedSignal
• General Electric • Black & Decker
• Dupont • Dow Chemical
• Polaroid • Federal Express
• Kodak • Boeing
• Sony • Johnson & Johnson
• Toshiba • Navistar
GE is a registered trademark of General Electric Corporation.

AlliedSignal and Honeywell are trademarks of Honeywell International, Inc.
VILLANOVA UNIVERSITY II - 10 LEAN SIX SIGMA HANDBOOK

Lean Enterprise *
The lean enterprise encompasses the entire production system, beginning with the
customer. It includes sales outlets, the final assembler, product or process design,
and all tiers of the supply chain (including raw materials). Any truly lean system is
highly dependent on the demands of its customers and the reliability of its suppliers.
No implementation of lean manufacturing can reach its full potential without
including the entire enterprise in its planning.
Lean Manufacturing I Lean Production

Lean techniques are, in their most basic form, the systematic identification and
elimination of wastes, the implementation of the concepts of continuous flow, and
customer pull. The touted benefits of lean production systems include lower
production costs, fewer personnel, quicker product development, higher quality,
higher profitability, and greater system flexibility. By continually focusing on waste
reduction, there is truly no end to the benefits that can be achieved.
Generally, five areas drive the lean producer: cost, quality, delivery, safety, and
morale. Just as mass production is recognized as the production system of the 20th
century, lean production is viewed as the production system of the 21 st century.
Typically, Japanese terms are used in defining lean principles in order to convey
broad concepts with iconic (representative) terminology. Once properly explained,
the term "kanban" can be more descriptive than "those little cards which help
control product moves." However, use of these terms can have a negative effect,
especially if the culture of a particular organization is predisposed against all things
non-American. One should choose carefully the training methods (and terms) for
conveying lean tools and methods.
Lean Techniques in Service

Are lean techniques applicable in a service-oriented industry or office environment?
Every system contains waste. Whether one is producing a product, processing a
material, or providing a service, there are elements which are considered waste. The
techniques for analyzing systems, identifying and reducing waste, and focusing on
the customer are applicable in any system, and in any industry. Any implementation
of lean techniques will be different, depending on various factors such as industry,
internal culture, and internal business considerations. The tools used to implement
lean operations, and the order in which one combines them, are highly dependent
on whether a company is a discrete manufacturer, continuous producer, or provider
of a service.
* Modified from Northwest Lean Networks, (200St1

LEAN PIONEERS
Lean Pioneers
The following is a list of major contributors to the concept of lean enterprise.
.
Lean Pioneer Contribution
Frederick W. Taylor Wrote Principles of Scientific Management
Divided work into component parts
Was the foremost efficiency expert c.f his day
Applied scientific methods to maximize output
Henry Ford Known as the father of mass production
Advocated waste reduction
Founded Ford Motor Company
Brought affordable transportation to the masses
Sakichi Toyoda Known as a hands-on inventor
Developed the jidoka concept
Initiated the Toyota Motor Company (TMC)
Kiichiro Toyoda Continued the work of his father Sakichi
Promoted mistake proofing concepts
Became president of Toyota Motor Company
Eiji Toyoda Was the cousin of Kiichiro Toyoda
Developed an automotive research lab
Hired outstanding people within TMC
Became the Chairman of TMC
Taiichi Ohno Created the Toyota production syste!m (TPS)
Integrated the TPS into the supply chain
Had the vision and focus to eliminatl~ waste
Shigeo Shingo Developed the SMED system
Assisted in the development of otherr TPS elements
James Womack Well-known promoters of lean enterprise
Daniel Jones Co-authors of major lean thinking beloks
Anand Sharma CEO of TBM Consulting Group
Author of prominent books on lean Emterprise
Michael L. George Widely known for lean six sigma boelks
Founder of the George Group
Figure 2.6 Matrix of Lean Pioneers and Their Contributions
VILLANOVA UNIVERSITY II - 12 LEJ!,N SIX SIGMA HANDBOOK

LEAN PIONEERS
Frederick Winslow Taylor (1856-1915)

Peter Drucker called Frederick Taylor "One of the most important influences of the
20th century and most lasting since the Federalist papers." In 1977, the American
Management Association considered the top 71 contributors to management thought
and practice and Taylor ranked first with 31 votes. Henry Ford received the second
most votes, with 3. After World War II, Mr. Taylor's books were considered vital to
the economic recovery in Japan.
Frederick Taylor was born into a wealthy Philadelphia family, but chose the career
of an engineer. He started working as an apprentice in a machine shop and became
a foreman. In his spare time, he obtained a Mechanical Engineering degree from
Stevens Institute of Technology in 1883. From the very beginning, he focused on
improving the work methods and the efficiencies of the shop. This characteristic no
doubt led him to develop his own system and thus be called "The father of scientific
management. "
Frederick Taylor was the first efficiency expert; the original time and motion study
specialist. He applied scientific methods to obtain maximum output. This was
accomplished by having management in control of the workplace and by detailing
the minute routine of the worker. Through operations analysis, Taylor took away job
complexity. That is, he could now take a person from the street and train that person
to do a simpler operation. Work now required less brains, less muscle, and less
independence. "Taylorism" was "the application of scientific methods to obtain
maximum efficiency in industrial work." (Kanigel, 1999)31
In his book, The Principles of Scientific Management, Taylor emphasized that the
employer and the employee must both "prosper" through his system. One can have
high employee wages and low manufacturing costs. Maximum prosperity can exist
as the result of maximum productivity. Some key Taylor concepts are:
• Understand each element of the task

• Select, train, and develop the worker
• Have a division of work between management and worker
• Cooperate with the worker to follow the procedures (Taylor, 1911, 1998)61
Some selected highlights of Frederick Taylor's career are:
• 1878-1890: Midvale Steel (earliest contributions)

• 1883: BSME, Stevens Institute of Technology
• 1898-1901: Consultant to Bethlehem Steel
• 1901-1915: Spokesman and author on scientific management
• 1906: President of ASME (American Society of Mechanical Engineers)

LEAN PIONEERS
Henry Ford (1863 -1947)

Henry Ford was born and grew up on a farm near Dearborn, Michigan. Early in his
career, Mr. Ford worked as an apprentice machinist, a sawmill operator, and an
engineer. In 1893, he became chief engineer for Edison Illuminating Company,
before leaving to found Ford Motor Company.
The Ford Motor Company was founded in 1903 with the introduction of the Model A.
By 1908, after 20 design changes, the Model T was created. Mr. I=ord had a vehicle
that was designed for both the ease of manufacture and use. ThE! vehicle had parts
with interchangeability and simplicity. The common man was able to drive and
repair his own car. IN 1927 a second Model A was launched to meet the features
offered by other U.S. competitors.
Henry Ford was the master of "mass production." The successful implementation
of the assembly line at the Highland Park, Detroit plant in 1913 reduced costs and
increased productivity for Ford Motor Company. The reduced malnufacturing costs
made cars more affordable for Americans. (The Henry Ford website)63
In 1908, the workers required an average station task time of 514 minutes. With
improved work techniques and time/motion studies, the average task was reduced
to 2.3 minutes. In 1913, the introduction of the assembly line pushed the average
task cycle time down to 1.19 minutes. This was accomplished by reducing the
complexity of the task. The operator did not need to be a skilled craftsman. It must
be noted that the use of the assembly line resulted in a labor turnover rate of 380%
in the beginning of 1913, and 900% by year's end. On January 4, 1914, wages were
doubled to $5.00 per day. The increased wages resulted in a much improved
retention rate. In 1915, the Highland Park Plant had 7,000 workers. There were 50
different languages spoken at the plant. Therefore, the reduced complexity of the
task aided in the training of new workers.
Mr. Ford went beyond just managing the internal resources of the plant. He sought
to reduce costs and increase productivity by controlling the costs of raw materials.
The River Rouge plant near Dearborn, Michigan was a great eJ(ample of vertical
integration. Ford Motor Company, had a steel mill for producing steel, a glass
factory for windshields, rubber plantations in Brazil and iron ore mines in Minnesota.
Ford owned the ships that carried the ore.
(Womack, Jones, & Roos, 1990)65, and (Kanigel, 1999)31
VILLANOVA UNIVERSITY II - 14 LE,e,N SIX SIGMA HANDBOOK

LEAN PIONEERS
Henry Ford (1863 - 1947) (Continued)

Some basic facts on the River Rouge plant for the year 1930:
1. 81,000 employees
2. 6,952,000 square feet of production space
3. $268,991,552 in investment costs (The Henry Ford websitet3
Mr. Ford was an advocate of reduced waste in every operational area. Some
examples include:
• Using straw from his farm to make "Fordite" for steering wheels
• Reworking and reusing worn steel rails
• Remelting scrap steel at the River Rouge plant
• Reworking broken tools and equipment
• Converting used paper, rags, and hardwood into binder board
Mass production techniques involved the interchangeability of parts,

interchangeability of workers, simpler tasks, and better organization. These
techniques were widely used for 60-70 years and were adopted by companies in
, North America and Europe.
Some highlights of Henry Ford:
1. 1896: Built his first automobile, the Quadricycle

2. 1903: Founded Ford Motor Company, serving as the Vice-President
3. 1913: Started the first moving assembly line at the Highland Park Plant
4. 1918: Constructed the worlds largest industrial complex (River Rouge Plant)
5. 1919: One of every three cars purchased is a Model T
6. 1927: The 15 millionth Model T was produced (Ford, 1926, 1988)19
Sakichi Toyoda (1867 - 1930)

Sakichi Toyoda was a businessman and was called the "king of inventors." He was
a carpenter by trade and, thus, able to work with his hands. He had his first patent
in 1890, and in 1897 invented the first Japanese power loom. Because other
members of the Toyoda family and friends were in the cottage industry of weaving,
this led him to try to reduce the amount of manual labor and effort required for
weaving. For his efforts he used a steam engine as the source of power for the
looms. As an engineer, he spent many hours working and reworking the steam
engine to operate properly and then to link to the looms to obtain a power loom.

LEAN PIONEERS
Sakichi Toyoda (Continued)

A prime concept used at the Loom Works was jidoka (automation with a human
touch). This invention was designed to stop the loom whenever a thread broke. A
human did not always have to be present to oversee each loom. This enabled
workers to handle more than one loom and provide more value-added work. He
followed up with the founding of Toyoda Automatic Loom Works in 1926. In 1929,
the sale of the patent rights to the Platt Brothers (England) for 100,000 British
Pounds (one million Yen or $500,000 US) provided the research funds for entry into
the automotive industry. {Liker, 2004)~'7, {Toyoda, 1987)62
Kiichiro Toyoda (1895 -1952)

Kiichiro Toyoda was the son of Sakichi Toyoda and Second President of Toyota
Motor Company. He was a mechanical engineering graduate of the Tokyo Imperial
University with a focus on engine technology. In 1929, Kiichiro Toyoda went to
England and negotiated the patent rights to the "mistake proof" loom. The funds
from the sale helped to finance the automotive efforts of ToyodiEl Automatic Loom
Works. Kiichiro Toyoda was able to make a study tour of US auto plants in 1929,
followed by his own research efforts on motor vehicles in 1!e30. In 1935, the
company was able to produce three model A-1 passenger cars. During that year, the
government mandated the building of trucks, causing the passenger car activities
to end. There were 18 Model G-1 trucks built by the end of 1935.
The Toyota Motor Company (TMC) was spun-off as a separate company in 1937.
From the beginning the concept of just-in-time production was u!,ed. Due to lack of
materials this concept had to be used for economics and to increase cash flow. Mr.
K. Toyoda was very much influenced by his trips to Ford plants and by seeing the
supermarket process of restocking goods on the shelves. Toy01ta Motor Company
faced bankruptcy during the post war years due to inflation and credit management
problems. The situation even led to the layoff of workers and to a series of strikes.
In a classic show of the sense of obligation and responsibility, Kiichiro Toyoda took
responsibility for this failure and resigned as President. By 1950, after 13 years of
manufacturing, 2,685 automobiles had been produced by TMC, compared to 8,000
per day from the Ford River Rouge plant. Kiichiro Toyoda was asked to return as
President in early 1952, but died suddenly within the year at the age of 57.
{Womack, 1990t5 , {Liker, 2004)37, {Toyoda, 1987t2

LEAN PIONEERS
Eiji Toyoda (1913 - 1999)

Eiji Toyoda was a younger cousin to Kiichiro Toyoda. He also attended Tokyo
Imperial University studying Mechanical Engineering (1933 -1936). Upon graduation,
he was persuaded to join his cousin's business and started a research lab called the
"car hotel." This garage, housed Eiji and his staff as they conducted research on
engines, repaired cars and worked on other special projects. Eiji Toyoda was
drafted into the army, but released back to industry within 2 months. He served in
the auto business during the war effort making trucks. He became a Director of
Toyota Motor Company in 1945, and Managing Director in 1950. That was a bad year
due to labor strife and the resignation of President Kiichiro Toyoda.
During 1950, he traveled to the United States for a 3-month tour of the auto plants
and their suppliers. This trip provided evidence to Eiji Toyoda that little Toyota
Motor Company could compete in the automotive arena, but not using the same
"mass production" techniques. There was waste in the system and TMC could build
a new system from that. (Note at this time that Toyota was producing 40 units per
day, while Ford Rouge was at 8,000 per day.) In 1955, Eiji Toyoda drove the first
"Crown" passenger car off the assembly line. The Crown is credited with
transforming TMC into a large company. E. Toyoda was President of Toyota Motor
Company from 1967 - 1982. During that time period he sponsored Taiichi Ohno's
hard work inside TMC. Upon the merger of Toyota Motor Company and Toyota
Motor Sales, he served as Chairman from 1982 - 1994.
(Liker, 2004)37, (Toyoda, 1987)62
Taiichi Ohno (1912- 1990)

Taiichi Ohno was the creator of the Toyota Production System. He graduated from
Nagoya Technical High School and joined Toyoda Spinning and Weaving in 1932.
In 1943, he transferred to Toyota Motor Company. By 1947 he managed the machine
shop where he experimented with parallel lines and/or L-shaped processes. Of
course, there was much resistance from the machine operators. Since he was from
the weaving company, he was aware of jidoka (automation with a human touch) and
used it productively in the auto company.
In the 1950s, he also toured the United States auto plants to view and evaluate the
"mass production" process. From the tour, Ohno learned that the mass production
system could achieve economies of scale and reduced costs, but the system was
still full of waste. The waste was present in the forms of over production, excess
inventory, long setup times, rework, etc. Earlier, Kiichiro Toyoda had set an
"impossible" goal for Toyota Motor Company to catch up with America. The initial
estimates of productivity were 9:1. That is, it took nine Japanese workers to equal
the productivity of one American. The adoption of the customized mass production
system with elimination of waste could be the method for catching up.

LEAN PIONEERS
Taiichi Ohno (1912- 1990) (Continued)

Some sample techniques of the Toyota Production System that Ohno and his team
developed:
1. Pull system (supermarket, 7-Eleven idea)

2. Muda (7 types of waste)
3. Quick die changes (from days and hours, to minutes and seconds)
4. Flexible job assignments
5. Removing non-value added work
6. Kanban methods
7. U-shaped cells
8. One-piece flow
9. Production leveling
Mr. Ohno had the vision and focus to uncover and eliminate waste both within
Toyota and their suppliers. He was fascinated by the obvious and would unravel
invisible problems. He stated that he could focus on a situation, mentally run the
process sequence forward and then in reverse to uncover probh~ms. Ohno would
immediately put his ideas to the test.
From 1950 on, as a manager and executive, and with the backing of President Eiji
Toyoda, he pushed and fought to install the concepts of lean throughout Toyota and
into the supply base. This was not an easy task. It was very difficult to overcome
the conventional wisdom that things are fine as they are. He had many clashes with
his superiors on the TPS, due to his "take-no-prisoners" approach. Upon extending
TPS throughout the supply base, Ohno retired from Toyota in 19~r8 as an Executive
Vice-President.
Fearing retribution on his assistants, Ohno formed the Shingijuts Consulting Group
as a way for his loyal assistants to leave Toyota. Shingijuts means "new
technology." The assistants included Yoshiki Iwata and Chihiro Nakao. The
consulting group has been active in the United States. They have maintained the
style of training used at Toyota. (Liker, 2004)37, (Ohno, 1988)42, (Womack, 1996)65
Some selected highlights of Mr. Ohno:
• 1947: Transferred to Toyota Motor Company

• 1954: Director of Toyota
• 1964: Managing Director of Toyota
• 1975: Executive Vice-President of Toyota
• 1950 - 1978: Relentless pursuit of the total elimination of waste
VILLANOVA UNIVERSITY II - 18 LE)!.N SIX SIGMA HANDBOOK

LEAN PIONEERS
Shigeo Shingo (1909 - 1990)

Shigeo Shingo was influenced by Frederick Taylor's book: The Principles of
Scientific Management. He first read a Japanese version in 1924. Shingo was one
of Japan's foremost consultants on manufacturing operations improvement. He has
written many books on improvement including:
• Revolution in Manufacturing: The SMED System

• Zero Quality Control: Source Inspection and the Poka-Yoke System
• Non-Stock Production
• The Toyota Production System from an Industrial Engineering Viewpoint
Shingo graduated in 1930 with a Mechanical Engineering degree from Yamanashi

Technical College and started to work at Taipei Railway Factory. In 1945, he became
a consultant to industry through the Japan Management Association (JMA). He
started performing quick die change work in 1950 at Toyo Industries.
By 1959, Shigeo Shingo formed his own consulting firm, Institute of Management
Improvements, and provided consulting throughout the Far East. Much of his work
centered on mistake-proofing, zero quality control, and supplier sourcing. It was
not until 1969, at the Toyota Motor Company when Taiichi Ohno demanded the
impossible, that the SMED (Single Minute Exchange of Die) concept really came to
life. Ohno's demand was to reduce setup changes from 1.5 hours to 3 minutes. It
had previously been 4 hours, so 3 minutes seemed impossible. But, within three
months the goal was accomplished.(Shingo, 1989)56, (Utah State University, 2006)8
Shigeo Shingo trained and consulted for TMC from 1954 to 1982. During that time
he conducted over 87 sessions involving over 2,000 students. While he was not a
Toyota employee, he was a consultant that assisted in the development of the
Toyota Production System. In 1988, he was awarded an Honorary Doctorate in
Business from Utah State University. The Shingo Prize was established by the
College of Business, Utah State University to promote leanl world-class business
practices to enable a company to compete globally. The first winner, in 1989, was
Globe Metallurgical, Inc., Cincinnati, Ohio.
Some selected career highlights of Mr. Shingo:
• 1930: Started work at Taipei Railway

• 1945: Became a consultant for Japan Management Association
• 1950: Conducted initial work on SMED
• 1959: Formed own firm, Institute of Management Improvement
• 1954 -1982: Consulted at Toyota Motor Company

LEAN PIONEERS
James Womack and Daniel Jones

James Womack and Daniel Jones have been linked together as researchers on the
capabilities of the automotive industry since 1979. MIT Professor Daniel Roos
recruited them for an aggressive study of the automotive industry titled: "The Future
of the Automobile." The "Future" was published in 1984 with one of the conclusions
showing a 3:1 productivity difference between Japanese and USA workers. This is
an incredible turnaround from Taiichi Ohno's initial estimate that the Japanese to
American worker productivity ratio was 1:9 in the 1950s.
Their report induced a total of 36 companies, agencies, and countries to support a

more intensive $5,000,000, 5-year study of the Toyota system and the rest of the
industry. This study led to the monumental book The Machine that Changed the
World. This book changed the world of mass production by detailing how lean
manufacturing produces products with perhaps one-half the resources as before
(human effort, space, investment, engineering, and time). Womack and Jones have
jointly published two more lean books: "Lean Thinking: Banish Waste and Create
Wealth in Your Corporation" and "Lean Solutions: How Companies and Customers
can Create Value and Wealth Together."
Dr. Womack received a Ph.D. in Political Science from MIT in 1982 and was a
research scientist at MIT from 1979 to 1991. Professor Jones was Professor of
Manufacturing at Cardiff University Business School in 1989 and Founding Director
of the Lean Enterprise Research Centre (1994-2001). Womacl< and Jones have
established a global network on lean manufacturing with individual networks in
America and in Europe. (Daniel T. Jones n.d.)27, (Kleiner 2006)35
Anand Sharma
Anand Sharma is President and CEO of TBM Consulting GroUlp, Durham, North
Carolina. He has been profiled by Fortune magazine as one of the "Heroes of U.S.
Manufacturing" (March 2001). In 2002, the Society of Manufa,cturing Engineers
awarded Mr. Sharma the Donald C. Burnham Manufacturing Award for achieving
manufacturing excellence without sacrificing human capital. His supporters state
that he is an expert who can figure out what is wrong with an organization by
walking the shop floor. He proclaims, "Where other people see complexity, I look
at how simple things can be." His company, TBM Consulting Group, employing over
70 employees, has worked with over 500 enterprises on improving manufacturing
productivity and profits. Mr. Sharma prides himself on refusing to work with firms
that will layoff workers due to use of his system.
VILLANOVA UNIVERSITY 11- 20 LEA,N SIX SIGMA HANDBOOK

LEAN PIONEERS
Anand Sharma (Continued)

Mr. Sharma is a graduate of the University of Roorke (India). After an initial position
in India, he moved to the U.S. working for a variety of companies. His last corporate
position, before consulting, was Vice President of Strategic Planning, at American
Standard. Mr. Sharma learned the Toyota Production System from the Shingijutsu
Group in Japan. Mr. Sharma has a M.B.A. from Boston University and has recently
co-authored two books on lean: The Perfect Engine: How to Win in the New Demand
Economy by Building to Order with Fewer Resources and The Antidote: How to
Transform Your Business for the Extreme Challenges of the 21st Century.
(Bylinsky, 2001)5, (Potts, 2001)46, (TBM Consulting Group)52, (Womack, 1996t5
Michael George
Michael George is Chairman and CEO of The George Group based in Dallas, Texas.
His company has worked with over 300 clients focusing on operational performance
and shareholder value through six sigma, lean six sigma, management of
complexity, and innovation efforts. Mr. George has a B.S. in Physics from the
University of California and a M.S. in Physics from the University of Illinois. His first
assignment was at Texas Instruments in 1964. In 1969, he founded International
Power Machine, which he sold to Rolls-Royce. The funds from the sale enabled him
to travel to Japan to study the Toyota Production System. The George Group was
formed in 1986. Mr. George is the holder of several patents on the reduction of
process cycle time and complexity. He has authored or co-authored a multitude of
lean six sigma books including: Fast Innovation, Lean Six Sigma, Lean Six Sigma for
Service, and Conquering Complexity in Your Business. (George Group, nd.)21
Shingijutsu CO., Ltd.

The premier consulting firm in lean manufacturing is Shingijutsu Co., Ltd. with main
headquarters in Gifu, Japan. In 1987, Taiichi Ohno helped form Shingijutsu Co.,
Ltd., as a business for his loyal deputies to gracefully leave Toyota. Yoshika Iwata,
Chihiro Nakao (President), and Akira Takenaka were the three founders of the new
firm. Shingijutsu in Japanese means "new technologies" The Shingijutsu mission
is to "find the way to combine people, material, and equipment most efficiently to
improve the financial condition of companies". A partial list of clients includes: Jake
Brake; United Technologies; Otis; Pratt & Whitney; Boeing; Porsche; Wiremold;
American Standard; and TBM Consulting. Shingijutsu strives to consult only with
companies that have a strong commitment to the struggle for continuous
improvement and will uphold "respect for people".
(Emiliani, 2006)17, (Sabatini, 2000)51, (Shingijutsu Co., Ltd., 2001)55

QUALITY AND SIX SIGMA GURUS
Quality and Six Sigma Gurus

Listed below are some well-known gurus and what they contriibuted to both the
business and technical foundations of six sigma. This list is far from inclusive.
I
Guru Contributi-o n I
Philip B. Crosby Senior management involvement
4 absolutes of quality management
Quality cost measurements
W. Edwards Deming Plan-do-study-act (wide usage)
Top management involvement
Concentration on system improveme!nt
Constancy of purpose
Armand V. Feigenbaum Total quality control/management
Top management involvement
Kaoru Ishikawa 4M (SM) or cause-and-effect diagram
Companywide quality control (CWQC)
Next operation as customer
Joseph M. Juran Top management involvement
Quality trilogy (project improvement)
Quality cost measurement
Pareto analysis
Walter A. Shewhart Assignable cause vs. chance cause
Control charts
Plan-do-check-act (as a design apprclach)
Use of statistics for improvement
Genichi Taguchi Loss function concepts
Signal to noise ratio
Experimental design methods
Concept of design robustness
Bill Smith First introduced the term "six sigma"
Mikel Harry The main architect of six sigma
Forrest Breyfogle '" Author of Implementing Six Sigma
Table 2.7 Major Contributors to the Six Sigma Knowledge Bank

(Wortman, 2001 t7
VILLANOVA UNIVERSITY 11- 22 LEt.N SIX SIGMA HANDBOOK

Philip B. Crosby (1928 - 2001)

Philip B. Crosby was vice-president of ITT for 14 years. In 1979, he founded Philip
Crosby Associates, Inc. in Winter Park, Florida. Mr. Crosby consulted, spoke, and
wrote about strategic quality issues throughout his professional life.
Philip Crosby started his career as a junior technician testing fire control systems
for B-47s. He eventually moved on to ITT and became one of the first corporate VPs
of quality in the country. He attributed his management training to Harold Geneen
and to the monthly general management meetings. It was Philip Crosby's deep
understanding of the concerns of management that made him akin to top
management. The other quality deep thinkers could be viewed as academicians, but
Crosby was considered a businessman. This explained the numbers of top
management that flocked to his quality college.
Crosby believed that quality was a significant part of the company and senior
managers must take charge of it. He believed the quality professional must become
more knowledgeable and communicative about the business. Crosby stated that
corporate management must make the cost of quality a part of the financial system
of their company.
Philip Crosby was a fellow and past president of ASQ. One of his most popular
statements on quality was: Quality is conformance to requirements.
A selection of some of Mr. Crosby's more popular books includes:
Quality Is Free: The Art of Making Quality Certain (1980)

The Art of Getting Your Own Sweet Way (1981)
Quality Without Tears: The Art of Hassle-free Management (Crosby, 1984)11
The Eternally Successful Organization: The Art of Corporate Wellness (1988)
Leading, the Art of Becoming an Executive (1990)
Completeness: Quality for the 21st Century (1992)
Running Things: The Art of Making Things Happen (1992)
Quality and Me: Lessons from an Evolving Life (1999)

Philip B. Crosby (Continued)

Philip Crosby preached four absolutes of quality management:
1. Quality means conformance to requirements
The requirements are what the customer says they are. There is a need to
emphasize a "do it right the first time" attitude.
2. Quality comes from prevention
Opportunities are available to correct problems in the system.
3. The quality performance standard is zero defects
You must insist on zero defects. Otherwise, it is acceptable to send out

nonconforming parts and goods. Ifthere is a nonconformance, then action
must be taken to eliminate and prevent it from occurring in the future.
4. Quality measurement is the price of nonconformance
A measurement of quality is needed to get managemEmt's attention,

prioritize problems, correct problems, and to measure prclgress.
The four absolutes of quality management are basic requirements for understanding
the purpose of a quality system. Philip Crosby also developed al 14 step approach
to quality improvement:
1. Management Commitment
2. Quality Improvement Team
3. Measurement
4. Cost of Quality
5. Quality Awareness
6. Corrective Action
7. Zero Defects Planning
8. Employee Education
9. Zero Defects Day
10. Goal Setting
11. Error Cause Removal
12. Recognition
13. Quality Councils
14. Do It All Over Again

Dr. W. Edwards Deming (1900 -1993)

Dr. Deming obtained a B.S. from the University of Wyoming, and a M.S. from the
University of Colorado and a Ph.D. in Physics from Vale. Dr. Deming was also
considered the founder of the third wave of the industrial revolution.
Dr. Deming was an honorary member of ASQ. He was awarded the ASQ Shewhart
Medal in 1955. During his life Dr. Deming published over 200 papers, articles, and
books. Notable books include:
Quality, Productivity, and Competitive Position (1982)

Out of the Crisis (Deming, 1986)14
W. Edwards Deming was the one individual who stood for quality and for what it
means. He is a national folk hero in Japan and was perhaps the leading speaker for
the quality revolution in the world. He did summer work at the Hawthorne plant
while working on his Ph.D. At the Hawthorne plant he became acquainted with W.
Shewhart and studied Shewhart's statistical methods.
The World War II effort enabled Deming to conduct classes in statistical methods to
thousands of American engineers, foremen, and workers. The statistical methods
were later credited to be a major factor in the war effort. But, as he would state it,
after the war, all traces of statistical methods were gone in a puff of smoke.
There were several visits to Japan between 1946 and 1948 for the purpose of census
taking. He developed a fondness for the Japanese people during that time. JUSE
invited Deming back in 1950 for executive courses in statistical methods. He refused
royalties on his seminar materials and insisted that the proceeds be used to help the
Japanese people. JUSE named their ultimate quality prize after him.
Deming would return to Japan on many other occasions to teach and consult. He
was well known in Japan, but not so in America. Only when NBC published its white
paper, "If Japan can, why can't we?" did America discover him. An overnight
success at age 80, W.E. Deming died at the age of 93. During his last 13 years,
Deming gave American industry a dose of strong medicine in quality. His message
to America is listed in his famous 14 points and 7 deadly diseases .
.
•

Dr. W. Edwards Deming (Continued)

The Fourteen Obligations of Top Management:
1. Create constancy of purpose for improvement of products and service
2. Adopt a new philosophy; we are in a new economic age
3. Cease dependence upon inspection as a way to achieve quality
4. End the practice of awarding business based on price tag
5. Constantly improve the process of planning, production, and service - this
system includes people
6. Institute training on the job
7. Institute improved supervision (leadership)
8. Drive out fear
9. Break down barriers between departments
10. Eliminate slogans/targets asking for increased productivity without providing
methods
11. Eliminate numerical quotas
12. Remove barriers that stand between workers and their pride of workmanship;
the same for all salaried people
13. Institute programs for education and retraining
14. Put all emphasis in the company to work to accomplish the transformation
Seven Deadly Diseases That Management Must Cure:

1. Lack of constancy of purpose to plan a marketable product and service to
keep the company in business and provide jobs
2. Emphasis on short-term profits
3. Personal evaluation appraisal, by whatever name, for people in management,

the effects of which are devastating
4. Mobility of management; job hopping
5. Use of visible figures for management, with little or no consideration of

figures that are unknown or unknowable
6. Excessive medical costs
7. Excessive costs of warranty, fueled by lawyers that work on contingency fees
Among other educational techniques, Deming promoted the parable of the red
beads, the PDSA cycle, and the concept of 94% management (system) causes versus
6% special causes. (Deming, 1986)14

"
Dr. W. Edwards Deming (Continued)

Deming's philosophy focused on individuals as fellow members of a system, and
treated people as partners, customers, neighbors, and friends. While the Deming
philosophy benefits the individuals in a firm, the company itself should be able to
reap the benefits from this new philosophy. Among the economic benefits to be
gained are:
• Reduction of the economic burden

• Expansion of markets
• Survival of organizations that serve customers
Deming's Chain Reaction

Deming shared the following chain reaction with Japan in the summer of 1950:
Improve quality ... Decrease costs (less rework, fewer delays) ... Productivity
improves ... Capture the market with better quality and price'" Stay in business ...
Provide jobs.
Deming's chain reaction is summarized by Delavigne and Robertson (1994)15 as the

following series of events:
1. The quality and productivity rise

2. Costs decrease
3. The time required for development and production is reduced
4. Management begins to know their cost, "they have a system"
5. Increased division of labor and specialization occurs
6. The near-term future is more predictable
7. The standard of living rises
8. The system has a future and can provide "jobs and more jobs"
As the above sequence is occurring, the marketplace is responding to the firm:
• The customer obtains reduced prices

• There is increased cooperation
• New products and services are provided to the customers
• There are higher levels of customer satisfaction
• There is a reduction of competition for share of the market
Thus, a "chain reaction" of good things can occur through the Deming philosophy.

Dr. Armand V. Feigenbaum (1920 - )

Dr. Feigenbaum is currently President of General Systems Company, Pittsfield, MA.
He was associated with General Electric for 26 years in engineering. Dr. Feigenbaum
holds a B.S. from Union College and an M.S. and Ph.D. from MIT.
Some of Dr. Feigenbaum's many distinctions include
Honorary Member, ASQ, 1986

E. Jack Lancaster Award, ASQ, 1981
Edwards Medal, ASQ, 1965
Fellow, American Association for the Advancement of Scienc:e
Life Member, IEEE and ASME
2-time President of ASQ 1961/63
Founding Chairman, International Academy for Quality
A few of Dr. Feigenbaum's many books:
Quality Control: Principles, Practice (1951)

Total Quality Control- Engineering and Management (1961)
Total Quality Control, 3rd ed (1983)
Total Quality Control, 40th Anniversary Edition (Feigenbaum, 1991)18
Mr. Feigenbaum is generally given credit for establishing the concept of "total
quality control" in the late 1940's at General Electric. His TQC statement was first
published in 1961, but at that time, the concept was so new, that no one listened.
A. V. Feigenbaum eventually formed his own consulting company in 1968 to provide

services in quality management and strategic planning to worldwide clients.
Feigenbaum states that the American industry must strive to bec::ome as strong as
it can be in its own marketplace. This has become valuable as global
competitiveness has spread into the U.S. Being strong at home via proper design,
production, selling and servicing will provide the potential for supremacy in the
marketplace.

Dr. Armand V. Feigenbaum (Continued)

The TQC philosophy maintains that all areas of the company must be involved in the
quality effort. The quality effort has generally only affected the shop floor people,
but must extend to all sections of the company. Products must not only be made
quicker and faster, but also sold faster. Feigenbaum noted that the quality
professional has an opportunity to become more than a functional specialist. The
opportunity is there to become a true businessman by providing valuable
information and direction.
The success of TQC includes these principles:
• TQC is a company wide process, all functions are involved

• Quality is what the customer says it is
• Quality and production costs are in partnership
• Higher quality will equate to lower costs
• Both individual and team zeal are required
• Management must provide a continuous and relentless emphasis on quality
• Quality and innovation can work together in product development
• All of management must be involved in quality, not just the specialists
• Continuous improvement requires the use of new and existing technologies
• Quality is the most cost-effective route to productivity
• Quality must be implemented with both customers and suppliers
Certain quality phrases of A.V. Feigenbaum:
"Quality does not travel under an exclusive foreign passport."
"Quality and costs are partners, not adversaries."
Failure driven companies ... "If it breaks, we'll service it." versus the quality
excellence approach ... "No defects, no problems, we are essentially moving
toward perfect work processes."
"Quality is everybody's job, but because it is everybody's job, it can become

nobody's job without the proper leadership and organization."

Dr. Kaoru Ishikawa (1915 - 1989)

Dr. Ishikawa held a B.S. in Chemistry and Doctorate of Engineering from the
University of Tokyo. In 1993, ASQ established the Ishikawa Award to recognize
outstanding contributions to the improvement of the human aspects of quality.
Some of Dr. Ishikawa's many awards include:
Deming Prize (1952)

Nihon Keizai Press Prize
Industrial Standardization Prize
Grant Award (ASQ)
Shewhart Medal (ASQ), first Japanese to receive this award
Honorary Member, ASQ (1986)
A few notable books include:
Authored the first Japanese book to define the word "TQC" in 1981
Guide to Quality Control (1982)26
What is Total Quality Control? The Japanese Way (Ishikawa, 1985)25
Kaoru Ishikawa was involved with the quality movement in its earliest beginnings
and remained so until his death in 1989. His father, Ichiro Ishikawa, President of the
Federation of Economic Organizations and of JUSE, invited Deming to speak before
top Japanese executives in 1950. A review of Ishikawa's training tapes, produced
in 1981, contain many of the statements of quality that are in vogue today. Subjects
such as total quality control, next operation as customer, trclining of workers,
empowerment, customer satisfaction, elimination of sectionalism (it's not our job),
and humanistic management of workers, are examples. It is am.azing to hear such
statements of quality on record from more than two decades ago.
Ishikawa stated that total quality control had been practiced in Japan since 1958.
The time for such a philosophy to take hold in a company can range from 2-5 years.
That time will depend on the commitment of top management. To reduce confusion
between Japanese-style total quality control and western-style total quality control,
he called the Japanese method the companywide quality controll (CWQC).
VILLANOVA UNIVERSITY 11- 30 LEJI.N SIX SIGMA HANDBOOK

Dr. Kaoru Ishikawa (Continued)

CWQC involves the participation of workers from the top to the bottom of an
organization and from the start to the finish of the product life cycle. CWQC requires
a management philosophy that has respect for humanity. There must be
acknowledgment that the worker can contribute to the success of the company
through suggestions, creativity, and worthwhile ideas.
One of the first concepts that western management took back to their own shores
was the quality circle. The quality circle concept represents the bottom up
approach. In Japan in 1988, there were one million quality circles, involving ten
million people. Quality circles were originally study groups that workers formed in
their department to study the quality concepts that were published in "Quality
Control for Foreman" (Ishikawa was the editor). Quality circles involve members
from within a department. The circle solves problems on a continuous basis. Circle
membership changes dependent upon the task or project under consideration.
Ishikawa also wrote that he originated the concept: "Next operation as customer"
in 1950 when he was working with a steel mill. Operators concerned about their own
defects were considered spies whenever they traveled to the next department to
view their original work. Departments were defensive when outsiders made tours,
thus, a concept of "Next operation as customer" was developed to remove those
fears. The separation of departments was referred to as sectionalism.
A man with many thoughtful concepts, Kaoru Ishikawa was known for his lifelong
efforts as the father of Japanese quality control efforts. The fish bone diagram is
also called the Ishikawa diagram in his honor.

Dr. Joseph M. Juran (1904 - )

Dr. Juran is the Founder and Chairman Emeritus of The Juran In:stitute. He holds a
B.S., from the University of Minnesota, a J.D. from Loyola Univer:sity and numerous
honorary doctorates degrees.
Dr. Juran's numerious awards include:
Edwards Medal, ASQ

Brumbaugh Awards, ASQ
Grant Awards, ASQ
Honorary Member, ASQ
Second Order Medal of the Sacred Treasure, 1981
Plus 30 other medals, fellowships, and honorary memberships
Notable, among Dr. Juran's 15 books, are:
Juran on Planning for Quality (1988)

Juran on Leadership for Quality (1989)
Juran on Quality by Design (Juran, 1992)28
Quality Planning & Analysis (Juran, 1993)30
Juran's Control Handbook, 5th ed (Juran, 1999)29
J.M. Juran started in quality after his graduation from engineering school with an
inspection position at Western Electric's Hawthorne plant in Chicago in 1924 (Walter
Shewhart and W.E. Deming were also at that plant). He left Westem Electric to begin
a career in research, lecturing, consulting, and writing that has lal;ted over 60 years.
An association with the American Management Association ha~, enabled Juran to
teach a course, "Managing for Quality," for 30 years to about 100,000 people in over
40 countries.
The publication of his book... Quality Control Handbook and his work in quality
management led to an invitation from JUSE in 1954. Juran's first lectures in Japan
were to the 140 largest company CEOs, and later to 150 senior managers. The right
audience was there at the start. Juran commented that no one was more surprised
than he to see CEOs at the seminars. His visit thus marked Japan's use of QC as a
management, rather than a specialist, technique.
VILLANOVA UNIVERSITY II - 32 LE.lIN SIX SIGMA HANDBOOK

Dr. Joseph M. Juran (Continued)
Dr. Juran has a prime basic belief that quality in America is improving, but it must
be improved at a revolutionary rate. Quality improvements need to be made by the
thousands, year after year. Only then will a company become a quality leader.
Juran's basics for success can be described as follows:
• Top management must commit the time and resources for success
• Specific quality improvement goals must be in the business plan and include:
• The means to measure quality results against goals

• A review of results against goals
• A reward for superior quality performance
• The responsibility for improvements must be assigned to individuals
• People must be trained for quality management and improvement
• The workforce must be empowered to participate in the improvement process
Juran Trilogy
Juran has felt that managing for quality requires the same attention that other
functions obtain. Thus, he developed the Juran or quality trilogy which involves:
• Quality planning
• Quality control
• Quality improvement
Juran sees these items as the keys to success. Top management can follow this
sequence just as they would use one for financial budgeting, cost control, and profit
improvement.
For any project, quality planning is used to create the process that will enable one
to meet the desired goals. The concept of quality control is used to monitor and
adjust the process. Chronic losses are normal in a controlled state, while the
sporadic spike will initiate an investigation. Eventually, only quality improvement
activities will reduce the chronic losses and move the process to a better and
improved state of control and that's the "last word."

Dr. Walter A. Shewhart (1891 -1967')

Dr. Shewhart held a B.S. and M.S. from the University of lIIinc)is and a Ph.D. in
Physics from the University of California.
Dr. Shewharts's awards include:
Holley Medal, ASME

Honorary Fellowship of the Royal Statistical Society
First Honorary Member of American Society for Quality
Honorary Professor of Statistical Quality Control, Rutgers University
The Shewhart Medal is awarded in his honor
Notable articles and books include:
Published a series of articles in Bell System Technical Journal

Economic Control of Quality of Manufactured Product (Shewhart, 1931 )54
Statistical Method from the Viewpoint of Quality Control (193!9)
Dr. Shewhart worked for the Western Electric Company, a manufac:turer of telephone
hardware for Bell Telephone, from 1918 until 1924. Bell Telephone's engineers had
a need to reduce the frequency of failures and repairs. In 1924, Shewhart framed the
problem in terms of "assignable-cause" and "chance-cause" variation and
introduced the control chart as a tool for distinguishing between the two. Bringing
a production process into a state of "statistical control," where the only variation is
chance-cause, is necessary to manage a process economically.
Shewhart worked to advance quality thinking at Bell Telephone Laboratories from

their foundation in 1925, until his retirement in 1956. Shewhart's charts were
adopted by the American Society for Testing and Materials (ASTM) in 1933. The
charts were used to improve production during World War II in the form of American
War Standards Z1.1, Z1.2, and Z1.3. W. Edwards Deming championed Shewhart's
methods, working as a consultant to Japanese industries from 1950 to 1990.
Walter Shewhart's statistical process control charts have become a quality legacy
that continues today. Control charts are widely used to monitor processes and to
determine when a process changes. Process changes are only made when points
on the control chart are outside acceptable ranges. Dr. De!ming stated that
Shewhart's genius was in recognizing when to act, and when to leave a process
alone. (Capitol Hill, 2001)6
VILLANOVA UNIVERSITY 11- 34 LE.tIN SIX SIGMA HANDBOOK

Dr. Walter A. Shewhart (Continued)

The Shewhart Cycle
The historical evolution of the PDCA problem solving cycle is interesting. Kolsar
(1994)36 states that Deming presented the following product design cycle (which he
attributed to Shewhart) to the Japanese in 1951:
1. Design the product (with appropriate tests)

2. Make the product (tested both in laboratory and production)
3. Put the product on the market
4. Test the product in service through market research
5. Redesign the product, in light of consumer reaction, and continue the cycle
Perhaps from this concept, the Japanese (Mizuno, 1984)39 evolved a general problem
solving process called PDCA. Both PDCA and PDSA are reviewed elsewhere in this
book.
Dr. Genichi Taguchi (1924 - )

Dr. Taguchi is currently a consultant for Ohken Associates (Tokyo), and was the past
director of the American Supplier Institute, Inc. He is called the "Father of Quality
Engineering." One of his profound statements on quality is that "Quality is related
to the financial loss to society caused by a product during its life cycle."
Dr. Taguchi's numerous distinctions include:
Deming Prize, 1960

Rockwell Award, 1986
MITI Purple Ribbon Award, 1989
Indigo Award, Japan, 1989
ASME Medal, 1992
Dr. Taguchi's notable books include:
System of Experimental Design, 2 volumes

Introduction to Quality Engineering (1986)
Off-line Quality Control (Taguchi, 1979)60

Dr. Genichi Taguchi (Continued)

Quality engineering techniques were developed by Genichi Taguchi in the 1950s.
The techniques enabled engineers to develop products and processes in a fraction
of the time as required by conventional engineering practices. He made his first visit
to the U.S. in the summer of 1980 to assist American industry in the pursuit of
quality. In 1983, Ford and Xerox began to promote the Taguchi's system both
internally and among suppliers. Taguchi's system was appealing because it was a
complete system that started with the product concept and continued into product
design, and into the manufacturing operations. It is a system to optimize the design
of products and processes in a cost-effective manner.
Taguchi's plan takes a different view of product quality:
1. The evaluation of quality
Use the loss function and signal-to-noise ratio as ways to evaluate the cost
of not meeting the target value. The traditional view is that a product is
either in specification or not. Taguchi feels the quality loss increases
parabolically as the product strays from a single target value.
2. Improvement of quality and cost factors
Use statistical methods for system design, parametelr design, and

tolerance design of the product. The methods could include quality
function deployment, signal-to-noise characteristics, and design of
experiments (using orthogonal arrays).
3. Monitoring and maintaining quality
Reduce the variability of the production line. Insist on consistency from

the floor. Take measurements of quality characteristics from the floor and
use the feedback.
Taguchi methods, and other design of experiment techniques, have been described
as tools that tell us how to make something happen, whereas most statistical
methods tell us what has happened. Taguchi methods are concepts that many
engineers can take out of a book and use.
VILLANOVA UNIVERSITY 11- 36 LEt\N SIX SIGMA HANDBOOK

Dr. Genichi Taguchi {Continued}

The concept of robust products is now being explored in the design phase to reduce
quality losses. Robustness derives from consistency. Robust products and
processes demonstrate more insensitivity to those variables that are either
ill-controlled or uncontrollable. Building parts to target (nominal) is the key to
success. One should work relentlessly to achieve designs that can be produced
consistently and demand consistency from the factory.
It has been published that about 50% of the practicing engineers in Japan are
competent in Taguchi methods. Dr. Taguchi has presented America with quality
engineering techniques that can work to produce better products and reduce costs.
It is more technical in nature and made for technical specialists. Top management
needs only to provide the training to learn the concepts and allow its use throughout
the corporation for it to be effective. The Taguchi approach does not call for an
internal revolution. His concepts do improve products and procedures.
Bill Smith {1929 - 1993}

Bill Smith was Vice President and Senior Quality Assurance Manager for Land
Mobile Products Sector, Motorola, when he introduced six sigma as a measure of
quality in 1986. At that time, he was handling an increasing number of complaints
from the field sales force regarding warranty claims. Motorola's product line was a
complex one and operating near three sigma levels. At that level, it was producing
too many defects. Mr. Smith determined that quality levels beyond 3 sigma were
necessary. He suggested a ± 6 sigma level allowing for a ± 1.5 sigma shift.
Mr. Smith helped Robert W. Galvin, Chairman and CEO of Motorola, recognize the
need to control variation and to work toward 3.4 defects per million or for six sigma
levels of quality. Later with Mikel Harry, Smith developed the initial four-step six
sigma stages: measure, analyze, improve, and control, to reduce the defect levels.
In 1988, Motorola won the first Malcom Baldrige National Quality Award. Mr. Smith's
six sigma efforts were credited with achieving that award. Upon his death in 1993,
Northwestern University and Motorola established a scholarship to honor Bill Smith.
(Barney 2002)\ (Chadwick, 2003f, (Ramberg, 2000)49

Mikel Harry
Mikel Harry and Richard Schroeder founded Six Sigma Academy in 1994 as a
consulting firm specializing in the six sigma methodology. Mikt!1 Harry has called
Bill Smith "The father of six sigma" and gave himself the title "The godfather of six
sigma". Many industry people have called Mikel Harry the main "alrchitect" of the six
sigma movement, as he has been the most widely known driver in the industry.
In 1985, Mikel Harry joined Motorola as a quality and reliability tmgineer where he
initially developed a problem solving program that included: Juran's quality journey,
SPC, Shainin's tools, and planned experimentation. He later teamed with Bill Smith
and developed the MAIC methodology with the "logic filters" approach. The logic
filters are a collection of tools to be used at each stage of tht! problem solving
approach. These originated with Harry's Ph.D. research at Arizona State University.
In 1989, Robert Galvin gave Harry the head position for the new Siix Sigma Research
Institute at Motorola University, where the emphasis and focus wlould be on dollars,
business transformation, and building a foundation for the six sigma process.
During the process of building the six sigma structure, Mikel Harry and a Unisys
Plant Manager derived the term "black belts" for the new bre!ed of statistically
trained problem solving experts. (They were both martial arts enthusiasts.)
(DeCarlo 2004)13!1 (MaGuire, 1999)38

(Process Quality Associates, n.d.)47, (Ramberg, 2000)49
Forrest Breyfogle III

Forrest Breyfogle III is the Founder and CEO of Smarter Soluticms, Inc., based in
Austin, Texas. The company was formed in 1992 to provide training and consulting
services for the six sigma methodology. His integrated system applies the principles
of strategic planning, metrics, ISO 9000, theory of constraints, le,an, and processes
in conjunction with six sigma. Forrest has a B.S. Mechanical Engineering from the
University of Missouri at Rolla and a M.S. in Mechanical Engineering from the
University of Texas. Mr. Breyfogle is a Professional Engineer, an ASQ Fellow, and
a recipient of ASQ's Crosby Medal for Implementing Six Sigma, 2nd ed. which made
a significant contribution to the understanding of six sigma.
(Smarter Solutions, Inc. n.d.t8, (SixSigma First, n.d.t7

ORGANIZATIONAL LEADERSHIP
Organizational Leadership
There are several ways to structure a lean six sigma strategy. However, successful
applications share a common core of management support, training, rewards, and
reinforcement.
Management Support
Effective lean six sigma programs do not happen accidentally. Careful planning and
implementation are required to ensure that the proper resources are available and
applied to the right problems. Key resources may include people trained in problem
solving tools, measurement equipment, analysis tools, and capital resources.
Assigning human resources may be the most difficult element, since highly skilled
problem solvers are a valuable resource and may need to be pulled from other areas
where their skills are also needed.
It has been said that there are two occasions when it is difficult to implement an
improvement program, when times are bad and when times are good. When times
are bad, profitability is low, resources are tight and "strategic" activities take a back
seat to "survival." When times are good, profitability is high, and resources are
focused on the current source of cash flow. Improvement may be last on the list of
things to do in order to take advantage of the current opportunity.
It has also been said that there are two times when an improvement program is
critical, when times are bad and when times are good. When times are bad, and
profitability is low, a company can not afford to continue losing money because of
poor quality and performance.
When times are good, and profitability is high, the costs of poor quality and internal
wastes are also likely to be high. Customers are not likely to repeat business with
a company that delivers a poor quality product or service, when a better option is
available. Unfortunately, many companies cruise along like the Titanic thinking they
are unsinkable because they are the market leaders.
Skilled managers must be willing to make significant commitments in order to

implement and support a successful six sigma initiative. Early successes must be
exploited to propel the company forward. This includes overcoming the hurdles that
appear at various transition points. At each of these transition points, the low
hanging fruit, that was previously available, has been picked and a more advanced
analysis is needed to reach the next level. The hurdles get progressively higher. If
an organization does not continue to persevere, then it is likely to take a step
backwards.

Organizational Leadership (Continued)

Training
The role of training, in the successful implementation of IE!an six sigma, is
fundamental. The skills necessary for breakthrough improvE!ments cannot be
developed without proper training. Companies that have implemented successful
lean six sigma programs found that training investments return siignificant benefits.
Paybacks of 10 to 1 have been reported. Extensive training is nec:essary for several
levels of individuals and basic training is required for the entire organization.
Potential black belts may undertake a 4 month training program consisting of one
week of instruction each month. A variety of software packages; are used to aid in
the presentation of projects, including Excel or MINITABTM for the! statistics portion.
Potential black belts will receive coaching from a master black Ibelt to guide them
through a project. The completed project will typically require thE! trainee to use the
majority of the tools presented during the training sessions.
Lesser amounts of training will qualify individuals for the green belt title. Some
companies include extensive lean training as part of their black belt and master
black belt programs. Other companies provide general lean instrlUction at the green
and black belt levels and then identify one or two master blac:k belts to receive
specialized lean training. The diagram below outlines a high levell training plan with
special training for executives and master black belts. The relative volume of each
diagram level represents the relative number of people receivin~1 training.
Senior Management - Sponsorship Training
Master Black Belt Candidates - Master Black lBelt Training
Management - Executive Training
Black Belt Candidates - Black Belt Training
Supervisors - Overview Training
Green Belt Candidates - Green Belt Tralining
Everyone - Lean Six Sigma Orientatioln Training
Figure 2.8 Six Sigma Training Pyramid
VILLANOVA UNIVERSITY 11- 40 LE.l!.N SIX SIGMA HANDBOOK

Organizational Leadership (Continued)

Training (Continued)
In some organizations, black belts are full-time positions that report directly to
management sponsors, who, in turn, assign specific projects to them. These
assignments mayor may not include a process improvement team. Green belts are
within the normal organizational structure and are typically assigned to process
improvement teams as needed. Black belts have specific mentoring responsibilities,
including the development of individuals assigned to them.
In other organizations, black belts continue in their normal assignments and

participate in process improvement teams as needed. In this structure, black belts
and green belts act as internal consultants and are pulled into teams when their
specific skills are needed. Black belts are typically responsible for mentoring 1 to
3 green belts or black belt candidates.
Many organizations have a structure that fits somewhere in between the two
previous models. Master black belts are responsible for coaching and training black
belts in order to make the best use of their skills. Master black belts also train and
coach management in order to help them support the lean six sigma program.
Rewards and Reinforcement
Rewards and reinforcement may be one of the hardest parts of successfully

institutionalizing a lean six sigma program. Black belts and green belts must have
positive career paths in order to encourage the best candidates to commit to the
extensive training and development required. Especially now that black belt skills
are in demand, it is important to recognize the accomplishments of black belts by
tangible and intangible means.
It is also important that green and black belts experience the rewards of achieving
significant savings for the company. At the same time, other team members must
be recognized for their contribution to performance improvements. To only reward
the black belts for improvements that were achieved by teams, creates resentment
and isolates the black belts from team members.

METRICS AND GOALS
Metrics and Goals

A fairly recent management term is "metrics." In the past, one pn)bably would have
stated that there is a need to measure the results of programs or ,activities via some
standard or evaluation criteria. The term "metrics" is now tak4~n to indicate that
numerical performance measurements are necessary. (Besterfield,1999)2
Installing the metrics for a certain activity means that a set of perf()rmance goals and
standards have been determined. Metric analysis should thenl provide effective
control feedback for reaching strategic goals. Organizational perfcnmance goals and
corresponding measurements are often established in the areas of:
• Profit
• Cycle times
• Marketplace response
• Resources
A company should develop metrics for each major performanc:e goal. A unit of
measure and a method of measurement must be defined for e,ach goal. For the
above performance goals, possible metrics include:
Profit
• Stockholder value • Return on investment

• Capital investments • Sales dollars
• Personnel costs • Profit margin on salles
• Community comparisons
Profit may be short-term (6 months or less) or long-term (2 yean; or more).
Cycle Times
• Existing cycle times

• Internal benchmarks
• External benchmarks
• Reduction in cycle times
Ten fold reductions in cycle times are possible.
VILLANOVA UNIVERSITY 11- 42 LEA.N SIX SIGMA HANDBOOK

METRICS AND GOALS
Metrics and Goals (Continued)

Marketplace Response
• Market survey of customers • Customer losses

• Analysis of returns • Courtesy ratings
• New product development • Facilities ratings
• Customer retention
Resources
• Number of improvement projects

• Return on improvement projects
• Process capability studies
• Variation reduction
• Cost of poor quality
• Percent defects
Use of Metrics
Metrics can be and must be developed in order to measure achievement of the
organizational goals. Dr. Deming discussed the problem of obtaining a "true value."
There is variation in all measurement and one must be skeptical of how data is
collected. The device used to collect the data must be accurate. Additionally, the
questions of "when," "where," and "how" will impact the accuracy and precision of
the data. According to Juran (1993)30, the development of any measurement system
should take into account the following factors:
• There should be a standardized meaning of the measurement

• The data should help the decision making process
• It should provide worthwhile information
• It should be easy to install
• It can be benchmarked or used elsewhere
Any mechanical and electrical instruments, gages, tools, etc., used for data
collection must undergo recognized calibration procedures. In many applications,
the appropriate metrics are qualitative based on customer, supplier, or appraisal
feedback forms.

METRICS AND GOALS
Use of Metrics (Continued)

After the metrics have been established, or in some cases aftE~r the goals are in
place, project teams, or natural work teams, must work toward implementing them.
Each team may have a portion of the metrics to achieve. The stratlegic business plan
includes strategic goals. The strategic goals are then deployed into smaller tactical
goals. Many parts of the organization work toward achieving similar tactical goals,
which collectively result in successful strategic goals. See the !;chematic below:
STRATEGIC
GOALS
I
STRATEGIC STRATEGIC STRATEGIC STRATEGIC
QUALITY OPERATING MARKETING SAFETY
GOALS GOALS GOALS GOALS
, , ,
TACTICAL QUALITY GOALS 1"'1-._--- METRICS USED TO EVALUATE PERFORMANCE
Figure 2.9 Strategic and Tactical Goal Metrics
The Balanced Scorecard

At times, a company's metrics pertain only to financial measures. Measures for
items such as customer satisfaction, internal business growth, or creativity are
neglected. Kaplan (1992)32, (1993)33, and (1996)34 describes the use of a system that
translates the company's vision and strategy into meanin!gful performance
measures. This technique is called the balanced scorecard.
VILLANOVA UNIVERSITY 11- 44 LE.Il.N SIX SIGMA HANDBOOK

METRICS AND GOALS
The Balanced Scorecard (Continued)

The balanced scorecard focuses the company's attention on four perspectives of the
vision and strategy:
• Financial: How should we appear to our shareholders?

Return on investment, cash flow, sales backlog
• Internal business process: What business processes must we excel at?

Reduce rework, cycle times, setup times
• Learning and growth: How will we sustain our ability to change and improve?
Employee surveys, employee suggestions, money spent on training
• Customers: How should we appear to our customers?

Customer surveys, complaints logged, market share
Observers and users of the balanced scorecard can see the strategy and goals of the
company and align themselves accordingly. Building steps include:
1. The scope of the balanced scorecard is defined.
2. A facilitator gathers information for the scorecard through interviews with

senior management.
3. The facilitator distributes the information at an executive workshop

designated to develop a draft of the measures for a balanced scorecard.
4. The facilitator generates a new report and rough draft scorecard.
5. A second workshop is held with senior and other levels of management. The
draft is refined and objectives are provided for proposed measures.
6. The third workshop finalizes the vision, objectives, and measures.
7. A new task team develops an implementation plan.
8. Periodic reviews of the balanced scorecard are conducted.

(Kaplan, 1993)33
Quite often, a substantial portion of the balanced scorecard measurement is

delegated to improvement teams.

METRICS AND GOALS
Performance Metrics
Effective business process management (BPM) requires an integrated system of
metrics in order to achieve the desired lean six sigma busine!ss improvements.
Pearson (1999)45 describes how this system of metrics might linlk all three levels of
the enterprise (process, operations, and business) with the KPOVs of each level of
the process becoming the KPIVs of the next.
Business Level Metrics

Business level metrics are typically financial (external) and op1erational (internal)
summaries for shareholders and management. The balanced scclrecard, previously
discussed, is widely used for business level metrics.
Breyfogle (2003)3 refers to these metrics as satellite metrics, the highest level
measures in business process management. Business (executive) level metrics
comprise summaries of detailed operations and financial results, reported monthly,
quarterly, or annually.
Traditional end-of-period cutoff reports are not sufficient for six sigma projects.
Other standard reporting practices, such as comparing year-to-date totals to the
same period last year, are also inadequate. It is important to remember that these
metrics are part of a complete system and should be treated with the same statistical
process monitoring and control techniques as operations data" Wheeler (1993t 4
provides excellent examples of using statistical methods for bu~,iness monitoring,
control and improvement.
Operations Level Metrics

Lean six sigma provides new metrics for managing complex operations. Business
effectiveness measures track how well products are meetingl customer needs
(external focus). Breyfogle (2003)3 calls them "30,000 feet" metriics, indicating that
they should have a longer-term perspective and reflect the totall variation that the
customer sees.
Operational efficiency measures relate to the cost and time required to produce the
products. They provide key linkages between detailed process measures and
summary business results, and help identify important relationships and root
causes. Senge (1990t 3 found that employees and teams who can see the impact of
their efforts on the overall business outcome, learn and make improvements more
effectively and efficiently.
VILLANOVA UNIVERSITY 11- 46 LE~~N SIX SIGMA HANDBOOK

METRICS AND GOALS
Process Metrics
Detailed process-level metrics include the data from production people and
machinery. This is the information that operators and supervisors need to run
normal operations. This information is also the subject of much of the measure,
analyze, improve and control phases (MAle) of lean six sigma once the improvement
project has been selected and defined. A number of process performance metrics
are reviewed later in this Handbook.
Performance Measurement Considerations

Brown (1996t provides helpful recommendations for effective process performance
metrics for the modern enterprise.
• The "vital few" versus the "trivial many": Large organizations may have
thousands of metrics, but no individual should have to focus on more than a
few. Overall business level metrics should be less than 20.
• Metrics should focus on the past, present, and future. Past history provides
context for decisions and builds organizational wisdom. The present data
provides real-time process control. Future predictions provide the basis for
estimates, improvement plans, and strategies.
• Metrics should be linked in a systematic way to meet the needs of

shareholders, customers, and employees.
• The key to an effective system is to have multiple metrics, not just one
important one. Success is about balance, not a mindless focus on quality,
shareholder value, profit, or any other individual measure.
• Metrics should be linked to shareholder needs on the business level.
• Metrics should be linked to the customer needs on the operations level.
• Metrics should be linked to the employee needs on the process level.
• Metrics should be consistent through all levels of the organization.
• Multiple measures can be combined (aggregated) into overall indices of

performance for higher levels.
• Metrics should evolve as strategy and situations evolve.

REFERENCES
References
1. Barney, M. (May,2002). "Motorola's Second Generation." Silx Sigma Magazine.
Accessed November 2, 2006 from
http://kellogg.northwestern.edu/faculty/savaskan/opnsquality/moCsix_sig
ma.pdf
2. Besterfield, D.H., et. al. (1999). Total Quality Management, 2nd ed. Upper Saddle
River, NJ: Prentice Hall.
3. Breyfogle, F.W., III. (2003). Implementing Six Sigma, 2nd ed. New York: John
Wiley & Sons, Inc.
4. Brown, M.G. (1996). Keeping Score, Using the Right Metrics to Drive World -
Class Performance, Quality Resources.
5. Bylinsky, G. (March 19, 2001). Heroes of U.S. Manufacturing. [Electronic

version]. Fortune, 143(6).
6. Capitol Hill. (2001, August 22). Retrieved from web site:

http://www.geocities.com/capitolhillnobby/4951/
7. Chadwick, G. (Sept., 2003). Remembering Bill Smith, Father of Six Sigma.

Accessed Nov. 2, 2006 from
http://www.isixsigma.com/library/content/c030915a.asp
8. College of Business, Utah State University. Shingo Prize for Excellence in

Manufacturing. http://www.shingoprize.org accessed October 24,2006.
9. Crabtree, R. (2004). "Peaceful Coexistence? Exploring the relationship between

the heavyweight methodologies of 'Six Sigma' and 'Le'a n'" APICS - The
Performance Advantage. Volume 14, Number 7, July/August.
10. Crabtree, R. (2006). Management Presentation on Lean Six Sigma.
11. Crosby, P.B. (1984). Quality Without Tears. New York: McGraw-Hili
12. Cutler, A.N. (2001, August 22). Retrieved from web site: http://www.sigma-
engineering.co.ukl, Sigma Engineering Partnership.
13. DeCarlo, N. (2004). Mikel J. Harry Ph.D. Accessed November 2, 2006 from
http://mikeljharry.com.

REFERENCES
References (Continued)
14. Deming, W.E. (1986). Out of the Crisis. Cambridge, MA: Massachusetts Institute
of Technology, CAES.
15. Delavigne, K.T. & Robertson, J.D. (1994). Deming's Profound Changes.
Englewood Cliffs: Prentice Hall.
16. DMAIC Process. (2001). Downloaded December 1, 2001 from web site:
http://www.ge.com/capital.vendor/dmaic.htm
17. Emiliani, M. (2006). Origins of Lean Manufacturing in America: The Role of

Connecticut Businesses. [Electronic version]. Journal of Management
History, 12(2).
18. Feigenbaum, A.V. (1991). Total Quality Control. 3rd ed., Revised, Fortieth
Anniversary Edition. New York: McGraw-Hili.
19. Ford, H. (1926), (1988). Today and Tomorrow. Cambridge, MA: Productivity
Press.
20. Gee, G., Richardson, W.R. & Wortman, B.L. (2005). CMQ Primer. Terre Haute, IN:
Quality Council of Indiana.
21. George Group. (n.d.). Michael George. Accessed November 5, 2006 from:
http://www.georgegroup.com.
22. Hahn, G., Hill, W., Hoerl, R. & Zinkgraf, S. (1999, August). The American
Statistician. 53 (3).
23. Harry, M. (1998, May). "Six Sigma: A Breakthrough Strategy for Profitability."
Quality Progress.
24. Harry, M. & Schroeder, R. (2000). Six Sigma. New York: Currency, Doubleday.
25. Ishikawa, K. (1985). What Is Total Quality Control? The Japanese Way.
Englewood Cliffs: Prentice Hall.
26. Ishikawa, K. (1982). Guide to Quality Control. White Plains, NY: Quality
Resources.
27. Jones, D. T. (n.d.). Lean Enterprise Institute. Accessed November 5,2006 from:
http://www.lean.org/whoweare/.

REFERENCES
28. Juran, J.M. (1992). Juran on Quality by Design. New York: Free Press.
29. Juran, J.M. (1999). Juran's Quality Handbook, 5th ed. New York: McGraw-Hili.
30. Juran, J.M. & Gryna, F.M. (1993). Quality Planning and Anc,'ysis, 3rd ed. New
York: McGraw-Hili.
31. Kanigel, R. (1999). The One Best Way: Frederick Winsllow Taylor and the
Enigma of Efficiency. New York: Penguin.
32. Kaplan, R.S. & Norton, D.P. (1992, January-February). "The Bi:1llanced Scorecard-
Measures That Drive Performance." Harvard Business Review.
33. Kaplan, R.S. & Norton, D.P. (1993, September-October). "Putting the Balanced
Scorecard to Work." Harvard Business Review.
34. Kaplan, R.S. & Norton, D.P. (1996, January-February). "U!;ing the Balanced
Scorecard as a Strategic Management System." Harvard I~usiness Review.
35. Kleiner, A. (Autumn 2006). Leaning Toward Utopia. Strategy + Business,

Special Issue.
36. Kolsar, P.J. (Fall, 1994). "What Deming Told the Japanese in 1950." Quality
Management Journal. Milwaukee:ASQ.
37. Liker, J. (2004). The Toyota Way. New York: McGraw-Hili.
38. MaGuire, M. (Oct 1999). "Cowboy Quality." [Electronic version]. Quality

Progress, 32(10).
39. Mizuno, S. (1984). Companywide Total Quality Control. Toyko:JUSE.
40. Missouri Enterprise. (2007). Management Presentation Materials. Columbia,

MO:MEP.
41. Northwest Lean Networks© (2006). Retrieved from

http://www.nwlean.netJleanfaqs.htm October 14, 2006.
42. Ohno, T. (1988). Toyota Production System. Cambridge, MA: Productivity

Press.

REFERENCES
43. Omdahl, T. (1997). Quality Dictionary, Terre Haute, IN: Quality Council of Indiana.
44. Pande, P.S., Newman, P.R. & Cavanagh, R.R. (2000). The Six Sigma Way. New
York: McGraw-Hili.
45. Pearson, T.A., (1999, May). "Measurements and the Knowledge Revolution."
presented at the ASQ Annual Quality Congress.
46. Potts, M. (March 16, 2001). "Fortune" Calls Anand Sharma Hero of U.S.
Manufacturing. [Electronic version]. India - West, 26(20).
47. Process Quality Associates, Inc. (n.d.). The Evolution of Six Sigma. Accessed
November 2,2006 from http://www.pqa.netlprodservices/sixsigma.
48. Quality Digest (November, 2006) "Six Sigma and Lean: Happily Ever After." No
author attributed.
49. Ramberg, J. {May, 2000). Six Sigma: Fad or Fundamental? [Electronic version].
Quality Digest.
50. Rother, M. & Shook, J. (2003). Learning to See. Cambridge, MA: The Lean
Enterprise Institute.
51. Sabatini, J. (October, 2000). Turning Japanese. [Electronic version].

Automotive Manufacturing & Production, 112{1 0).
52. TBM Consulting Group. (n.d.). Accessed November 6, 2006 from:

http://www.tbmcg.com/en/
53. Senge, P.M. (1990). The Fifth Discipline, The Art and Practice of the Learning
Organization, New York: Currency/Doubleday.
54. Shewhart, W.A. (1931 reprinted 1980). Economic Control of Quality of

Manufactured Product. Milwaukee: ASQ Quality Press.
55. Shingijutsu Co., Ltd. (2001). Shingijutsu website. Accessed November 8, 2006
from: http://Shingijutsu,co.jp/.
56. Shingo, S. (1989). A Study of the Toyota Production System from an Industrial
Engineering Viewpoint. (Revised). Cambridge, MA: Productivity Press.

REFERENCES
57. Six Sigma First. (n.d.). A Conversation with Forest Breyfogle. Accessed
November 8,2006 from: http://www.sixsigmafirst.com/forrest.htm
58. Smarter Solutions, Inc. (n.d.). Forrest Breyfogle, III. AccE!ssed November 8,
2006 from: http://www.smartersolutions.com.
59. Snee, R.D. (1999, September). "Why Should Statisticians Pay Attention to Six
Sigma?" Quality Progress.
60. Taguchi, G. & Wu, Y. (1979). Off - Line Quality Control. Nagaya: Central Japan
Quality Control Association.
61. Taylor, F. (1911,1998). The Principles of Scientific Management. Norcross, GA:

Engineering & Management Press.
62. Toyoda, E. (1987). Toyota, Fifty Years in Motion: An Autobiography. Tokyo &
New York: Kofansha International.
63. The Henry Ford website. "The Life of Henry Ford." Retrieved October 20, 2006
from http://www.hfmgv.org.
64. Wheeler, D.J., (1993). Understanding Variation, The Key to Managing Chaos,
Knoxville, TN: SPC Press.
65. Womack, J., Jones, D., & Roos, D. (1990). The Machine that Changed the World.
New York: Harper Perennial.
66. Womack, J., & Jones, D. (1996). Lean Thinking: Banish Waste and Create
Wealth in Your Corporation. New York: Simon & SChustE!r.
67. Wortman, B.L. (2001). CSSBB Primer. Terre Haute, IN: Quality Council of Indiana.

QUESTIONS
2.1. Lean and six sigma share in common all of the 2.6. Kaplan and Norton have outlined a business
following issues, EXCEPT: planning process that gives consideration to
factors other than strictly financial ones. It
a. They both focus on continuous provides a greater perspective for stakeholder
improvement interests. This approach is referred to as:
b. They both require top management
commitment a. Balanced scorecard
c. They both focus on customer satisfaction b. Strategic planning
d. They both require long learning curves c. Five forces of competitive strategy
d. Quality function deployment
2.2. The most important element in lean six sigma
deployment would be considered: 2.7. Increasing performance in a lean six sigma
corporation from 3 sigma to 4 sigma would
a. Training reduce defects per million by a factor of:
b. Organizational structure
c. Management support a. 2
d. Reward and recognition b. 8
c. 10
2.3. Which of the following concepts is mostly d. 16
associated with Taiichi Ohno?
2.8. In a nutshell, lean six sigma is considered:
a. SPC
b. TOC a. A business improvement approach
c. CTQ b. A focus on critical customer items
d. TPS c. An elimination of mistakes and defects
d. A concentrated focus on business outputs
2.4. Which of the following is the LEAST acceptable
reason for the deployment of lean six sigma 2.9. What guru is MOST widely associated with
projects? DOE?
a. A focus on cost savings a. Shingo

b. A focus on customer satisfaction b. Juran
c. A focus on internal problems c. Ishikawa
d. A focus on design improvements d. Taguchi
2.5. Lean six sigma project benefits could include all 2.10. The term "metrics" most frequently refers to:
of the following, EXCEPT:
a. A unit of measurement
a. Increased profrts b. The metric system
b. Improved process capability c. The science of weights and measurements
c. Increased defects d. An evaluation method
d. Reduced warranty claims

QUESTIONS
2.11. If one chose to look at any business enterprise 2.16. Strategic goals will be subdivided into:
on a main level basis, which of the following
categories would NOT have either KPIV (key a. Major benchmarks;
process input variables) or KPOV (key process b. Loss functions
output variables)? c. Numerous tactical goals
d. Appropriate metric:s
a. Process
b. Operations 2.17. Of the key elements 011 an organizational plan,
c. Business which of the following would be most likely to
d. Technological contain numbers and dates?
2.12. What luminary is generally recognized as being a. Vision statements

the creator of the control chart? b. Tactical objectives;
c. Guiding principles
a. Deming d. Mission statements
b. Shewhart
c. Harry 2.18. Who was the first CEO to understand the need
d. Ishikawa to control variation instead of the process
average as a way te) significantly improve
2.13. The defect levels, as reported by Motorola in quality?
their six sigma program, are higher than one
might expect from use of a standard normal a. Jack Welch
table or traditional capability calculations. Why b. Robert Galvin
is this true? c. Henry Ford
d. Bill Smith
a. Motorola found their processes followed
the exponential distribution 2.19. Which of the followingl would be considered a
b. Motorola allowed for failure on one-tail only strategic quality goal?
c. Motorola allowed for a 1.5 sigma shift in the
mean a. Commitment to thc! customer
d. Motorola found that six sigma efforts b. Reducing the scmp rate in the finishing
increased process variation department by 1%
c. Improved perform;ance inspection checks
2.14. A lean enterprise approach encompasses which on work in progre!ls
of the following departments? d. Training 6 greeln belts in statistical
techniques this quarter
a. Manufacturing only
b. All company departments plus suppliers 2.20. The concept most closely associated with lean
c. Manufacturing, quality, and the supply production is:
chain
d. Manufacturing and quality departments a. Better quality
b. Faster production
2.15. From an upper management perspective, what c. Flexible productioll
has been the principal motivating factor in d. Elimination of waste
embracing lean six sigma?
a. Bottom line results

b. Market share growth
c. Defect reductions
d. Customer focus
VILLANOVA UNIVERSITY 11- 54 LEtlN SIX SIGMA HANDBOOK

QUESTIONS
2.21. Why is lean six sigma called TQM on steroids? 2.26 What would occur if the quality goals were not
a part of the strategic plan?
a. Because of the extensive training element
required a. There would be no strategic goals
b. Because of the inclusion of statistical and b. There would not be as much emphasis on
lean tools quality
c. Because of the heavy impact of top c. The total quality effort would not suffer
management support d. The quality department would still maintain
d. Because of the impact of cost savings on the quality goals
the bottom line
2.27. Which American figure is seen as the earliest
2.22. Strategic goals must be subdivided. Thus, they advocate of waste reduction?
are:
a. Henry Ford
a. Delegated b. Frederick W. Taylor
b. Distributed c. W. Edwards Deming
c. Accountable d. James Womack
d. Deployed
2.28. A company has just started a lean six sigma
2.23. Many tools can be used in either lean or six initiative. Which set of tools would be better
sigma projects. A problem solving approach suited for initial projects?
that unifies project follow-up is:
a. Lean tools because they provide stability
a. SIPOC and repeatability for future projects
b. DOE b. Six sigma tools because they provide a
c. DMAIC larger number of available options
d. TPM c. Lean tools because they provide a more
reliable and accurate picture of the root
2.24. Which of the following quality gurus is most cause
closely associated with the term "total quality d. Six sigma tools because they provide more
management?" measurement data
a. Crosby 2.29. Which of the following quality luminaries would

b. Feigenbaum be most clearly identified as a proponent of
c. Deming improvement and breakthrough projects?
d. Juran
a. Ishikawa
2.25. The difference between strategic goals and the b. Deming
strategic business plan is that: c. Juran
d. Crosby
a. Strategic goals are often a lower tier than
the strategic business plan 2.30. Strategic plan implementation at the functional
b. They are determined by top management level requires:
only
c. They may offer conflicting priorities a. Functional level metrics
d. They are based on priorities given by all b. A company quality policy with everyone
levels of the company understanding it
c. Functional short and long-term strategic
goals
d. A customer focus
VILLANOVA UNIVERSITY II· 55 LEAN SIX SIGMA HANDBOOK

QUESTIONS
2.31. Which of the following legendary lean thinkers 2.36. Metrics of the eftecti'ileness of the strategic
was never officially on Toyota's payroll? quality plan include:
a. Sakichi Toyoda a. DMAIC

b. Taiichi Ohno b. Customer satisfaction
c. Shigeo Shingo c. The CSI number
d. Eiji Toyoda d. Drennik Index
2.32. Which of the following are resource 2.37. The extension from I,ean production to lean
requirements needed to achieve strategic office is possible because:
goals?
a. The office producEIS a variety of services
b. The concept of ""aste applies to every
I. Infrastructure and support for the
business environment
projects c. Offices are mOire data driven than
II. Training manufacturing
III. Administration of the programs d. There is little difference between
IV. Evaluation of the projects production and ofl'ice environments
a. I, II, III, and IV 2.38. If metrics were used to track marketplace

b. I and IV only response as part of strategic plan deployment,
c. II and III only which of the following measures is LEAST likely
d. I, II, and IV only to be included?
a. Analysis of returml
2.33. Who created the initial problem solving
b. Cost of quality
framework that would later become c. Customer market !Iurveys
DMAIC? d. Customer retention
A: Mikel Harry 2.39. Who should be the ultimate recipient of lean six
B: Forrest Breyfogle, III sigma project results?
C: Robert Galvin
0: Michael George a. Top management
b. Employees
2.34. After the development of a viable corporate c. Customers
strategy, the next logical step would be: d. Project sponsors
a. PDCA 2.40. One of Dr. Deming's 14 points for management

b. Deployment states, "Cease dependl!nce upon inspection as
c. SWOT analysis a way to achieve quality." The underlying tenet
d. Benchmarking of this statement is which of the following?
2.35. If a metrics format were being developed to a. Many American c:ompanies employ too
track marketplace response, which of the many inspectors; perhaps 5% - 10% of the
following items would be included? workforce
b. Quality should be built into the product, not
a. Cost of quality inspected in
b. Customer retention c. In most cases, the worker should perform
c. Cycle time reduction his/her own inspt!ction and not rely on
d. Profit margin on sales someone else
d. Most manual inspection will miss 10% - 20%
of defects under typical working conditions
The answers to all questions are located at the end of Section XIII.
VILLANOVA UNIVERSITY 11- 56 LE~\N SIX SIGMA HANDBOOK

,I
III. LSS PROJECT MANAGEMENT
IMPROVEMENT COMES ON A PROJECT B¥

PROJECT BASIS, AND IN NO OTHEB WAY.
JOE JURAN
VILLANOVA UNIVERSITY III - 1 LEAN SIX SIGMA HANDBOOK

LINKING PROJECTS TO ORGANIZATIONAL GOALS
Lean Six Sigma Project Management
Lean Six Sigma Project Management is presented in the following topic areas:
• Linking projects to organizational goals

• Problem solving methodologies
• Project selection criteria
• Project risk analysis
• Project management techniques
Linking Projects to Organizational GOlals

Pande (2000)25 suggests that embarking on a six sigma initialtive begins with a
management decision to embrace "A better way to run our organization." The
readiness assessment includes a review of the following areas:
• Assess the outlook and future path of the business:
• Is the strategy course clear for the company?

• Can our company meet financial and growth goals?
• Does our organization respond effectively to new circLilmstances?
• Evaluate the current organizational performance:
• What are our current overall business results?

• How effectively does our company meet customer requirements?
• How effectively is our company operating?
• Review the capacity for systems change and improvement:
• How effective is our company in managing system changes?

• How well are our cross-functional processes managed?
• Are there conflicts in our current improvement efforts'?' (Pande, 2000)25
The above assessment will go a long way towards deciding if c:urrent efforts are
sufficient, or whether the timing is appropriate to undertake a lean six sigma effort.
Lean six sigma can also be applied as a targeted approach. A number of companies
have improvement techniques and teams in place, and only assign black belt
assistance as needed.
VILLANOVA UNIVERSITY 111- 2 L.EAN SIX SIGMA HANDBOOK

Linking Projects to Organizational Goals (Continued)

Pande (2000)25 states that a decision on six sigma might be negative if the following
conditions exist:
• The company already has a strong, effective performance and process

improvement effort in place.
• Current changes are already overwhelming the company's people and

resources.
• The potential gains aren't sufficient to finance the investments necessary to

support six sigma.
The lean six sigma approach achieves the best results if implemented by high
performance organizations. Medium and low performance companies should
consider some building block steps, in order to take advantage of the "low hanging
fruit" that can be picked with these more basic techniques. Examples include:
• Using process simplification

• Applying cost management
• Engaging customers in product innovation
• Using basic problem solving teams
• Concentrating on standard operating procedures
According to Michael George (2002)11 an organization engaging in lean six sigma

should ask itself questions like:
• In which processes should LSS be first applied?

• How can the biggest cost improvements be achieved?
• How can quality and lead times be improved? To what degree?
George (2004)12 indicates that some key themes of lean six sigma include:
• Customers are important

• Speed, quality, and low cost are linked
• Variation and defects must be minimized
• Time traps must be found and removed
• Data is critical to making sound business decisions
• All employees must work together

Linking Projects to Organizational Goals (Continued)

Companies often find it desirable to create some form of benefit/E!ffort matrix for use
with project selection. Refer to Figure 3.1 below.
High
1 2
Benefit
3 4
Low
Low High
Effort
Figure 3.1 A Simplified Benefit/Effort Matrix
Obviously, those projects fitting into quadrant 1 should receive immediate attention.
Generally, projects in quadrant 4 should be avoided. Quadrant 3 projects are
sometimes beneficial for initial team activities. The most difficullt decision involves
the quadrant 2 category. These projects are potentially very d1esirable, but often
require careful analysis to ensure strategic fit and adequate resource availability.
Value stream mapping is often used to identify projects with the' highest impact.
Harry (2000)15 details a methodology to focus the deployment ()f (lean) six sigma
projects. There are a considerable number of options, dependE!nt upon the goals
and objectives of the organization. Considerations include:
• Focus on project cost savings

• Focus on customer satisfaction deliverables
• Focus on processes
• Focus on problems
• Focus on a targeted area
• Focus on design
• Focus on supplier processes (Paraphrased from Harry, 2000)15
The typical methodology that is followed for lean six sigma projects is either Define-
Measure - Analyze - Improve - Control, or some variation of this approach. This
assumes that a key business problem can be clearly defined, iand that it can be
addressed by data measurement and/or other improvement techniques.
VILLANOVA UNIVERSITY 111- 4 LE,I\N SIX SIGMA HANDBOOK

PROBLEM SOLVING METHODOLOGIES
Project Selection Process

With any problem solving methodology, attention should be directed towards the
initial project selection and an adequate definition of the problem or project. When
selecting an improvement team's initial project, the following factors should be
considered:
• Have broad appeal to team members, co-workers, and management

• Be fairly simple - but not trivial
• Be selected to show some quick benefit (within 3 to 4 months)
• Be within the group's control
• Be selected using time and resource constraints
The problem or project statement must be clearly defined:
• Problem statements are often "fuzzy." Consider the following examples:
• Poor communications
• Low recovery
• A car runs rough
• Excessive downtime
• Lack of training
• Too much scrap
• The true problem location must be clearly identified. There is often a

tendency to work on a downstream symptom of an upstream problem.
• A problem is the gap between:
• "What is?" and "What should be?"

• Current results and desired results
• A clearly defined problem statement that is measurable should be the initial

product. Frequently, a target timetable is included.
• The team should be working on a problem that is scheduled for solution.
• The team's two major activities are project resolution and learning to work
effectively together.

Selection of Problem Solving Methodo,logy

A green or black belt directed project team should select the approach and tools
appropriate to the task. On the following pages is a discussion of six
methodologies: PDCA, PDSA, a classical approach, DMAIC, IDEA, and the Ford 80
system. These methodologies generally use the seven original quality tools:
checksheets, Pareto diagrams, process flow diagrams, scatter diagrams, run (or
control) charts, histograms, and fishbone diagrams. In addition, a multitude of lean
concepts and basic team tools (brainstorming, consensus, multivoting, etc.) are
often employed.
The use of these basic approaches can resolve many problems a~ nd complete many
projects. In some cases, more powerful tools are necessary. In these instances, the
team would be wise to utilize the DMAIC approach because of the implied support
of professionals trained in the use of statistical software programs and techniques
such as ANOVA, DOE, confidence intervals, process capabilitiE!s, and hypothesis
testing.
PDCA
The PDCA cycle is very popular in many problem solving situatilons because it is a
graphical and logical representation of how most individu,als already solve
problems. Refer to Figure 3.2 below:
;f.:
1.P~
Act (A): Implement Plan (P): Establish a
necessary reforms when plan for achieving
the results are not 4. ACT a goal.
as expected.
Check (C): Measure 3. CHECK 2. D~ Do (D): Enact the

and analyze the plan.
results.
Figure 3.2 The PDCA Cycle
It is helpful to think that every activity and every job is part of .:1 process. A flow
diagram of any process will divide the work into stages and these stages, as a
whole, form the process. Work comes into any stage, changes are affected on it,
and it moves on to the next stage. Each stage has a customer. The improvement
cycle will send a superior product or service to the ultimate customer.

PDSA
Deming (1986)5 was somewhat disappointed with the Japanese PDCA adaption. In
1951 Deming presented a four or five step product design cycle to the Japanese, and
attributed the cycle to Shewhart. Deming proposed a Plan-Do-Study-Act continuous
improvement loop (actually a spiral), which he considered principally a team
oriented, problem solving technique. The team objective is to improve the input and
the output of any stage. The team can be composed of people from different areas
of the plant, but should ideally be composed of people from one area of the plant's
operation.
1. Plan - What could be the most important accomplishment of this team? What
changes might be desirable? What data is needed? Does a test need to be
devised? Decide how to use any observations.
2. Do - Carry out the change or test decided upon, preferably on a small scale.
3. Study - Observe the effects of the change or test.
4. Act - Study the results. What was learned? What can one predict from what
was learned? Will the result of the change lead to either (a) improvement of
any, or all stages and (b) some activity to better satisfy the internal or external
customer? The results may indicate that no change at all is needed, at least
for now.
5. Repeat step 1 with the new knowledge accumulated.
6. Repeat step 2, and onward. (Deming, 1986)5
As noted with other problem solving techniques, everyone on the team has a chance
to contribute ideas, plans, observations and data which are incorporated into the
consensus of the team. The team may take what they have learned from previous
sessions and make a fresh start with clear ideas. This is a sign of advancement.
Both PDCA and PDSA are very helpful techniques in product and/or process
improvement projects. They can be used with or without a special cause being
indicated by the use of statistical tools.

Classic Team Problem Solving Steps

1. Identify business or customer problems; select one to work on.
Brainstorming Customer feedback relPorts

Check sheets Pareto diagrams
Plan/Do/Check/Act Process flow diagrams
2. Define the problem; if it is large, break it down to smaller onles and solve these
one at a time.
Fishbone diagrams Value stream mapping

Process flow diagrams Check sheets
Pareto diagrams Systematic troubleshooting
3. Investigate the problem. Collect data and facts.
Data sheets Graphs

Histograms Control charts
Process capability Scatter diagrams
4. Analyze the problem. Find all the possible causes; decide whiich are major ones.
Brainstorming Check sheets

Fishbone diagrams Graphs
Hypothesis testing Systematic troublesho1oting
Design of experiments Value stream mapping
5. Solve the problem. Choose from available solutions. Select the one that has the
greatest organizational benefit. Obtain management appr10val and support.
Implement the solution.
Brainstorming Check sheets

Pareto diagrams Consensus
Management presentations Work flow improvemerut
6. Confirm the results. Collect more data and keep records on the implemented
solution. Was the problem fixed? Make sure it stays fixed.
Control plans Control charts

Pareto diagrams Histograms

Classic Team Problem Solving Steps (Continued)

The flow chart below shows how the quality tools and other techniques can be
utilized to solve a problem or improve a process.
r----------- SELECT CONDITION(S) Inn

_____~-=--_.....J
1...L1lllll.n..a.
Y
NEEDING ITROVEMENT
1 -_ _ _ _ _ _ _N_o_S....T~UDYCURRE T PROCEDURES w-o-I

DETERMINE THE POTENTIAL CAUSES ------<...1
(CANDIDATE FOR DATA COLLECTION)
r:r r:r D
~~
COLL+ATA 1'"-C"; ; A; ; T; ;'r"'; O; ; R; ; X- F-i~-EQ-uU-E-N-cY- '1

ANAL XlE DATA - - - - - ,
XES
.-: .
:....
..
NO L
IMPL EMENT SOLUTION
Figure 3.3 A Problem Solving Flow Chart Showing Use of Quality Tools

DMAIC Process
Each step in the cyclical DMAIC process is required to ensure, the best possible
results from lean six sigma team projects. The process steps are detailed below:
Define the customer, their critical to quality (CTQ) issues, and the core business
process involved.
• Define who the customers are

• Define customer requirements and expectations
• Define project boundaries - the stop and start of the proct~ss
• Define the process to be improved by mapping the process flow
Measure the performance of the core business process involved.
• Develop a data collection plan for the product or process

• Collect data from many sources to determine the current ,s tatus
• Collect customer survey results to determine shortfalls
Analyze the data collected and process map to determine root causes of defects and
opportunities for improvement.
• Identify gaps between current performance and goal perf()rmance

• Prioritize opportunities to improve
• Identify excessive sources of variation
• Identify objective statistical procedures and confidence limits
Improve the target process by designing creative solutions to fix and prevent
problems.
• Create innovative solutions using technology and discipline

• Develop and deploy improvement implementation plans
Control the improvements to keep the process on the new course.
• Prevent reverting back to the "old way"

• Develop an ongoing monitoring plan
• Institutionalize the improvements through system modifications
The above information is modified from GE® Capital Services web site (DMAIC,
2001)6. Note how closely the DMAIC process steps parallel the six classical team
problem solving steps presented on the previous two pages.
VILLANOVA UNIVERSITY 111- 10 LEj~N SIX SIGMA HANDBOOK

IDEA Process
The IDEA problem solving loop is similar in nature to the PDCA and DMAIC process
cycles. IDEA stands for Investigate, Design, Execute, and Adjust. The process
consists of basic step-by-step questions to help guide the problem solving team
toward new and innovative solutions. The detailed IDEA process steps are:
• Investigate: Provide a definition of the problem, provide some facts about the
problem, and provide a root cause.
• Design: Envision the idealized future state and create a list of options to
achieve the idealized state.
• Execute: Establish the specific metrics for success, test the best solution,
and determine a measurable project impact.
• Adjust: Reflect on the outcome of the project (the Japanese word is hansei).
This is an after action review and is also conducted for successful projects.
The IDEA report is formatted so that the four steps are concisely and clearly
displayed on one simple page. A condensed example is shown below.
Project purpose - A short project statement

Investigate
• Provide a problem statement
• Provide background information
• Determine the cause of the problem
Design
• Present the idealized future state
• Suggest several solutions
• Indicate the potentially best solution
Execute
• Detail key goals of the project
• Outline the implementation steps
• Indicate the expected project impact
Adjust
• Reflect on the outcome of the project
(May, 2007)24
Table 3.4 An Example IDEA Report

Ford 80 Problem Solving (Eight Oiscipllines)

Ford Motor Company has developed a method of root cause ;:malysis for use in
group problem solving. It is a staged problem solving methodology. Each stage is
verified by answering a checklist. The champion of the project uses the checklist
for each step (D), to verify that the questions are answered satisfactorily before the
next D can progress. Completion of the steps through a discipllined approach will
result in a completed solution. The 8D process has the following 8 steps:
D1. Establish the team: Use cross~functional team membership. This is a diverse
team with the knowledge, time, authority, and skill to solve the problem and
implement the corrective action. There will be a team champion.
D2. Describe the problem: Identify the problem in terms of the internal/external
customer problem. Define the problem in terms of "what is wrong with what?," use
of the 5W 2H method (who, what, where, when, why, how, how many). Use
quantifiable terms to define the problem. Some other useful tOlols include: "is/is
not," cause~and~effect diagrams, and flow charts.
D3. Develop a containment action: Determine a short~term (intelrim) solution to the

immediate problem. This prevents problems from affecting the customer.
D4. Identify the root cause: Search for all possible causes of thl! problem. Update
"is/is not," cause~and~effect diagram, process flow charts, etc. A FMEA (failure
mode effect analysis) could also be used here. Test each pc,tential root cause
against the problem description and test data for elimination of the problem.
D5. Develop alternative solutions: Identify alternative corrective clctions to eliminate

the root cause. One must establish a link between the alternative solution and the
root cause. The corrective action should not have any undesirable side affects.
Analyze the risks involved.
D6. Implement a permanent corrective action: Provide a plan to install a permanent

corrective action, with control monitoring to ensure elimination. Determine a
schedule necessary to measure the problem. Evaluate and verify the effectiveness
of the permanent corrective action.
D7. Prevent recurrence: Modify the existing systems, practices, and procedures to
prevent recurrence of the problem. One must be able to state that this is the best
possible long~term solution for the customer.
08. Recognize team and individual contributions: Recognize te!am and individual
contributions, and celebrate.
VILLANOVA UNIVERSITY III ~ 12 LEAN SIX SIGMA HANDBOOK

PROJECT SELECTION CRITERIA/PROCESS ELEMENTS
Project Selection
Improvement projects may be selected from a variety of sources. Many of the same
tools and techniques discussed later in Section V can be employed. A number of
items will be discussed here: .
• Process elements
• Stakeholder analysis
• Customer data
• Quality functional deployment
• Benchmarking
Process Elements
A business process is the logical organization of people, materials, energy,
equipment, and information into work activities designed to produce a required end
result (product or service). (Juran, 1999)21
Business process management (BPM) is a fundamental concept of lean six sigma.

Efforts to improve individual (local) process components are replaced by systematic
methods to understand, control and improve overall business results. These
methods have evolved from the basic tenets of quality, waste elimination, value flow,
and continuous improvement to address specific business objectives.
The SIPOC diagram is a foundation technique for lean six sigma improvement.
SIPOC is an acronym for the five major elements in the diagram:
Supplier: The person or organization providing resources (information, materials,

or service) to the process of concern
Input: The information, materials, or service provided
Process: The set of action steps that transforms the inputs into outputs by
adding customer value
Output: The final product or service resulting from the process
Customer: The person, process or organization that receives the output
(Pan de, 2000)25

PROJECT SELECTION CRITERIA/PROCESS ELEMENTS
Process Elements (Continued)

Lean six sigma relies on the SIPOC model to create, monitor and improve closed-
loop business systems for process management, process improvement, and
process design/redesign. Refer to Figure 3.5 below.
Suppliers Inputs Processes Outputs Customers
Figure 3.5 SIPOC Diagram
SIPOC can help everyone "see" the business from an overall pr10cess perspective
by:
• Displaying cross-functional activities in simple diagrams (process flow charts)

• Providing a framework applicable to processes of all sizes
• Helping maintain the big picture business perspective
• Providing methods for adding additional detail as needed
When process flow charts are used with the SIPOC model, business process
monitoring, control, understanding, and improvement are greatl), enhanced. When
lean specific projects are under construction, the detail created by value stream
mapping may be required. This technique is described in Secticln V.
Stakeholder Analysis
A lean six sigma project with high impact will bring about major changes to a system
or to the entire company. The change can affect various people iinside and outside
of the system. Major resistance to the change can develop. Attempts to remove or
reduce the resistance must be made. The stakeholders involved should be identified
and then a plan to convert or enroll them in the change process must be developed.
This should provide for the needed buy-in, alternate solutionl;, and removal of
pitfalls. Stakeholders can be identified as:
• Managers of the process • Customers

• People in the process • SUppIiE!rS
• Upstream people in the process • Financ«:!
• Downstream people in the process (Rath, 2000)26

PROJECT SELECTION CRITERIA/STAKEHOLDER ANALYSIS
Stakeholder Analysis (Continued)

Rath (2000)26 proposes developing a communication plan for the stakeholders
involved. The communication plan should identify the level of commitment or
resistance that the stakeholder is perceived to have. A communication plan should
then be developed to reduce the resistance to the change.
Businesses have many stakeholders including stockholders, customers, suppliers,

management, employees and their families, the community, and society. Each
stakeholder has unique relationships with the business. The SIPOC model explains
the classic supplier - process - customer relationship, but this is only one of the
relationships that must be addressed by business process management. Figure 3.6
illustrates some typical business - stakeholder relationships.
SOCIETY
STOCKHOLDERS OR OWNERS
en en
a: a:
w INTERNAL COMPANY w
:::i :1E
D.. 0
D.. ~
PROCESSES (/)
~
~
en 0
MANAGEMENT AND EMPLOYEES
Figure 3.6 Illustration of Company Stakeholder Interactivity
Lean six sigma identifies a process owner as a senior manager in charge of a

process. A black belt or green belt serves as a business process improvement
project manager, working across multiple processes to identify the process
stakeholders, understand their requirements and process interdependencies. The
objectives are to improve the individual process configurations and settings, in
order to improve the overall outcome for all stakeholders. This involves all
stakeholders in a variety of ways.
Stockholders choose to invest based on expected returns in the near term

(dividends or increased stock price) or longer term (growth). If the business
produces the revenues and profits expected, they may choose to invest further in the
business. If not, they may sell and reduce the company's financial resources.

PROJECT SELECTION CRITERIA/STAKEHOLDER ANALYSIS
Stakeholder Analysis (Continued)

When customers choose to purchase goods or services, they provide financial
resources to the process. If the delivered goods or services provide the desired
value, the customer may be motivated to provide reinforcing feedback in the form
of additional orders, positive word-of-mouth, new product ideas, referrals, etc. that
help build the business. However, if the perceived value is not up to par, negative
feedback to the process can retard the business.
Activities which fail to meet their stated objectives will have neg()ltive effects on the
stakeholders. To the stockholders, the net worth of the company will be reduced.
Suppliers may have payments delayed or never paid in full. Management and
employees may see wage levels frozen or diminished, and the number of employees
may be reduced. Customers may react to unsuccessful activities by seeking other
companies with which to deal, or they may impose penalties stated in the contract.
Organizational performance and the related strategic goals and objectives may be
determined for:
• Short-term or long-term emphasis • Marketplace response

• Profit • Resources
• Cycle times (FI!igenbaum, 1991)8
Goals may be set for either short-term or long-term results. The profit margin
required to operate a business should be optimized for all stakeholder requirements.
Projects and programs initiated by the company usually require a return on
investment (ROI). The maximum profits are usually not taken because of internal
stakeholder interests. If stockholder returns are maximized, then items such as re-
investment in the company, purchases of new machinery and equipment, wages and
salary increases must be turned down. An optimal level of stockholder dividends,
investments, personnel costs, etc. must be maintained.
For maintaining competitiveness, a reduced product cycle time must be emphasized.

This applies to both new product development and existing product lines. Reduced
cycle times will affect such things as the company's inventory', WIP, waste, and
operational efficiency.
The marketplace response is an organizational performance measure. The ability

to respond quickly to a competitor's quality, technology, product designs, safety
features, or field service are collectively very important. As a perfl)rmance measure,
proper resource utilization will result in reduced waste, reduced costs, and a more
effective organization.
VILLANOVA UNIVERSITY 111- 16 LE.II.N SIX SIGMA HANDBOOK

PROJECT SELECTION CRITERIA/CUSTOMER DATA
Customers
Lean six sigma is built around the customer. Everything starts and ends with
customers. They define quality and set expectations. They rightfully expect
performance, reliability, competitive prices, on-time delivery, service, and clear and
accurate transaction processing. (Harry, 2000)15
At times, the customer of a project may not be as evident as initially thought. The
receiver of the next operation, an internal department, could be thought of as a
customer. The external customer of a process could be the purchaser. But yet, if
the purchaser is a distributor, then they may not really be the true customer.
Pande (2000)25 points out that the primary customer of the process will or should
have the highest impact on the process. The primary customer is of utmost
importance to the process. The sorting out of the primary customer may take some
discussion on the team's part. The question of "Who is the customer?" may bring
out discoveries of "Which customers make money?". That is, are there certain
customers that make up the bulk of company revenues? Are there a small
proportion of customers that simulate the Pareto law? The case being that 80% of
the revenues come from 20% of the customers, or that 80% of net profit comes from
20% of the customers. See the discussion of Pareto analysis later in this Handbook.
External customers are the most important part of any business. If one can identify
them and understand their requirements, we can design products (goods and
services) that they will want to buy. Every business has many potential customers,
and each customer has their own decision criteria. They attempt to weigh the overall
value of goods and services by considering ~ost, guality, !eatures, and ~vailability
factors (CQFA). Businesses compete for customers on this CQFA value grid, and
must excel in at least one category in order to succeed.
Clearly, the voice of the customer is critical to business success. Both internal and
external customers should be identified. Their requirements must be identified in
order to understand and improve the business process. The relationship that
management can develop with either basic customer type will affect the company's
ability to be effective in delivering customer satisfaction.

Collecting Customer Data

Collecting data to gain "the voice of the customer" is a multi-level task. Business,
operations, and process levels of business process managemEmt were described
earlier in this Section. When collecting data from customers, it hE!lps to consider the
levels where these customers impact the business.
Business Level
Customers at this level are primarily shareholders and top mana~,ement employees.
The data of interest is primarily financial data such as stock price, market share,
revenues, earnings, return-on-investment (ROI), return-on-net assets (RONA), etc.
Typical analysis tools are financial. Typical measurement intervals may be quarterly
or annually.
Operations Level
Customers at this level are primarily those who purchase the product (external) and
those who manage production operations (internal). Data of interest measures
overall process performance with the focus on customer satisfaction (external
measures of operational effectiveness), and internal operations lefficiency (internal
measures such as rolled throughput yield, sigma levels, WIP inventory, etc.). Typical
analysis tools come from six sigma methods and lean manufacturing, industrial
engineering, and various forms of operations analysis. Typical measurement
intervals may be daily or weekly.
Process Level
Customers at this level are primarily internal, including employ'ees and the "next
process" in the operation. External customers include suppliers for detailed
material specification questions. Data of interest primarily invl)lves key process
variables. Typical analysis tools are statistical methods for process control,
capability, and improvement. Typical measurement may vary from hours to fractions
of a second, depending on production rates.

Instruments to Gather Data

There are instruments or tools available to everyone for the purposes of collecting
customer information. Some of the common instruments are described below:
• Surveys: A properly designed questionnaire gathers data using a consistent

set of standardized questions. Usually, a sample is selected for use.
Interviewers can be used or it can be self-administered.
• Focus groups: A small group (3 to 12 typically) of individuals is assembled

to explore specific topics and questions. A time of 1 to 2 hours is normal.
• Face-to-face interviews: Individual interviews of 30 to 60 minutes in length

may be used. This can be very time consuming.
• Satisfaction/complaint cards: The return of a card prompts a reaction by the

company. These could serve as feedback forms.
• Dissatisfaction sources: Some methods which voice dissatisfaction include:

complaints, claims, refunds, recalls, returns, repeat services, litigation,
replacements, downgrades, warranty work, misshipments, etc.
• Competitive shopper: Shoppers evaluate the company and competitors.

CEOs may call their own offices to measure the ease of customer access.
The information gathered should allow the organization to identify customer

requirements and to spot upcoming trends. The trends will be new ways for the
company to gain or retain customers.
Voice of the Internal Customer

Internal information should be gathered on the following factors:
• State of the company: What are the employees' perception of the company?
• State of quality efforts: Are the quality efforts worthwhile?
• State of the processes: Are there improvements?
• Reaction to policies: What dumb things were implemented?
• Rating of company satisfaction: Is the company a good place to work?
• Rating of job satisfaction: Do I like my job, my boss, etc.? (Snee, 1995)27

/

Customer Surveys
A few customer survey themes are noted below:
• To determine what quality is

• Find out what competitors are doing
• Define quality performance measures for use
• Identify factors to give a competitive edge
• Identify urgent problems
In the evaluation of customer information, not all attributes and transactions should
be treated equally. Some are much more important than others. As customers'
needs change, the evaluations will change. Griffin (1995)14 conducted a study on
best customer satisfaction practices and recommended the use of multiple
instruments to collect customer satisfaction data. The opportunity to collect
misleading or useless "iriformation is possible with just one instrument. Validation
of the initial results can be accomplished via multiple measurements.
Customer survey sample sizes and frequency can have significant cost implications,
and should be chosen to balance business resources and the need to monitor
changes in the business environment. Surveys can be developed in questionnaire
form. An adequate sample would range from 25 to 30 questions. For an L-Type
matrix survey, the use of a numerical scale from 1 (very dissatisfied) to 10 (very
satisfied) can make it easier to quantify the results, as shown in Figure 3.7.
Customer Satisfaction
Very Dissatisfied Very Satisfied
Task 1 2 3 4 5 6 7 8 9 10
On Schedule
Good Product
Friendly
Prompt
Figure 3.7 L-Type Survey Matrix
Breyfogle (2000)4 recommends using questions that measure relative changes in

customer attitudes from one survey period to the next, or from one product to
another. He suggests using a Likert scale to evaluate changes in customer attitudes
and determine shifts in the business environment.

Survey Pitfalls
Surveys are a method to gather data, but care should be taken with that data. A well
designed and properly executed survey can be a help to the company. The survey
can show what resources do not satisfy customers, identify opportunities for growth
or correction, and focus on customer issues (Futrell, 1994)10. Problems which can
occur in the use of surveys include:
• Improper survey form design

• Poorly defined survey issues
• Sampling errors or poor sampling techniques
• Ignoring nonresponses
• Treating customer perceptions as objective measures
• Using incorrect analysis methods
• Treating surveys as an event, not a process
• Asking nonspecific questions
• Failing to ask the right questions
• Ignoring the results or using them incorrectly
• Failing to provide feedback when necessary
• Using too many questions (25-30 questions are typical)
• Using a temporary employee to conduct interviews
Customer Service Measurement

Customer service measurements can be obtained through the various instruments
mentioned above. Some of the more common techniques include:
• Customer surveys measuring quality, service, performance, etc.

• Customer visits
• Customer service engineer feedback
• Complaint analysis: rejects, Pareto analysis
For a consumer survey, Hayslip (1994)17 suggests there may be some differences:
• A potentially larger customer population

• Small purchase volumes
• A small purchase transaction
• The supplier may know more than the customer

Customer Expectations
Customers ultimately determine the value of any product (goods and services) with
their decision to buy or not buy. These decisions are made based on a complex
system of critical customer requirements. In order to manage (ccmtrol and improve)
any business process, one must be able to determine the critical customer
requirements that influence these decisions.
A deeper knowledge of the customer is required in order to properly serve them.

There is a need to go beyond the sale, to uncover the subjective! factors of why the
product was purchased. The emphasis on listening to the customer, results in
information on customer expectations, priorities of expectation:s, and needs.
The customer's expectations of the product can be described tlhrough an analogy

similar to Maslow's hierarchy of human needs:
• Basic: The bare essential attributes of the product or service should be

present. A new personal computer should be assembll!!d, formatted, and
loaded with some basic software. A rental car should be serviced and
operating.
• Expected: Some attributes will be provided as a part I)f the product. A

knowledgeable technician will provide general operating instructions for a
personal computer. A few minutes would be provided as a matter of course
to help in the initial computer startup. The rental car is cOlnvenient to get to,
some features are explained, and policies clarified to the customer.
• Desired: These are attributes that are worthwhile to have, but not necessarily
provided as part of the package. A few extra hints by the technician on
operating procedures or hookups. The person at the rental car agency gives
good directions to your location and helpfully tells you how to save some
money when returning the rental car.
• Unanticipated: These are surprise attributes that go beyond what the

customer expects from a purchase. The computer technician calls for an
appointment to deliver your personal computer to your home at your
convenience. The rental car agency drops the car by your place at your
requested time and will pick it up when you are finished.
(Albrecht, 1992)1

Customer Needs
Customer needs are not stable, but are continually changing. A product or service
that satisfied a certain need may generate new needs for the customer. Maslow's
hierarchy of needs would be a good illustration of the advancement of needs
(physical, safety, social, esteem, and self-awareness). As an individual's needs are
fulfilled, he/she advances up the hierarchy. As the customer obtains a suitable
product or service (the basic needs are fulfilled), they will look for new attributes.
Juran (1999)21 lists customer needs as follows:
• Stated needs: What the customers say they want (a car)

• Real needs: What the customer really wants (transportation)
• Perceived needs: What the customer thinks is desired (a new car)
• Cultural needs: Status of the product (a BMW)
• Unintended needs: The customer uses the product in an unintended manner
(a BMW used to haul concrete blocks)
There are customer needs related to the use of a product:
• Convenience: Technology in today's world can bring about new products and
services that were not dreamed of. There are sections of society that limit the
extent of technology in their lives (Amish for example).
• Safety needs: Products or services become available that customers need.

The thinning of the ozone layer is creating a greater need for sun protection
lotions and creams.
• Product simplification features: New products can be complicated to use.

Products or services should help ease the conversion to their use.
• Communications: The need to be informed and to be given access to rightful

information. Open door meetings are a must for many government meetings.
• Service for product failures: When a product fails, what recourse (warranties,
returns, exchanges, etc.) does the customer have? A newly purchased tennis
racket cracks. What is the replacement policy?
• Customer service: Customers are expecting companies to have properly

trained personnel on hand to handle complaints. The agent of the company
should be empowered to satisfy the customer at that point.

PROJECT SELECTION CRITERIAlQFD
Quality Function Deployment*

Quality function deployment (QFD) is a tool that is sometimes referred to as the
"voice of the customer," or as the "house of quality." Quality function deployment
has been described as a process to ensure that customers' wants and needs are
heard and translated into technical characteristics.
The technical characteristics are handled by the company through the design
function, or better still, through a cross-functional team that includes sales,
marketing, design engineering, manufacturing engineering, andl operations. This
activity should focus the product or service on satisfying custol11er requirements.
QFD is a tool for the entire organization to use. It is flexible and cListomized for each
case and works well for manufactured products and in the service industry.
QFD was first applied in the Kobe shipyards in 1972 by Yoji Akao and his associates.
It met with great success and was introduced to the United States by Don Clausing
in the mid 1980s. Various United States companies (mostly automotive) have
applied the principles of QFD to their product design process. Hauser (1988)16
provides an illustration concerning the location of the emergency parking brake
lever for an American sports car. Engineering initially wanted 1to place the brake
between the seat and the door, but this caused a problem for a woman driver
wearing a skirt. Could she get in and out gracefully? Would this eventually cause
dissatisfaction?
As an advantage to using QFD, Besterfield (1999)2 cites an example of Honda and

Toyota reducing the new product cycle time to two and a half to three years by
means of QFD. The U.S. auto makers, during the late 1980s to early 1990s, were on
a cycle time of five years. QFD could be described as a concurrent engineering tool.
QFD provides a graphic method of expressing relationships between customer

wants and design features. It is a matrix that lists the attributes a customer wants
and compares it to the design features (services that satisfy customer wants).
(Hunter, 1994)18
The collection of customer wants and expectations are expressed through the
methods available to most any organization: surveys, focus group!;, interviews, trade
shows, hot lines, etc. The house of quality is one technique to 4lrganize the data.
The house of quality is so named because of the image used in its Iconstruction. The
use of matrices is the key to building of the house. The primary matrix is the
relationship matrix between the customer's needs or wants and the design features
and requirements.
* QFD, as a design tool, will be discussed in Section X.
VILLANOVA UNIVERSITY III - 24 LEtlN SIX SIGMA HANDBOOK

PROJECT SELECTION CRITERIA/QFD
Quality Function Deployment (Continued)

The construction of the house follows, similar to Hauser (1988)16:
• The left side of the house has the customer needs

• The ceiling has the design features and technical requirements
• The right side contains the customer priorities (comparisons)
• The foundation contains the benchmarking, target values
• The roof describes the relationship between design features
HOW TO MEET THE

CUSTOMER WANTS
(DESIGN FEATURES)
CUSTOMER COMPARISON
NEEDS OF CUSTOMER
(WANTS) PRIORITIES
BENCHMARKED
TARGET VALUES
Figure 3.8 Basic House of Quality Description
The possible benefits for using the QFD process are:
• Creates a customer driven environment

• Reduces the cycle time for new products
• Uses concurrent engineering methods
• Reduces design to manufacture costs (fewer changes)
• Increases communications through cross-functional teams
• Creates data for proper documentation of engineering knowledge
• Establishes priority requirements and improves quality

PROJECT SELECTION CRITERIAlQFD
Quality Function Deployment (Continued)
P = POSITIVE INTERACTIONS
N = NEGATIVE INTERACTIONS
DESIGN FEATURES
I-
z C/)
COMPETITION
w Z
I- 0 COMPARISON
z i=
0 w 0 ~
0 N C/)
w 0
w >-
....I l-
en 0c:: w C/) m
m ::::i I-
c:: 0 w I-
0 ....I Z
<I:
I- m <I: c:: ii: m w w
c.. w c.. <I:
c:: en
c.. <I:
CUSTOMER <I: c:: :!: 3: 3: > c:: a:
o
NEEDS 0 ::::l 0 W 0 ::::l
0
<I: 0 0 LL. ....I :!: 0 ~
COMPREHENSIVE 4 5 1 2 1 1 3 5
LOW COST 2 0 4 3 1 5 3 0
UP-TO-DATE 5 4 1 0 3 0 2 5
EASILY AVAILABLE 4 0 1 3 0 4 0 0
TEST QUESTIONS 5 5 0 2 5 1 5 5
RATINGS RANKINGS
5 = MOST IMPORTANT
I- Z 0= NO IMPORTANCE
::::l 0
c.. i= I-
::::l
~ 0 :I: c..
c:: w 0
LL. C/)
Z ~
0 enc:: c::
::::l
TARGET I-
0 ....
0
.... c.. W
VALUES ::::l 0 w 3:
c:: + C/) c:: ....I W
I- i:i: m c:: 10 0 :>
C/) ....I
0 W r-- :I: W
EA-
~
CD
<? V M V M c::
Figure 3.9 A Hypothetical QFD for an Exam Preparation tiandbook
The house of quality is flexible and customized to each situation. lEach organization
will develop guidelines that will modify the above image. However, the basics of
QFD will remain the same: to hear the voice of the customer and to be proactive in
its design of products in order to meet customer needs. (Besterfield, 1999)2
VILLANOVA UNIVERSITY III - 26 LE,iIN SIX SIGMA HANDBOOK

PROJECT SELECTION CRITERIAIBENCHMARKING
Benchmarking
Benchmarking is the process of comparing the current project, methods, or
processes with the best practices and using this information to drive improvement
of overall company performance. The standard for comparison may be competitors
within the industry, but is often found in unrelated business segments.
Benchmarking has gained tremendous influence and usage in the 1990s.

Correspondingly, front-line employees and operating managers have applied basic
benchmarking skills in scores of different business situations. These applications
can be grouped into distinctive types. Examples are explained below:
Process Benchmarking
Process benchmarking focuses on discrete work processes and operating systems,

such as the customer complaint process, the billing process, or the strategic
planning process. This form of benchmarking seeks to identify the most effective
operating practices from many companies that perform similar work functions.
Performance Benchmarking
Performance benchmarking enables managers to assess their competitive positions

through product and service comparisons. This form of benchmarking usually
focuses on elements of price, technical quality, ancillary product or service features,
speed, reliability, and other performance characteristics.
Project Benchmarking
Benchmarking of project management is easier than many business processes,

because of the opportunities for selection outside of the group of direct competitors.
Areas such as new product introduction, construction, or new services are activities
common to many types of organizations. Although the project objectives are
different, the projects will share the same constraint factors of time, costs,
resources, and performance. Project management benchmarking is useful in
selecting new techniques for planning, scheduling, and controlling the project.
Strategic Benchmarking
In general terms, strategic benchmarking examines how companies compete.

Strategic benchmarking is seldom industry-focused. It moves across industries
seeking to identify the winning strategies that have enabled high-performing
companies to be successful in their marketplaces.
(Bogan, 1994)3

PROJECT SELECTION CRITERIA/BENCHMARKING
Benchmarking (Continued)
Benchmarking activities often follow the following sequence:
• Determine current practices
• Select the problem area

• Identify key performance factors
• Understand your own processes and the processes of others
• Select performance criteria based on needs and priorities
• Identify best practices
• Measure the performance within the organization

• Determine the leader(s) in the criteria areas
• Find an internal or external organization to benchmark with
• Analyze best practices
• Visit the organization as a benchmark partner

• Collect information and data of the benchmark leader
• Evaluate and compare current practices with the benchmark
• Note potential improvement areas
• Model best practices
• Drive improvement changes to advance performance IE!vels

• Extend performance breakthroughs within the organizeltion
• Incorporate the new information in business decision making
• Share results with the benchmark partner
• Seek other benchmark leaders for further improvement
• Repeat the cycle
Juran (1993)20 presents the following examples of benchmarks (sl ightly modified) in
an advancing order of attainment:
• The customer specification

• The actual customer desire
• The current competition
• The best in related industries
• The best in the world
VILLANOVA UNIVERSITY III - 28 LE.l~N SIX SIGMA HANDBOOK

PROJECT SELECTION CRITERIA/BENCHMARKING
Benchmarking (Continued)
Some companies attempt to achieve a higher performance level than their
benchmark partner. Shown below is a comparison between a typical and a
breakthrough benchmark approach.
Typical Benchmark Breakthrough Benchmark

~
Time Time
Figure 3.10 Benchmarking Comparison
The competitive benchmarking partner presents some interesting options. For

example, Xerox Corporation has used IBM and Kodak (direct competitors in many
product lines) as benchmarks for many Xerox operations. However, Xerox chose
L.L. Bean Company, a catalogue sales distributor of clothing and consumer
products, as a benchmark for warehousing and distribution activities (Juran, 1993)20.
They did this because they felt L.L. Bean to be world class.
In some cases, benchmarking against the best-in-class is not possible because:
• The best in the world is not known (should be rare)

• There is no related process available (rare)
• The best-in-class is not willing to partner
• The best-in-class is not accessible due to geography or expense
(Goetsch, 2000)13
It should be noted that organizations often choose benchmarking partners who are
not best-in-class because they have identified the wrong partner or simply picked
someone who is handy.

PROJECT RISK ANALYSIS
Risk Analysis and Management

Risk analysis is an important part of every project. Detailing the possible project
risks can become a very complex task. Therefore, having a systE!matic approach to
manage risks and uncertainty becomes crucial to the future success of the project.
The reason for performing a risk analysis is to inform the appropriate stakeholders
(sponsors, customers, team members, others) about the magnitude of the risks
associated with the project, as well as the contingencies that will be developed in
order to mitigate or set the risks to acceptable levels (Martin, 1997)23.
Risk can be defined as a measure of the probability of an event and the costs
associated with not achieving an expected purpose.
risk =(probability of occurrence of an event) x (cost of that event)

To understand whether an event is truly risky requires an understanding of the
potential consequences of the occurrence or non-occurrence of that event. Failure
of a project or venture can be experienced in any combination of 1the following three
ways:
• The product or activity does not meet expected performance levels
• The actual costs are higher than budgeted costs
• The delivery or timing of the product or activity is too late
Additionally, performance, cost, and schedule risks can be segml:mted into five risk
areas. These are:
• Technical performance
• Supportability risks
• Environmental risks
• Cost risks
• Schedule risks
In most risk management discussions, the project approach us normally used.

However, risk can be applied to many business decisions like major equipment
purchases, changes in work flow, changes in pricing, relocaltion of sales and
manufacturing locations, etc. In many companies, the decision often comes down
to the lesser of two risks: moving too quickly or not moving quic:kly enough.
VILLANOVA UNIVERSITY III - 30 LEtlN SIX SIGMA HANDBOOK

Risk Analysis and Management (Continued)

Risk analysis, as such, is a part of a larger methodology called risk management.
Risk management is the more inclusive process of dealing with risk. According to
Kerzner (2001)22, risk management can be divided in the following four phases: risk
planning, risk assessment, risk handling and risk monitoring. Table 3.11 provides
a description of these phases.
Phase Description Inputs

Risk Develop and document a • Needs
Planning systematic methodology for • Resources
identifying, analyzing, and tracking • Focal points
risk issues. Develop contingency • Techniques
plans, perform assessments, and • Responsibilities
assign adequate resources. • Requirements
Risk Identify and analyze project risks to • Expert interviews
Assessment increase the ability of meeting cost, • Similar systems
performance, and schedule • Review of plans
requirements. • Lessons learned
• Assessments
Risk Identify, evaluate, select and • Simulations
Analysis implement actions to fix risk at the • Life cycle models
lowest possible levels given project • Reaction models
constraints and goals. • Decision analysis
• Danger lists
• Performance tracking
Risk A continuous evaluation and • Avoidance
Handling tracking system to ensure that • Control
implemented actions to mitigate • Acceptance
risk are working as planned. • Transference
• Mitigation
Table 3.11 Risk Management Phases and Descriptions*
* Adapted from (Kerzner, 2001)22 and (Frank, 2002)9

Risk Analysis and Management (Continued)

The risk management process can also consist of the following steps:
• Identify: Search for and locate risks before they become problems.
• Analyze: Transform risk data into decision making information.
• Plan: Translate risk information into decisive actions.
• Track: Monitor risk indicators and actions taken throughclut the project.
• Control: Adjust for deviations from planned actions.
• Mitigate: Reduce the impact of any unforeseen event.
• Communicate: Provide visibility and feedback data, internal and external to

the project or activity, on current and emerging risks. This is the KEY element
to a successful risk aversion program.
The management of risk requires the specific actions identified in Figure 3.12.
(Frank, 2002)!~
Figure 3.12 Risk Management Continuous Process
For major projects or activities, a risk management plan is a smart way to guide the
risk management process and to document the results.

PROJECT MANAGEMENT TECHNIQUES/PLAN ELEMENTS
Project Plan Elements

A project is a series of activities and tasks with a specified objective, starting and
ending dates, and resources. Resources consumed by the project include time,
money, people, and equipment. Project management includes project planning and
implementation to achieve:
• The specified goals and objectives,

• At the desired performance or technology level,
• Within the time and cost constraints,
• While utilizing the allocated resources.
The stages of project management are:
• Planning - deciding what to do

• Scheduling - deciding when to do it
• Controlling - ensuring that the desired results are obtained
Key project management elements include:
• Identifying schedule time limits

• Allocation of functional responsibilities
• Establishing continuous reporting methods
• Selecting applicable trade-off methodologies
• Measuring accomplishments against plans
• Identifying problems early
• Applying corrective action to problems
• Knowing when objectives will be met or exceeded
• Improving capabilities for future projects
Project Scope
The project scope refers to the boundaries of the project. It is an attempt to outline
the range of the team's activities. In the area of product development, the team may
decide to limit itself to the launching of a new product at a single manufacturing site.
Issues or problems regarding market research, prototype development, or financial
investments would be outside the scope of the team's activities. Eckes (2001 r
suggests that each team work very hard in its first meetings to clarify the project
scope. The team champion, the team leader, and the team will all be involved in this
process.

PROJECT MANAGEMENT TECHNIQUESIPLAN ELEMENTS
Goal Statement
The goal statement will be created and agreed to by the team and team champion.
Hopefully, the goals will be achievable within a 120 to 160 day period. Eckes {2001f
indicates that a typical "rule of thumb" for six sigma goals is a requirement of a 50%
reduction in some initial metric (or improvement of 50%). For example, reduce the
collectibles from 120 days to 60 days; reduce the scrap from 5% to 2.5%.
Milestones/Deliverables
For any well managed project, a set of stages or milestones are used to keep the
project on track and to help bring a project to completion. Again, Eckes (2001)7
points out that initial team projects should be at the 120 day length. Only half of the
project would be allocated to the define and measure stages. Assigning teams an
initial project with lengths of more than 160 days will lower the anticipated success
rate. A typical milestone chart might be:
Day 0: Start team activities

Day 1: Start the define portion of the project
Day 40: Begin the measure portion of the project
Day 80: Start the analysis portion of the project
Day 120: Start the improvement phase of the project
Day 160: Conclude the project with a management presentation
>Day 160: The bulk of project control elements are in progress
Team Composition
The composition of the team is of great importance, especially for critical projects.
Teams should be composed of qualified people with sufficient expertise to achieve
the team's charter. The team should not be staffed with people just interested in
improvement. To carry out high impact lean six sigma projects, highly qualified,
highly trained team members will best serve the team champion. {Eckes, 2001 f
Required Resources
The resources required for a project must be detailed. Typical Iproject resources
include:
• Qualified people • Machine time

• Equipment • Phones and faxes
• Machinery • Computer equipment
• Lab or office space • Utilities, etc.
VILLANOVA UNIVERSITY III - 34 LE.lIN SIX SIGMA HANDBOOK

PROJECT MANAGEMENT TECHNIQUESIWBS
Work Breakdown Structure

The work breakdown structure (WaS) is a detailed plan which expands the project
(statement of work) into a detailed listing of activities required to complete the
project. The project team leader is usually responsible for completion of the work
breakdown structure, including an assignment of responsibilities for each task to
individuals or team groups. More information on work breakdown structure may be
found in Kerzner (2001 )22.
Each project task is divided into smaller activities, and then elements, until the level
is reached in which each element is under one identifiable individual or group
responsibility. Another way to say this is: you take a long journey one step at a time.
Each activity is assigned a duration, along with the interrelationships between

activities. If one activity must be completed before another can begin, this is called
a predecessor event. After the material, equipment, and personnel needs are
established for each activity, the project network can be scheduled. The schedule
is a balance between time constraints, resource constraints, and costs.
If the time constraints are fixed, such as a set deadline, then the resource
constraints must be flexible to accommodate variations in the project. Time delays
during the project require offsetting increases in resources and costs to maintain a
set deadline. One example of a time constrained project is the construction and
delivery of a new product. These projects have fixed completion dates, often with
penalties for every day it is late. The contract is often won on the basis of the
company that plans to meet the deadline date, and is willing to risk the losses if the
project is late.
Time constrained projects are assumed to have unlimited resources. In a practical

sense, this means that resources must be acquired until they reach a level that will
ensure an on time project completion.
Most projects have relatively fixed levels of resources including manpower and
equipment. In resource constrained projects, the objective is to meet the project
duration requirements, without exceeding the resource limits. If these resources are
shared within an organization, resource scheduling and coordination between
projects becomes vital.
For those resources with time conflicts, it may become necessary to schedule
planned parallel activities as sequential tasks, using existing slack time and possibly
delaying the project completion date.

PROJECT MANAGEMENT TECHNIQUES/PLANNING TOOLS
Planning Tools
Project planning tools include developing and analyzing the project timeline,
determining required resources, and estimating costs. Common techniques for
evaluating project timelines include PERT charts, Gantt charts, the Critical Path
Method (CPM), and activity network diagrams (AND). The work breakdown structure
(WBS) helps identify detailed activities for the plan and enables estimation of project
costs.
Network Planning Rules

Common applications of network planning include the Program Evaluation and
Review Technique (PERT), the Critical Path Method (CPM), and Gantt Charts. The
following network rules are widely followed:
• Before an activity may begin, all activities preceding it must be completed.
• Arrows imply logical precedence only. The length and compass direction of
the arrows have no meaning.
• Any two events may be directly connected by only one activity.
• Event numbers must be unique.
• The network must start at a single event, and end at a sin~lle event.
Program Evaluation and Review Technique! (PERT)

The Program Evaluation and Review Technique (PERT) requirements are:
• All individual project tasks must be included in the network.
• Events and activities must be sequenced in the network to allow

determination of the critical path.
• Time estimates must be made for each activity in the netw1ork, and stated as
three values: optimistic, most likely, and pessimistic elapsied times.
• The critical path and slack times for the project are calculated. The critical
path is the sequence of tasks which requires the greatest I!xpected time.
VILLANOVA UNIVERSITY III - 36 LE.lI.N SIX SIGMA HANDBOOK

PERT (Continued)
The slack time, 5, for an event is the latest date an event can occur or can be
finished without extending the project, (TL)' minus the earliest date an event can
=
occur (TE). For events on the critical path, TL TE' and 5 O.=
(Kerzner, 2001 )22
Advantages of using PERT include:
• The planning required to identify the task information for the network and the
critical path analysis can identify interrelationships between tasks and
problem areas.
• The probability of achieving the project deadline can be determined, and by

developing alternative plans, the likelihood of meeting the completion date is
improved.
• Changes in the project can be evaluated to determine their effects.
• A large amount of project data can be organized and presented in a diagram

for use in decision making.
• PERT can be used on unique, non-repetitive projects.
Disadvantages of using PERT include:
• The complexity of PERT increases implementation problems.
• More data is required as network inputs.
Each starting or ending point for a group of activities on a PERT chart is an event,
also called a node, and is denoted as a circle with an event number inside. Events
are connected by arrows with a number indicating the time duration required to go
between events. An event at the start of an arrow must be completed before the
event at the end of the arrow may begin. The expected time between events, te is
given by:
Where: to is optimistic time, tm is most likely time, and tp is pessimistic time.

PERT Chart Example

An example of a PERT chart for a company seeking ISO 9001 (2000)19 certification
is shown below. Circles represent the start and end of each tclsk. The numbers
within the circles identify the events. The arrows represent task:s and the numbers
along the arrows are the task durations in weeks.
Figure 3.13 PERT Chart Example
TASK ACTIVITY DURATION

0 ISO 9001 Certification Objectiive
0-1 Planning 4 weeks
1-2 Select Registrar 4 weel(s
1-3 Write Procedures 8 weeks
1-4 Contact Consultant 3 weeks
2-6 Schedule Audit 6 weeks
3-5 Write Quality Manual 4 weel(s
4-7 Consultant Advising 12 weeks
5-6 Send Manual to Auditor 1 week
5-7 Perform Training 6 weeks
6-8 Auditor Review Manual 4 weeJes
7-8 Internal Audits 2 weeks
8-9 ISO Audit 1 week
9 -10 Corrective Action 3 weeks
10 Certification Milestone
Table 3.14 PERT Chart Tasks and Durations
VILLANOVA UNIVERSITY 111- 38 LEP,N SIX SIGMA HANDBOOK

PERT Chart Example (Continued)

Event 1 on the chart is called a burst point because more than one task (1-2, 1-3, and
1-4) start at that event. Event 5 is also a burst point. Points 6, 7, and 8 are sink
points because more than one task ends at that event.
To calculate the critical path, add the durations for each possible path through the
network. Which path is the critical path, and how long is it?
The possible paths and total times are:
PATH TOTAL TIME

0-1-2-6-8-9-10 22 weeks
0-1-3-5-6-8-9-10 25 weeks
0-1-3-5-7-8-9-10 28 weeks
0-1-4-7-8-9-1 0 25 weeks
Table 3.15 Comparison of Possible Event Paths
The critical path is 0-1-3-5-7-8-9-10. Obviously, with larger and more complex
networks, this calculation could be tedious by hand, but computer software is
available to do these calculations. During the project implementation, tasks which
are late in ending may delay the project, and can modify the remaining tasks' critical
path. Projects not on the critical path may be delayed by an amount equal to the
slack time without delaying the completion of the project.
What is the slack time for event 6? First one observes that event 6 is not on the
critical path calculated above. The earliest, T E' that event 6 can occur is at the 17th
week, found by path 0-1-3-5-6. Event 8 is on the critical path and occurs at the 24th
week, and since task 6-8 takes 4 weeks, the latest, TL' that event 6 can take place is
the 20th week. Using the formula for slack time:
S = TL - TE = 20 - 17 = 3 weeks for event 6.

Critical Path Method (CPM)

The critical path method (CPM) is very similar to PERT, eXCE!pt PERT is event
oriented, while CPM is activity oriented.
Unique features of CPM include:
• The emphasis is on activities

• The time and cost factors for each activity are considered
• Only activities on the critical path are contemplated
• Activities with the lowest crash cost (per incremental time savings) are
selected first
• As an activity is crashed, it is possible for a new critical path to develop
For each activity, there is a normal cost and time required for completion. To crash
an activity, the duration is reduced, while costs increase. Cralsh, in this sense,
means to apply more resources to complete the activity in a !,horter time. The
incremental cost per time saved to crash each activity on the critical path is
calculated. To complete the project in a shorter period, the activity with the lowest
incremental cost per time saved is crashed first. The critical path is recalculated.
If more reduction in project duration is needed, the next least expensive activity is
crashed. This process is repeated until the project can be completed within the time
requirements.
Using information from the PERT chart example, and adding crash times and costs,
we have:
Figure 3.16 CPM Example
Note that each activity arrow on the PERT chart example becomes a circle on the
CPM example. The letter indicates the activity and a number. The number, in this
example, is the normal activity duration in weeks. The critical p~lth is indicated by
the thicker arrows, along path A-C-F-I-K-L-M.

CPM Example
TASK ACTIVITY DURATION COST COSTI
weeks $ WEEK
0 ISO 9001 Certification normal crash normal crash CRASH
A Planning 4 3 2,000 3,000 1,000
B Select Registrar 4 3 1,000 1,200 200
C Write Procedures 8 6 12,000 15,000 1,500
D Contact Consultant 3 1 500 700 100
E Schedule Audit 6 5 200 1,000 800
F Write Quality Manual 4 3 800 1,200 400
G Consultant Advising 12 9 9,600 14,400 1,600
H Send Manual to Auditor 1 1 100 100 -
I Perform Training 6 4 9,000 12,000 1,500
J Auditor Review Manual 4 3 1,000 1,250 250
K Internal Audits 2 1 600 750 150
L ISO Audit 1 1 10,000 10,000 -
M Corrective Action 3 2 1,600 2,000 400
10 Certification Milestone 48,400
Table 3.17 Time-Cost Detail for CPM Example
Table 3.17 above shows the cost of crashing an activity, and the activity duration in
weeks if it is crashed.
What is the cost and project total duration if done in a normal manner? The time is
calculated adding the normal durations for events on the critical path A-C-F-I-K-L-M.
The normal time is 28 weeks. The total normal cost is the sum of the normal cost for
each activity, or $48,400.
If we wish to complete the project in 27 weeks, we must crash an activity on the

critical path. It does no good to crash activities off of the critical path, because the
total project duration would not be reduced. The lowest crash cost per week saved
for an item on the critical path is activity K at $150/week. The total project cost
increases to $48,550, and the time is reduced to 27 weeks. If K is crashed, we must
recalculate the critical path, using a duration of 1 for activity K. In this case, the
critical path does not change.

CPM Example (Continued)

To shorten the project to 26 weeks, we must next crash either ac:tivity F or M, each
at $400/week. Activity F is the better selection because it is earlielr in the project and
if other events are late, we can crash additional activities. The total cost increases
to $48,950.
Rearranging the table information, in order of activities to be crashed, we can

develop a relationship between project completion times and t01tal costs.
Task Sequence of Total Cost Total

Crash Activities Duration $ Cost
- ISO 9001 Certification WK SAVE normal crash $
- Normal 28 0 - - 48,400
K Internal Audits 27 1 600 750 48,550
F Write Quality Manual 26 1 800 1,200 48,950
M Corrective Action 25 1 1,600 2,000 49,350
A Planning 24 1 2,000 3,000 50,350
C Write Procedures 22 2 12,000 15,000 53,350
D Contact Consultant 22 0 500 700 53,550
I Perform Training 20 2 9,000 12,000 56,550
J Auditor Review Manual 20 0 1,000 1,250 56,800
B Select Registrar 20 0 1,000 1,200 57,000
E Schedule Audit 20 0 200 1,000 57,800
G Consultant Advising 20 0 9,600 14,400 62,600
H Send Manual to Auditor 20 0 100 100 62,600
L ISO Audit 20 0 10,000 10,000 62,600
10 Certification Milestone 48,400
Table 3.18 Priority Arrangement of CPM Activitie~s
The next activity to be crashed is A at a cost of $1,000 per week. After task C is
crashed, there are two critical paths, A-C-F-I-K-L-M and A-D-G-K-L-M, each 23 weeks
long. Both D and I must be crashed to shorten the critical path.
After task I is crashed, there are four critical paths, A-B-E-J-L-M, A-C-F-H-J-L-M,
A-C-F-I-K-L-M, and A-D-G-K-L-M, each 20 weeks long. Crashing any additional
activities provides no further time savings.
VILLANOVA UNIVERSITY 111- 42 LEt\N SIX SIGMA HANDBOOK

CPM Example (Continued)

The CPM cost versus time trade off for this example can be represented graphically
as shown below:
65000
~
80000
---
u; 55000
I
"-
~
o
(.)
50000 ...........
--
45000 I I I I I I
18 20 22 24 26 28 30
Time (weeks)
Figure 3.19 CPM Cost Versus Time Trade Off Example
The graph in Figure 3.19 illustrates that crashing activities beyond activity I,
increases cost without further reduction in time. If this is done, it is a useless waste
of resources. The assumption made in crashing an activity is that it is independent
of other activities. This may not be a valid assumption, for example if the same
resource is needed to crash different activities in overlapping time periods.
For calculations of more complex projects, linear programming methods are used
to determine the optimal cost-time point and activities to be crashed, which satisfy
the project time constraints. The cost-time curve shown above is a convex shape
in this example. Various algorithms are used to deal with convex and concave
curves, as well as those that are neither convex nor concave, but follow a more
complex relationship.
Similar to the PERT chart, CPM includes the concept of slack time for activities.
Without crashing, activity J has a slack time of 20 - 17 3 weeks. =

Gantt Charts (Bar Charts)

Gantt charts (bar charts), named after Henry Gantt, display activities or events as a
function of time (or cost). Each activity is shown as a horizontal bar with ends
positioned at the starting and ending dates for the activity.
Advantages of Gantt charts include:
• The charts are easy to understand.
• Each bar represents a single activity.
• It is simple to change the chart.
• The chart can be constructed with minimal data.
• Program task progress versus date is shown.
Disadvantages of Gantt Charts include:
• They do not show interdependencies of activities.
• The effects of early or late start of an activity are not shown.
• There is no means to indicate the variation in expected time to complete an

activity.
• The details of an activity are not indicated.
• There is little predictive value to this presentation of data.
The bar charts indicate only an ambiguous description of how the project, as a
system, reacts to changes. The network relationship between activities which are
indicated in PERT and CPM charts, are not shown in the Gantt chart.
An example of a Gantt chart for the ISO 9001 certification project is shown in Figure
3.20 on the following page.
VILLANOVA UNIVERSITY 111- 44 LEt,N SIX SIGMA HANDBOOK

<
j= I I I
r
l>
Z
0 ."r-
< :DC/)
l>
c:
Gantt Chart Example oC/)
z c..."
<: m:D
m Schedule ISO 9001 Certification Project As of Date : March 20, 200X 00
:II
(J)
-4c..
::::j Project Manager Quality Assurance Manager File name : IS09001.TLX 3: m
-< »0
z-4
Mar Apr May Jun Jul Aug Sep »3:
G)>
Task Start End Duration Path L ~ L L ~ L ~ mz
3:»
ISO 9001 Certification 6-Mar 20-Sep 2Swks C=========================== mG)
Planning
Select Registrar
6-Mar
3-Apr
31-Mar
2S-Apr
4wks
4wks
C
•••••
•••••••••••
Zm
-43:
-4m
mz
Schedule Audit 1-May 12-Jun 6wks 0-4
Auditor Review Manual
Write Procedures
5-Jul
3-Apr
1-Aug
26-May
4wks
Swks C
•••••••••••••• ••••• :::E:
z
is
c:
- Write Quality Manual 30-May 26-Jun 4wks C m
se."
~
U1
I
Send Manual
Perform Training
27-Jun
27-Jun
3-Jul
S-Aug
1 wk
6wks C ••••••• ~
Contact Consultant
Consultant Advising
3-Apr
24-Apr
21-Apr
1S-Jul
3wks
12 wks
•••••••••••••••
•
-
z
z
Z
Internal Audits 9-Aug 22-Aug 2wks C ••• G)
ISO Audit
Corrective Action
23-Aug
30-Aug
29-Aug
20-Sep
1 wk
3wks
C
C
•
•••• -4
0
0
r-
Certification 21-Sep 21-Sep Milestone C .0 C/)
r
m Key: = = = = Summary Task === = Summary Progress • • • • Detail Task •••• Task
l>
z Progress
(J)
><
(J) • •_ Slack 0 Milestone I Current Date ~ ~ ~ ~ Conflict . .•• Delay
i5 Path: C indicates critical path Scale: Each character is 5 work days or 1 week
s::
l>
:I:
l>
Z Figure 3.20 Gantt Chart Example
C
III
o
o
~
Activity Network Diagrams

The arrow diagram is the original Japanese name for this tool. The activity network
diagram describes a methodology that includes program evaluation and review
techniques (PERT), critical path method (CPM), activity on nodt! diagrams (AON),
precedence diagrams (PDM), and other network diagrams.
The activity network diagram incorporates a lot of PERT and CPM techniques in its
usage. The activities, milestones, and critical times must be developed and then
drawn onto a chart. The chart will then provide a tool to help monitor, schedule,
modify, and review the project.
As with other methods, the use of Post-it® notes or 3" x 5" cards will help in the
preparation stage of the chart. A planning meeting to uncover the required activities
is required. Various creative methods may be used to generate the activities or
milestones.
With the identification of activities, the following would occur:
• Write activities on cards and arrange the cards in sequenc:e

• Identify links to other activities
• Record times for each activity
• Verify the critical path
• Calculate the earliest start and finish times for each activity
• Calculate the latest start and finish times for each activity
• Calculate the slack times
• Review the activity network diagram
• If needed, "crash" the network, find ways to reduce the time needed
• If approved, put the diagram on paper and distribute
Some definitions for use with the activity network diagram:
Event, node: The junction point of an activity
Job, activity: The activity or task
Dummy node: A node inserted to combine the timing of pan:tllel operations
Critical path: The path with the longest time
Slack time (SL): The difference between the latest time and the earliest finish time
"Post-it"® is a registered trademark of 3M
VILLANOVA UNIVERSITY III - 46 LEj~N SIX SIGMA HANDBOOK

Activity Network Diagrams (Continued)

Key
10
~
Earliest
Task Start Finish 10 1 + - - - - Determine need
Length o 10 for LSS Handbook
(days)
Latest
Start Finish
10
10 20 ....- - - Determine LSS BOK
Critical
110 120 I
Path
20 40
EE 20
120 140 1
1+---- Locate five
outside authors
~-10
40 50 Make
Costa Rican
150 160 I assignments
Costa Rican
..4--authors
complete
30 1451751 ...- material
50 80
50 70
20
60 80
Indiana and Michigan
authors complete material 10 ISOl90 1 Proof major
80 90 content
90 120
Develop 30 1--+--4 90 110
test 90 ~20
questions 110120
@
" o 0... ... '"
... ' ! 1 ... 120125 Obtain

5 approvals
120125
12 Publish material
Figure 3.21 Activity Network Diagram Example

PROJECT MANAGEMENT TECHNIQUES/PROJECT DOCUMENTATION
Project Documentation
The initial project documentation is the project proposal. The prc)posal is usually in
response to meeting an improvement objective. The proposal should include the
objective(s), project plan, and budget. Approval of the proposal is management's
indication of support for the project objectives and commitment to provide funding
and resources. During implementation of the project, status reports are the
communication vehicle to management (or the customer) on the progress and health
of the project.
Measurement of Project Activity

During the planning stage of project management, the monitoring and measuring
requirements should be defined. In most cases, upper manag1ement will require
scheduled briefing sessions during the project. These sessions lean range in depth
from an overview of the project milestones to comprehensive reports.
The project monitoring plan should address the following areas::
• What is being monitored

• The purpose of the monitoring
• Timing or frequency of reporting
• Method of reporting (written reports, verbal summaries, fClrms used)
• Procedure for indicating a need for assistance
• Criteria for reporting of unusual events or urgent information
• The channel for feedback (to whom and how the information is sent)
• Assignment of feedback loop responsibilities
• Action to be taken when performance differs from requirements
The feedback loop defines the methods for monitoring and adjusting the process if
results are different than desired. Planning for feedback is analogous to designing
an automatic control system. The success or failure of a project i!s measured in the
following dimensions:
• Were the specified goals and objectives achieved?

• Within the time deadlines?
• At or below cost constraints?
• Utilizing the allocated resources?

Project Documentation (Continued)

Measurement of Project Activity (Continued)
Well executed project plans meet all of the previous criteria. As project complexity,
duration, or innovative technology increase, the more likely the project will not meet
the desired time target. Crash programs, to return a project to the desired time
schedule, are done at the expense of higher costs and resource usage.
It is possible for a project to be considered a success, even when the project is late,
over budget, or does not meet the stated objectives. An example of this type of
success is when the project accomplishes a significant feat.
Nearly every project encounters unanticipated events or problems, but this is not an
acceptable excuse for failing to meet the performance standards. The skillful project
leader will manage the resources to resolve the issues and maintain the project
schedule and budgets. Performance is measured on results, not effort.
The project time line is the most visible yardstick for measurement of project
activities. The unit of measurement is time in minutes, hours, days, weeks, months,
or years, and is readily understood by all participants on a project. The overall
project has definite starting and ending dates, both planned and attained.
From a quality viewpoint, both early and late projects have the opportunity for poor
quality compared to the project on schedule. For projects ahead of schedule, the
skeptical question is asked, "What corners were cut?" For projects behind
schedule, an appropriate pessimistic question is, "What is not being done properly
in an effort to regain lost time?"
Methods for planning, monitoring, and controlling projects range from manual
techniques (using plain paper, graph paper, storyboards, grease boards, and colored
magnetic markers) to computer software.


Measurement of Project Activity (ContinuE~d)
Advantages of manual project management methods include:
• Ease of use
• Low cost
• Best for monitoring schedules and timing of events
• A hands-on feel for the status of the project
• Easily customized for the specific project needs
• Minimal training requirements
Disadvantages of manual project management methods include:
• May not be transportable, depending on format used

• Project status information is only available at one site
• Large or complex projects may be difficult to display
• Potential problem activities may be overlooked
• Costs might not be tracked
• Resource conflicts may not be apparent
• Requires manual summarizing if interim status reports arE~ needed
• If the status board is erased the data may be permanently lost
• If information is incorrectly posted, it may not be noticed IUntii it is too late
• Project analysis is difficult regarding what was done (both right and wrong)
• Not flexible if reporting formats change during the project
• Not flexible if new monitoring controls are identified
Advantages of computer/automated project management methods include:
• Able to model what-if scenarios

• Able to show the impact of alternate options
• Can present information in a variety of formats and detail
• Schedules are automatically calculated
• Variances from plan are known in almost real time
• Project status reports are easier to generate
• People at different locations can input data, and share the same information
• Projects can be easily summarized
• Some data collection activities can be automated


Measurement of Project Activity (Continued)
Disadvantages of computer/automated project management methods include:
• A learning curve for the user of the software program

• Higher costs (although these may be low compared to overall project costs)
• Data entry and updating can be a time consuming task
• The computer will accept inaccurate data leading to incorrect decisions
• The software may not include options needed for the specific project
• The manager may focus on the computer and lose touch with the project
• Environmental conditions and/or location may be unsuitable for computer use
• Unusual events or task interrelationships may not fit into the software model
Software packages accept dates, durations, cost information, available resources

and can determine resource conflicts, project estimated costs, and can present the
data in a variety of formats including tables, Gantt (bar) charts, and networks.
Whichever method is used by the project team, keep in mind that the method is only
a tool to organize and summarize the data. The completion of project is the
objective, not the status boards or bar charts.
Milestones Reporting
Milestones are significant points in the project which are planned to be completed
at specific points in time. Intermediate milestones serve the purpose of refocusing
priorities on the longer range objectives, and at the same time providing status of
progress. Milestones typically occur at points where they act as a gate for a
go/no go decision to continue the project.
The project team leader would be expected to make a presentation to management

at each major milestone. The status of the project relative to the milestone, any
potential roadblocks for the completion of the project, and the plans for dealing with
roadblocks, would be presented.
The date and time for the milestone and the milestone activity are set very early on
in the project planning phase. Once set and approved, the milestones are not
normally subject to change or negotiation. If the project is late on meeting a
milestone, this fact will reach the visibility of upper management quite quickly.

Final Report
The project final report is the report card on performance for completion of
objectives, comparison of actual benefits and costs with budgets, and measures of
major activity completion dates versus milestones.
Lessons Learned
The next project closure step is the postmortem analysis (also called lessons
learned, autopsy, and post-project appraisal). The analysis of what went well and
what went wrong is used as a learning tool for future projects. The intent is to avoid
making the same mistakes, and to benefit from effective proces!.es. This review is
a formal and documented critique conducted by a committee of qualified company
personnel. The project review extends over all phases of development, from
inception to completion. Some of the fundamental review topics include:
• The adequacy of personnel, time, equipment, and money

• The effectiveness of the entire project
• How well the project was tracked
• How well top management and the project sponsor were informed of status
• How well the project team performed together
• How well the project team was recognized for their efforts.
• The effectiveness and reliability of corrective actions
• The true quality level of the delivered product and/or service
Results of the project review will be retained, along with the other project
documentation, and archived for future reference.
Document Archiving
The final project stage is document archiving. This includes test data, traceability
of materials, key process variables, and reports generated during the project. The
documents must be complete and organized. Storage requirements include:
• Protection from damage, including fire, water, and other deterioration

• Security of access
• Retrievability within a reasonable period, e.g. 3 days
• Adequate markings and indication of storage location
• Consideration of duplicate copies at different sites
• Use of a medium with a life longer than the record retention period
Software tapes or disks may be in a format which is obsolete or the equipment to

read them is no longer available. This requires retention of older reading equipment
or a media review on an annual basis.
VILLANOVA UNIVERSITY III - 52 LEJ~N SIX SIGMA HANDBOOK

REFERENCES
References
1. Albrecht, K. (1992). The Only Thing That Matters. New York: Harper Collins.
2. Besterfield, D.H., et at (1999). Total Quality Management, 2nd ed. Upper Saddle
3. Bogan, C., & English, M. (1994). Benchmarking for Best Practices. New York:
McGraw-Hili.
4. Breyfogle, F.W. III, et al. (2000). Managing Six Sigma. New York: John Wiley and
Sons.
5. Deming, W.E. (1986). Out of the Crisis. Cambridge, MA: Massachusetts Institute
of Technology, CAES.
6. DMAIC Process. (2001). Retrieved August 19, 2001 from

http://www.ge.com/capital.vendor/dmaic.htm
7. Eckes, G. (2001). The Six Sigma Revolution. New York: John Wiley & Sons.
8. Feigenbaum, A.V. (1991). Total Quality Control, 3rd ed. Revised, Fortieth
Anniversary Edition. New York: McGraw-Hili.
9. Frank, B., Marriott, P., & Warzusen, C. (2004). CSQE Primer. Terre Haute, IN:
10. Futrell, D. (1994, April). "Ten Reasons Why Surveys Fail." auality Progress,
27(4).
11. George M.L. (2002). Lean Six Sigma. New York: McGraw-Hili.
12. George M.L., Rowlands, D. & Kastle, B. (2004). What is Lean Six Sigma? New
York: McGraw-Hili.
13. Goetsch, D.L., & Davis, S.B. (2000). Quality Management, Introduction to Total
Quality Management for Production, Processing, and Services, 3rd ed. Upper
Saddle River, NJ: Prentice Hall.
14. Griffin, A., Gleason, G., Preiss, R., & Shevenaugh, D. (1995, Winter). "Best
Practice for Customer Satisfaction in Manufacturing Firms." Sloan
Management Review, 36(2).

REFERENCES
15. Harry, M., & Schroeder, R. (2000). Six Sigma, The Breakthr10ugh Management
Strategy. New York: Currency/Doubleday.
16. Hauser, J.R., & Clausing, D. (1988, May-June). "The House of Quality." Harvard
Business Review, 66(3), 63-73.
17. Hayslip, W.R. (1994, April). "Measuring Customer Satisfaction in Business

Markets." Quality Progress, 27(4).
18. Hunter, M.R., & Van Landingham, R.D. (1994, April). "Listening to the Customer
Using QFD." Quality Progress, 27(4).
19. ISO 9001 :2000. Quality management systems - Requirements. Geneva:

International Organization for Standardization.
20. Juran, J.M., & Gryna, F.M. (1993). Quality Planning and Anarlysis, 3rd ed. New
York: McGraw-Hili.
21. Juran, J.M. (1999). Juran's Quality Handbook, 5th ed. New Y,ork: McGraw-Hili.
22. Kerzner, H. (2001). Project Management: A Systems Approach to Planning,

Scheduling and Controlling, 7th ed. New York: Wiley and Sons.
23. Martin, P. & Tate, K. (1997). Project Management Memory Jogger: A Pocket
Guide for Project Teams. Salem, NH: Goal QPC.
24. May, M. (2007). The Elegant Solution: Toyota's Formula for Mastering
Innovation. New York: Free Press
25. Pande, P.S., Neuman, P.R., & Cavanagh, R.R. (2000). The Six Sigma Way. New
York: McGraw-Hili.
26. Rath & Strong. (2000). Rath & Strong's Six Sigma Pocket Guide. Lexington, MA:
Rath & Strong Management Consultants.
27. Snee, R.D. (1995, January). "Listening to the Voice of the Employee." Quality
Progress, 28(1).

QUESTIONS
3.1. Which two of the following categories typically 3.6. A project has more than one critical path. This
provide feedback to business systems and means that:
processes?
a. Crashing an event might shorten the project
I. Customers time
II. Suppliers b. The critical path was not calculated
IJI. Inputs correctly
IV. Outputs c. Delaying an event on a critical path may not
delay the project
a. I and II only d. Shortening anyone event cannot shorten
b. IJI and IV only the project duration
c. II and IJI only
d. I and IV only 3.7. Using the DMAIC approach to lean six sigma
improvement, at what step would the root
3.2. Modifying or redesigning a product would most causes of defects be identified?
likely occur during which two of the PDCA
phases? a. Measure
b. Control
a. Plan and Do c. Improve
b. Check and Act d. Analyze
c. Do and Act
d. Plan and Act 3.8. Which of the following statements can NOT be
made regarding the objective: "Quality will be
3.3. Other than project development, what other improved next year."
business decisions can be handled with risk
management? a. It does not define the measures of quality
b. The statement needs a defined ending date
I. Major equipment purchases c. How the improvements will be made should
II. Changes in work flow be stated
IJI. Changes in priCing d. It will not be possible to determine if the
IV. Relocation of facilities objective was met
a. I and II only 3.9. Project documentation should ideally be done

b. IJI and IV only by:
c. I, IJI, and IV only
d. I, II, III, and IV a. Maintaining detailed notes during the
project and then converting them to a
3.4. The major elements of project management do formal written report after the project is
NOT include: completed
b. Getting the work done first, and then
a. Modeling summarizing the events later
b. Planning c. Distributing plans at the project start, status
c. Scheduling reports during, and the final report at the
d. Controlling end of the project
d. The project leader who issues the project
3.5. Using a PDCA process to design a customer proposal and final summary, and has
survey, while implementing it customer interim information available upon request
feedback and improvement process, is an
example of: 3.10. When using the SIPOC model, any change in
process outputs will NOT be related to one or
a. The critical path method more changes in:
b. A customer driven company
c. A PDCA process within a PDCA process a. System inputs
d. A reactive versus a proactive approach b. Process inputs
c. Process actions
d. Customer actions

QUESTIONS
3.11. According to Juran, what the customer thinks 3.16. In a project network, if the earliest an event can
he/she desires, would be classified as: take place is week 45, the most likely time it will
take place is week 49, find the latest it can take
a. Stated needs place, without delaying the project completion,
b. Real needs is week 51. What is thE! event slack time?
c. Perceived needs
d. Cultural needs a. 2weeks
b. 4weeks
3.12. A team would be operating in which phase of c. 6weeks
the Shewart cycle if they were in the process of d. Cannot be determined from the above
conducting a pilot program test activity? information
a. Plan 3.17. If a company were to evaluate why some

b. Act customers refused to purchase a newly released
c. Check product, what Deming cycle phase would
d. Do apply?
3.13. Benchmarking should be done in the following a. Plan

sequence: b. Do
c. Study
I. Measure competitive performance d. Act
II. Implement significant improvements
III. Understand your own processes 3.18. Identify the problem solving methodology that
IV. Identify improvement criteria formally recognizes team and individual
contributions.
a. I, III, IV, II
b. III, IV, I, II a. DMAIC
c. IV, III, II, I b. IDEA
d. I, II, III, IV c. Ford 8D
d. PDSA
3.14. In consumer products, the complaint rate is
most directly a measure of: 3.19. A technique for tran:slating the customers'
wants and needs into quantifiable technical
a. Product quality characteristics is called:
b. Customer satisfaction
c. Market value a. Customer ombudsmanship
d. Rejection rate b. Quality function de,ployment
c. Pareto analysis
3.15. The slack time along a critical project path: d. The fifth discipline
a. Is the difference between the earliest and 3.20. Risk analysis planning should:
the latest start dates
b. Is equal to the ·crash" time I. Be done at the beginning of the project
c. Is neither less than nor greater than zero planning phase
d. Is the highest risk area II. Identify associated contingency plans
III. Identify and implement a metrics program
IV. Be reviewed and u~ldated at each milestone
a. I, II, and IV only

b. I and III only
c. I, II, and III only
d. I, III, and IV only
VILLANOVA UNIVERSITY 111- 56 LE."N SIX SIGMA HANDBOOK

QUESTIONS
3.21. List the following project scheduling techniques 3.26. Identify the tool that is most similar to project
in order of complexity, from least to most management and concurrent engineering:
complex.
a. Activity network diagram
I. CPM b. Praeto chart
II. Gantt c. Plan do check act
III. Pert d. Affinity diagram
a. I, II, III 3.27. When considering stakeholder groups, which of

b. I, III, II the following terms is most closely identified
c. II, III, I with the term "community?"
d. III, II, I
a. Society
3.22. An organized and disciplined approach to b. Government
problem solving in most six sigma c. External customers
organizations is called: d. Suppliers
a. SIPOC 3.28. Which of the following conditions are driving

b. DMAIC companies to adopt a lean approach?
c. PDCA
d. DPMO I. Higher performance requirements
II. Eliminating waste from the process
3.23. The right hand side of a completed house of III. Reducing total costs
quality displays ran kings and values for: IV. Increasing capacity
a. Customer needs or desires a. I and IV only

b. Competitive assessments or comparisons b. II and III only
c. Design feature measurements and c. I, II, and IV only
importance d. I, II, III, and IV
d. Design feature interactions
3.29. The company's management team is
3.24. What is a major distinction between the CPM formulating a five year strategy for internal
and PERT methods in the evaluation of project manufacturing and assembly capabilities. The
performance? best companies for them to benchmark are:
a. Only the PERT method can be displayed on a. Large U.S. companies with like
a Gantt chart manufacturing processes
b. The PERT technique allows for easier b. Small U.S. firms making similar products
crashing of project time c. Latin American organizations with intensive
c. The PERT technique permits network labor utilization
relationships but CPM does not d. Japanese "world-class" conglomerates with
d. The PERT technique is event oriented, while diverse operations
CPM is activity centered
3.30. Why is the PDCA cycle so readily accepted by
3.25. In what areas would upper management be most American teams and individuals?
most helpful in the initiation of a lean six sigma
effort? a. It is the natural way that most people
already approach problems
a. Providing direct training to black belts b. It was promoted by Dr. Deming who has a
b. Standardizing business operations wide American following
c. Providing key resources to the organization c. It has been widely used in Japan with
d. Directing the improvement projects success
d. It requires much less work than comparable
improvement techniques

QUESTIONS
3.31. Focus groups can best be defined as: 3.36. Which of the following is NOT a primary reason
for periodic project reviews?
a. Small groups with a specific topic interest
b. A segmented group of suppliers a. To highlight the prc>ject team's efforts
c. A segmented group of intermediate users b. To update goal achievement
d. A pre-selected group of users c. To review the schedule
d. To review the cost!1
3.32. Stakeholders that could help define a project
charter include: 3.37. Which of the following techniques has proven
useful in translating customer needs into
I. Customers product design feature:.?
II. Suppliers
III. Management a. Changing percepticms
IV. Benchmark partners b. Customer service principles
c. Confrontation and problem solving
a. I and II only d. Quality function deployment
b. I and III only
c. I, II, and III only 3.38. Outstanding service companies perform which
d. I, II, III, and IV activity first:
3.33. Internal customers of a retail store generally are: a. Treat employees as external customers
b. Train employees extensively in customer
a. People who shop inside a store instead of service skills
using the internet c. Give communicaticln skills training
b. Buyers that receive special pricing d. Provide good reco"ery skills training
discounts
c. Purchasers of the goods who buy in bulk or 3.39. One advantage of project management is that it
volume lot sizes does NOT require:
d. Store employees who are next in the
processing sequence a. Planning
b. Objectives
3.34. Planning for feedback by the project leader c. Unlimited resources
during a project would NOT address: d. People
a. The method of reporting 3.40. What type of expectation is met when a product
b. What is to be communicated has attributes that are worthwhile, but not
c. When reports are to be made necessarily expected?
d. Project milestones
a. Desired
3.35. Properly designed surveys should avoid all of b. Expected
the following, EXCEPT: c. Basic
d. Unanticipated
a. Asking specific questions
b. Ignoring non-responses
c. Poorly defined survey issues
d. Asking too many questions
The answers to all questions are located at the end of Section XII.

1
IV. TEAM ORGANIZATION & DYNAMICS
IT'S EASY TO GET THE PLAYERS. GETTIN'EM

TO PLAY TOGETHER, THAT'S THE HARD PART.
CASEY STENGEL
VILLANOVA UNIVERSITY IV -1 LEAN SIX SIGMA HANDBOOK

INITIATING TEAMS
Team Organization and Dynamics
Team Organization and Dynamics are presented in the following topic areas:
• Initiating teams • Conflict resolution

• Team roles • Team tools
• Team stages • Team performance evaluation
• Team dynamics
Initiating Teams *
A participative style of management is the best approach to ensure employee
involvement in the improvement process. Today's work force has higher
educational levels and is eager to participate in the decision making process
affecting them. There is no better way of motivating employees than to provide them
with challenging jobs which make use of their talents and abilities.
In spite of all the obvious advantages, team participation is one of the key areas
where most American companies fail. Dr. Ishikawa, a leading Japanese quality
professional, said of team involvement, "A people-building philosophy will make the
program successful, a people-using philosophy will make the program fail."
Company Team Benefits

Usually team members have diverse skills and experience and may represent
various departments and functions in the organization. What they share in common
is their involvement in the problem to be addressed.
The benefits of a team approach to issues are numerous. Consider the role an
individual plays in the organization. Team members may represent the role of
supplier, processor, and customer. On a team, each member often brings different
experiences, skills, know-how, and perspectives to the issues.
Such diversity is important for most improvement teams. A single person trying to
remove a problem or deficiency, no matter how skilled, has rarely mastered the
intricacies of an entire work process. The most significant gains are usually
achieved by teams - groups of individuals pooling their talents and expertise.
* Considerable content from this Section is derived from CMQ Primer (Gee, 2005)13
and CQE Primer (Wortman, 2006)41 material.
VILLANOVA UNIVERSITY IV - 2 LEAN SIX SIGMA HANDBOOK

INITIATING TEAMS
Company Team Benefits (Continued)

Improvement teams:
• Can usually solve larger issues than individuals working alone
• Can build a fuller understanding of the process needing improvement
• Can have immediate access to the technical skills and knowledge of all team
members (plus green, black, and master black belts, in the typical lean six
sigma arrangement)
• Can rely on the mutual support and cooperation that arises among team
members, as they work on a common project
Team Member Benefits

Teamwork offers some obvious benefits to team members, including:
• An opportunity for greater understanding of the issues affecting their work
• A chance to be creative and share ideas
• The opportunity to forge stronger working relationships with colleagues
• The opportunity to learn new skills and enhance existing ones
• A chance to work on a project with the full support and interest of upper
management
• The satisfaction of solving a chronic problem, which may attract and/or retain
more customers, increase revenues, and reduce costs
Team Resources
Resources are time, talent, money, information, and materials. The development of
productive teams will use considerable resources. Management must optimize the
resources available to teams. The team charter is the best place to establish the
team's expectations concerning available resources.

INITIATING TEAMS
Team Objectives
The team process can be a highly effective, people-building, potential-releasing,
goal-achieving social system that is characterized by:
• A climate of high support • Creative problem solving

• An open communication process • Individual achievement
• Organizational goal achievement • Commitment
The fundamental purpose of establishing teams is to improve the internal and

external efficiencies of the company. This is done through the efforts of the team
members to improve quality, methods, and/or productivity. If teams are properly
functioning, they will:
• Improve employee morale

• Remove areas of conflict
• Develop creative skills of members
• Improve communication and leadership skills of members
• Develop problem solving techniques
• Improve attitudes of both management and team members
• Indicate to team members that management will listen
• Demonstrate that employees have good ideas
• Improve management/employee relationships
Listed below are some of the reasons that teams have been successful in many
companies:
• If management has sanctioned teams in the company, this means that

management will be more apt to listen to employees and believe they have
ideas worthy of implementation.
• The team procedure allows all team members to communicate and exercise
creative expression.
• The concept of teams is supported by modern motivational theory:
• Maslow's higher level of human needs

• McGregor's Theory V, which recognizes the worth of an individual
• Herzberg's theory that true motivation is found in the work itself

INITIATING TEAMS
Team Empowerment .
Most power is derived from the organization's management authority. A team is
empowered by virtue of that power that is granted to it by management. A team
charter is a very useful tool for helping a team and management understand just
exactly what the team is empowered to do.
Team members have control over the team's performance and behavior. Control is
one source of power. Information is another source of power. To be effective, teams
need information. Teams should be told everything that could possibly help them
to achieve their objectives. They should be aware of financial conditions,
organizational changes, market conditions, etc. Access to resources is a third
source of power. A team's ability to succeed will depend in part on how free it is to
use organizational resources.
Management Support
Management must give more than passive team support. This means that
management, especially middle management, must be educated to the degree that
they are enthusiastic about the team concept. The implementation of project
schedules and solutions originating from teams should be given precedence. In
order for teams to be successful, management must recognize that there will be
additional work created by their efforts. Leaders, facilitators, and team members
should be thoroughly trained in six sigma and other improvement techniques, as
required.
In spite of the potential benefits, some people are skeptical of the long-term success
of teams. These people point out that the traditional style of management in the
typical American industry carries with it such momentum that the team approach will
have little appreciable long-term effect.
There are reasonable arguments that can be expressed either for or against teams.
The important questions that need answers are: (1) Does the company have the
proper environment in which teams can survive and thrive? and (2) Does
management fully comprehend the value of teams?
Management supports the team process by:
• Ensuring a constancy of purpose • Giving people a sense of mission

• Reinforcing positive results • Providing direction and support
• Sharing business results • Developing an integrated plan

INITIATING TEAMS
Types of Teams
The following types of teams are used by industries throughout the world today:
Lean Six Sigma Teams
The structure and functional roles of lean six sigma teams closely follow the
description of project and ad hoc teams that follow, with the addition of black and
master black belt support.
It should be noted that not all companies, lauded to have effective lean six sigma
programs, follow the same structure. Many companies use a variety of team
arrangements and provide black belt and master black belt support as necessary.
Improvement Teams
A group belonging to any department chooses to solve a quality/productivity
problem. It will continue until a reasonable solution is found and implemented. The
problem may be management selected, but the solution is team directed.
Process Improvement Team
For a process improvement team, employees may be drawn from more than one
department to look into the flow of material and semi-finished goods required to
streamline the process.
Project Teams/Task Forces/Ad Hoc Teams

Members are selected based on their experience and directed by management to
look into specific areas such as the modernization of a piece of equipment or
solution to a customer complaint. These teams are generally ad hoc and disband
upon the completion of their assignments.
Team membership can be all management, all work area, or a composite of the two.
Usually, the boundaries of the assignment are tightly drawn for project teams or ad
hoc teams. Some task forces may have broader mandates.

INITIATING TEAMS
Types of Teams (Continued)

Self Directed Teams
After a team understands its charter and has worked through its norms, it is ready
to get down to the business of solution building and improvement. Ideally, the team
should select its own leader to interface with other teams and coordinate team
activities. As the team meets and works together, the team leader should assume
an equal position with the other team members. Some teams find it helpful to rotate
team leadership to give everyone experience. At the pinnacle of performance,
anyone on the team should be able to lead the team.
This type of team operates with minimal day-to-day direction from management. Self
directed teams are asked to accomplish objectives within time frames that are truly
stretch objectives. Management must give the team the maximum latitude possible
to achieve their objectives.
Cross Functional Teams
Cross functional teams are made up of individuals who represent different

departments or functional areas in the organization. Individuals who represent a
department or functional area should be subject matter experts. That is, they should
be very knowledgeable about the policies, practices, and operations of their
department or functional area. The thoughtful selection of the members is an
important aspect of building an effective team. This becomes even more critical if
the team is to work only on a single project for several months or more.
Team involvement promotes sharing of the problem and minimizes "fingerpointing."

Representation from various departments also promotes the acceptance and
implementation of change throughout the organization. Solutions designed with the
active participation of affected departments tend to be technically superior and
accepted more readily by those who must implement them.

INITIATING TEAMS
Types of Teams (Continued)

Qual ity Ci rcles
The concept of circles originated in Japan after WW II. They were so successful in
Japan that many managers in the United States tried to duplicate them. The circle
is a means of allowing and encouraging people on the production floor to participate
in decisions that will improve quality and/or reduce manufacturing costs.
Quality, while it is a major consideration of the circle, is only a part of circle

involvement. Most ideas worthy of implementation must be justified on the basis of
cost savings in some way. Department members voluntarily par1ticipate to improve
departmental performance. Quality circles are effective forlUms to exchange
suggestions and find solutions. Since membership is voluntary, people are highly
motivated to continue the improvement process.
Quality Teams
The quality circle approach has been on the decline in the USA for some time. A
variety of quality team nomenclatures have replaced the term "quality circle." The
major reasons for the shift appear to be two-fold. First, the term ";quality circle" has
a strong Japanese connotation. And secondly, most circle projects tend to be
employee selected, while most team efforts are management selected, but team
directed.
The fundamental purpose of establishing quality teams is to improve the internal

efficiencies of the company, and internal and external products and service quality.
This is done through the efforts of the team members to improve quality, methods,
and/or productivity.
Natural Work Team Organization
In natural work teams, leadership is usually given to the area supervisor. Members
of teams come from the supervisor's work force. Outside members, from
specialized organizations, can be included in the membership, either as active
members or as contributing guests. Often, a facilitator is another important person
in this team organizational structure. He or she is specifically trained to coordinate
multiple team activities, oversee team progress, document results, and train team
members in their assorted duties.

INITIATING TEAMS
Synopsis of Team Applications
Team Types Structures Best Applications

Improvement May be 8 to 10 Can work on quality or productivity
Teams members from a issues. A process improvement team can
single department consist of multi-department membership
and focus on process flow and product
issues.
Quality May be 8 to 10 May initially work on quality topics or
Teams members from a overall department performance. Can
single department evolve into self directed teams.
Project Can have broad or Works on specific projects such as the
Teams specific member installation of a conveyor system. Can
selection and may also focus on material related items like
consist of all or part an improved inventory control system.
management Usually disbands upon the completion of
a project.
Lean Six Generally 8 to 12 Works on specific process or customer
Sigma members with black based projects of importance. Usually
Teams belt or master black disbands upon project completion.
belt support
Cross 8 to 12 members Members are carefully selected.
Functional from different areas, Knowledgeable people are required. Very
Teams departments, or similar to project teams. Tends to deal
disciplines more with policies, practices and
operations.
Self Directed 6 to 15 members. Requires considerable training and
Teams Generally a natural exposure. Can be given objectives or
work area team and develop their own. Some companies
may need staff select people with co-operative skills to
support help with success.
Table 4.1 Synopsis of Team Types, Structures and Applications

INITIATING TEAMS
Typical Team Operating Guidelines
Topic Ideas
Team agenda Who sets? When published? Input invited? etc. Rolling
agenda with priorities. Recorder to publish.
Attendance Excused absences only. How are latecomers handled?
Minimum attendance to conduct business? An obligation
to be present; excused absence permitted through team
leader; latecomers to be updated on critical issues only.
Meetings Time, frequency, place? Which meeting room? The time
and frequency must be determined.
Decision process Consensus, collaborative, majority? Can one person
remove an item from the agenda?
Minutes and Select a recorder. How are minutes approved? Where
reports posted? Who types? How distributed? Use a flip chart
for minutes? Is the recorder a volunteer or appointed by
chairperson? Time keeper to maintain agenda timing?
Recorder to transcribe, type, and distribute minutes.
Leader role How defined? How selected? Expectations? Leader
keeps things on track and moving, makes housekeeping
decisions, monitors participation, attendance and
timeliness, helps manage conflict.
Behavioral norms Listening; interruptions; radios, cell phones and pagers
off; no smoking; breaks called at members discretion;
limited cheap shots; empathetic listening; common
courtesy is expected; feedback should be constructive,
specific, and timely.
Confidentiality What goes outside the group?
Guests How invited? How excused?
Meeting audits How frequent? Who is responsible?
Facilitator How selected? Expectations? How will this role differ
from the leader?
Conflict Expected? How managed?
Recommendations How initiated? How routed? Who is informed?
Commitments Follow through on commitments, analysis, word
processing, etc.

INITIATING TEAMS
Team Meeting Structure

Any effective team meeting needs logical structure. Listed below is an example
format.
1. Develop an agenda
• Define goal(s)
• Identify discussion items
• Identify who should attend
• Allocate time
• Set time and place (semi-permanent if possible)
2. Distribute the agenda in advance
3. Start on time
4. Appoint a recorder to record minutes
5. Use visual aids liberally (flip chart, chalkboard)
6. Reinforce:
• Participation
• Consensus building
• Conflict resolution
• Problem solving process
7. Summarize and repeat key points throughout
8. Put unfinished items on the next agenda or table them
9. Review assignments and completion dates
10. Finish on time
11. Distribute minutes promptly
12. Critique meeting effectiveness periodically

INITIATING TEAMS
Sample Meeting Forms

Shown below are some simplified team based forms. A group accomplishment
record is very beneficial, since team progress in a single sessioln may be slow.
Agenda/Minutes
Team name _ _ __ Facilitator _ _ _ _ __ Date _ __

Leader _ _ __ Recorder
Members Present Others Present

1. 4. 1.
2. 5. 2.
3. 6. 3.
I. Current problem or project: III. Decision/acti()n items:

II. Ke discussion oints: IV. Next week's a enda:
Meeting Attendance Log

Team name: _ _ _ _ __ Leader: _ _ _ _ __ _
Date'.
I Members I I I I I I I I I I
Key: P-Present L-Leave A-Absent V-Vacation
Group Accomplishments
I Date- I ActivitY or Accom~lishment I Savings or Results I
Facilitator's Log
Date Comments Concerns
II
II I
Figure 4.2 Sample Meeting Forms

INITIATING TEAMS
Selecting Team Members

When selecting a team, upper management identifies those parts of the organization
that are associated most closely with the problem. There are four places to look:
• Where the problem is observed or the pain is felt

• Where sources or causes of the problem might be found
• Among those with special knowledge, information, or skill
• In areas that can be helpful in developing the remedy
Often a cross functional team is assembled to accomplish significant results in a

short period of time. The best and brightest people the organization has to offer
should be chosen. What is required are people who believe that two or more minds
are better than one, and who will contribute a diversity of perspective, experience
and knowledge. Obviously, lean six sigma teams are technically supported by green
belts, black belts, master black belts, or lean specialists.
Adding New Team Members

Care must be taken when adding new people to existing teams. The rule is to not
impose an individual on a team. This can be handled by involving the entire team
in the selection process. Team members interview prospective new team members
either one at a time or collectively. The team should be questioned about the skills
they feel a new team member should bring to the team. When the team has a
significant role in deciding on any new team member, the team will be much more
committed to making sure the decision was the right decision.
Removing Team Members

Sometimes, despite everyone's best efforts, a team member may need to be taken
off the team. There are any number of reasons why this situation could occur.
Perhaps one of the members lacks the required skills and shows little interest in
developing them. Personality conflicts may exist between team members. Perhaps
a team member is too stretched or stressed by other projects or personal problems,
and can't keep his/her commitments to the team.
The result is a very delicate situation for the team leader or sponsoring manager.
Both the team and the manager should have a series of frank discussions with the
individual. The conversations should center on what's expected, What's at stake,
and what's not happening that needs to happen, or what is happening that shouldn't
be happening. If the situation doesn't improve, the team member must be removed.

INITIATING TEAMS
Team Size
A team can consist of members from only one area or can be made up of a group of
representatives from different parts of the organization. Each person may be a
subject matter expert who understands the processes and activities at issue. It is
usually impractical to include every person who could be involvled.
Conventional wisdom is that teams over 20 people, some think over 15, become too
unwieldy and lose the active participation of all team members. Teams of 4 people
or less may not generate enough ideas. A major change management principle
embraces the notion that people will more readily accept and support a change, if
they are included in the development of the solution. This presents a major dilemma
for teams: How can the team be kept small enough to effectively 'Work together and
at the same time involve everyone? This is not a trivial matter in large organizations
that may have several hundred people actively supporting a work process.
Extending the group to customers of the process generates an even larger group of
people whose collective buy-in is needed to ensure successful c:hange.
Special efforts have to be used to involve the larger group in the understanding of
the initial team's charter and the collection of needed information. Input and ideas
should be sought from the larger group as the solution set is develloped. Successful
teams organize, develop, and implement a communication plan to gain the
participation, support, and ultimately, the commitment of an en1tire department or
operation.
Team Diversity
To achieve optimum performance, a team often needs diversity in the orientation of
its individual team members. Some team members are needed who are primarily
oriented towards task and target date accomplishment. Other team members will
be needed who hold process, planning, organization and methods in the highest
regard. Teams also need members who nurture, encourage and communicate well.
Teams need some members who are creative and innovative. This quality is helpful
when product design, inspiration, optimism or humor is needed.
It is important to understand that the above characteristics are not normally an

assigned role. People naturally tend to orient their thinking along one or more of the
desired traits. A good understanding of the above strengths and the value each
brings to a team provides the much needed guidance for team selection. Productive
teams are sensitive to each other's viewpoints.

TEAM ROLES
Team Roles
While some organizations choose to use different names and definitions, most
successful organizations have implemented the following roles in their black belt
program.
• Black belts • Executive sponsors

• Master black belts • Champions
• Green belts • Process owners
The definitions in this portion of the Handbook are a combination of the author's
experience and references at the end of this Section.
Black Belts
Black belts are most effective in full-time process improvement positions. The term
black belt is borrowed from the martial arts, where the black belt is the expert who
coaches and trains others as well as demonstrates a mastery of the art. In a similar
way, lean six sigma black belts are individuals who have studied and demonstrated
skill in implementation of the principles, practices, and techniques of lean six sigma
for maximum cost reduction and profit improvement.
Black belts typically demonstrate their skill through significant positive financial
impact and customer benefits on multiple projects. Black belts may be utilized as
team leaders, responsible for measuring, analyzing, improving and controlling key
processes that influence customer satisfaction and/or productivity growth. Black
belts may also operate as internal consultants, working with a number of teams at
once. They may also be utilized as instructors for problem solving and statistics
classes. Black belts are encouraged to mentor green belt and black belt candidates.
Master Black Belts
Lean six sigma master black belts are typically in full-time process improvement
positions. They are, first and foremost, teachers who mentor black belts and review
their projects. Selection" criteria for master black belts includes both quantitative
skills and the ability to teach and mentor. For master black belt recognition, an
individual must be an active black belt who continues to demonstrate skill through
significant, positive, financial impact and customer benefits on projects. The ability
to teach and mentor is evaluated by reviewing the number and caliber of people they
have developed. Teaching may also be demonstrated in classroom environments.

TEAM ROLES
Team Roles {Continued}

Green Belts
Green belts are not usually in full-time process improvement positions. The term
green belt is also borrowed from the martial arts. Green belt refers to an individual
who has mastered the basic skills. Green belts may be black belts in training,
having less experience than full black belts. Green belts must demonstrate
proficiency with the core statistical tools by using them for positi"e financial impact
and customer benefits on a few projects. In some organizations, individuals may
remain a green belt for several years. Green belts operate under the supervision and
guidance of a black belt or master black belt.
Executive Sponsors
Executive sponsorship is a key element in an effective lean six sigma program.

Executive leadership sets the direction and priorities for the ()rganization. The
executive team is comprised of the leaders that will communicate, lead, and direct
the company's overall objectives towards successful and profitable lean six sigma
deployment. Executives typically receive training that includes a lean six sigma
program overview, examples of successful deployment and strategies, and tools and
methods for definition, measurement, analysis, improvement, and control.
Champions
Lean six sigma champions are typically upper level managers that control and
allocate resources to promote process improvements and black belt development.
Champions are trained in the core concepts of lean six sigma and deployment
strategies. With this training, champions lead the implementation of the lean six
sigma program. Champions also work with black belts to ensure that senior
management is aware of the status of six sigma deployment. Champions ensure
that resources are available for training and project completion. They are involved
in all project reviews in their area of influence.
Process Owners
Key processes should have a process owner. A process owner coordinates process
improvement activities and monitors progress on a regular basis. Process owners
work with black belts to improve the processes for which they are responsible.
Process owners should have basic training in the core statistical tools, but will
typically only gain proficiency with those techniques used to improve their individual
processes. In some organizations, process owners may be lean six sigma
champions or sponsors.

TEAM ROLES
Team Roles (Continued)

Process Owners (Continued)
Process owners should:
• Be comfortable with the team's capabilities

• Believe in the team's objectives
• Support team members with resources and information
• Share information with the team
• Understand the team's mission
• Participate in project reviews
• Believe that personal goals are aligned with the team's goals
• Be knowledgeable of lean six sigma core elements
Lean Six Sigma Structure

Companies have differing duties and terminologies for organizational roles that
support lean six sigma improvement. Some of the common functions and optional
structures are listed in Table 4.3 below. (Adapted from Pande, 2000)30
Functions Structure Options

Executive direction • Lean six sigma steering committee
• Quality council
• Executive steering council
Lean six sigma management • Lean six sigma manager
• Lean six sigma director
• Master black belt
Process owner • Champion
• Sponsor
Sponsor • Process owner
• Champion
Coach • Master black belt
• Black belt
Team leader • Trained supervisor/facilitator
·• Black belt
Green belt
Team member • Associate with team training
• Associate with process knowledge
• Green belt
Table 4.3 Functions in a Lean Six Sigma Organization
VILLANOVA UNIVERSITY IV ·17 LEAN SIX SIGMA HANDBOOK

TEAM ROLES
Steering Committee Role*

Establishing a steering committee is a logical first step wheln an organization
launches a lean six sigma improvement initiative. The steering committee is usually
composed of upper management. In some companies, middle management and
hourly employees are also represented.
Some of the steering committee key roles include:
• Setting goals. Top management identifies opportunities, improvement needs

and sets strategic goals for the organization.
• Identifying projects. The steering committee selects those major

improvement projects critical to meeting quality and other goals.
• Selecting teams. Once a project has been identified, the steering committee
appoints a team to see the project through the remaining steps of the
improvement process.
• Supporting project teams. Some lean six sigma teams are generally required
to make significant improvements. It is up to the steering committee to see
that improvement teams are well prepared and equipped to carry out their
mission. Steering committee support may include:
• Providing lean six sigma training to black belts and green belts
• Providing training in team tools and techniques to other team members
• Providing a trained leader or facilitator to help the team work efficiently
• Reviewing team progress
• Approving revisions of the project mission
• Identifying/helping with team-related problems
• Helping with logistics, such as meeting sites
• Providing expertise in data analysis and/or survey design (black belts)
• Furnishing resources for unusually demanding data collection
• Communicating project results throughout the organization
• Monitoring progress. The steering committee is generally responsible for

keeping the improvement process on track, evaluating progress, and making
mid-course corrections to improve the effectiveness of the entire process.
* Various companies call this committee the lean six sigma steering committee, the
management council, or the executive steering committee.

TEAM ROLES
Team Facilitation
As noted earlier, the team leader in lean six sigma and other team arrangements is
often the facilitator. However, many companies find facilitators useful both for team
start-ups and for a variety of other team arrangements. The team leader and/or
facilitator must understand group dynamics and how a group moves through
developmental stages (for example: forming, storming, norming and performing).
Facilitators are useful in assisting a group in the following ways:
• Identifying members of the group that need training or skill building

• Avoiding team impasses before the task is completed
• Providing feedback on group effectiveness
• Summarizing points made by the group
• Balancing group member activity so each member is able to provide inputs
• Helping to secure resources that the team needs
• Providing an outside neutral perspective
• Clarifying points of view on issues
• Keeping the team on track with the process
• Helping with interpersonal difficulties that may arise
• Focusing on progress
• Assessing the change process
• Assessing cultural barriers (attitudes, personalities)
• Assessing how well groups are accomplishing their purpose
• Asking for feelings on sensitive issues
• Helping the leader to do his/her job more easily
• Coaching the leader and participants
If there is no facilitator, the team leader, an assigned black belt, or a coach must
assume many of the above duties.
The facilitator must avoid:
• Being judgmental of team members or their ideas and opinions

• Taking sides or becoming caught-up in the subject matter
• Dominating the group discussions
• Solving a problem or giving an answer
• Making suggestions on the task instead of on the process

TEAM ROLES
The Leader Role

Some teams have both leaders and facilitators. This is common for manufacturing
line teams. As a general rule, the team leader focuses on the team product (the
results) and the facilitator is most concerned with the team process. Teams
consisting of staff personnel will often have only one of these roles; generally a
facilitator. Within the lean six sigma framework, the team leader iis normally trained
both as a facilitator and a black belt. The leader will:
• Provide direction and suggest assignments

• Act as a communication hub and as a liaison with management
• Handle administrative details like meeting sites and scheduling
• Ensure that individual needs and expectations are considered
• Recommend meeting agendas and conduct meetings
• Assess group progress to plan, evaluate and initiate action
• Take the steps necessary to ensure success
• Possess an ability to encourage participation
• Be genuinely concerned about people and be a good listener
• Be encouraging and supportive
• Be accepting and tolerant of mistakes
• Work with, not over participants
• Stick to the task at hand
The leader's role is not to "boss" the team, but to ensure implementation of the
team's mission and charter. Facilitation and leadership requirements often diminish
as capability is developed within the team. Refer to the diagram below:
1
TEAM MEMBERS I
(COMBINED) I
,..-
_-------1I
,," !
,." I
I-
Z
-- ,. I
w /"__ I
:E
w ,. -- f
>
-I --__ I
o /
,. ,.
> / .......... LEADER I
~
.......... I
.......... 1
-- ,.. " "
TIME - - - -..
J'
Figure 4.4 Plot of Team Roles over Time

TEAM ROLES
The Team Member Role

Each team member is responsible for:
• Participating in training to become an effective team member

• Attending team meetings, as required
• Completing assignments between meetings
• Participating actively during meetings by contributing information and ideas
• Encouraging active participation by other team members
• Benefitting from the experience, expertise and perspectives of others
• Applying the steps of the improvement process
The Recorder Role

The team recorder/secretary is normally a full-fledged team member. The recorder
maintains the team's minutes and agendas. Often selected by team members, the
recorder also coordinates the preparation of letters, reports, and other documents.
Often, this duty is rotated among team members. He/she also distributes relevant
materials to team members. The recorder:
• Mayor may not participate as a member

• Takes clear notes including project responsibilities
• Publishes and distributes the minutes
• May ask for clarification of issues (for the record)
The Timekeeper Role

The timekeeper's role is an optional responsibility. This function sometimes
becomes the responsibility of the facilitator, when a facilitator is assigned to a team.
The timekeeper:
• Advises the team of the remaining time to review a project

• Enforces any time "norms" of the team

TEAM ROLES
Belbin Team Roles

Meredith Belbin describes a pattern of behavior that characterizes relationships in
team performance. People often have a mixture of roles and will have dominant and
sub-dominant traits. The Belbin roles can be used to identify types when starting
up teams and ensure there is a good balance (or that the imbalances are known and
can be managed). (12 Manage, 2007)1 (Loog, 2007)25
Action-oriented roles:
• Shaper: Shapers are highly motivated people with a lot of drive, energy, and
need for achievement. They often seem to be aggressive ,extroverts.
• Implementer: Implementers are well organized and have practical sense. They
favor hard work and tackle problems in a systematic fashion.
• Completer: Completers have a great capacity for follow-through and attention

to detail. They seldom start what they cannot finish.
People-oriented roles:
• Coordinator: Coordinators have an ability to cause others to attain shared

goals. They spot individual talents and use them to pursue group objectives.
• Team worker: Team workers are the most supportive members of a team.
They are sociable and adaptive to different situations and people.
• Investigator: Investigators are good communicators both inside and outside

the organization. They are extroverted and enthusiastic.
Problem solving roles:
• Plant*: Plants are innovators and can be highly creative. They provide the
seeds and ideas from which major developments spring.
• Evaluator: Evaluators are serious, prudent, individuals,. They are slow

deciders and possess a critical thinking ability.
• Specialist: Specialists are self-starting professionals. They pride themselves

in acquiring technical skills and specialized knowledge.
* A plant can be considered a planter or innovator.

TEAM ROLES
Belbin Team Roles (Continued)

Individuals are seldom strong in all nine Belbin team roles. Additionally, many of the
contributing role strengths have associated weaknesses. Listed in Table 4.5 are
typical strengths and "allowable" weaknesses.
Role Team Strengths Allowable Weaknesses

Shaper Shapers bring the drive and courage to Shapers offend people and will
overcome obstacles. The shaper is display aggression in the
committed to achieving ends. pursuit of goals. Two or three
shapers in a group can lead to
conflict.
Implementer Implementers turn ideas into practical Implementers are conservative,

actions. They tend to work for the team in a inflexible, and slow to respond
practical and realistic way. to new possibilities.
Completer Finishers finds errors and omissions. They Finishers worry unduly and are
Finisher deliver their contributions on time and pay reluctant to delegate. They tend
attention to details. to be over anxious.
Coordinator The coordinator is a positive thinker who Coordinators can be seen as

supports goal attainment and effort in others. manipulative. They might not
They clarify goals and delegate well. stand out in a team.
Teamworker Team workers tend to keep team spirit up and They tend to be indecisive in
allow other members to contribute. They moments of crisis and are
bring cooperation and diplomacy to a team. reluctant to offend.
Resource The resource investigator explores They are over-optimistic and

Investigator opportunities and develops contacts. They may lose interest quickly. They
are good negotiators. are not the sources of original
ideas.
Plant A plant brings creativity, ideas, and Plants ignore incidentals and
Innovator imagination to a team. They can solve may be too preoccupied to
difficult problems. communicate effectively.
Monitor The monitor evaluator is not deflected by The evaluator may appear dry,
Evaluator emotional arguments. They are serious boring, and overcritical. They
minded and bring objectivity and judgment to are not good at inspiring others.
options.
Specialist Specialists bring dedication and initiative. They may contribute only on a
They provide needed knowledge and narrow front and dwell on
technical skills. technicalities.
Table 4.5 Description of Nine Belbin Team Roles
Table 4.5 is modified from 12Manage (2007)\

TEAM STAGES
Team Stages
Most teams go through four development stages before they become productive:
forming, storming, norming, and performing. These stages can also be cyclical.
Individuals may be storming with one teammate and performing with another. Bruce
W. Tuckman (1965)39 first identified the four development stages.
Forming
Forming is the beginning of team life. Expectations are unclear. Members test the
water. Interactions are superficial. This is the honeymoon stage. When a team
forms, its members typically start out by exploring the boundaries of acceptable
group behavior. As each member makes the transition from individual to team
member, each looks to the team leader (or facilitator) for guidance as to his or her
role and responsibilities.
Storming
The second phase consists of conflict and resistance to the group's task and
structure. There are healthy and unhealthy types of storming. CClnflict often occurs
in the following major areas: authority issues, vision and values dissonance, and
personality and cultural differences. However, if dealt with appropriately, these
stumbling blocks can be turned into performance later.
This is the most difficult stage for any team to work through. Teams realize how
much work lies ahead and feel overwhelmed. They want the project to move forward
but are not yet expert at team improvement skills. They often cling to their own
opinions, based on personal experience, and resist seeking the opinions of others.
This can lead to hurt feelings and unnecessary disputes. A disciplined use of the
quality improvement process and the proper tools and commUinication skills can
assist teams members to express their various theories, lower their anxiety levels,
and reduce the urge to assign blame.
Norming
During the third phase, a sense of group cohesion develops. Team members use
more energy on data collection and analysis as they begin to test theories and
identify root causes. Members accept other team members and develop norms for
resolving conflicts, making decisions, and completing assignments. Norming takes
place in three ways. First, as storming is overcome, the team becomes more relaxed
and steady. Conflicts are no longer as frequent and no longer throw the team off
course. Second, norming occurs when the team develops a ro,utine. Scheduled
team meetings give a sense of predictability and orientation. Third, norming is
cultivated through team-building events and activities. Norming is a necessary
transition stage. A team can't perform if it doesn't norm.

TEAM STAGES
Team Stages (Continued)

Performing
This is the payoff stage. The group has developed its relationships, structure, and
purpose. The team begins to tackle the tasks at hand. The team begins to work
effectively and cohesively. During this stage, the team may still have its ups and
downs. Occasionally, feelings that surfaced during the storming stage may recur.
Refer to Figure 4.6 for a graphical display of the performance of teams as they
advance through the team evolutionary stages.
Performing
Members:
show maturity
focus on the process
achieve goals
operate smoothly
Norming
Members:
cooperate
talk things out
Q)
(.)
focus on objectives
r:: have fewer conflicts
ns
E
L.. Storming
.g Members:
Q) have confrontation
~
think individually
are learning roles
have divided loyalties
Forming
Members are:
inexperienced
excited
anxious
proud
Time
Figure 4.6 Schematic of Team Development Phases
Adjourning
At the end of most lean six sigma projects the team disbands. This step is called
adjourning to rhyme with four other team stages (forming, storming, norming and
performing). Adjourning is also a very common practice for other project teams,
task forces, and ad hoc teams.

TEAM STAGES
Team Stages (Continued)
Stage 1 FORMING Stage 2 STORMING
BEHAVIORS BEHAVIORS
Lack of task focus Problem solving is sup,erficial
Difficulty in defining problems There is petty arguing
Uneven participation Hidden agendas and cliques emerge
Ineffective decision making Decisions don't come easily
Resistance to team building Plenty of uncertainties persist
FEELINGS FEELINGS
Excitement, anticipation, and pride Resistance is seen
Shaky alliance to the team Individual attitudes vary widely
Suspicion, fear, and anxiety Anger and jealousy abc)und
Roles and responsibilities are unclear
HOW TO IMPROVE
HOW TO IMPROVE Follow a problem solving format
Take time to become acquainted Clearly define roles
Establish mission and goals Debrief meetings for cClntent and process
Establish team ground rules Deal openly with conflilct
Add structure to meetings Work to expose hidden agendas
Train members in team concepts Focus team on goals
Encourage equal participation
Stage 3 NORMING Stage 4 PERFORMING
BEHAVIORS BEHAVIORS
Attitudes improve Members to work through problems
Trust and commitment grow Members manage the glroup process
Some goals and objectives are achieved There is creativity and unformality
Feedback becomes regular and objective High levels of unity and spirit are seen
Conflicts are dealt with and resolved Close bonds form
The leader receives respect
Some leadership is shared by the team FEELINGS
Self improvement is noted
FEELINGS Acceptance of weakness
Comfort with giving feedback Appreciation of strengths
Comfort with receiving feedback Satisfaction with team progress
Sense of cohesion and spirit Team knows clearly what it is doing
Friendlier and more open exchanges
HOW TO IMPROVE
HOW TO IMPROVE Promote openness
Evaluate team performance Permit more self direction
Periodic summaries of progress Establish new goals
Create ties outside of the team (Tuckman, 1965)39
Table 4.7 Details on the Four Stage Team Process

TEAM STAGES
Team Life Cycle Characteristics

Shown below is another representation of team development stages.
Effectiveness Optimize
Build
Time
Figure 4.8 Team Life Cycle Characteristics
Build Phase (Forming/Storming)
• Group will be uncertain

• Group lacks cohesiveness
• Group will not easily develop consensus
• Leader exhibits a high task/ high relationship style
Develop Phase (Norming)
• Task related work is assumed by the group

• The group must work to involve any non-participating members
• Leader exhibits a low task/high relationship style
• Team focuses on presentations, tasks, and relationships
Optimize Phase (Performing)
• Members prioritize and perform tasks

• Members work out decisions in a caring way
• Conflict is accepted, but cooperation is preferred
• Team leader is a delegator and exhibits a low-task/low-relationship style
• Team exhibits a high-task/high relationship style

TEAM DYNAMICS
Team Dynamics
On this and the following pages are a number of issues (in addition to team
development stages) that fall into the category of team dynamics. These include
team recognition, a number of team problem areas, communication challenges, and
negotiating techniques.
Recognition and Reward

The ultimate reason that rewards and recognition are given is te) provide positive
reinforcement for good performance or correct behavior, with the expectation that
this performance will be repeated in the future. The effect of the rleward will depend
on the perception of the person receiving the reward. Recognitio,n and rewards for
teams and team members can be grouped into the following types:
Material items of significant value or equivalent:
• Cash or gift greater than $1,000

• Vacation time or trip
• Patent award
• Bonus or percent of savings/profit
Material items of incidental value or equivalent:
• Cash less than $1,000 • Picture on bulletin board

• Trophy, plaque, certificate • A meal with the boss
• Special parking space • Name on a list
Intangible Items:
• Satisfaction • Thanks
• Pleasure • Admiration
• Friendship • Notoriety
• Learning experience • Prestige
Individual rewards and recognition are best received when they are personal to the
individual receiving them. If the award is unique, it has greater value than the same
recognition a second or third time.
Team rewards should be the same for all members of the team. Intangible rewards
are generally not given from one person to another, but yet people may receive them
as a result of their activities. Probably one of the best rewards is "thank you" when
it is sincerely meant.

TEAM DYNAMICS
Groupthink
One aspect of group cohesiveness can work to a team's disadvantage. Members of
highly cohesive groups may publicly agree with actual or suggested courses of
action, while privately having serious doubts about them. Strong feelings of group
loyalty can make it hard for members to criticize and evaluate other's ideas and
suggestions. Desiring to hold the group together and avoid disagreements may lead
to poor decision making. Psychologist Irving Janis (1971 )23 calls this phenomenon
"groupthink," the tendency for highly cohesive groups to lose their critical
evaluative capabilities. Janis ties a variety of well-known historical blunders to
groupthink, including the lack of preparedness of the U.S. naval forces for the 1941
Japanese attack on Pearl Harbor, the Bay of Pigs invasion under President Kennedy,
and the many roads that led to the USA's involvement in Vietnam.
Irving Janis (1971 )23 describes groupthink as: "A mode of thinking that people
engage in when they are deeply involved in a cohesive in-group, when the members'
strivings for unanimity override their motivation to realistically appraise alternative
courses of action." "The more amiability and esprit de corps among members of a
policy- making in-group, the greater is the danger that independent critical thinking
will be replaced by groupthink."
Eight Symptoms of Groupthink

1. Illusion of invulnerability: Feeling that the group is above criticism.
2. Belief in inherent morality of group: Feeling that the group is inherently

"right" and above any reproach by outsiders.
3. Collective rationalization: Refusing to accept contradictory data or to consider

alternatives thoroughly.
4. Out-group stereotypes: Refusing to look realistically at other groups.
5. Self-censorship: Refusing to communicate personal concerns to the group as

a whole.
6. Illusion of unanimity: Accepting consensus prematurely, without testing its

completeness.
7. Direct pressure on dissenters: Refusing to tolerate a member who suggests

the group may be wrong.
8. Self-appointed mindguards: Protecting the group from hearing disturbing

ideas or viewpoints from outsiders.

TEAM DYNAMICS
Risky - Shift
Many people think that the proposed solutions to team projects would be fairly
conservative. That is, any proposed solution is an "averaged" r'emedy. However,
those experienced with team mechanics and dynamics have found the opposite to
be the case. Teams often get swept up with expansive and expensive remedies.
There are ways to combat this tendency. One way is to discuss risky-shift openly
in the initial training. Another approach is to have a team member (after full
discussion of the issue) ask the question, If this were our personal money, would we
still risk it on the proposed solution?
Additional Team Problem Areas

The following team problem areas are frequently encountered and must be
addressed by team leaders, facilitators, sponsors, and upper management:
• Waning management support. Symptoms of this problem are:
• The supporting champion leaves

• There is an increasing need for top management stimuli
• There is lip service without action by mid-managemen1:
• There are delays in implementation of projects
• There is passive, rather than active, management support
• There is inadequate documentation of meeting results

• There is inadequate time or training given to team members
• Teams expose problems that may be viewed as a threat by mid-management
• Controversies can develop between facilitators and leadel's
• People with good facilitation skills may be hard to find
• If there is an existing suggestion program, how should it function with teams?
• If there is a labor union, there may be resistance to team involvement
• Teams may tackle problems outside the realm of their responsibility
• Crisis management always creates team scheduling problems
• A company's measurement and reward system may be inconsistent
• Unproductive competition and conflict may occur among 1team members
• Idea generation and evaluation are not kept separate
• Facts and opinions are not distinguished
• There is a failure to assign team members specific responsibilities
VILLANOVA UNIVERSITY IV - 30 LEJ~N SIX SIGMA HANDBOOK

TEAM DYNAMICS
Common Team Problems
Problem Examples How to Fix

Floundering
· Team direction is unclear
· Leader must provide clarity
·· ··
Members seem overwhelmed Review the team purpose
Decisions are postponed Ask "How can we proceed?"
Dominant
Participants
·· Members interrupt others
Members dominate the conversation
•
·
Promote equal participation
Structure the discussion
Overbearing
Participants
·· A member has excessive influence
A member has legitimate authority
·• Reinforce team concepts
Ask the expert to lead the group
· A member is an "expert"
· Have a private discussion with "expert"
Negative
Nellies
·· Members say "We tried that already"
Members defend their turf
·• Reinforce the positive
Ask for other points of view
• Members are negative of suggestions
· Separate idea generation from criticism
···
Opinions as Members present opinions as facts • Ask for support data
··
Facts Members make unfounded assumptions Question opinions and assumptions
Self assurance seen as unquestionable See groupthink discussion
Shy Members
·· Members are reluctant to speak
Members afraid of making mistakes ·· Structure group participation
Direct conversation their way
Jumpto
Solutions
··
• Members rush to accomplish something
·
Members avoid data collection and analysis •
Reinforce the need for data analysis
Ask for alternate solutions
Members want immediate decisions
· Slow the process down
Attributions
·· Members make casual inferences
Members don't seek real explanations
·· Challenge assumptions
Challenge judgments
· Members make psychological judgments
· Ask for data to support conclusions
Put-downs
(Discounts &
···
A member's comments are ignored
Members are not listening
·· Encourage active listening
Encourage equal participation
Plops) The meaning of a suggestion is missed
· Talk to parties privately
· Sarcasm is noted
· Promote uniform idea consideration
Wanderlust
(Tangents & ·· Conversations stray from the main topic
Sensitive issues are avoided
·· Follow a written agenda
Reinforce team operating guidelines
Digressions)
· Group pursues tangents
· Redirect the discussion
Feuding
· Win-lose hostilities emerge
· Confront the adversaries alone
·· ··
The team takes entrenched sides Reinforce team operating guidelines
Some members become spectators Replace the guilty parties if necessary
Risky-Shift
· Expansive and expensive remedies are
suggested (using company money)
• Ask "If this were my personal money
would I still spend it?"
Table 4.9 Common Team Problems
All of the above problem areas can be minimized with proper team training and team
member awareness. A portion of the above Table was modified from Lorber (2001 )26.
Many of the problem areas are identified by Scholtes (1996)34.

TEAM DYNAMICS
Communication Techniques
For any organization to survive, information must continually now vertically and
horizontally across the company. A good team leader must be able to understand
the process flow and how to use it. Motivating people, collaborating with people,
and getting things done by people are all accomplished with communication. An
effective leader will operate as a:
• Monitor of external information from contacts, peers, or experts

• Monitor of internal information
• Disseminator or distributor of information
• Spokesperson to outsiders
• Decision maker using gathered information
Some basic purposes of communications are:
• To influence the work of peers or organizational employees

• To inform employees of necessary job performance information
• To control the organization's progress toward the objectives
• To inspire employees through displays of values, attitudes, and modeling
Downward Flow of Communication
Team leaders must relay information and give orders and directives to team
members and peers. Typical information includes:
• Instructions for team members

• Rationale for the instructions to the team members
• Policies and procedures of the company
• Team member performance feedback (Bateman, 1993)4
A detractor in downward communications is the filtering process. An original

message from the highest officer of the company will often be distorted by the time
it reaches floor level employees.

TEAM DYNAMICS
Communication Techniques (Continued)

Upward Flow of Communications
Upward communication consists of information relayed from the bottom or
grassroots, to the higher levels of the company. This gives the higher levels a
chance to learn about what is happening in the lower levels. If management can act
on the information from below, it can motivate the lower levels. As the information
regarding overall production plans and goals moves upward, each manager gets to
put his/her "spin" on the data. To encourage more reliable upward communications,
top level managers can have open door policies, surveys, questionnaires,
suggestion systems, breakfast meetings, shift meetings, and the like.
Misleading Information
Misleading information has many origins and causes. There are three important
reasons why management may not receive accurate and complete information:
• In vertical communication, subordinates may withhold information that tends

to discredit them. Thus, the information undergoes a screening process
which allows only the positive to be reported.
• There may be a tendency to tell a supervisor, particularly the authoritarian

type, what he or she wants to hear, shielding the negative. This practice
results in concealed errors and avoidable delays and costs.
• An incumbent manager or leader is not always surrounded by allies. In a

typical organization, there may be competition and rivalry.
Horizontal Communications
Horizontal communications refers to the sharing of information across the same

levels of the organization. The production engineer shares information with
production planning. The planning group, in turn, shares information with
manufacturing, and so forth. It is a very important part of the communications
process.
VILLANOVA UNIVERSITY IV·33 LEAN SIX SIGMA HANDBOOK

TEAM DYNAMICS

Formal and Informal Communications
Formal communications are official company sanctioned methods of communicating

to the employees. The methods can occur up, down, or across the organization.
The informal communication link in a company is the grapevine, rumor mill, etc.
Information on many things is discussed. Clarification regarding the latest policy
can be made; just as the clarification on Joe's "transfer or retire!ment."
Special Roles
Gatekeepers are described as individuals who are at the crossroads of
communications channels. They are centers of information, normally because of
their jobs. People and groups pass information to the gatekeeper due to this
positioning. Boundary spanners are individuals who have positions that link them
with others outside of their work units. They exist to exchange in1formation between
groups. (Gordon, 1991 )14
Appropriate Vehicles for Different Situatiolns
Reprimanding an employee should be done in private and is typically done verbally.

Second or third offenses should be done in person with the employee and
documented with a written letter given to the employee. Business agreements may
be made verbally, but must always be followed with a letter to ensure both parties
have the same understanding of the agreement. Strategic initiatives, such as lean
six sigma, should be communicated by the CEO. The vehiclle must match the
situation.
Other Communication Forms
The spoken word via the telephone, face·to·face, formal briefings, videotapes, and
even internet are forms of oral communications. Examples of written
communications include letters, reports, computer messages, and e-mails. The
written forms can be described as one-way channels. That is, in one-way channels
feedback or response to a report or a posting is not immediate. Face-to-face
meetings generally allow for immediate feedback from the receiver to the sender
(two-way communications).

TEAM DYNAMICS

Other Communication Forms (Continued)
There are a variety of nonverbal communication signals from people that the
manager might be able to pick up from face-to-face or group meetings. The
nonverbal signals include:
• Hand movements • Head movements

• Eye contact • Use of interpersonal space
• Body posture • Leg positions
In the use of interpersonal space (proxemics), certain cultures are comfortable with
a set spacing. Note that not all "non-verbal" signals really mean what others claim
them to mean. For instance, folded arms across the front of the body is supposed
to indicate a defensive posture by the person. It might also mean that the arms are
comfortable in that position. (Schermerhorn, 1993)33
Important factors always present in motivating people are verbal and written
communication skills. The ability to explain and clarify has been long recognized.
Speaking and writing abilities are essential for leadership success.
Questioning Techniques
The skillful use of questioning is of great value to the team leader. Scholtes (1996)34
suggests the following seven key questions that managers should ask:
• Why? Use the Japanese technique of asking why, five times.

• What is the purpose?
• What will it take to accomplish the project?
• Will someone care?
• What is your theory (hypothesis) on the subject?
• What data do you have?
• Where did your data come from?
Auvine (1978)3 provides some additional ideas for the art of asking questions:
• Avoid leading questions: let the group or individual draw their own conclusion
• Phrase questions in a positive manner
• Prepare questions in advance of the interview
In line with this, the use of open ended questions will allow for some discussion and
probing rather than just a simple "yes" or "no" answer.

TEAM DYNAMICS

Listening Strategies
Listening, the other half of the communication concept, has received far too little
attention. Effective project leaders have learned the art of listening. Verbal
information can often be very difficult to understand, even when active listening
takes place. A project leader will spend up to 45% of the time being a listener.
Active listening is recommended. Active listening is defined als helping find the
source of problems or meaning. A passive listener will respond in a manner that will
discourage the message sender from saying more.
Tips for good listening include:
• Put the message sender at ease

• Show that you want to listen
• Remove listening distractions
• Empathize with the person
• Be patient with your response
• Hold your own temper
• Avoid arguing and criticism
• Ask questions
• Stop talking (Schermerhorn, 1993)33
Many of us would rather hear ourselves speak than listen to another person. The
good news is listening skills can be learned and developed by practice.
Negotiation Techniques
Nierenberg (1986)28 states that negotiating is the act of exchanging ideas or
changing relationships to meet a need. As common and as important as negotiating
is in everyday life, most people learn to negotiate through trial and error.
Negotiating should not be a process of using overwhelming and irresistible force on
the other party. Some degree of cooperation must be employed in the process.
Many negotiating methods are used successfully. However, in dealing with people
in a business context, the best approach is to think win-win. The concept of win-win
negotiating is for both sides to emerge with a successful deal.

CONFLICT RESOLUTION
Conflict Resolution
Conflict is the result of mutually exclusive objectives or views, manifested by
emotional responses such as anger, fear, frustration, and elation. Some conflicts
are inevitable in human relationships. When one's actions may be controlled by the
actions of another, there is opportunity for conflict. Common causes of conflict
include:
• Organizational structure • Status threats

• Value differences • Personality clashes
• Role pressures • Differences in ideals
• Perceptual differences • Changes in procedures
• Divergent goals • Discrepancies in priorities
Conflicts may be categorized as to the relationship between the parties involved in

the conflict. The relative power or influence between parties is a factor both in the
cause and the resolution of the conflict. Categories of conflicts are:
• Intrapersonal - within an individual

• Interpersonal- between any two people
• Intragroup - within a group
• Intergroup - between groups
• Interdepartmental - between departments
• Intercompany - between companies
The results of conflicts may be positive in some instances, negative in some, and
irrelevant in others. Irrelevant conflicts occur when the outcome has neither positive
nor negative effects for either party.
Positive conflicts result in:
• A combined desire to unite and improve

• Win - win situations
• Creative ideas brought forth
• Better understanding of tasks, problems
• Better understanding of other's views
• Wider selection of alternatives
• Increased employee interest and participation
• Increased motivation and energy

CONFLICT RESOLUTION
Conflict Resolution (Continued)

Negative conflicts result in:
• Hostile, impulsive drives to destroy • Undesirable consequences

• Win - lose situations • Isolation
• Lose - lose situations • Loss of productivity
Each individual uses a number of ways to deal with conflicts depending upon the
circumstances and the relationships between the people invc)lved. Whether a
conflict resolution method is appropriate or effective will also depend on the
situation. The ways of dealing with conflicts can be depicted in a two dimensional
model for conflict handling behavior, adapted from Thomas-Kilmann Conflict Mode
Instrument: (Thomas, 1975)37
Competing Collaborating
G)
>
t:
G)
U)
U)
~
Compromising
t
~
t:
G)
U)
U)
as
£:
::J Avoiding Accommodating
Uncooperative'" cooperative
Figure 4.10 Conflict Resolution Matrix
• Avoiding is unassertive and uncooperative - the individual withdraws from the

situation. (You lose, I lose).
• Accommodating is unassertive but cooperative - the individual yields to the

wishes of others. (You win, I lose).
• Competing is assertive and uncooperative - the individual tries to win, even

at the expense of others. (You lose, I win).
• Collaborating is assertive but cooperative - the individual wants things done

their way, but is willing to explore solutions which satisfy the other person's
needs as well. (You win, I win).

CONFLICT RESOLUTION
Conflict Resolution (Continued)

• Compromising is intermediate in both assertiveness and cooperativeness-
the individual is willing to partially give in to reach a middle position, splitting
the differences, and partially satisfying both parties. (Neither win or lose).
There is no specific right or wrong method for handling conflicts. The method that
works best depends upon the situation and the interactions between the affected
parties. The project leader must be able to understand and use all of the conflict
resolution techniques, as is appropriate for the situation. The following are general
applications for the various conflict handling methods:
• Avoiding is appropriate for less important issues or when the potential

damage from conflict outweighs the benefits of the goal.
• Accommodating is suitable when one party is wrong or the issue is more

important to the others than it is to yourself.
• Competing is applicable when quick decisions are needed and a stronger

influence is held by one side.
• Collaborating is used when both views are important and an integrated

solution is desired.
• Compromising is used when two opponents have equal power and the goals
are not worth the effort or disruption of mutually exclusive solutions.
According to Tjosvold (1984)38 the following guidelines can be used by project

leaders to resolve conflict:
• Determine how important the issue is to all involved

• Determine if the issue can be discussed by all involved
• Select a private meeting place
• Make sure that all parties understand their responsibilities
• The parties must deal with both the problem and solution
• Let all parties make opening comments
• Let parties express their concerns, feelings, ideas, etc.
• Guide all parties toward a clear problem definition
• Encourage participants to propose solutions
• Examine the problem from a variety of perspectives
• Discuss any and all proposed solutions
• Evaluate the costs versus the gains for all proposed solutions
• Choose the best solution and ask participants how it might be improved

TEAM TOOLS
Team Tools
Often parties need to arrive at a decision or problem resolution using team tools.
Several useful techniques are presented below.
Nominal Group Technique

The nominal group technique (NGT) was developed at the University of Wisconsin
in 1968 by Andre Delbecq and Andrew Van deVen. (Eitington, 1989)8. This technique
brings people together to solve problems but limits initial interaction among them.
The concept is to prevent peer or social pressures from influencing the generation
of ideas. Hence the term "nominal" is used to describe the limiting of
communications. To conduct a NGT problem solving meeting:
• A facilitator or moderator leads the discussion.
• A group of five to nine individuals are assembled for idea generation.
• A problem is presented.
• Before any discussion is held, all members create ideas silently and
individually onto a sheet of paper for about 5 to 10 minutes.
• The facilitator then requests an idea from each member in sequence. Each
idea is recorded until ideas are exhausted.
• Like brainstorming, no discussion is allowed at this point.
• After the exhaustion of ideas, the clarification, support, and evaluation of

ideas is permitted. Hitchhiking on the ideas of others is encouraged.
• Voting for the best solution idea is then conducted (rank ordering, priority
ratings, etc.). Several rounds of voting may be needed before a "best" idea
is found. One voting method employs the use of cards and a Pareto
breakdown of favored ideas.
The facilitator should allow about 60 to 90 minutes for a problem solving session.
As with brainstorming sessions, the facilitator should avoid trying to influence the
problem solving process. The chief advantage of this technique is that the group
meets formally, and yet encourages independent thinking. The authors of this
Handbook feel that exposure to other problem techniques is usefUlI for the NGT team
members.

TEAM TOOLS
Multivoting
Multivoting is a popular way to select the most popular or potentially most important
items from a previously generated list. Quite often, in a team environment, a list of
ideas is generated by simple brainstorming. In many cases, brainstorming can be
used to segregate potential causes into their 4 or 5 M components in a cause-and-
effect diagram. The list that is generated can consist of either ideas for improvement
or potential causes for a problem (too much scrap, too much inventory, excessive
downtime, etc.).
Having a list of ideas does not translate into action. Often, there are too many items
for a team to work on at a single time. Additional data or experimentation can help
identify significant items. However, it is often worthwhile to narrow the field to a few
items worthy of immediate attention.
According to Scholtes (1996)3\ multivoting is useful for this objective and consists
of the following steps:
1. Generate and number a list of items
2. Combine similar items if the group agrees
3. If necessary, renumber the list
4. Allow members to choose several items that they feel are most important. A
suggested guide is to permit each member a number of choices equal to at
least one-third of the total items on the list.
5. Members may make their initial choices silently and then the votes are tallied.
This is usually done by a show of hands as each item is announced.
6. To reduce the list, eliminate those items with the fewest votes. Group size will
affect the results. The facilitator may chose to eliminate items receiving 0-4
votes.
It should be noted that most problem solving teams can only work on two or three
items at a time. The items receiving the largest number of votes are usually worked
on or implemented first. The original list should be saved for future reference and/or
action.

TEAM TOOLS
Brainstorming
Brainstorming is an intentionally uninhibited technique for generCllting creative ideas
when the best solution is not obvious. The brainstorming technique is widely used
to generate ideas when using the fishbone (cause-and-effect) diagram.
Generate a large number of ideas: Don't inhibit anyone. Just let the ideas out. The
important thing is quantity, but record the ideas one at a time.
Free-wheeling is encouraged: Even though an idea may seem half-baked or silly, it

has value. It may provoke thoughts from others.
Don't criticize: There will be ample time after the session to sift through the ideas
for the good ones. During the session, do not criticize ideas because that might
inhibit others.
Encourage everyone to participate: Everyone thinks and has, ideas. So allow

everyone a chance to speak. Speaking in turn helps.
Record all the ideas: Appoint a recorder to write down everything suggested. Don't
edit the ideas, just jot them down as they are mentioned. Keep a permanent record
that can be read.
Let ideas incubate: You are freeing the subconscious mind to be creative. Let it do
its work by giving it time. Don't discontinue brainstorming sessions too soon.
Consider adding to the list at another meeting.
Select an appropriate meeting place: A place that is comfortablle, casual, and the
right size will greatly enhance a brainstorming session.
Group size: The ideal group size is 4-10 people.
Brainstorming, just like the cause-and-effect diagram, does not necessarily solve
problems or create a corrective action plan. It can be effectively used with other
techniques, such as multivoting, to arrive at a consensus as to an appropriate
course of action. It is a participative method to help work teams achieve their goals
and objectives.
VILLANOVA UNIVERSITY IV - 42 LEJI\N SIX SIGMA HANDBOOK

TEAM PERFORMANCE EVALUATION
Team Performance Checklist

A very simple team process checklist is presented below. Similar forms can be used
by team members or facilitators to evaluate some of the process factors of team
performance.
Establish agenda
Clarity about goals Uncertainty about
and content II I I II goals and content
Stick to subject
Always to the point II I I II Frequently gets off track
Only one person talks
No interruptions II I I I Multiple conversations
Build on positive
Comments are additive I I I II Focuses on what's wrong
Active listening
No misunderstandings II I I I[ Comments that show lack

of clarity
Participation
Active participation II I I I Opinions only, no data
Consensus
Buy-in on all issues II I I I People are stampeded
Figure 4.11 Illustrative Team Process Checklist

Team Performance Factors

Listed below are some illustrative relationship, process, goal, environmental, and
role factors that can be used (or expanded upon) to evaluate team performance.
Well Functioning Teams Poorly Functioning Teams

Relationship Factors
There is a firm team identity Members do not identify with the team
Conflict is openly discussed There are open or covert F)ersonality conflicts
Team members support each other Relationships are competitive
Members enjoy each other Members are defensive
Process Factors
Decisions are made by consensus Decisions are made by a few members
Meetings are efficient and task oriented Meetings are unproductivl3
There is growth and learning Minor points are debated
All members participate in discussions Members are late, passive, or do not attend
Members are kept informed. Minutes are kept No record of progress is kept
There is ongoing performance feedback Feedback awaits the end of the project
Members listen well There are frequent interruptions
Goal Factors
Team members help set objectives Most goals are predetermined for teams
Objectives are understood by all members Goals are unclear or poorly communicated
Objectives are realistically set and met Members are oblivious to team goals
Environmental Factors
Team members are in close physical proximity Physical separation prevents attendance
There are adequate skills and resources Resources are inadequate
There is management and member support There is a lack of organizational recognition
Role Factors
There is strong effective leadership No clear leader is identified
Clear responsibilities are defined Members engage in power plays
Roles are understood and supported by all There is buck-passing of rt~sponsibility
Members work as a team Members act independently
Table 4.12 Team Performance Factors

Management Presentations
Another useful technique in team performance evaluation are formal management
presentations.
Management presentations are opportunities for lean six sigma and other
improvement teams to:
• Display skills
• Show accomplishments
• Summarize projects
• Gain any necessary management approvals
• Keep lines of communication open to upper management
• Demonstrate an understanding of the customer's true needs
Teams should follow the general presentation guidelines:
• Have a leader and a handout report
• Introduce team members and give everyone a chance to speak
• Capture key project action steps
• Indicate both the costs and the benefits of recommendations
• Present an implementation plan
• Use visual aids - be professional
• Emphasize achievements and accomplishments
• Start and end on time (30 minutes is a typical target)
• Present no sudden surprises!

REFERENCES
References
1. 12Manage. (2007). "Belbin Team Roles." Downloaded January 5, 2007 from
http://www.12manage.com/description_belbin_team_roles_theory.html
2. Albrecht, K. (1992). The Only Thing That Matters. New York: Harper Collins.
3. Auvine, B., Densmore, B., Extrom, M., Poole,S., & Shanklin, M. (1978). A Manual
for Group Facilitators. Madison, WI: The Center for Conflict Resolution.
4. Bateman, T.S., & Zeithaml, C.P. (1993). Management: Function & Strategy, 2nd
ed. Homewood, IL: Irwin.
5. Besterfield, D.H., et at (1999). Total Quality Management, 2nd ed. Upper Saddle
6. Breyfogle, F.W. III, et al. (2000). Managing Six Sigma. New York: John Wiley and
Sons.
7. Deming, W.E. (1993). The New Economics, MIT Center for Advanced Engineering
Study.
8. Eitington, J. (1989). The Winning Trainer, 2nd ed. Houston: Gulf Publishing.
9. Feigenbaum, A.V. (1991). Total Quality Control. 3rd ed. Revised, Fortieth
Anniversary Edition. New York: McGraw-HilI.
10. Furlong, C.B. (1993). Marketing for Keeps Building Your Busfness by Retaining
Your Customers. New York: John Wiley.
11. Futrell, D. (1994, April). "Ten Reasons Why Surveys Fail." Quality Progress,
27(4}.
12. Geddes, L. (1993). Through the Customer's Eyes. New York: AMACOM.
13. Gee, G., Richardson, W.R., & Wortman, B.L. (2005). CMQ Primer. Terre Haute, IN:
14. Gordon, J.R. (1991). A Diagnostic Approach to Organizational Behavior, 3rd ed.
Boston: Allyn and Bacon.

REFERENCES
15. Griffin, A., Gleason, G., Preiss, R., & Shevenaugh, D. (1995, Winter). "Best
Practice for Customer Satisfaction in Manufacturing Firms." Sloan
Management Review, 36(2}.
16. Harry, M., & Schroeder, R. (2000). Six Sigma, The Breakthrough Management
Strategy, New York: Currency/Doubleday.
17. Hauser, J.R., & Clausing, D. (1988, May-June). "The House of Quality." Harvard
Business Review, 66(3}, 63-73.
18. Hunter, M.R., & Van Landingham, R.D. (1994, April). "Listening to the Customer
Using QFD." Quality Progress, 27(4}.
19. Hayslip, W.R. (1994, April). "Measuring Customer Satisfaction in Business

Markets." Quality Progress, 27(4}.
20. Heskett, J.L., Jones, T.O., Loveman, G.W., Sasser, Jr., W.E.S., & Schlesinger, L.
(1994, March-April). "Putting the Service-Profit Chain to Work." Harvard
Business Review.
21. Industrial Engineering. (1995, February). "Emphasize Customer Retention over

Customer Acquisition." Industrial Engineering, 27(2}.
22. Ishikawa, K. (1985). What Is Total Quality Control? The Japanese Way.
Englewood Cliffs, NJ: Prentice Hall.
23. Janis, I. (1971, June). "Groupthink," Psychology Today.
24. Juran, J.M. (1999). Juran's Quality -Handbook, 5th ed. New York: McGraw-Hili.
25. Loog, K. (2007). Downloaded January 5, 2007 from

http://www.hot.ee/tarkvaraprojektlbelbin_teamroles.pdf
26. Lorber. (2001). "Fixing Group Problems." Downloaded from

http://www.lorober.com/resources/documents/groupproblems10.htm
27. Lowenstein, M.W. (1995). Customer Retention: An Integrated Process for

Keeping Your Best Customers. Milwaukee: ASQ Quality Press.
28. Nierenberg, G. (1986). Complete Negotiator. New York: Nierenberg & Zeif.

REFERENCES
29. Osborne, D., & Gaebler, T. (1992). Reinventing Government: How the
Entrepreneurial Spirit is Transforming the Public Sector. New York: Addison-
Wesley.
30. Pande, P.S., Neuman, P.R. & Cavanagh, R.R. (2000). The Six Sigma Way. New
York: McGraw-HilI.
31. Pearson, T.A. (1999, May). "Measurements and the Knowledge Revolution."
presented at the ASQ Annual Quality Congress.
32. Schaaf, D., & Zemke, R. (1989). The Service Edge. New York: Plume.
33. Schermerhorn, J.R., Jr. (1993). Management for Productivity, 4th ed. New York:
John Wiley & Sons.
34. Scholtes, P.R., Joiner, B.L., & Streibel, B.J. (1996). The Team Handbook, 2nd ed.
Madison, WI: Oriel, Inc.
35. Simmerman, S.J. (1993, November). "Achieving Service Quality Improvements."

Quality Progress, 26(11).
36. Snee, R.D. (1995, January). "Listening to the Voice of the Employee." Quality
Progress, 28(1).
37. Thomas, K. (1975). The Handbook of Industrial and Organizational Psychology,

edited by Marvin Dunnette, Chicago: Rand McNally.
38. Tjosvold, D. (1984). "Making Conflict Productive." Personnel Administrator, 29.
39. Tuckman, B. W. (1965). "Development Sequence in Small Groups," Psychological

Bulletin.
40. Whiteley, R.C. (1991). The Customer Driven Company - Moving from Talk to
Action. Reading: Addison-Wesley.
41. Wortman, B.L., Carlson, D.R., & Richardson, W.R. (2006). CQE Primer. Terre
Haute, IN: Quality Council of Indiana.
VILLANOVA UNIVERSITY IV - 48 LE,"N SIX SIGMA HANDBOOK

QUESTIONS
4.1 . Which of the following places should NOT be 4.6. Identify the communication technique that has
considered when selecting improvement team the widest universal acceptance.
members?
a. The use of pictures and charts
a. Where the sources or causes of the b. The use of language translation tools
problem may be found c. The standardization to English as a
b. Among those with special knowledge, business language
information, or skills d. The invention of Spanglish which combines
c. In areas that can be helpful in developing English and Spanish
remedies
d. Among those who reach consensus easily 4.7. When selecting potential team members, it is
important that candidates:
4.2 The normal order of the stages of a team's
development are often referred to as? a. Are from the same work group
b. Are at the same level within the
a. Forming, norming, storming, performing organization
b. Develop, build, optimize c. Can discontinue their normal duties while
c. Forming, storming, norming, and participating on the team
performing d. Have a reason to work together
d. Forming, building, optimizing, and
performing 4.8. Prior to participating on a team, each team
member should:
4.3. A skilled team facilitator will:
a. Be trained in problem solving techniques
a. Dominate group discussions b. Have undertaken a study of group dynamics
b. Correct the group when their ideas are c. Have participated in a team motivation
wrong seminar
c. Take sides when one side is correct d. Have a knowledge of the team's intended
d. Make sure all opinions are heard purpose
4.4. Which two of the following teams are likely to 4.9. Divided member loyalties, having personal
disband upon completion of a specific job or confrontations, thinking individually, and
application? learning team members roles are indicators that
a team is in which stage of its evolution?
I. Self-directed
II. Project a. Forming
III. Cross-functional b. Storming
IV. Improvement c. Performing
d. Norming
a. I and II only
b. I and III only 4.10. Which lean six sigma role is most likely to
c. II and III only define objectives for an improvement team?
d. III and IV only
a. Leader
4.5. The proper sizing of a team is: b. Sponsor
c. Facilitator
a. Something that senior management should d. Member
consistently determine
b. Best based upon the inclusion of
representatives from all affected areas
c. A straightforward management decision
d. A complex balancing of considerations of
inclusion and manageability

QUESTIONS
4.11 . It is noted that the involvement and participation 4.16. A team facilitator or leader can use all of the
of which of the following team roles expands following tactics to combat digression and
and increases over time? tangents within a worl< team, EXCEPT?
a. Team members c. Sponsor a. Using a written aglenda

b. Facilitator d. Leader b. Directing the conversation back on track
c. Engaging in a team groupthink activity
4.12. One of the human dynamics that occurs within d. Reinforcing team ,operating guidelines
work teams is conflict. The advantages of
having some internal team conflict include all of 4.17. When developing the potential for win-win
the following, EXCEPT: agreements, wha1 strategy sequence is most
appropriate?
a. It increases a team's energy level
b. It minimizes the confidentiality issue I. Verify areas of agl'eement
c. It offers diverse perspective for problem II. Resolve any differences
solutions III. Confirm each party's goals
d. It provides the opportunity for greater IV. Consider alternate solutions
creativity
a. III, IV, I, II
4.13. When a team finds itself divided into two b. I, III, IV, II
groups, and they have reached an impasse on a c. I, III, II, IV
solution, then: d. III, I, IV, II
a. The team leader should disband the team 4.18. Natural work teams share in common:
b. The facilitator should guide the team
through the process a. Their attitudes and experiences
c. The sergeant at arms should ask one group b. Their motivation and training
to excuse themselves c. Their involvement in the problem to be
d. The team should split and have each group addressed
work on solutions d. Their concern about the large issues of the
organization
4.14. With regard to conditions on the work floor, in
what order should upper management consider 4.19. A team commissioned to design and install an
the value of hislher information? automated part degreaser is which type of
team?
I. Upward flow (from below)
II. Downward flow (from the top) a. Self-directed
III. Horizontal flow b. Autonomous
IV. Informal networks c. Improvement
d. Project
a. I, IV, III, II
b. II, I, III, IV 4.20. Cross-functional membership would NOT
c. III, I, IV, II typically be used for which of the following?
d. I, II, III, IV
a. Self-directed teams
4.15. Valid reasons for removing a team member from b. Six sigma teams
a team include: c. Ad hoc improvement teams
d. Project improvement teams
I. Personality conflicts
II. Lack of skills by the member
III. Poor team meeting attendance
IV. Member needed for other activities
a. II and IV only c. II and III only

b. I, III, and IV only d. I, II, III, and IV

QUESTIONS
4.21. Which of the following would be considered the 4.26. The main reason to analyze behavior patterns
three most important team roles? among improvement team members is to:
a. Member, supervisor, manager a. Ensure a healthy number of team workers

b. Leader, member, timekeeper b. Ensure a good balance of behavior types
c. Facilitator, consultant, member c. Ensure adequate specialist skills
d. Facilitator, member, recorder d. Minimize the number of shaper members
4.22. Which of the following statements is a true 4.27. The cross-functional team approach to quality
description of a team during the storming improvement is effective because:
stage?
a. Nobody is ultimately responsible for the
a. The team wants the project to move forward results
b. The team works cohesively b. "Finger pointing" has an opportunity to get
c. The team members are very cooperative sorted out once and for all
d. The team operates smoothly c. The diversity of team members brings a
complete working knowledge of the
4.23. The role of the team recorder is to: processes
d. Problems are solved on the spot
a. Assign responsibilities
b. Take clear notes of the meetings 4.28. The most important resource that upper
c. Select team members management can provide a natural work team
d. Restrain disruptive team members is:
4.24. To discourage a team from selecting an a. Money

expensive problem solution, which is not cost b. Consultation
effective, team members should be trained to c. Time
ask the question: d. Outside training
A. Why not do it this way, since it is only a 4.29. A successful team effort should produce all of
"drop in the bucket?" the following benefits, EXCEPT:
B. If this were my money, would I spend it on
this solution? a. Improved worker morale
C. Is this what a "world-class" company would b. A decreased need for management efforts
do? c. Improved communications between
D. Since the organization has considerable managers and team members
resources, can we solve it quickly? d.' Cost savings from participative problem
solving
4.25. The best way to disseminate information about
a new quality program is to: 4.30. Identify the INCORRECT statement from the
choices below:
a. Send an e-mail to all employees
b. Post the announcement on the company a. The team leader is always the supervisor
bulletin board b. The recorder maintains the team's minutes
c. Send a memo to all department heads and agendas
d. Use a combination of media c. The facilitator should have training in
improvement techniques
d. Each team member is responsible for
completing assignments between meetings

QUESTIONS
4.31. The greatest value a team can provide to an 4.36. A team sponsor or chslmpion typically:
organization is:
a. Is a liaison betwE!en the team and upper
a. Implementing solutions management
b. Having well planned meetings b. Attends all team meetings
c. Identifying organization wide problems c. Directs the team on implementing solutions
d. Meeting at least once a week d. Fulfills the facilitator role
4.32. Techniques useful for team facilitators, when 4.37. The tendency for highly cohesive groups to lose
narrowing a list of potential problem areas to their critical evaluative capabilities is what:
investigate, include all of the following,
EXCEPT: a. Irving Janis called "groupthink"
b. Frederick Taylor ISlbeled as "cognition"
a. Brainstorming c. Abraham Maslow c:ited as "human needs"
b. Nominal group technique d. B.F. Skinner definl!d as "the Skinner box"
c. Voting
d. Multivoting 4.38. When a problem arises in the performance of a
team member and it has a negative impact on
4.33. The classical number of development stages for the performance of the team, which of the
team growth is: following actions would be considered
inappropriate?
a. 3
b. 4 a. The team leader !:;ummarily removes the
c. 5 individual from thE! team
d. 7 b. The leader or members advise the
individual of what is expected
4.34. Facilitators do NOT normally: c. If action to remOVE! the individual from the
team is warranted, do so respectfully
a. Act as group leaders d. The team leader should clarify expectations
b. Summarize group ideas with the individual
c. Provide feedback to the group
d. Know problem solving techniques 4.39. At which evolutionary stage would a team
typically select a leader?
4.35. You are working through a conflict situation
with a customer. You try your best to reduce a. Performing
the tension and to have the customer take his b. Storming
time on the problem. You are practicing which c. Forming
conflict resolution model? d. Norming
a. Changing perceptions 4.40. If an improvement team needs a member with

b. Avoiding conflict innovative ideas, then the necessary Belbin
c. Smoothing team role model would be a:
d. Forcing
a. Plant
b. Teamworker
c. Specialist
d. Coordinator
VILLANOVA UNIVERSITY IV -52 LEJ~N SIX SIGMA HANDBOOK

I
I
v. DEFINING OPPORTUNITIES
THE STARTING POINT FOa IVI ••OVEMENT IS TO

RECOGNIZE THE NEED.
IMAI
VILLANOVA UNIVERSITY V -1 LEAN SIX SIGMA HANDBOOK

V. DEFINING OPPORTUNITIES
PROJECT CHARTER
Lean Six Sigma Methodology - Define
Defining Opportunities is presented in the following topic are1as:
• Project charter (boundaries) • Customer inputs

• Metrics selection • Lean thinking
• A3 report • Cycle time reduction
• Definition tools • Value stream mapping
• Affinity diagrams • Process mappin~J
• Cause-and-effect diagrams • Process mapping vs. VSM
• Pareto diagrams • Spaghetti diagrams
The initial step of the lean six sigma problem solving methodolog~' is the define step.
Properly defining the problem is the most important part of solving the problem.
Many of the above tools and techniques are explained in other portions of this
Handbook because they are also very applicable to other lean six s;igma areas. Many
of the tools in this Section are also beneficial in other problem s,olving areas.
Project Charter
A critical element in the establishment of an improvement team is the development
and acceptance of a charter. A charter is a written document that defines the team's
mission, scope of operation, objectives, time frames, and consequences. Charters
can be developed by top management and presented to teams, or' teams can create
their own charters and present them to top management. Either way, top
management's endorsement of a team's charter is a critical factor in giving the team
the direction and support it needs to succeed.
Teams need to know what top management expects of them. The team must have
the authority, permission, and blessing from the necessary levels of management
to operate, conduct research, consider and implement any changes needed to
achieve the expected results.
The charter begins with a purpose statement. This is a one or two line statement
explaining why the team is being formed. The purpose statement should align with,
and support, the organization's vision and mission statements. The charter should
also identify the objectives the team is expected to achieve.
VILLANOVA UNIVERSITY V-2 LE,lI.N SIX SIGMA HANDBOOK

PROJECT CHARTER
Project Charter (Continued)

Objectives should always be stated in measurable terms. The charter should also
define the operating scope. This is an opportunity to identify the organizational or
operational boundaries within which the team is expected and permitted to operate.
Defining boundaries is crucial to avoid energy draining and time delaying turf wars.
The project charter is very useful for several reasons:
• The team is aware of the project goals and boundaries

• The team will remain focused on the original goals
• The team will work on projects that align with the organization's goals
• The team champion will support the team and their goals
A good charter should contain a section describing top management's support and
commitment. This is important because some team members may feel that they are
taking a personal risk by becoming a member of a high profile team. A charter
provides the following advantages:
• Eliminates any confusion

• Defines the subject boundaries
• Identifies areas which should not be addressed
• Identifies the deliverable product
• Provides a basis for team goal setting
• Authorizes the team to collect relevant data
• Provides access to necessary resources
• Approves time for team members to address problems
Moen (1991)17 suggests that a team project charter should contain the following key
points:
• Business case (financial impact)

• Problem statement
• Project scope (boundaries)
• Goal statement
• Role of team members
• Milestones and deliverables (end products of the project)
• Resources required
Identifying the above details, in written form, will provide a constant and consistent
target for the team. In this Handbook, the authors will address the business case
and problem statement as part of the project charter and the remaining subjects as
part of the project scope.
VILLANOVA UNIVERSITY V-3 LEAN SIX SIGMA HANDBOOK

PROJECT CHARTER
Business Case
The business case is a short summary of the strategic reasons fClr the project. The
general rationale for a business case would involve subjects like waste elimination,
quality, cost, or delivery of a product with a financial justification. Moen (1991)17
suggests that there are four basic activities:
• Design of a new product

• Redesign of an existing product
• Design of a new process
• Redesign of an existing process
Eckes (2001)5 reports that a common problem for many projects is the lack of a
company impact measurement. For example, if the existing quamy defective rate is
at 5,000 defectives per million opportunities; the possible justification is a reduction
to 250 defectives per million opportunities with a cost savings of $1,000,000. A
project improvement team should follow typical financial department justification
guidelines. Projects which do not show a significant financial impact to the
company should be stopped or eliminated as soon as possible.
(Rath & Strong, 2000)25, (Eckes, 200'1t, (Moen, 1991)17
Problem Statement
A problem statement will detail the issue that the team wants tC) improve. Eckes
(2001)5 explains the problem statement should be crafted to be as descriptive as
possible. That is, how long has the problem existed, what mleasurable item is
affected, what is the business impact, and what is the performance gap. The
problem statement should be neutral, to avoid jumping to conclusions. A sample
problem statement might be: "The ABC Company, in 2006, has experienced a 25%
drop in sales, with a 40% drop in net profits." The problem statemEmt should include
a reference to a baseline measurement. A baseline measurement is the level of
performance of the particular metric at the initial start of a project. (Pande, 2000)19
The collection of good data and process performance measurement will provide a
picture of the areas in the company that have the greatest need for improvement.
In addition, the measurement system will provide a foundation for c)ther teams to use
to pursue other projects. If baseline measures differ from the a!isumptions of the
team or the company, more clarification may be necessary. The measures may not
be wrong, but more data may be necessary to understand the situation (Pande,
2000)19. The problem statement should detail the amount of anticipated
improvement and include a firm completion date.
VILLANOVA UNIVERSITY V-4 LE.lI,N SIX SIGMA HANDBOOK

PROJECT CHARTER
Charter Negotiation
The project charter can be created and presented by upper management. However,
the project team might be closer to the actual facts and might propose a different
mode of attack than envisioned by management. Hence, charter negotiation may be
required.
Consider the following examples:
1. Objectives: The final customer product or internal process may require a

substantial redesign that was not envisioned.
2. Scope: The boundaries of the project could require expansion. The project
may be sufficiently large and require dividing into more manageable pieces.
This could require two or three projects in succession by a single team or
additional project teams working on a portion of the original project.
3. Boundaries: The project team may discover that additional areas

(engineering, maintenance, finishing, etc.) should be included in the solution.
4. Resources: Excessive project resources are rarely provided. Generally, an

oversight of some key resource component is encountered. These needed
resources can be internal or external to a company. Management may be
called upon to prioritize certain fixed resources outside of the team's control.
5. Project transition: The transition of a project to normal company controls may

necessitate either a time extension or additional monitoring on the part of
another group. This mayor may not require lengthier team involvement.
6. Project closure: The improvement team could discover that related processes
or products need the same type of effort as that which was undertaken for the
initial project. In some cases, the project closure date might be moved up
because of such diverse events as unexpected success or shifts in customer
preference.
Obviously, both the project team and upper management are interested in success,
not failure. There should be a willingness on the part of both parties to negotiate a
number of project details. Generally, it is best to handle charter negotiations at the
start of a project. However, all pertinent information may not be apparent at this
point. Charter negotiations can be required at any point during the project.

METRICS SELECTION
Metrics Selection
The primary and more detailed metrics are developed, but are [not finalized in the
definition step. That is left to the measure step. The primary metrics for
consideration in a project could come from several sources:
• Suppliers
• Internal processes
• Customers
The metrics that will affect projects involving suppliers, internlal processes, and
customers would be: quality, cycle time, cost, value, and labor. (Eckes, 2001)5.
Garvin (1988t and Besterfield (1998)2 suggest nine dimensions of quality
measurement:
• Performance: Primary features of the product

• Features: Secondary features added to the product
• Conformance: Obtaining a product that meets fit, form, alnd function
• Reliability: The dynamic quality of a product over time
• Durability: Useful life
• Service: Ease of repair
• Response: Human interface
• Aesthetics: Product appearance
• Reputation: Based on past performance
Hill (1993)10 presents similar measurements that are important h, the marketplace.
Hill framed his characteristics to answer the question: "How do products win orders
in the market place?" His suggested measurements are:
• Price • Product range

• Conformance quality • Design
• Delivery speed • Brand name
• Reliability of delivery • Technical support
• Demand increases • After sales support
• Color range (Hill, 1993)10
The secondary or consequential metrics would be derived from the primary metrics.
For example, if cycle time were determined to be a key metric, the next step would
be to establish the numerical measurement. Examples of measurements include:
• Defects per unit (DPU)

• Defects per million opportunities (DPMO)
• Average age of receivables
• Lines of error free software code
• Reduction in scrap
VILLANOVA UNIVERSITY V-6 LEj!~N SIX SIGMA HANDBOOK

A3 REPORT
A3 Report
The A3 report is a concise summary of a project described in five or six parts on one
piece of paper. The name A3 relates to the metric paper size used for the report,
approximately equivalent to 11" x 17" paper. This can be the final report presented
to management for approval of a project. The process removes the fluff and flashy
showmanship from the presentation. It may only take 3 to 5 minutes for the speaker
to present the business case. A project is thus judged on its merits.
Thirty minutes of PowerPoint slides can be enough to put an audience to sleep. The
A3 report is lean production (at its finest) in report presentations. Toyota makes
extensive use of the A3 style. Variations of the one page project reports are also
used at Komatsu and Cummins, Inc.
There are four basic types of story lines used in the A3 report. Each type is similar
in structure, but the purposes are different. Extensive information is provided in
graphic form. The basic types are:
• Proposal stories: a business case for a new venture

• Problem solving stories: an analysis and implementation report
• Status stories: a project status report
• Information stories: a transmittal of information
The report format fits the problem solving stages onto a 11" x 17" sheet. For
example, the problem definition and analysis stages would appear on the left side
of the paper, while the action plans, results of activities, and future steps will appear
on the right side. Some key concepts for A3 report writing:
• Use concise wording

• Use charts to illustrate points
• Restrict statements to 3 or 4 bullets per section
• Provide substance, not fluff
(Liker, 2006)15, (May, 2007)16
A hypothetical A3 report is presented on the following pages, but there certainly is

flexibility in the format. The concept is to present a summary of the project in a
concise, clear, and understandable way.
Please note that the following A3 example is presented on two 8-1/2" x 11" sheets.
Each sheet represents approximately one-half of an A3 report.

A3 REPORT
"
A3 Report (Continued)
Improve Productivity of Air Filter Line
1. Define the Problem: Volume

120
goal
100
1. Units are below monthly goals 80
2. Revenues are below goal U)
:!::60
3. The condition is not getting better !::
=>
4. Rework is increasing 40
20
0 I
J F M A M J J
Revenue Months
150
Rework
goal 80
'"~ 100
60
,?!.
...
II)
40
..!!! 50 -
(5 II)
0 :!:
s::::: 20
:::l
0 0
J F M A M J J J J
J F M A
Months
Months
2. Problem Analysis: Loss Categories

Seven Month Period
Problem Causes 80
1. Material variation = different suppliers ;;- 60

2. Machine knife blades get dull o
.....
3. Adhesive problems
-.0
><
t J)
40
20
Root Causes ...J
1. Lack of supplier control o

2. Lack of data on knife blade usage Machine Trimming
Material Variation Adhesive
3. Lack of data on material usage
4. Environmental adhesive conditions
VILLANOVA UNIVERSITY LEAN SIX SIGMA HANDBOOK

A3 REPORT
3. Action Plans:
Action Item Responsibility Due 10 20 30

Install SOA for suppliers SOA Mgr 30
•
20
--
Design DOE for materials Sr.O.E.
Collect data on knife blade life P.M. Engineer 10

Investigate adhesive conditions Sr.O.E. 10
Check in-house environment Sr.O.E. 20
Volume
4. Results of Actions: 120
goal
100
1. SQA program installed 80

2. Suppliers are more consistent III
~ 60
3. DOE performed :J
40
4. Knife blades changed more frequently
5. New adhesive used 20 -
6. Environmental controls installed 0

N 0 J F M A M
Months
Revenue
150 Rework
80
'0
T'"
100 60
~
...
If)
40
~
0
50 If)
;::
C c 20
::J
0
0 I I I
N 0 J F M A M N D J F M A M
Months Months
5. Unresolved issues and future steps:
1. Perform evolutionary EVOP

2. Look for even better blades
3. Need more robust adhesives properties
4. Maintain SQA program

A3 REPORT
The A3 report has also been used as a problem solving methodollogy. It is a format
replacement for other options such as PDCA, PDSA, and DMAIIC. The followin!~
steps are used:
1. Background: What is the context and importance of the probllem?
2. Current Condition: Diagram the current process or situation. What is the

measured extent of the problem? Identify problem areas with storm bursts.
3. Cause Analysis: List the problems. What is the most likely r04lt cause? Use the
5 whys technique.
4. Target Condition: Diagram the proposed new process. What are the measurable
target values? Countermeasures are noted as fluffy clouds.
5. Implementation Plan: This is the proposed who, what, whEm, and where for
actions, and the people, times, and locations to correct the pl'oblem.
6. Follow-up Plan: Define how and when the outcomes will be checked. Are the
actual results as predicted? Sobek (2007)24
To:
0 THEME: What are we trying to do? By:

!Jate:
Background I
• Background of the problem
t Target Condition I
• Diagram of proposed new proc:ess
• Importance of the problem ~ • Countermeasures noted as fluffy clouds
• Measurable targets (quatity , time)
Current Condition I 'f Implementation Plan I H
0 •
•
•
Diagram of current situation or process
Highlight problem(s) with storm burst
What about the system is not ideal
What?
Actions
Who?
Person
When?
Times,
Where?
Location
• Extend of the problem(s) taken Dates
Cost:
, (
Cause Analysis
• List problem(s)
I
Follow-Up Plan I
,,.
• Most likely root causes ~ Plan Actual
• Use 5 whys
• How are effects checked? • C(Jmparison to predicted
• When are they checked? • Date check done

DEFINITION TOOLS/AFFINITY DIAGRAMS
Problem Definition Tools

Problem Definition Tools is presented in the following topic areas:
• Affinity diagrams
• Cause-and-effect diagrams
• Pareto diagrams
Affinity Diagrams*
Affinity diagrams are good tools in organizing complex information into logical
categories. Because of its format, the affinity diagram is helpful in the problem
definition stage of lean six sigma.
Unfamiliar problems may present difficulties to the improvement team. Affinity

diagrams are especially beneficial for new or complicated problems. The affinity
diagram appears similar to the mind mapping technique, in that generated ideas are
linked to other ideas to form thought patterns.
The affinity diagram uses an organized method to gather facts and ideas to form
categories of thought. The normal implementation steps are explained below:
• Define the problem under consideration
• Use 3" x 5" cards or Post-it® notes
• Enter ideas, data, facts, opinions, etc. on the cards or notes
• Place the cards or notes on a table or on a wall
• Arrange the groups into similar thought patterns or categories
• Develop a main category or idea for each group and that becomes the affinity
card
• Once all of the cards or notes have been placed under a proper affinity card,
borders can be drawn around the affinity groups for clarity
The affinity diagram can also be referred to as the KJ Method. It was developed by
Dr. Kawakita Jiro, founder of the Kawayoshida Research Center.
* This material is adapted from Gee (2005)8.
VILLANOVA UNIVERSITY V ·11 LEAN SIX SIGMA HANDBOOK

DEFINITION TOOLS/AFFINITY DIAGRAMS
Example Affinity Diagram
HOW TO PREPARE FOR A LSS EXAIM
IRESOURCESI
rI, OBTAIN KNOWLEDGE J-
IGET LSS HANDBOOKI I WATCH VIDEO PRESENTATION I
ENROLL IN
I TAKE OTHER LSS COUFlSES I
VILLANOVA ON-LINE
ISTUDY RELATED MATERIALS I

I GET OTHER
I LSS TEXTBOOKS ISTUDY IN GROUPSIIHAVE A TUTORI
TAKE UNIVERSITY LEVEL.

COURSES IN LEAN SIX SIGMA
i PREPARATION 1
TEACH LSS
IHAVE A Q & A SOURCEI
SUBJECTS
USE PRACTICAL IMOTIVATION:

EXPERIENCE
ISTUDY INTENSIVELY I IMOTIVATE SELFI IGET EIONUSI
ISTART EARLY I LISTEN TO SUCCESSIFUL

LSS STUDENTS
STUDY 1 SUBJECT BE AROUND OTHERS

AT A TIME WHO ARE POSITIVE
IPRIDEI
ISTUDY LSS TESTS I
MAKE YOUR OWN IPUMP YOURSELF UP I
LSS EXAMS
Figure 5.1 An Affinity Diagram Example
VILLANOVA UNIVERSITY V -12 LEJ\N SIX SIGMA HANDBOOK

DEFINITION TOOLS/C & E DIAGRAMS
Cause-and-Effect (Fish bone) Diagrams

Key process variables should be defined and measured using metrics like percent
defective, operation costs, elapsed time, backlog quantity, and documentation
errors. Process workers and owners are the best sources for identification of the
critical variables. Once identified, the relationships between the variables can be
depicted using a tool called a cause-and-effect diagram.
A cause-and-effect diagram:
• Breaks problems down into bite-size pieces

• Displays many possible causes in a graphic manner
• Is also called a fishbone, 4-M, or Ishikawa diagram
• Shows how various causes interact
• Follows brainstorming rules when generating ideas
A fish bone session is divided into three parts: brainstorming, prioritizing, and
development of an action plan. The problem statement is identified and potential
contributing factors are brainstormed into categories. To prioritize problem causes,
polling is often used. The three most probable causes may be circled for the
development of an action plan.
Generally, the 4-M (manpower, material, method, machine) version of the fishbone
diagram will suffice. Occasionally, an expanded version must be used. In a
laboratory environment, measurement is a key issue. When discussing the brown
grass in the lawn, environment is important. A 5-M and E schematic is shown in
Figure 5.2.
Problem
Statement
Figure 5.2 Basic Fishbone 5 - M and E Example

DEFINITION TOOLS/C & E DIAGRAMS
Cause-and-Effect (Fish bone) Example
MATERIAL MACHINE MAN
VARIATION IN INSUFFICIENT TRAINING

'rOLE RANCE
1. WORN NUMBERS KEYPUNCH ERRORS
1. PLATING ON SCALE KEYS
OVER ISSUE UPDATES
2. MATERIAL 2. CONTAINERS NOT MADE
THICKNESS BROKEN
PULLED WRONG PARTS
3.SCRAPAND FROM LOCATION
FOREIGN ELEMENTS
4. LENGTHS
REDUCE INCOMING RECEIPT

ERRORS FROM
4% TO 1% OF
TRANSACTIONS
SUSPECT PAN
TARE WEIGHTS # - - - AIRFLOW
VENDOR COllNTS ACCEPTED
TARE WEIGHTS 1 - - - - DEBRIS
NOT ON PANS NON-STANDAFID SAMPLING
PROCEDURE (iNADEQUATE
SAMPLE QUA~ITITY)
SCALE CALIBRATION
WRONG PART NUMBERS
THREE DIFFERENT FROM DEPARTMENTS
SCALES
SCALE# 2 MORE
ACCURATE THAN
SCALE# 1
MEASUREMENT ENVIRONMENT METHOD
Figure 5.3 Actual Fishbone Example
For additional examples of cause-and-effect or Ishikawa diagrams refer to: Ishikawa,

K., Guide to Quality Control, (1982)12. Dr. Ishikawa attributed the ·first application of
a cause-and-effect diagram to Tomiko Hashimoto's article, "Elimination of Volume
Rotation Defects Through QC Circle Activities," Factory Work and QC. No. 33.
{Hashimoto)9.

DEFINITION TOOLS/PARETO DIAGRAMS
Pareto Diagrams
Pareto diagrams are very specialized forms of column graphs. They are used to
prioritize problems (or opportunities) so that the major problems (or opportunities)
can be identified. Pareto diagrams can help lean six sigma teams get a clear picture
of where the greatest contribution can be made.
History
There is a very interesting story behind the name of Pareto diagrams. The word
"Pareto" comes from Vilfredo Pareto (1848-1923). He was born in Paris after his
family had fled from Genoa, Italy, in search of more political freedom. Pareto, an
economist, made extensive studies about the unequal distribution of wealth and
formulated mathematical models to quantify this maldistribution. He also wrote a
political book on nationalism which helped lead to Fascism in Italy.
Dr. Joe M. Juran, world renowned leader in the quality

field, was preparing the Quality Control Handbook13 in
the late 1940's. He needed a short name to apply to the
phenomenon of the "vital few" and the "trivial many." He
depicted some cumulative curves in this manuscript and
put a caption under them, "Pareto's principle of unequal
distribution ... " The text makes it clear that Pareto only
applied this principle in his studies of income and
wealth; Dr. Juran applied this principle as a "universal."
Thus, the diagram could be named a "Juran diagram."
To complicate matters more, the cumulative curve
diagram itself was first used by M. O. Lorenz in 1904.
Briefly stated, the principle suggests (in most situations) Vilfredo Pareto
that a few problem categories (approximately 20%) will 1848-1923
present the most opportunity for improvement (approximately 80%).
Pareto diagrams are used to:
• Analyze a problem from a new perspective

• Focus attention on problems in priority order
• Compare data changes during different time periods
• Provide a basis for the construction of a cumulative line

Typical Pareto Diagram

The defects for a book product are shown in Pareto form below:
100
300
I/J
E
Q)
::s0 75
...
Q.
en
c
'"0
c
200 Cumulative Line
-
cQ)
50 Q)~
-
OJ
...0
Q)
.a
100
25
Q.
E
::l
Z
o
ABC D E F G H J K L M N
Problem Categories
A. Emulsion - glue .....•.••.... 67 H. Square variation ............ 17

B. Grease/oil/dirt ..........•••• 59 I. Head bands ........•........ 6
C. Hot melt - glue ......•....... 30 J. Case damage - unknown ....•. 5
D. Sewing thread ......•....... 29 K. Case damage - area II ••.•.•... 4
E. Gilding defects .......••..•. 28 L. Upside down books •...•..... 2
F. End sheet problems .•....... 25 M. Torn pages .•...•............ 0
G. Case damage - area I • . . . . . . . . 17 N. All others .................. 23
Figure 5.4 Typical Pareto Diagram
"First things first" is the thought behind the Pareto diagram. Attention is focused
on problems in priority order. The simple process of arranging data may suggest
something of importance that would otherwise have gone unnoticed. Selecting
classifications, tabulating data, ordering data, and utilizing the Pareto diagram, have
proved to be useful in problem investigation.
Note that the "all others" category is placed last. Cumulative lines are convenient
for answering such questions as, "What defect classes constitute 70% of all
defects?"

Weighted Pareto Analysis

Pareto analysis is very helpful in assisting improvement teams with the selection of
serious problems. The Pareto method assumes that there will be segregation of the
significant few from the trivial many. In many cases, the Pareto diagram is
constructed based upon the number of event occurrences. However, criticality (or
potential safety or economic loss) factors might result in a different Pareto
alignment. Consider the following audit report results: (Anderson, 2001)1
Deficiency Description Occurrences
A Critical test equipment overdue for calibration 1

B Inadequate training of employees 1
C Marginal internal auditing 1
D Marginal procedures 12
E Obsolete drawings in use in manufacturing 3
F Problems with record retention 8
G Unauthorized changes on controlled documents 4
H Questionable criteria for inspection judgments 2
I Inadequate corrective action 1
J Unidentified material in stockroom 6
K Employees not following procedures 2
L Nonconforming material mixed with good 2
The traditional Pareto diagram, based on occurrences, would look like so:
12 12
9 -
8
r---
- 6
r----
~
,--L
-
2 2 2 (All Others)
.... I
o i
D F J G E H K L
Figure 5.5 Audit Occurrence Pareto Diagram

Weighted Pareto Analysis (Continued)

Assume that an objective assessment of the criticality of the same deficiencies
yielded the following weighting:
Critical (100 demerits each) A, G, L Minor (10 demerits each) B, 0, I

Major (25 demerits each) C, E, K Incidental (1 demerit each) F, H, J
The accumulated, weighted audit data now becomes:
A 1 X 100 = 100 G 4 X 100 = 400

B 1X10=10 H 2X1=2
C 1 X 25 = 25 I 1X10=10
0 12X10=120 J 6X1=6
E 3 X 25 = 75 K 2X25=50
F 8X1=8 L 2 X 100 = 200
The composite Pareto diagram suggests a different priority than the previous one:
400 400
B~ 300
s::.-
ns(1)
t::3:
&.>< 200
E en 200
_(1)
-
(1)(.)
> s::
._ (1)
ns ~~ 120
100
Q; ~ 100
0::(.) I 75
61
o 50 (All Others)
J • I
o G L o A E K
Figure 5.6 Criticality Weighted Pareto Diagram
Dollars can also be used instead of demerits. In addition, the original data could be
sorted by shift, the day of the week, etc.

CUSTOMER INPUTS
Voice of the Customer (VOC)

An understanding of the needs of the customer is critical to the survival of most
companies. The pace of change in today's economy mandates that a company not
rely on their past knowledge of the customer. A detailed plan to gather and collect
customer needs and customer perceptions can be described as listening to the voice
of the customer (VOC). This enables the organization to:
• Make decisions on products and services

• Identify product features and specifications
• Focus on improvement plans
• Develop baseline metrics on customer satisfaction
• Identify customer satisfaction drivers
An interesting concept to consider in this technique is to determine the reactive

approaches being used and then replace them with proactive approaches. Rath &
Strong (2000)25 suggests the following process for collecting VOC data:
1. Identify customers and their needs
2. Collect and analyze reactive data (complaints, service calls) and then consider
proactive approaches (interviews, surveys)
3. Convert collected data into customer needs
4. Sort out the most important attributes (the critical characteristics)
5. Obtain specifications from the critical to quality characteristics
Not only can VOC input be useful in product and process design, it can also be
critical in lean six sigma team project selection and measurement.
Critical to Quality (CTQ) Tree

One technique that is useful in the lean six sigma problem definition stage is the
construction of a critical to quality tree. This tool focuses on the key metrics of
customer satisfaction. A CTQ tree will translate the initial customer requirements
to numerical or quantified requirements for the product or service. These are the
detailed critical requirements that the organization should satisfy. These can be
regarded as key results, or the "Ys" of the process. (Eckes, 2001)5. The
development of a CTQ tree would go from the general requirement to the specific,
or from "hard to measure" to "easy to measure."

CUSTOMER INPUTS
Critical to Quality (CTQ) Tree (Continued)

Generally, it will require two or three levels to migrate from need to drivers to CTas.
See Figure 5.7 below:
1 General -------------------IIJj. . 1 Specific
INeed -----..._.1 Drivers 1 CTas
Figure 5.7 The CTa Tree Concept
The creation of the CTa tree involves the following steps:
1. Identify the customer. Example: A customer at a fast food restaurant.
2. Identify the customer's need. Example: A hungry customer orders a meal.
3. Identify the first set of basic requirements of the customer. Example: The
measures could be promptness of delivery, price, and gocld taste.
4. Progress further with more levels as needed. Determine i1 additional details

are needed for breakdown of the measures.
5. Validate the requirements with the customer. The CTIJ tree should be
reviewed with the customer to ensure that key requirements are known.
A sample CTa is shown below:
I ~~_~_IN_~__~--~.~I_S_~ __~
Order Meal ..
~
Price
...... E1conomical
...... Taste ... Good
Figure 5.8 Sample CTa for Meal Performance
Eckes (2001)5 points that out the exact metrics are not determined at this stage. The
measure stage is where the precise measurements will be determined.
VILLANOVA UNIVERSITY v - 20 LEAN SIX SIGMA HANDBOOK

CUSTOMER INPUTS
Customer Data Analysis

A black belt or green belt can analyze customer data in order to determine when and
where customer attitudes are different or are changing. Comparing customer
attitudes over time, or between groupings, can provide insights into market niches
and changes. The results of customer feedback data collection can be analyzed
using a variety of tools:
Statistical tests: A large number of tests and tables can be used to determine, with
identified confidence levels, that customer preferences have shifted. In addition,
most normal statistical tests may be used on many of the numerical survey results
such as the Likert scale (0-5 or 0-10 ranking) surveys described earlier.
Line graphs: Line graphs can graphically show if either discrete or continuous
characteristics of a product or service are changing. In most cases, a visual
assessment can be made as to whether the product or service is getting better,
worse, or staying the same. A discrete chart follows:
Good
Product
or Service
Characteristic I----''''br-:----- Average
Bad
Time ..
Control charts: A variety of variable or attribute charts can also be used to display
customer feedback data. This tool offers an advantage over line charts because the
addition of calculated control limits facilitates the ability to detect special or
assignable causes of variation. An attribute chart is shown below:
# of
failuresl
1000
customers
~-~~~~~~------Average
Time ..

CUSTOMER INPUTS
Customer Data Analysis (Continued)

Matrix diagrams: A variety of matrix diagrams can be used for examination of
customer return defects or complaints. Data from matrix diagrams can be used to
create a Pareto chart or can be used directly for project selectiion, where a large
number of occurrences are noted. See the hypothetical exam ph! below:
Defect Type
A
Customer B C
~ E F G H I J Totals Number of
A 1 1 - 12 2 3 - 6 3 1 29 occlllrrences
B 5 - 1 4 2 1 - 1 1 - 15 per fixed
time
C 2 2 1 2 - 1 4 1 - h 13
D
lA. - - 6 - - - - - ~ 21
E
~ - - 8 2 - - - 4 - 32
F - 1 1 1 - - - - - 1 4
4 - - - - - 1
o
G 4 1 1 11
H
~- 1
~ 2 1 - - 1 - 13
Problem
Totals
~ 3 4 4~ 8 6 5 9 9 15 138 BlreaS
Pareto analysis: Snapshots of customer defects or rejects (Parteto charts) can be

displayed at selected time intervals to answer such questions as: Are reject
categories still of the same magnitude? Are reject categories, still in the same
sequence of magnitude? See the illustration below:
Time Period I Time Period"
50 50
.c 40 .c 40
....c ....c
o 30 o 30
~ :l!:
All .....
]j 20 .l!l 20
(J u
<ll <ll
'(i) 10 "(i) 10
0::: c::
0 0
A B C D E F G 0 A C B E F G
Categories Categories
Other comparative analyses: The comparative Pareto analysis illustrated above is
a powerful tool for analyzing customer data. In the same way, other charts (control
charts, line graphs, histograms, and even matrix diagrams) can be compared from
one time period to another, from one supplier to another, etc., to provide real insight
into the needs of the customer and the changes in the market. Vi!»ual comparisons,
however, are risky. A significance test may be required.

CUSTOMER INPUTS
Kano Model
The Kano model is also referred to as Kano analysis. It is used to analyze customer
requirements. Noriaki Kano is a Japanese engineer and consultant, whose work is
being used by a growing number of Japanese and American companies. The model
is based on 3 categories of customer needs:
1. Dissatisfiers (basic requirements, or "must be's"):
The customer expects these basic requirements as part of the total package.
If the basic requirements are not present, the customer will be unhappy. For
American tourists traveling to a foreign country, say China, the expectations
are that travel facilities there are as good as in the United States. There is
definitely a big let down as travel facilities are quite primitive in many
locations. American tourists can probably expect to see a dramatic
improvement in travel facilities in China as the 2008 Beijing Games approach.
2. Satisfiers (variable requirements, or "more is better"):
When the requirements of the customer are met, the more they are met, the
better. The tourist taking a Caribbean cruise expects a week of entertainment
and food at a reasonable price. The more personal attention the tourist
receives on the cruise, the better it is. In fact, the entire staff and crew on the
ship are out to make your travel experience a great one.
3. Delighters (latent requirements):
These are features or services that go beyond the expectations of the

customer. For the tennis fan who purchases a ticket to the U.S. Open, the
basic requirement would be to see some players in compete, not get
sunburned, and to see some thrilling matches. Imagine meeting these
expectations, and then being able to have lunch or dinner with some players
as part of the experience. This would be a delighter.
Competition in today's environment raises the basic expectations of all customers.

The standards of a happy customer continue to rise. What was once considered a
delighter, may later become a basic satisfier. The organization has to be aware of
customers' changing needs and move to improve their own performance.
Improvement projects can often be selected from among the satisfier and delighter
categories. Most companies, in a competitive environment, would not be around
long enough to tackle a basic requirement issue.
(Pan de, 2000 19), (Rath & Strong, 200025)

LEAN THINKING
Lean Thinking
Womack (1990)26 introduced the term lean production to the Western World in 1990
with the publication of The Machine that Changed the World. The book describes
the basis of lean production (lean manufacturing) as practiced by the best
companies in the world.
After several years, Womack (1996)27 wrote Lean Thinking: Banish Waste and Create
Wealth in Your Corporation. This book describes the concepts of converting a mass
production plant into a lean organization. Womack (1996)27 clffers five guiding
principles for consideration:
• Specify value by product

• Identify the value stream for each product (value chain)
• Make value flow
• Let the customer pull value from the producer
• Pursue perfection
Value
Value is defined by the customer. Quite often, manufacturers cannot provide

appropriate customer value. In many cases, senior business managers in the United
States have lost touch with the customer. They desire customers to purchase
products at reasonable prices in order to keep the company profitable. They attempt
to increase the performance of the firm, while reducing costs.
The German mind set is more product feature and process orien1ted. The technical
people (engineers) are in control of the businesses. Thus, the Germans are very
strong technically. Therefore, features and enhancements are of utmost importance.
However, some of the new, complex enhancements have failed to attract the
customers' interests. Often, the German mind set is that thE! customer is not
sophisticated enough to understand the new features.
The Japanese define value in the context of where value is created. As the
proportion of the product made on the Japanese homeland increases, the greater the
value that is retained at home for their society. Customers, in general, do not define
value based on where it is made. Customers want their needs satisfied quickly. As
the yen strengthened, the previous advantages of Japanese companies using
Japanese suppliers disappeared. This has resulted in a weakening of many
Japanese companies. (Womack, 1996)27
VILLANOVA UNIVERSITY V -24 LEJ~N SIX SIGMA HANDBOOK

LEAN THINKING
Lean Thinking (Continued)

Value (Continued)
In the long run, value must be defined by the customer. The customer wants
specific products, with specific capabilities, at specific prices. Specifying value is
the first step in lean thinking. New methods must be developed to communicate
with customers, to get closer to them, and to find out what they want. Once value
is determined, a firm must go back again and again to determine if they really have
the right answers.
The target cost of the product may be determined after defining customer value.
This target cost is more than the "market cost" of the product. The market cost is
typically the manufacturing cost of the product plus the selling expenses and profit.
In lean thinking, the target cost is the mixture of the current selling prices by
competitors and the examination of and the elimination of muda (waste) through
lean methods. This analysis results in a target price which is below the current
selling price. The firm then applies lean thinking to its processes.
(Womack, 1996)27
Value Stream (Value Chain)

The benefits of reducing waste can be magnified many times by concentrating on
the set of activities that link a process together. If the emphasis is on Kaizen in
various individual operations, then islands of improvement exist. Real improvement
may be hard to identify. Improving the effectiveness of those activities that produce
products is only part of the answer. There are three streams or chains that should
be considered:
• Problem solving
• Information management
• Physical transformation
The problem solving stream for a business includes: solve the concept design,
develop the prototype, plan reviews, and determine the mechanism for product
launch. The information management stream consists of customer order taking, raw
material sequencing from suppliers, in-house scheduling, and delivery to the
customer. The physical transformation stream (or product realization in ISO 9001
terms) proceeds from the conversion of raw materials to finished goods.
Value streams can be constructed for each major product (or process) that an
individual organization or plant utilizes. Womack (1996)27 describes applications of
a value stream beyond the boundaries of a typical plant, involving the suppliers, the
organization, and the customers. Efforts must be made to eliminate the muda
(waste) in this value stream.

LEAN THINKING

Value Stream (Continued)
Each firm in the value stream is reluctant to disclose too much about itself to the
other firms, for fear of price concession requests. A voluntalry alliance of the
members of the value stream must be developed in order to elliminate the muda
hidden in the value stream.
The value chain, as described by Porter (1985)20 in his work on competitive

advantage, is at a higher operating level. That is, the generic vallue chain involves
the components: human resources, technology, procurement, iinbound logistics,
operations, outbound logistics, marketing, sales, and service. Of course, the
discussion and analysis centers on methods to differentiate the firm. An
organization gains competitive advantage by performing certain lstrategic activities
better or at a lower cost than its competitors. The value stream (or chain), as used
in lean manufacturing efforts, goes into greater detail. It involves a single product
stream, which is analyzed for reduction of waste, reduction in cycle time, or
improvement in quality.
A value stream map is created to identify all of the activities invol"ed in the product.
This value stream can include the various suppliers, production activities, and the
final customer. The activities are viewed in terms of the following criteria:
• It adds value, as perceived by the customer

• It adds no value, but is required by the process
• It adds no value, and can be eliminated
Conner (2001)3 identifies the steps for documenting the value str'eam mapping as:
1. Product development: Identify customer requirements, mE~thods of delivery,

and typical quantities.
2. Process design: Perform a walk through of the process, rec::ording each step.
Start from the shipping dock and work back through the process to the
receiving dock. Make a note of machine times, cycle times, operators,
changeover times, WIP, available time, scrap rate, machinle reliability, etc.
3. Record current status on 11" x 17" sheet of paper.
4. Planning: Develop a future state map.
Cycle time reduction and value stream mapping are discussed la1ter in this Section.
VILLANOVA UNIVERSITY v - 26 LEJ~N SIX SIGMA HANDBOOK

LEAN THINKING

Value Flow
Traditional mass production is often accomplished by the batch technique. The
objective is to produce many units of a specific part at a given time, in order to
maintain the production efficiency of the machines and the overall efficiency of the
departments. However, optimization of the individual operation unknowingly leads
to sub-optimization of the process as practiced by non-lean companies.
The lean effort requires the conversion of a batch process to a continuous flow
process. In some cases, converting the batch process to a one piece flow is ideal.
Some of the obstacles to overcome include:
• The plant has always done it in batches or lots
• The plant has a multitude of departments and functions
• The plant can't afford quick changeover tooling
• The plant has high momentum, inflexible machinery
• The plant machinery would be expensive to move
Ideally, in a continuous flow layout, the production steps for single piece flow,
without WIP, are arranged in a sequence, straight line, U-shaped, or cellular. Inside
this flow, the work of each station and operator must be performed with reliability.
When the machinery is performing as expected, there are zero breakdowns. This is
the concept of TPM (total productive maintenance). The quality level of each
operation is very high, near perfect, using a variety of defect elimination and
detection techniques.
The activities needed for production should be in a steady, continuous flow, with no
wasted motions, no batches, and no WIP. There should be flexibility to meet the
present needs. The work of people, functions, departments, and firms will require
adjustments to the value stream to make it flow, and to create value for the
customer. (Womack,1996)27

LEAN THINKING

Pull Value
Instead of creating product in response to an estimated sales 1forecast, the plant
manufactures product as the customer requires it. This is the pulll system in action.
This results in many positive things for the organization:
• Cycle times decrease (concept to launch, sales to delive'1" etc.)

• Finished inventories are reduced
• Work-in-process (WIP) is reduced
• The customer stabilizes their ordering
• Pricing is stabilized
{Womack, 1996)27
Most mass production manufacturing firms are in the push production mode. Each
operation produces as much as possible and sends it onto the nl!xt operation. The
goal is to maximize machine efficiency with a maximum amount of in process
inventory sitting around the plant.
Contrast the above manufacturing firm with the factory that is dependent on the pull
of the market. The receipt of a customer order initiates activitel;. Each operation
produces parts as needed through a signal from downstream. 1rhere is a minimal
amount of WIP in the process stream.
This arrangement enables flow through the plant, using the !principles of lean
thinking. Quality, machinery downtime, absenteeism, etc., are all of concern.
Problems in any area will stop the process, disrupting the pull plrocess. Problems
of any sort are magnified and must be immediately corrected. {Imai, 1997)11
Womack (1996)27 presents reduction in cycle times due to lean thi nking methods as
shown in Table 5.9:
Process Percent Reduction

Product development 50%
Order processing 75%
Physical production 90%
Table 5.9 Throughput Reductions

LEAN THINKING

Perfection
The customer is searching for a value-added product. The pursuit of the first four
principles of lean thinking allows the firm to move toward perfection. Solving
customer value problems, working the value stream, converting to flow, and making
pull occur, all help eliminate muda (waste). As the process continues, more muda
is eliminated. Perfection seems possible.
Perfection is accomplished via:
• Product teams working with the customer to find better ways to specify value,
enhance flow, and achieve pull
• Using collaboration between the value stream partners (suppliers,

distributors, customers, employees) to uncover more value
• Using technologies to eliminate muda
• Developing new products (Womack, 1996)27
Lean thinking principles are the cornerstones to higher performance and economic
growth. Perfection is a journey.
Some possible improvement results from lean thinking are:
Improvement Area Reduction/Improvement

Labor productivity 100% increase
Throughput times 90% reduction
Inventories 90% reduction
In-house scrap 50% reduction
Safety injuries 50% reduction
Product development time 50% reduction
Capital investment Modest
Table 5.10 Lean Thinking Results
It may take years to apply lean thinking principles in a company, and even more time
to apply lean thinking in the entire value chain.
(Womack, 1996)27

CYCLE TIME REDUCTION
Cycle Time Reduction

Cycle time is a critical component of the operation of a lean enterprise. Cycle time
is defined as the amount of time needed to complete a single task and to move a
product or activity forward in the process. Cycle times may differ by task, but to
make the line flow, all operations must be completed under a !~iven takt time. A
kaizen event or a line improvement activity can be utilized to analyze the operations
and reduce the cycle times to be below the required takt time. The standard work
is analyzed for value and non-value-added work.
(Gee, 1996f, (Womack, 1996)27
A reduction in cycle time is undertaken for many of the followin!~ reasons:
• To please a customer
• To reduce internal or external wastes
• To increase capacity
• To simplify operations
• To reduce product damage (improve quality)
• To remain competitive
Training
Some of the cycle time training principles and topics are listed below:
• Introduction to the total systems concept

• Problem solving tools such as the "5 Whys"
• Importance of the next process as the customer
• Non-judgmental attitude to problem solving
• Identification of value and non-value-added work
• Identification of muda (the seven wastes)
• Principles of motion study
• Work flow patterns (straight, T-shaped, U-shaped)
• Standard operations
• 5S workplace organization
• Visual management principles
• Just-in-time (JIT) production
• Poka-yoke principles
• Team dynamics (Gee, 1996f
VILLANOVA UNIVERSITY v - 30 LEJIN SIX SIGMA HANDBOOK

Cycle Time Reduction (Continued)

To discuss the concept of cycle time reduction, consider an example of a line with
5 stations and observed cycle times as provided in Table 5.11.
Station Work Time (seconds)

1 45
2 40
3 60
4 70
5 50
Total 265
Table 5.11 Work Time Example
The process required a takt time of 60 seconds, however only station 3 is at 60

seconds. It is good that stations 1, 2, and 5 are each below 60 seconds. Station 4
is above the takt time, so something must be done to reduce the cycle time there.
An improvement team is observant enough to conclude that with a total cycle time
of 265 seconds (or 66.25 seconds per station) a reduction in stations (or operators)
might be achieved. One idea is to have 4 stations at 60 seconds each. The options
are illustrated in Table 5.12.
Station Work Time Option 1 Option 2 Option 2

(seconds) after Kaizen
1 45 53 66.25 60
2 40 53 66.25 60
3 60 53 66.25 60
4 70 53 66.25 60
5
Total
50
265
53
265
>< ~265 240
Table 5.12 Work Time Example Options
VILLANOV A UNIVERSITY v - 31 LEAN SIX SIGMA HANDBOOK


A kaizen team could be empowered to improve the 5 station line. 1rhe team facilitator
would guide the team to a significant cycle time improvement within 5 days using
a kaizen event. There is a general format to the kaizen event:
• There is a time limit of 5 days to accomplish the change (~:;ome use 3 days)
• Two days of training are provided on lean manufacturing techniques
• Two and one-half days are alloted for collecting data and making changes
• The last one-half day consists of a presentation on the results to the
workforce (Gee, 1996f
Team members actively participate in collecting and analyzing dalta. The operators,
technical staff, supervisors, and maintenance staff can all be team members and be
involved in the analysis. The team will perform work sampling, pace studies, line
balancing, elemental analysis, motion studies, and takt time calc:ulations.
Work sampling provides a picture of the work content of the station. This reveals
the content and ratio of work, inspection, walking, and other factc~rs. See Table 5.13
for content ratios.
Station Work Inspection Delay Walking Other

1 75 5 10 10
2 80 5 10 5
3 65 5 5 20 5
4 50 5 5 35 5
5 65 15 5 10 5
Total 335 35 35 80 15
Table 5.13 Work Content Ratio (%)
The line value added activities, at this point, only comprise 50% to 80% of the work
content of each station. The inspection, delay, walking, and other categories are
considered muda. This provides the team with information to ccmsider for station
combinations. The team will investigate ways to eliminate (or reduce) the four muda
elements. It appears possible to reduce from the number of staltions from five to
four.


Further analysis of the work content ratio is displayed in Table 5.14.
Station Original station Work content Actual work time

time (seconds) (%) (seconds)
1 45 75 34
2 40 80 32
3 60 65 39
4 70 50 35
5 50 65 33
Total 265 65.3 173
Table 5.14 Main Work Analysis
As indicated in Table 5.14, the actual work time for the 5 stations amounts to 173
seconds. If there is presently 1 operator per station; given that the takt time is 60
seconds; the 173 seconds divided by 60 seconds suggests that 3 operators will be
sufficient. There may still be a need for 5 stations, but only 3 operators. Additional
data collection will be necessary to confirm this analysis. Pace studies of each
station will provide a clearer picture of the cycle times. Usually up to 25 cycles of
the line are studied in order to determine the average cycle time. A line balancing
chart can be made and compared to the desired takt time. (Gee, 1996f
A study of the stations reveals the motions used by the operators. In this study, an
exacting industrial engineering approach to human motions is not used. An
approximation of the operator effort will suffice. Robinson (1990)22 describes the
Shingo technique of classifying human motions. It is divided into 4 grades:
1. Assemble, disassemble, and use (true value-added)
2. Transport empty, grasp, transport loaded, and release load (non-value-added)
3. Search, find, select, reposition, hold, inspect, and pre-position (non-value-

added, lower on the grading scale than above)
4. Rest, frequent planning, unavoidable delays, avoidable delays (non-valued-

added elements of the lowest rank)


The concept in studying human motions is to reduce the stress Bind strain upon the
operators, creating a more efficient operation. The task activity I)f the operator can
be classified and motion elements eliminated or reduced. Perhapl; some mechanical
device can be used to eliminate a non-valued element.
The team will prepare a workplace layout of the line. This layout will include
operators, WIP inventory, raw materials, and equipment in the workspace. A
charting of the current flow of the product may reveal a "spaghetti-like" flow. Thus,
it can be termed a spaghetti chart. In many cases, rework and questions add many
more lines than shown in Figure 5.15.
. In
-
J
Station 4 Station 1 .....
Operator 4 Operator 1
-
, f , f
"
Station 2 Station 5 - Station 3
operator 2 operator 5 ' fOut operator 3

Figure 5.15 Flow Chart of Existing Production Line
The idea is to arrange the production line using either aU-shaped, L-shaped, C-
shaped, or straight line arrangement, in orderto create continuous flow. The various
lines must reduce the distance traveled by the part, reduce the amount of WIP
inventory between stations, and still meet the required takt time.


Refer to Figure 5.15 in which there is 1 operator per station. The process can be
improved to reduce the distance traveled and to promote a smoother flow of work.
Perhaps a U-shaped line can be constructed as shown in Figure 5.16 .
~ In
.J
Station 2 ...... Station 1
"
Station 3
.A
;
""
~
Ope~ator
•
2 r
Operator 1
Operator 3
V 1
'"
,. Station 4 '"
~ Station 5 -. Out
Figure 5.16 A Proposed U-Shaped Line
Figure 5.16 illustrates a U-shaped line with 3 operators. An analysis of the walking
distance and material flow reveals significantly less wasted motion.
A kaizen event is a very stressful time for the whole team (especially for the
facilitator), since no one at the beginning will know exactly how it will end up. There
are many opportunities for innovation and creativity in the composition of line work
load and layout. The final results are almost always pleasantly surprising in terms
of achieving team goals. The team goals are usually:
• Reducing cycle times

• Meeting takt times
• Reducing space
• Reducing inventory
• Increasing line balance
• Maintaining a safe work environment

VALUE STREAM MAPPING
Value Stream Mapping

A value stream map is created to identify all of the activities involved in product
manufacturing from start to finish. This value stream may include suppliers,
production operations and the end customer. VSMs typically foclJs on material and
information flow. For product development, value stream mapping includes the
design flow from concept to product launch. This is the large view, looking at the
entire system for improvement opportunities.
(Rother, 1999)23
Benefits of a value stream map include:
• Seeing the complete process flow

• Identifying sources of waste in the value stream
• Providing common terminology for process discussions
• Helping make decisions about the flow
• Tying multiple lean concepts and techniques together
• Providing a blueprint for lean ideas
• Showing the linkage between the information and materiall flows
• Describing how the process will change
• Determining effects on various metrics
(Rother, 1999)23
The value stream mapping process is shown in Figure 5.17.
Define product family Use a product equipment matrix
Draw current state map Do this personally
Create future state map Use creative concepts
Implementation planning Can take months or years
Figure 5.17 The Value Stream Mapping Proces!;
VILLANOVA UNIVERSITY V - 36 LEAN SIX SIGMA HANDBOOK

Value Stream Mapping (Continued)

Define Product Family
The recommended value stream approach is to map one product family. A product
family is defined as a group of products that pass through similar processing steps
and over common equipment. A product and equipment matrix can be used to
indicate common features. See Table 5.18 for an example of the matrix.
Equipment
Product Shear Press Weld Assembly
A x x x x
B x x x
C x
D x x x
E x x x
Table 5.18 Example of Product and Equipment Matrix
The matrix shows products that go through a series of common processes. A work
cell could be formed to handle a particular flow. Another method is to create a
Pareto chart of the various products. The product with the highest volume should
be used for the model line. {Rother, 1999)23, {Conner, 2001)3
Value Stream Manager

The value stream for a product family may cross department boundaries in the
company. This creates the potential for difficulties in coordinating an effective value
stream project. Such problems call for creation of the position of value stream
manager. This manager must have the authority to make things happen and should
report to the plant manager.
It is recommended that a production person handle the job of value stream manager.
This manager would monitor all aspects of the project. Being a hands-on person,
the manager should be on the floor on a regular basis. {Rother, 1999)23


Current State Mapping
A current state map of the process is developed to facilitate proce:ss analysis. Basic
tips on drawing a current state map include:
• Start with a quick orientation of process routes

• Personally follow the material and information flows
• Map the process with a backward flow, from shipping docl< to the beginning
• Collect the data personally, do not trust the engineering standard times
• Map the whole stream
• Create a pencil drawing of the value stream
Some of the typical process data includes: cycle time (CT), changleover time (COT),
uptime (UT), number of operators, pack size, working time (minus breaks, in
seconds), WIP, and scrap rate. An analysis of the current status can provide the
amount of lead and value-added time.
In many situations teams take on the task of data collection. B01th individuals and
teams find it beneficial to develop a VSM data box in advance. Examples of data
boxes are shown in Table 5.24 later in this Section and in the service industry case
study in Section XI.
Value stream mapping definitions worth noting include:
Value-added time (VAn - The amount of time spent transforming the product, which
the customer is willing to pay for.
Lead t ime CUT) - The time it takes one piece of product to mo"e through all the
processes.
Cycle time (Cm - The time a piece is completed by an individual process.
See Rother (1999)23 and Conner (2001)3 for further details on valLie stream maps.


Future State Map
The future value stream map is an attempt to make the process lean. This involves
creativity and teamwork on part of the value stream manager and the lean team to
identify creative solutions. Everything the team knows about lean manufacturing
principles is used to create the process of the future.
Questions to ask when developing a future state map are:
• What is the required takt time?
• Do manufactured items move directly to shipping?
• Are items sent to a finished goods supermarket for customer pull?
• Is continuous flow processing applicable?
• Where is the pacemaker process? (This process controls the tempo of the
value stream.)
• Can the process be leveled?
• What is the increment of work to be released for kanban use?
• What process improvements can be used: changeover, machine uptime,

kaizen events, SMED, etc.? (Rother, 1999)23
Implementation Planning
The final step in the value stream mapping process is to develop an implementation
plan for establishing the future state. This includes a step-by-step plan, measurable
goals, and checkpoints to measure progress. A Gantt chart may be used to illustrate
the implementation plan. Several factors determine the speed of the plan. These
include available resources and funding. The plan could take months or years to
complete, and even then, there may be a need to improve upon it in the future.
(Rother, 1999)23
Examples of VSM icons and hypothetical current and future state maps are shown
on the following pages.

Value Stream Mapping Icons

The following icons may be used with value stream mappingl, kanban, or lean
manufacturing. (Lean, 2001)14
Electronic Flow
-FIFO-
FIFO
"---->
Finished Goods
Movement
txI
Go See
Inventory Kaizen Burst Kanban Batches Kanlban Post
D ~ [ox OX I
Kanban Production Kanban Signal Kanban Load Leveling
Withdrawal
Manual Information
Flow Operator Process Box
G
Pull Arrow
Pull Circle
IIIIIII~
Push Arrow
D Schedule Box Source
Supermarket Truck Shipment Buffer Stock Data Box

0.5 days
I I 12.7rrin I
Time Line - - - - - - -....... Lead Time Cycle Time
VILLANOVA UNIVERSITY V - 40 LEJI~N SIX SIGMA HANDBOOK

Current State Map
aCI Metal Forming, Inc. Tripod Weld Assembly
Customer
(7) Primary rustomers order

tripod~ in black or red
Order quantity and order
method varies among
customers Typicallead
tme to ,hip is 7 days
Currem annual sales =
3700 units
UPS
FPY = 98% FPY=999% FPV = 98% FPY =99 9% FPY=995%
REL =98% REL=99 9% REl =96% REL =100% REL = 990/.
ClO=O C/O =6 min
FPY = 99 9% FPV =99%

REl=99f~ REL =96%
CURRENT STATE VELOCITY =25.4 days to go from raw material to shipped product
Figure 5.19 Hypothetical Current State Map

Future State Map

QCI Metal Forming, Inc. Tripod Weld Assembly
Material Purchasing Customer Service Customer
(7) Primary Customers

order tripods in black or red
Order quanlity and order
S<lt40«1orcai1m;;ll!oo method varies among
if e-mail a :ldress is cuslo1'11efS Typicallead
provi::led time to ship is 7 days
Curren!: amual sales =
3700 units
Future state will ship
within 1 day from
P[(,,cUctforl order receipt
Contre!
One e91L hour shift per day

% hOlr unpaid lunch/shift
(2) 10 minute beaks/shift
430 m nutes/sh ft
250 work days :>er year
148 uits per day
Design capacit) for 20 unnsJdsy
430J2C = 21 5 Minutes/units TAKT
ups
Weld
RED FIFO
Assembly.
Grind and
Clean
') Paint
8lK FIFO
Pack and
Ship
err \.1.2.2 min =47 min
CiO= 10 m n
FPY;:99 9% FPY = 98% FPY=98% FPY =9a~ FPY = 99 9%
REL =96% REL = 98% REL = 98% REL = 96"1< REL = 100%
@@
FPV=99% 50%
REL ==96% reduction!
same day
shipping
.
18 Days 0.5 , Days 3.0 , Days 1.5 , Days
.
D,S Min ,
1 Min 12.2 , Min 2.1 , Min
FUTURE STATE VELOCITY =5.3 days to go from raw material to shipped product
Figure 5.20 Hypothetical Future State Map
VILLANOVA UNIVERSITY V - 42 LE.II.N SIX SIGMA HANDBOOK

FLOW CHARTS/PROCESS MAPPING
Flow Charts/Process Mapping

A flow chart, or process map, is useful to people familiar with a process, as well as,
to those that have a need to understand a process. A flow chart can depict the
sequence of products, containers, paperwork, operator actions, or administrative
procedures. A flow chart is often the starting point for process improvement. Flow
charts can be used to identify improvement opportunities as illustrated by the
following sequence:
• Organize a team for the purpose of examining the process

• Construct a flow chart to represent each process step
• Discuss and analyze each step in detail
• Ask the key question, "Why do we do it this way?"
• Compare the actual process to an imagined "perfect" process
• Is there unnecessary complexity?
• Does duplication or redundancy exist?
• Are there control points to prevent errors or rejects? Should there be?
• Is this process being run the way it should?
• Improvement ideas may come from substantially different processes
Listed below are examples of flow chart applications.
Purchasing Processing purchase orders, placing actual purchases,

vendor contract negotiations
Manufacturing Processing returned goods, handling internal rejections,
production processes, training new operators
Sales Making a sales call, taking order information, advertising
sequences
Administration Correspondence flow, processing times, correcting
mistakes, handling mail, typing letters, hiring employees
Maintenance Work order processing, p.m. scheduling
Laboratory Delivery of samples, testing steps, selection of new

equipment, personnel qualification sequence, management
of workflow
Table 5.21 Process Flow Application Examples

Flow Charts/Process Mapping (Continued)

There are advantages to depicting a process in a schematic fc.rmat. The major
advantage is the ability to visualize the process being described.
Process mapping or flow charting has the benefit of describing a process with
symbols, arrows, and words without the clutter of sentences. Mal11Y companies use
process maps to outline new procedures and review old procedurtes for viability and
thoroughness.
Most flow charting uses standardized symbols. Computer flow charting software
may contain 15 to 185 shapes with customized variations extending to the 500 range.
Many software programs have the ability to create flow charts ,or process maps,
although the information must come from someone knowledgeable about the
process.
Some common flow chart or process mapping symbols are shown below:
I ~ B
Alternate
Process
Process I
Pre :iefi ~ed Pr ce {oo::ent I 6-erminator )
Manual
Input
000 Off Page ~
Connector Connector
Extract
Figure 5.22 Common Flow Chart Symbols

Flow Chart Example

There are a number of flow chart styles including conceptual, person-to-person, and
action-to-action. Figure 5.23 is an action-to-action flow chart. This example comes
from the Quality System Handbook(2002)4 by Neville Edenborough and is used with
permission.
Start
Visual
inspection
Inform purchasing
No of rejection.
Generate
corrective
action report
Return to
supplier
Yes Dimensional
>----+1 inspection End
No
End
Figure 5.23 Action-to-Action Flow Chart Example

PROCESS MAPPING VERSUS VSM
Process Mapping Versus VSM

There seems to be a number of questions around the differences between process
mapping (commonly used in six sigma) and value stream mapping (associated with
lean). Team members often ask "Which is better?" and "Which should be used in
a given situation?" The answer: Both tools are completely valid and have their
place in documenting and improving processes. Let's examine the two tools.
Process mapping creates process visualizations to understand their technical and

organizational interactions. It is heavily used in six sigma and other methodologies
such as ISO and standard operating procedures. These maps identify all the factors
that can have an effect on the process or problem area. Process mapping is very
effective for complex inter-related processes that cross many functions.
By having all of the important aspects of the overall process on a single page, it is
much easier to understand everything that should be considered before making any
changes. If changes are being considered, one can re-draw the process map and
have an easy way to compare the "before" and "after" representations.
Process maps are a wonderful way to provide a clear understanding of all the
interrelationships in complex processes - like software codinSI, call processing,
product and process design cycles, chemical and pharmacelJtical processing,
clinical trials, etc. where there are many interactions and "cause-and-effect"
relationships. Six sigma, as a technique, heavily uses process mapping and SIPOC
together to attack process variation and to improve quality.
Value stream mapping (VSM) - Also known as "Information and Material Flow Maps"
is credited to Toyota Motor Company. It is best known in Leamin,g to See by Rother
and Shook. Value Stream Mapping is a scalable approach to create a visual
representation of what is happening in a process. It includes de!tailed information
at each step and across the whole value stream. VSMs are powerful tools to help
identify where changes are needed to improve system performance. The two major
things that VSMs focus on are material and information flow. Information flow is
often independent on the "thing" being mapped in the material f~ow.
If an organization generates information (say processing insur,ance claims), one

must imagine the claim as the "material" in the VSM, just as if it were a physical
product. By doing this it is possible to use a VSM approach succe:ssfully in a service
or office process. Using a process mapping approach to the infc)rmation flow may
also make sense, coupled with a material flow databox as iIIustri)ted in Table 5.24.

Process Mapping Versus VSM (Continued)
Description of process step:

Available time per day:
(shifts, net work time less paid breaks)
Number of people:
(Include specific types of skills if applicable)
Cycle time for this step:
(People and equipment time for each unit)
Unit of measure for this step:
(In some VSMs the UOM will change within the value stream)
Details about the process steps:
1.
2.
3.
Top 3 defect or rework issues/per step:
1. frequency:
2. frequency:
3. frequency:
Number of units waiting for this process:
Time value of WIP:
Observations about variation and waste in this process step:
1.
2.
3.
Table 5.24 Value Stream Map Data Box - Insurance Claim Processing
The VSM data box collects a great deal of information. The content is decided by the
VSM team at the onset of the work. The process step name, available time, who is
involved, and cycle time (Crr) are recorded. Cycle time should not be confused with
takt time, which is the production "pace" time to keep up with demand.
After populating the data boxes, a current state is developed. The team can then use
Pareto analysis to identify the relative few data boxes where the biggest
opportunities for value stream improvements lie.

Process Mapping Versus VSM (Continued)

There is one more critical item that is typically included in a VSM: cl calculation of the
total time the "thing" is in the value stream compared to the amount of time it is
being worked on. By adding all of the delays and time the thing is waiting in
inventory and then dividing this into the time it is actually bein'9 worked on, one
obtains the "value-added time" for the value stream. Often this number is very small
- much less than 1%. This is an indicator of system velocity and a useful metric to
compare the results to certain benchmarks. Some experts state that a 10% value-
added time would be approaching a world-class level of performance.
VSM approaches are extremely rich in factual and quantifiable diata that describes
the value stream's performance. In situations where making improvements is the
main goal, VSM is a stronger tool than process mapping. This assumes that the
level of complexity allows one to use VSM appropriately.
There seems to be some confusion when making a decision on which tool should
be used for a given situation. Read each of the scenarios below and form an opinion
with respect to which tool should be applied.
Scenario #1: A team has been charged with improving a process in an organization
that is under performing in profitability. Another big concern is the process lead
time. The process is fairly well understood, and it seems most of the issues are
driven by communication problems and high levels of human error.
Scenario #2: A team has been tasked with improving a process where the quality of
results is the single biggest concern. The process is very difficult to understand and
the reasons for the problems are hard to pin-point. The problenl seems technical
and the "people element" does not seem to be a big driver.
Scenario #1 focuses on velocity (lead time) and waste reduction (profitability) in a

process that seems easy to understand. Scenario #2 suggests a great deal of
complexity and a focus on quality. Scenario #1 points to VSM as the likely choice.
Process mapping is most likely the best choice for Scenario #2 because one must
determine the most likely sources of variation and build an initial hypothesis to
identify where to focus improvements. In Scenario #2, the process is complex and
the factors that may be impacting it are hard to understand, making it a good
candidate for process mapping tools.
There are cases where process mapping and VSM should happen concurrently on
the same piece of paper. For example, there may be a good caSE~ to use a process
mapping approach for the decision-making and information flow portion of the
visualization and VSM for the "thing" being processed. Taking advantage of the best
attributes of both tools, at exactly the same time, may just be thE~ ticket.

SPAGHETTI CHARTS
Spaghetti Diagrams
Spaghetti diagrams can be useful in describing the flow of people, information, or
material in almost any type of process. They are not as diagnostic or definitive as
VSM in traditional manufacturing operations, but certainly have a utility for a number
of service, administrative, and light production situations.
Most applications consider people, information, or material flows. The layout

follows a traffic or information route as though an imaginary line of string were being
deployed. The string accumulation does not have to continue for a full 8 hour shift.
Often a representative time interval is sufficient. The result normally resembles a
plate of spaghetti (hence the name). Shown below are examples of two of the three
major applications:
BINDER STORAGE
:----------j
_I '---_
COORS
PRlNTERA
+
...J
...J
~
W
~
CI)
~
0
tii
SERVER
!
a:
§I,
PRINtERS
SHIPPING
DOOR
PAPER
STORAGE
INACTIVE
DOOR
STORAGE
ACTIVE ACTIVE
DOOR DOOR
Figure 5.25 Work Area with Employee Traffic Flows
The spaghetti chart in Figure 5.25 shows individual movements (for a representative
shift interval). Included in this schematic are social activities, trips to the bathroom,
visits to the break area, etc. Normally, spaghetti diagrams focus on material or
information flow.

SPAGHETTI CHARTS
Spaghetti Diagrams (Continued)

Consider Figure 5.26 for a diagram of information flow. Employee A receives order
information via computer or other means. This is shown by path 1. He/she looks at
manual and comb bound inventories and communicates excepticmal needs directly
to employee B. This is shown by path 2. If additional manlJals are required,
employee B fills the manual buffer (path 3). If comb bound b()oks are required,
employee B provides this product and information via path 4. When electronic
media or other books are required, employee A follows path 5. The bundle is sealed,
boxed, and shipped by employee A following path 6.
BINDER STORAGE
--------------------,
I
_ _ _ II
I
Il __ _
ACTIVE
•
DOORS
8c
PRiNTERS
SHIPPING
DOOR
PAPER
STORAGE
INI\C'T1VE
, - -____- - , DOOR
STORAGE 1 --
ACTIVE ACTIVE
DOOR DOOR
Figure 5.26 aCI Work Area Showing Information FII)wS
Included in the discussion for Figure 5.26 are some items that could be considered
material flow. However, the 6 presented paths all contain information that is
important to the customer or to aCI.
Often spaghetti diagrams are used with computer generated or drafting layouts
before physical flow changes are made (particularly in the case of heavy equipment).
Refer to the first case study in Section XI for examples of ma1terial flows using
spaghetti diagrams.

REFERENCES
References
1. Anderson, P.F., & Wortman, B.L. (2004). CQA Primer. Terre Haute, IN: Quality
Council of Indiana.
2. Besterfield, D. (1998). Quality Control, 5th ed. Upper Saddle River, NJ:
Prentice Hall.
3. Conner, G. (2001). Lean Manufacturing for the Small Shop. Dearborn, MI:
Society of Manufacturing Engineers.
4. Edenborough, N.B. (2002). Quality System Handbook. Terre Haute, IN: Quality
Council of Indiana.
5. Eckes, G. (2001). The Six Sigma Revolution. New York: John Wiley & Sons.
6. Garvin, D. (1988). Managing Quality: The Strategic and Competitive Edge. New
York: The Free Press.
7. Gee, G., McGrath, P., & Izadi, M. (1996, Fall). "A Team Approach to Kaizen."
Journal of Industrial Technology.
8. Gee, G., Richardson, W.R., & Wortman, B.L. (2005). CMQ Primer. Terre Haute,
IN: Quality Council of Indiana.
9. Hashimoto, T. "Elimination of Volume Rotation Defects Through QC Circle

Activities." Factory Work and ac.
No. 33.
10. Hill, T. (1993). Manufacturing Strategy, 2nd ed. Burr Ridge, IL: Irwin.
11. Imai, M. (1997). Gemba Kaizen: A Commonsense, Low-Cost Approach to

Management. New York: McGraw-Hili.
12. Ishikawa, K. (1982). Guide to Quality Control. White Plains: Quality Resources.
13. Juran, J.M. (1999). Juran's Quality Handbook, 5th ed. New York: McGraw-Hili.
14. Lean Enterprise Institute. (2001). Mapping Icons. Retrieved November 6, 2001
from web site http://www.lean.org.
15. Liker, J., & Meier, D. (2006). The Toyota Way Fieldbook: A Practical Guide for
Implementing Toyota's 4Ps. New York: McGraw-Hili.
16. May, M. (2007). The Elegant Solution: Toyota's Formula for Mastering
Innovation. New York: Free Press.

REFERENCES
17. Moen, R., Nolan, T., & Provost, L. (1991). Improving QualitJif through Planned
Experimentation. New York: McGraw-Hili.
18. Omdahl, T. (1997). Quality Dictionary. Terre Haute, IN: (~uality Council of
Indiana.
19. Pande, P.S., Neuman, P.R. & Cavanagh, R.R. (2000). The Sb( Sigma Way. New
York: McGraw-Hili.
20. Porter, M. (1985). Competitive Advantage: Creating and Sustaining Superior

Performance. New York: The Free Press.
21. Productivity Development Team. (1999). Cellular Manufacturing: One-Piece

Flow for Workteams. Portland, OR: Productivity Press.
22. Robinson, A. (1990). Modern Approaches to Manufacturing Improvement: The

Shingo System. Portland, OR: Productivity Press.
23. Rother, M., & Shook, J. (1999). Learning to See: Value Stream Mapping to Add
Value and Eliminate Muda. Brookline, MA: The Lean Enterprise Institute.
24. Sobek, D.K., II. (2007). Montana State University. Retrived January 22,2007
from http://www.coe.montana.eduIlElfaculty/sobeklA3
25. Rath & Strong. (2000). Rath & Strong's Six Sigma Pocket Guide. Lexington,
MA: Rath & Strong Management Consultants.
26. Womack, J., Jones, D., & Roos, D. (1990). The Machine that Changed the
World. New York: HarperPerennial.
27. Womack, J., & Jones, D. (1996). Lean Thinking: Banish 'Waste and Create
Wealth in Your Corporation. New York: Simon & Schuster.

QUESTIONS
5.1. If one were to summarize in one word the 5.S. What is the major advantage of the A3 report?
ground breaking work of Womack, in
introducing the term "lean production" to the a. It is more thorough than other problem
Western world, that word would be: solving tools
b. It is a concise problem presentation format
a. Muda c. It requires less preparation time than other
b. Value reports
c Distance d. It minimizes the need for a follow·up plan
d Rhythm
5.7. Value stream mapping means:
5.2. When the term "pull" is used in lean thinking,
who is ultimately doing the pulling? a. Flow charting techniques
b. An identification of inputs, tasks, and
a. Downstream operations outputs
b. The takt time c. A pictorial view that identifies process
c. The customer steps
d. The cycle time d. A graphical flow-charting technique that
shows material and information flows
5.3. A breakdown of customer requirements is
accomplished using a diagram as depicted 5.8. Process flow improvement steps normally do
below: NOT include:
a. Asking "why" we do it this way

b. Asking what would make it "perfect"
c. Analyzing each step in detail
d. The use of Pareto diagrams
5.9. What does the following icon represent on a

What is the structure called? value stream map?
a. A Kano model
b. A CTa tree diagram
c. A CTa Gantt chart a. Buffer stock
d. A aFD matrix diagram b. Load leveling
c. Schedule box
5.4. The most widely used technique for d. Data box
distinguishing between chronic and
insignificant problems is: 5.10. Which of the following quality tools would be
LEAST important in the problem definition
a. A Pareto diagram phase?
b. A control chart
c. A cause-and-effect diagram A. Fishbone diagrams
d. A scatter diagram B. Control charts
C. Process flow diagrams
5.5. A team project charter is essential for all of the D. Pareto diagrams
following reasons, EXCEPT:
5.11. Management has dictated that inspection
a. The team will be focused on the appropriate locations be established for a new product.
problem What tool would customarily be applied to
b. The solution will be aligned towards the assist in the selection of the necessary
organization's goals inspection locations?
c. The team champion will be supportive
d. The charter provides a complete history of a. Histograms
the project b. c charts
c. Flow charts
d. Pareto diagrams

QUESTIONS
5.12. Which of the following statements can be safely 5.18. Which of the followin!~ management tools is
made about Pareto diagrams? most similar to the cau:se-and-effect diagram?
a. They have little application outside of the a. Pareto diagrams

quality area b. Affinity diagrams
b. They reflect an observation of fact c. Activity network dii:tgrams
c. They are bound by a universal set of laws d. CTQ Tree diagram!1
d. They have no validity for discrete data
5.19. In a cause-and-effect (l'ishbone) diagram what
5.13. The principal lean diagnostic tool for a complex position is the effect?
manufacturing operation would be a:
a. The problem or opportunity element
a. Process flow chart b. The environment or method element
b. Affinity diagram c. The machine or manpower element
c. Value stream map d. The measurement 4)r material element
d. Spaghetti diagram
5.20. Lean thinking enables c~ompanies to:
5.14. If a project charter was found to lack a company
impact measurement, what is missing in most I. Find the best way to specify value for the
situations? customer
II. Identify the value stream for each product
a. The importance of the improvement team to III. Permitthe customerto pull value as needed
the company from the producer
b. The SIPOC analysis IV. Reduce waste
c. Customer satisfaction improvements
d. Financial impact or cost savings a. I and IV only
b. II and III only
5.15. Lean manufacturing would be most concerned c. I, III, and IV only
with: d. I, II, III, and IV
I. Reducing waste 5.21. A team has been asked to improve the small
II. Reducing people purchase process in a company. They decide to
III. Reducing management layers create a process map of the existing process
IV. Eliminating bottlenecks in a process because it will help:
a. I and IV only a. Visualize the process

b. I, II, III, and IV b. Identify delays
c. I, II, and IV only c. Prioritize delays
d. I and III only d. Rank costs
5.16. The 5M model would typically NOT include 5.22. What is the main disadvantage of presenting a
which of the following options? team with an initial pr4)ject lasting more than
160 days?
a. Machines
b. Materials a. Excessive costs
c. Feedback b. A lowered expectation of success
D. People c. Too much time away from regular duties
d. The possibility that the team will expand the
5.17. Which of the following process mapping project boundaries
symbols would NOT be associated with a
decision point? 5.23. Affinity diagrams are useful tools to help
analyze and solve what type(s) of problems:
a. c.
<> a.
b.
c.
Unfamiliar problems
Structured problems
Mathematical modl!ls
b. o d. o d. Project flow problems

QUESTIONS
5.24. Considering that all of the following terms have 5.29. There are several varieties of A3 reports. They
benefits, which would LEAST likely affect can often have five or six key steps. Which of
product quality? the following would be a major step?
a. Muda a. An action or implementation plan

b. 5S b. A simplified benefit/effort matrix
c. Lean manufacturing c. The key brainstorming factors
d. Total productive maintenance d. Plans for management presentations
5.25. Identify the lean thinking statement that is NOT 5.30. Which of the following is NOT a critical element
true? in cycle time reduction?
a. Lean concepts include value, value stream, a. JIT training

flow, pull, and perfection b. Middle management commitment
b. Lean spans product development, c. Motion studies
production, purchasing, and customer d. Kaizen
support
c. Lean thinking streamlines processes for 5.31. In the six sigma define step, a critical-to- quality
better output tree is used by a project team. The various
d. Lean thinking often means reductions of levels of the tree are determined for all of the
people following, EXCEPT?
5.26. When comparing spaghetti diagrams, which a. The exact metrics for the customer
were constructed before and after a work area b. The needs of the customer
improvement, one would expect to find: c. The basic drivers for the customer
d. The potential third level CTQ metrics
a. Less traffic after the change
b. More exchange of information after the 5.32. The business case portion of a project charter
change would be likely to contain:
c. Less necessary equipment after the change
d. Fewer operators after the change a. A summary of the strategic reasons for the
project
5.27. Potentially, the most difficult area to obtain b. A problem or goal statement
meaningful information, during the six sigma c. The resources available to the project team
define phase, would be: d. The boundaries of the project team
a. From internal manufacturing processes 5.33. An Ishikawa diagram is also known as a:

b. From primary customers
c. From principal suppliers a. Box plot
d. From internal process input areas b. Process map
c. Scatter diagram
5.28. The affinity diagram gets its American name d. Fishbone diagram
from the fact that:
a. Objectives and solutions are drawn

together
b. Similar patterns of thought are grouped
together
c. The "whys" and "hows" attract each other
d. The "goals" and "means" are unified

QUESTIONS
5.34. Using the Kano model, improvement teams 5.39. Which of the following are principle reasons for
most often select improvement projects from NOT utilizing process mapping?
among which of the following customer need
categories? a. To identify where IJnnecessary complexity
exists
I. Satisfiers b. To visualize the pr,ocess quickly
II. Dissatisfiers c. To eliminate the planning process
III. Delighters d. To assist in work !iimplification
a. I only 5.40. According to Womack, the most important lean

b. I and II only principle is:
c. I and III only
d. I, II, and '" A. Pursue quality
B. Make value flow
5.35. When comparing push versus pull concepts, C. Pursue cost reduc1tions
who or what is really doing the push? D. Satisfy customers
a. The customer 5.41. A study of human moticms would most likely be

b. The cycle time applicable for:
c. Company management
d. The takt time a. Kano model analy!~is
b. Mapping the current value stream
5.36. One could say that the main improvement c. Cycle time reductions
between a current and future state map is: d. Pareto analysis
a. A reduced number of machine operations 5.42. Historical data indicate!~ that defective seat belts
b. A reduced number of operators are due to the following: 17% stitching, 14%
c. Less machine downtime metal corrosion, 23% mounting bolts, 38%
d. An increased work flow velocity foreign objects, and 8% other. Using a Pareto
diagram, one can conclude:
5.37. A criticality based Pareto analysis would most
likely focus on: a. The greatest cos'ls are due to foreign
objects
I. Internal scrap categories b. Mounting bolts cause more defects than
II. Potential safety risks stitching
III. Both real and potential economic losses c. The design causes more problems than the
IV. Occurrences of customer complaints consumer
d. Solving metal cormsion is insignificant
a. I, II, and III only
b. II, III, and IV only 5.43. A spaghetti chart would be helpful in tracking all
c. II and III only of the following, EXCEPT:
d. I, II, and IV only
a. Information flow
5.38. What is the basic objective of an A3 report? b. Material flow
c. Downtime activity
a. To simplify the problem solving process d. Operator movemenlt
b. To save valuable management time
c. To provide uniformity to project cause 5.44. Which of the followin,g is the LEAST likely
analysis element to be contained in a project charter?
d. To summarize a project in a concise and
understandable way a. Identification of thEI team members
b. The role of the team members
c. The quality and type of resources provided
d. The project scope
VILLANOVA UNIVERSITY V-56 LEJ~N SIX SIGMA HANDBOOK

1
VI. MEASUREMENT TECHNIQUES
IF YOU C,A N'T MEASURE A PROCESS OR SYSTEM, YOU

CERTAINLY CAN'T MAKE IT BETTER.
SOURCE OBSCURE
VILLANOVA UNIVERSITY VI-1 LEAN SIX SIGMA HANDBOOK

PROCESS ANALYSIS/PROCEDURES
Measurement Techniques
Measurement Techniques are presented in the following topic areas:
• Process analysis
• Data collection
• Measurement systems
• Process capability analysis
Process Analysis
Process Analysis is presented in the following topic areas:
• Procedures
• Work instructions
• Takt time
• LSS metrics
Procedures
ISO 9001 :200013 states that internal procedures shall control nonccmforming product
so that it is prevented from inadvertent use or installation. In many companies this
requirement is the responsibility of the quality department, although the actual
functions are performed by various other departments.
For most operations, a procedure can be created in advance by the appropriate

individual(s). Consider the situation where a process exists, but has not been
documented. The procedure should be developed by those having responsibility for
the process of interest. As an example, suppose the results of the interviews
revealed the following process to control nonconforming materii:lI:
1. The nonconformance is discovered.
2. The nonconforming material is segregated from conforming material.
3. The nonconformance is documented.
VILLANOVA UNIVERSITY VI-2 LE.lIN SIX SIGMA HANDBOOK

PROCESS ANALYSISIWORK INSTRUCTIONS
Procedures (Continued)
4. A material review board reviews the nonconformance to determine
disposition. The possible dispositions are:
• Scrap the part, which ends part usage and requires no further action.
• Accept the part for use "as is" or "as repaired."
• Rework the part to its original configuration requirement.
5. The actual disposition is made.
6. The product is returned to normal flow.
7. The paperwork is cleared.
The finished chart might look like that shown in Figure 6.1 on the following page.
Note, in this example, the customer contact element which is a substantial part of
ISOfTS 16949 (2002)15 has not been highlighted. This could certainly be made into
a separate process flow diagram as part of the deviation report process, however,
this example focuses on the internal nonconforming material flow. As one can see,
the process sounds simple in generic description. However, it may take several
twists and turns to show what really happens. The flow charting process helps to
visualize the necessary actions.
Work Instructions
Procedures describe the process at a general level, while work instructions provide
details and a step-by-step sequence of activities.
Flow charts may also be used with work instructions to show relationships of
process steps. Controlled copies of work instructions are kept in the area where the
activities are performed. Some discretion is required in writing work instructions,
so that the level of detail included is appropriate for the background, experience, and
skills of the personnel that would typically be using them.
The people that perform the activities described in the work instruction should be
involved in writing the work instruction. The wording and terminology should match
that used by the personnel performing the tasks.

PROCESS ANALYSIS/wORK INSTRUCTIONS
Nonconforming Material Flow Chart
Production or
Product is held
in special
holding area
Quality
Supervisor
reviews product
Yes
_ _ _ _ _ _..... Return product

to How
Clear DR and FilE'
File in DR Quality Records
Figure 6.1 Hypothetical Flow Chart for Control of Nonconforming Product
(Eden borough, 2002f, usedl with permission.
VILLANOVA UNIVERSITY VI-4 LEJl~N SIX SIGMA HANDBOOK

PROCESS ANALYSISITAKT TIME
Takt Time
Takt time is the available production time divided by the rate of customer demand.
In order to understand the measurement of takt time, and how it can be improved,
a discussion of a hypothetical process is presented below.
Consider a sequential operation with five operators or stations. The times allocated
for each station are indicated in Table 6.2.
Station Time (sec)

1 45
2 40
3 60
4 70
5 50
Total 265
Table 6.2 Work Time Example
If the takt time for the line is 60 seconds, the immediate observation is that station
4 exceeds the takt time and will not be able to maintain the pace. One option would
be to have some of the time eliminated by moving work to another station.
Upon further examination, some thought has to be given to stations 1, 2, and 5.

These stations do not align with the takt time because of the existing slack time.
Only station 3 has a full 60 seconds of work. A keen lean six sigma specialist will
seize upon this opportunity to review the entire line for another redesign.
The total work time used at the current time is 265 seconds. With 5 operators, this
equates to 53 seconds on a balanced line. Another perspective indicates that 4
operators will require 66.25 seconds. This is slightly more than the desired takt time
of 60 seconds. However, the initial task will be to re-examine the content of the
operations by looking at the value added and non-value added elements.
A thorough study may reveal that a significant portion of time can be eliminated,
leading to a reduction of 1 operator and possibly a reduction in floor space. (This
analysis was discussed in cycle time reduction presented in Section V. Refer to Gee
(1996)8 for a similar case study.
VILLANOVA UNIVERSITY VI- 5 LEAN SIX SIGMA HANDBOOK

Takt Time (Continued)

Table 6.3 illustrates the available options. Option 1 is to balance the line at 53
seconds. This will provide 7 seconds of slack time for each operator. This is
certainly a viable option. Option 2 is the theory of constraints approach showing
more slack time as one approaches the end of the line. Option 3 rE!duces the number
of operators and redistributes the work load. The redesign has an initial takt time
of 66.25 seconds. This is 6.25 seconds too much. A kaizen blitz might eliminate the
non-value added time reducing it to 60 seconds or lower, as shclwn in option 4.
Station Work time Option 1 Option 2 Option 3 Option 4

(seconds) after kaizen
1 45 53 59 66.25 60
2 40 53 56 66.25 60
3 60 53 53 66.25 60
4 70 53 50 66.25 60
5 50 53 47
Total 265 265 265 265 240
Table 6.3 Takt Time Options
The choice of the best option depends upon how well the proc:ess is controlled,
whether operations can be shifted from one station to another, and whether available
process improvement opportunities exist. Processes exhibiting Iclrge variation work
best using Option 2 conditions. For well controlled processes with small variatiol1J,
Option 1 is a better choice, and Option 3 is better still. If the process can be
improved to reduce cycle time, Option 4 is the best choice of all.
It should be emphasized that the use of lean manufacturing principles should not
result in the direct layoff of employees. It is best to reassign the ,e mployees to new
responsibilities, such as kaizen teams. Normal employee attrition is another option.
Expansion of the business into new areas is also a great solution.
The takt time is defined as the time needed to produce to customer requirements.
Takt time = Net operating time per period

Customer requirements per period


An example of the calculation of takt time is shown below:
Item Description Time

Net operating time Available: 480 minutes per 480
available per shift shift (1 shift per day)
Breaks: 2 @ 10 minutes each -20
Lunch @ 30 minutes -30
Net operating time (minutes) 430
Customer Monthly requirements in units 6000

requirements
Working days per month 20
Units per day required 300
Table 6.4 Takt Time Data
. 430 minutes per day . .

Takt time = = 1.433 mmutes per Unit
300 units per day
The takt time will be 1.433 minutes or about 86 seconds per unit. The ideal pace of
each operation is set at 86 seconds. The takt time can be listed either in minutes or
seconds. (Conner, 2001)6, (Rother, 1999)18, (Sharma, 2001)20
The perfect time to implement CFM may never arrive, but one can start with interim
solutions. If the appropriate equipment will take longer than desired to arrive,
perhaps an alternative solution can be found. If a one-pallet flow is required, a one-
container flow is better than no flow at all.
If a plant does not have the lUxury of mass production to establish a single takt time,
the solution may be to develop multiple takt times, breaking the requirements into
smaller components. (Conner, 2001)6


Illustration of Batch vs. One-piece Flow
The following three cases assume a series of three operations, which each process
one unit per minute.
Operation 1 1..----1...1
Operation 21. . ---I.~1 Opera ion 3 ~
Case 1: Orders are manufactured in batches of 100 units.
• Only batches of 100 units are transferred from operation to operation

• The total processing time is 201 minutes before the first Ulnit is available
• Total order time through the entire process is 300 minutes for 100 units
• Only batches of 10 units are transferred from operation to' operation

• The total processing time is 21 minutes before the first unlit is available
• Total order time through the entire process is 30 minutes for 10 units
• Each unit is transferred from operation to operation

• The total processing time is 3 minutes before the first unit is available
• Total order time through the entire process is 3 minutes f()r 1 unit
The above three cases illustrate the power of one-piece flow. If a c:ustomer changes
their requirements, the shop will not have 300 units in queue in partial stages of
production. The shop will be able to shift production requirements and provide the
first units rapidly. (Conner, 2001)li, (Sharma, 2001)20
The necessary quality, machine, personnel, materials, and suppliter resources must
be coordinated and made available as needed. The layout of the line or cell is a
starting point. The line should be examined as necessary to:
• Improve cycle times

• Reduce product defects
• Correct long changeover times
• Address equipment reliability issues
(Pr,oductivity, 1999)17

PROCESS ANALYSIS/LSS METRICS
Lean Metrics
The reader should note that some lean metrics are located elsewhere in this
Handbook, for example:
• General metrics and goals are discussed in Section II

• Takt time was reviewed earlier in this Section
• Total productive maintenance metrics are covered in Section IX
The TPM metrics include availability, operating speed rate (OSR), net operating rate
(NOR), performance efficiency (PE), and overall equipment effectiveness (OEE).
Some additional lean measurements that can prove beneficial:

1
Throughput Rate = - - - -
Cycle Time
Little's Law named after John D. C. Little (in 1961) states:
WIP Inventory = Throughput x Flow Time
Since the throughput rate is equal to 1/(cycle time), Little's Law can be written as:
Flow Time =WIP x Cycle Time

George (2005)9 further customizes Little's Law as:
· Number Items in Process

T ot aI Lea d T Ime =
Average Completion Rate
Note in this version of Little's Law that:
Total Lead Time = Flow Time

Items in Process = WIP Inventory
Average Completion Rate = Throughput Rate
George (2005)9 also defines process cycle efficiency as:
PCE = Value Added Time

Total Lead Time

Lean Metrics (Continued)
Example 6.1: If there are 30,000 units scheduled for production by ACME Stamping
and 3,000 units can be produced per day, what is the total required lead time?
TLT = WIP/Completion Rate
TLT =30,000 units/(3,000 units/day) =10 days
Example 6.2: Quality Shipping has a backlog of21 outstanding qu,otes. They provide
their customers 3 days maximum quotation service. What must their completion
rate be to meet their internal goal?
. WIP 21 quotes
Completion Rate =--
TLT
= 3 days
=7 quotes/day
Example 6.3: Better Brass Corporation requires 8 weeks to fulfilll a typical bearing
order. The value added time is 18 hours. Assuming the cc)mpany works 24
hours/day, seven day per week, what is the process cycle efficiEmcy expressed as
a percentage?
PCE = VAT x 100%

TLT
18 hours
peE = 1,344 hours
x 100% =1.34%
Example 6.4: If the throughput rate for an operation is 7,200 unit!; per hour, what is
the cycle time in seconds?
Cycle Time = 1
Throughput Rate
Cycle Time = 1 x 3,600 sec/hr = 0.5 sec

7,200/hr
Note that 1 hour equals 3,600 seconds.

Six Sigma Metrics

Harry (2000)12 introduced a new group of metrics for six sigma that serve to:
• Measure customer opinions
• Determine customer critical to quality (CTQ) factors
• Measure product outcomes (throughput yield, rolled throughput yield,

normalized yield)
• Correlate process outcomes to CTQs (measure processes with metrics that

correlate to the company's fundamental economics)
Breyfogle (2003)5 defines a large number of six sigma measurements with the
suggestion that some are controversial and an organization need not use them all.
The authors of this Handbook have presented only those that are widely used:
Widely Used Symbols
• Defects D =
• =
Units U
• Opportunities (for a defect) = 0
• =
Yield Y
Defect Relationships
• Total opportunities: TO =TOP =U x 0

• Defects per unit (see yield): DPU =DU =-In(Y)
• Defects per normalized unit (see yield): =-In(Y
norm )
• Defects per unit opportunity: DPO = DPU = D

o UxO
• Defects per million opportunities: DPMO = DPO x 106

Six Sigma Metrics (Continued)

Defect Relationships (Continued)
Example 6.5: A matrix chart indicates the following information for 100 production
units. Determine the DPU:
Defects 0 1 2 3 4 5
Units 70 20 5 4 0 1
DPU =D = 0(70} + 1(20) + 2(5} + 3(4} + 4(0} + 5(1) =: 0.47

U 100
One would expect to find 0.47 defects per unit.
Example 6.6: Assume that each unit in Example 6.5 had 6 opportunities for a defect
(i.e. characteristics A, B, C, D, E, and F). Determine DPO and DPMO.
DPO= DPU = 0.47 =0.78333

o 6
DPMO =DPO x 10 =78,333
6
Yield Relationships
Note that the Poisson equation is normally used to model defect occurrences. If
there is a historic defect per unit (DPU) level for a process, the probability P(x} that
an item contains x flaws is described mathematically by the equation:
e· OPU DPU x
P(x) =--x'--
Where: x is an integer greater or equal to 0
DPU is greater than 0
Note that 01 (zero factorial) = 1 by definition.


Yield Relationships (Continued)
If one is interested in the probability of having a defect free (as most people are),
then use x = 0 in the Poisson formula and the math is simplified:
P(O) =e- DPU
Resulting in the following common yield formulas:
Vield or first pass yield: v = FPV = e-DPU

Defects per unit: DPU = -In (V) (In means natural logarithm)
n
Rolled throughput yield: Vrt = RTY = II V; = (V )(V 1 2) ••• (Vn)
1=1
Normalized yield: Vnorm = ~RTY Where n = number of steps
Total defects per unit: TDPU = -In (Vrt)
Example 6.7: Assume that a process has a DPU of 0.47. Determine the yield.
V = e -DPU = e -0.47 = 0.625 = 62.5%
Example 6.8: Assume that a process has a first pass yield of 0.625. Determine the
DPU.
DPU = -In(V) = -In(0.625) = 0.47
Example 6.9: A process consists of 4 sequential steps: 1, 2, 3, and 4. The yield of

each step is as follows: V1 = 99%, V2 =98%, V3 = 97%, V4 = 96%. Determine the rolled
throughput yield and the total defects per unit.
Yrt = RTY = rry;

n
'-1
= (0.99)(0.98)(0.97)(0.96) = 0.90345 = 90.345%
TDPU = -In(RTY) = -In(0.90345) = 0.1015


Sigma Relationships
Probability of a defect = P( d)
P(d) = 1-Y or 1 - FPY
also P(d) = 1 - YRT (for a series of operations)
P(d) can be looked up in a Z table (using the table in reverse to determine Z).
Example 6.10: The first pass yield for a single operation is 95%. What is the
probability of a defect and what is the Z value?
P(d) =1 - 0.95 =0.05

Using the Z table for 0.05 approximates 1.645 sigma.
The Z value determined in Example 6.10 is called Z long-term or Z equivalent.
Z short-term is defined as: ZST =ZLT + 1.5 shift

Example 6.11: If Z long-term = 1.645, what is Z short-term?
ZST = ZLT + 1.5 = 1.645 + 1.5 = 3.145
Schmidt (1997)19 reports that the 6 sigma quality level (with the 1,, 5 sigma shift) can
be approximated by:
6 Sigma Quality Level = 0.8406 + .J29.37 - 2.221 In (ppm)
Example 6.12: If a process were producing 80 defectives/million, what would be the

6 sigma quality level?
6 Sigma Quality Level = 0.8406 + .J29.37 - 2.221 In(ppm)
6 Sigma Quality Level = 0.8406 + .j29.37 - 2.221 In(80)
6 Sigma Quality Level = 5.272
This answer can be also looked up in Appendix Table II, which (by interpolation)
a ears to be about 5.3.

DATA COLLECTIONITYPES OF DATA
Data Collection
Data Collection is presented in the following topic areas:
• Types of data
• Collection methods
• Measurement scales
• Data accuracy
Types of Data
Data is objective information that everyone can agree on. Measurability is important
in collecting data. The three types of data are attribute data, variable data, and
locational data. Of these three, attribute and variables data are more widely used.
Attribute Data
Attribute data is discrete. This means that the data values can only be integers, for
example, 3, 48, 1029. Counted data or attribute data are answers to questions like
"how many," "how often" or "what kind." Examples include:
• How many of the final products are defective?

• How many people are absent each day?
• How many days did it rain last month?
• What kind of performance was achieved?
Variable Data
Variable data is continuous. This means that the data values can be any real
number, for example, 1.037, -4.69, 84.35. Measured data (variable data) are answers
to questions like "how long," "what volume," "how much time," and "how far." This
data is generally measured with some instrument or device. Examples include:
• How long is each item?

• How long did it take to complete the task?
• What is the weight of the product?

Types of Data (Continued)

Variable Data (Continued)
Measured data is regarded as being better than counted data. It is more precise and
contains more information. For example, one would certainly know much more
about the climate of an area if they knew how much it rained ea,ch day rather than
how many days it rained. Collecting measured data is often difficult and expensive,
so one must often rely on counted data.
In some situations, data will only occur as counted data. For example, a food
producer may measure the performance of microwave popcorn by counting the
number of kernels of unpopped corn in each bag tested.
For information which can be obtained as either attribute or variable data, it is

generally preferable to collect variable data.
ICharacteristics I measurable
Variable
I
countable
Attribute
I
continuous discrete units or occurrences
may derive from counting good/bad
Types of data length number of de!fects
volume number of de!fectives
time number of scrap items
Examples width of door gap audit points lost
lug nut torque paint chips per unit
fan belt tension defective lamps
Data examples 1.7 inches 10 scratches
32.06 psi 6 rejected PCllrtS
10.542 seconds 25 paint runs
Table 6.5 A Comparison of Variable and Attribute Data
Locational Data
The third type of data does not fit into either category above. Thi~s data is known as
locational data which simply answers the question "where." Charts that utilize
locational data are often called measles charts or concentration charts. Examples
are a drawing showing locations of paint blemishes on an automobile, or a map of
the United States with the locations of the sales and distribution offices indicated.

Types of Data (Continued)

Conversion of Attribute Data to Variable Measures
Some data may only have discrete values, such as this part is good or bad, or I like
or dislike the quality of this product. Since variables data provides more information
than attribute data, for a given sample size, it is desirable to use variable data
whenever possible.
When collecting data, there are opportunities for some types of data to be either
attribute or variable. Instead of a good or bad part, the data can state how far out of
tolerance or within tolerance it is. The like or dislike of product quality can be
converted to a scale of how much do I like or dislike it.
Referring back to Table 6.5, two of the data examples could easily be presented as
variables data: 10 scratches could be reported as total scratch length of 8.37 inches,
and 25 paint runs as 3.2 sq. in. surface area of paint runs.
Even part failures can be reported as failure after 2,133 hours or cycles of operation,
etc., instead of the number of rejected parts.
Consideration of the cost of collecting variable versus attribute data should also be
given when choosing the method. Typically, the measuring instruments are more
costly for performing variables measurements and the cost to organize, analyze, and
store variables data is higher as well. A golno go ring gage can be used to quickly
check outside diameter threads. To determine the actual pitch diameter, a slower
and more costly process is required.
Variable data requires storing of individual values and computations for the mean,
standard deviation, and other estimates of the population. Attribute data requires
minimal counts of each category and hence requires very little data storage space.
For manual data collection, the required skill level of the technician is higher for
variables data than for attribute data. Likewise, the cost of automated equipment for
variables data is higher than for attribute data.
The ultimate purpose for the data collection and the type of data are the most
significant factors in the decision to collect attribute or variable data.
VILLANOVA UNIVERSITY VI·17 LEAN SIX SIGMA HANDBOOK

DATA COLLECTION/COLLECTION METHODS
Collection Methods
Collecting information is not cheap. To help ensure that the data is relevant to the
problem, some prior thought must be given to what is expected. Some guidelines
are:
• Formulate a clear statement of the problem
• Define precisely what is to be measured
• List all the important characteristics to be measured
• Carefully select the right measurement technique
• C,o nstruct an uncomplicated data form
• Decide who will collect the data
• Arrange for an appropriate sampling method
• Decide who will analyze and report the results
Without an operational definition, most data is meaningless. The ability to control

quality requires measurement. Both attribute and variable specifications must be
nailed down.
Data collection includes both manual and automatic methods. Data collected
manually may be done using printed paper forms or by data entry at the time the
measurements are taken. Manual systems are labor intensive and subject to human
errors in measuring and recording the correct values.
Automatic data collection includes electronic chart recorders and digital storage.
The data collection frequency may be synChronous, based on a set time interval, or
asynchronous, based on events. Automatic systems have higher initial costs than
manual systems, and have the disadvantage of collecting both "good" and
"erroneous" data. Advantages of automatic data collection systems include high
accuracy rates and the ability to operate unattended.
A large amount of data is difficult to analyze unless it is presen1ted in some easily

digested format. Graphs, charts, histograms, tallies, and Pareto diagrams are used
to analyze and present data.
VILLANOVA UNIVERSITY VI-18 LE.II.N SIX SIGMA HANDBOOK

Collection Methods (Continued)

Data Coding
The efficiency of data entry and analysis is frequently improved by data coding.
Problems due to not coding include:
• Inspectors trying to squeeze too many digits into small areas on a form
• Reduced throughput and increased errors by clerks at keyboards reading and

entering large sequences of digits for a single observation
• Insensitivity of analytic results due to rounding of large sequences of digits
Coding by adding or subtracting a constant or by multiplying or dividing by a factor:
Let the subscript, lowercase c, represent a coded statistic; the absence of a

subscript represent raw data; uppercase C represent a constant; and lowercase f
represent a factor. Then:
Code: Xc =X+C Decode: X = Xc - C and G = Gc

Code: Xc = X - C Decode: X = Xc + C and G = Gc
- -
Code: Xc = fX Decode: X = Xc/f and G = Gc/f
Code: Xc = X1f Decode: X = fXc and G = fG c
Coding by substitution:
Consider a dimensional inspection procedure in which the specification is nominal

plus and minus 1.25 inches. The measurement resolution is 1/8 inch and inspectors,
using a ruler, record plus and minus deviations from nominal. A typical recorded
observation might be 32-3/8 inches cramped in a check sheet space designed for a
width ofthree characters. The data can be coded as integers expressing the number
of 1/8 inch increments deviating from nominal. The suggestion that check sheet
blocks could be made larger could be countered by the objection that there would
be fewer samples and plot points per page.
Coding by truncation of repetitive place values:
Measurements such as 0.55303, 0.55310, 0.55308 in which the digits 0.553 repeat in
all observations can be recorded as the last two digits expressed as integers.
Depending on the objectives of the analysiS, it mayor may not be necessary to
decode the measurements.


Two common data collection methods are check sheets and chE~cklists.
Check Sheets
Check sheets are great tools for organizing and collecting facts and data. By
collecting data, individuals or teams can make better decisions, solve problems
faster, and earn management support.
Recording Check Sheets
A recording check sheet is used to collect measured or counted data. The simplest
form of the recording check sheet is for counted data. Data is c()lIected by making
tick marks on this particular style of check sheet. (Wortman, 2005)25
Dais of Week
Errors 1 2 3 4 5 6 Total
Defective Lights 1fH.1fH. 1fH. 1fH. 1fH.1II 1fH.1I 1fH. 40
Loose Fasteners I III 1fH. 1fH. II 16
Scratches I III III I III 1fH.1fH. 21
Missing Parts I I I 3
Dirty Contacts 1fH.1 1111 1fH.1II 1111 1fH. 1fH. 32
Other I III II III 9
ITotal I 19 I 19 I 16 I 19 I 23 I 25 I 121 I
Figure 6.6 Typical Recording Check Sheet
The check sheet can be broken down to indicate either shift, day, or month.
Measured data may be summarized by the means of a check sheet called a tally
sheet. To collect measured data, the same general check sheet form is used. The
only precaution is to leave enough room to write in individual mleasurements.


Recording Check Sheets (Continued)
Illustrated below is an example of a process check sheet used to evaluate the

productivity of a meeting. This is subjective data, but is very useful in some
circumstances. Measured data is physically measured information, such as: the pH,
the air pressure in psi, or the amount of downtime in hours.
Scale: (1 =Poor/1 O=Excellent) Member 1 Member 2 Member 3 Member 4 MemberS

On Track
Participation
Listening
Leadership
Decision Quality
Figure 6.7 Meeting Process Check Sheet
Checklists
The second major type of check sheet is called the checklist. A grocery list is a
common example of a checklist. On the job, checklists may often be used for
inspecting machinery or product. Checklists are also very helpful when learning
how to operate complex or delicate equipment.
Visual Washed Once Supervisor

Date: Area (Operator)
Inspection per Week
Preheat House
#2 Tub Room
#3 Tub Room
A Press Room
B Press Room
Finishing Room
Roll Mill Room
Figure 6.8 Housekeeping (24 hour) Checklist Example
A locational check sheet called a measles chart could be used to show defect
locations using a schematic of the product.

DATA COLLECTION/MEASUREMENT SCALES
Measurement Scales*
Table 6.9 details four measurement scales in increasing order of statistical
desirability.
Scale Description Example

Nominal Data consists of names or A bag of candy contained the
categories only. No ordering following colol"s:
scheme is possible. Yellow 1S
Red 10
Orange ~I
Green j'
Ordinal Data is arranged in some order Product defecjts, where A type

(Ranking) but differences between values defects are mOire critical than D
cannot be determined or are type defects eire tabulated as
meaningless. follows:
A 16
B 32
C 42
D 30
Interval Data is arranged in order and The tempera1tures of three

differences can be found. ingots were 2()O°F, 400°F, and
However, there is no inherent 600°F. Note, that three times
starting point and ratios are 200°F is not the same as 600°F
meaningless. as a temperature measurement.
Ratio An extension of the interval Product A costs $300 and
level that includes an inherent product B costs $600. Note,
zero starting point. Both that $600 is twice as much as
differences and ratios are $300.
meaningful.
Table 6.9 The Four Measurement Levels
* For a more expansive treatment of the measurement scales and probability, refer
to Triola (1994)23.

DATA COLLECTION/MEASUREMENT SCALES
Measurement Scales (Continued)
I Scale I Central Location I Dis~ersion I Significance Tests I

Nominal Mode Information Only Chi-square
Ordinal Median Percentages Sign or Run Test
Interval Arithmetic Standard or t test
Mean Average F test
Deviation Correlation Analysis
Ratio Geometric or Percent *
Harmonic Mean Variation
Table 6.10 Statistical Measures for Measurement Scales
*Note: Many of the interval measures may be useful for ratio data as well.
Data Examples
Example 6.13: Examples of continuous data, discrete data, and measurement

scales:
1. Continuous data: A station wagon weighs 3,478.6 Ib
2. Discrete data: On the last exam 200 people failed
3. Ordinal scale: Defects are categorized as critical, major A, major S, and minor
4. Nominal scale: A print-out of all shipping codes for last week's orders
5. Ratio scale: The individual weights of a sample of widgets
6. Interval scale: The temperatures of steel rods (OF) after one hour of cooling

DATA COLLECTION/DATA ACCURACY
Data Accuracy
Bad data is not only costly to capture, but corrupts the decision making process.
Some considerations include:
• Avoid emotional bias relative to targets or tolerances when counting,

measuring, or recording digital or analog displays.
• Avoid unnecessary rounding. Rounding often redu(~es measurement

sensitivity. Averages should be calculated to at least one more decimal place
than individual readings.
• If data occurs in time sequence, record the order of its calpture.
• If an item characteristic changes over time, record the, measurement or

classification as soon as possible after its manufacture, as well as after a
stabilization period.
• To apply statistics which assume a normal population, detE!rmine whether the

expected dispersion of data can be represented by at least 8 to 10 resolution
increments. If not, the default statistic may be the count of observations
which do or do not meet specification criteria.
• Screen or filter data to detect and remove data entry errors such as digital
transposition and magnitude shifts due to a misplaced decimal point.
• Avoid removal by hunch. Use objective statistical tests tel identify outliers.
• Each important classification identification should be recorded along with the

data. This information can include: time, machine, auditc)r, operator, gage,
lab, material, target, process change, conditions, etc.
It is important to select a sampling plan appropriate for the purpose of the use of the
data. There are no standards as to which plan is to be used for data collection and
analysis, therefore the analyst makes a decision based upon e:l(perience and the
specific needs. A few sampling methods are listed on the nex1t page. There are
many other sampling techniques that have been developed for specific needs.

DATA COLLECTION/DATA ACCURACY
Data Accuracy (Continued)

Random Sampling
Sampling is often undertaken because of time and economic advantages. The use
of a sampling plan requires randomness in sample selection. Obviously, true
random sampling requires giving every part an equal chance of being selected for
the sample.
The sample must be representative of the lot and not just the product that is easy to
obtain. Thus, the selection of samples requires some up front thought and planning.
Often, emphasis is placed on the mechanics of sampling plan usage and not on
sample identification and selection. Sampling without randomness ruins the
effectiveness of any plan. The product to be sampled may take many forms: in a
layer, on a conveyor, in sequential order, etc. The sampling sequence must be
based on an independent random plan. The sample is determined by selecting an
appropriate number from a hat or random number table.
Sequential Sampling
Sequential sampling plans are similar to multiple sampling plans except that
sequential sampling can theoretically continue indefinitely. Usually, these plans are
ended after the number inspected has exceeded three times the sample size of a
corresponding single sampling plan. Sequential testing is used for costly or
destructive testing with sample sizes of one and are based on a probability ratio test
developed by Wald (1947)24.
Stratified Sampling
One of the basic assumptions made in sampling is that the sample is randomly
selected from a homogeneous lot. When sampling, the "lot" may not be
homogeneous. For example, parts may have been produced on different lines, by
different machines, or under different conditions. One product line may have well
maintained equipment, while another product line may have older or poorly
maintained equipment.
The concept behind stratified sampling is to attempt to select random samples from
each group or process that is different from other groups or processes. The
resulting mix of samples can be biased if the proportion of the samples does not
reflect the relative frequency of the groups. To the person using the sample data,
the implication is that they must first be aware of the possibility of stratified groups
and second, phrase the data report such that the observations are relevant only to
the sample drawn and may not necessarily reflect the overall system.
VILLANOVA UNIVERSITY VI-2S LEAN SIX SIGMA HANDBOOK

MEASUREMENT SYSTEMS/MEASUREMENT SYSTEM ANALYSIS
Measurement Systems
Measurement Systems are presented in the following topic areas:
• Measurement system analysis • Measurement terms

• Measurement error • Repeatability and reproducibility
Measurement System Analysis

The following are simplified summaries of what must be accomplished to meet the
requirements for a measurement quality system.
• Measuring equipment (devices) - All measuring equipment (company or

employee owned) must be identified, controlled, and calibrated. Records of
the calibration and the calibration traceability must be kept.
• Confirmation system - The system by which the measuring equipment is

evaluated to meet the required sensitivity, accuracy, and reliability must be
defined in written procedures.
• Periodic audit and review - The calibration system must be evaluated on a

periodic basis by internal audits and by management reviews, as well as, any
other audits and reviews the system manager chooses.
• Planning - The actions involved with the entire calibration system must be
planned. This planning must consider management system analysis.
• Uncertainty of measurement - The uncertainty of measurement must be

determined. Generally, the determination of the uncertainlty of measurement
involves gage repeatability and reproducibility, as well as, other statistical
methods. Where uncertainties accumulate, the method of calculation of the
uncertainty must be specified for each case.
• Environmental conditions - Gages, measuring equipment, and test equipment

will be used, calibrated, and stored (when not in use) in conditions that ensure
the stability of the equipment. Ambient environmental cc)nditions must be
maintained. These conditions include temperature, vibration noise and
lighting control. Laboratories must also control dust and humidity.
VILLANOVA UNIVERSITY VI-26 LEj!~N SIX SIGMA HANDBOOK

MEASUREMENT SYSTEMS/MEASUREMENT ERROR
Measurement System Analysis (Continued)

• Records - Records must be kept on the operations that are used to calibrate
measuring and test equipment. The retention time for these records must be
specified. A gage status record is required.
• Nonconforming measuring equipment - Suitable procedures must be in place

to ensure that nonconforming measuring equipment is not used to make
determinations that qualify material or products.
• Confirmation labeling - A labeling system must be in place that shows the

unique identification of each measuring device and its status.
• Intervals of confirmation - The frequency that each measuring device is

recalibrated must be established and documented. The frequency of
calibration must be based upon the type of measuring equipment and the
severity of wear while in use.
• Sealing for integrity - Where adjustments may be made that may logically go
undetected, sealing of the adjusting devices or case is required.
• Use of outside products and services - Procedures must define controls that
will be followed when any outside calibration source or service is used.
• Traceability - Calibrations must be traceable to national standards. If no

national standard is available, the method of establishing and maintaining the
standard must be documented.
• Storage and handling - Measuring equipment, when in use, will be handled

according to established procedures, and in accordance with operator
training. When the measuring equipment is not in use, it will be in storage as
prescribed by procedures to ensure unwanted use.
• Personnel - Documented procedures are required for the qualifications and

training of personnel that make measurement or test determinations.
Measurement Error
The error of a measuring instrument is the indication of a measuring instrument
minus the true value. (Grant, 1988)10
a~RROR = a~EASUREMENT - a~RUE
or
a~EASUREMENT = a~RUE + a~RROR

MEASUREMENT SYSTEMS/MEASUREMENT ERROR
Measurement Error (Continued)

The precision of measurement can best be improved through the correction of the
causes of variation in the measurement process. However, it is frequently desirable
to estimate the confidence interval for the mean of measurements which includes
the measurement error variation. The confidence interval for the mean of these
measurements is reduced by obtaining multiple readings according to the central
limit theorem using the following relationship.
_ aREADINGS
a MEASUREMENT - In
Where n equals the number of readings
The formula states that halving the error of measurement requires quadrupling the
number of measurements. There are many reasons that a measuring instrument
may yield erroneous variation, including the following categories:
Operator This error occurs when the operator of a measuring instrument

variation obtains measurements utilizing the same equipment on the same
standards and a pattern of variation occurs.
Operator to This error occurs when two operators of a measuring instrument
operator obtain measurements utilizing the same equipment on the same
variation standards and a pattern of variation occurs betwlaen the operators
about the bias between them.
Equipment This error occurs when sources of variation within the equipment
variation surface through measurement studies. The reasons for this
variation are numerous. As an example, the equipment may
experience an occurrence called drift.
Material This error occurs when the testing of a sample destroys or
variation changes the sample prohibiting retesting. This same scenario
would also extend to the standard being used.
Procedural This error occurs when there are two or more methods to obtain a
variation measurement resulting in multiple results.
Software With software generated measurement programs, variation in the
variation software formulas may result in errors, even after identical inputs.
Laboratory This error is common when procedu res for measurement vary from
to laboratory laboratory to laboratory. The advent of standardized testing such
variation as the ASTM procedures have been developed to correct this type
of error.
VILLANOVA UNIVERSITY VI-28 LE.4'N SIX SIGMA HANDBOOK

MEASUREMENT SYSTEMS/MEASUREMENT TERMS
Measurement Terms
The following are not the official measurement definitions, but are meant to provide
the basic concepts. The following terms are primarily from (Stein, 2003)22 and (AIAG,
2002)3.
Accuracy Accuracy is the lack of bias between the user's current

measurement process and the same process using an accepted
standard as a reference. ISO VIM (1993)14 defines accuracy of
measurement as the "closeness of the agreement between the
result of a measurement and a true value of the measurand."
Accuracy is a parameter that changes slowly.
Bias The difference between the observed average of measurements
and a reference value. The definition for accuracy almost
always includes the phrase "lack of bias."
Precise, but Unbiased and Unbiased and

Biased not Precise Precise
Figure 6.11 Graphical Distinction Between Bias and Precision (Repeatability)

(Gryna, 2001 )11
Bias or Offset The systematic difference between the measurement results

from two different processes attempting to perform the same
measurement. The two processes could be different ways of
measuring the parameter, or they could be the same way of
measuring but differing as to some factor such as operator,
environment, or reference used. Bias (or offset) is usually
visualized as constant because it doesn't change much over
time. Bias is a parameter that changes slowly.

Measurement Terms (Continued)

Drift or Stability Drift is a change in bias, which means thj3 bias isn't really
constant, just changing on a slower time scale than the
measurement. Changes characterized as drift are systematic
in nature. Many times standards experience a predictable,
linear drift over many years. If supported with data, this
property can be used to reduce uncertainty by correcting for
the known part of the drift. Drift is a paramt!ter that changes
slowly.
Linearity The difference in bias (offset) values throughout the expected
operating ranges of a gage or measurement system. Linearity
is a description of measurement bias indicating how the value
of the bias varies over the entire capability range of a
measurement system.
Precision, Precision describes how close in vcllue successive
Variability, or measurement results fall when attempting tOI repeat the same
Noise measurement. This variability is often well described by a
normal distribution and may be quantified lJIsing a statistical
measure of spread such as standard deviatiion. Precision is
usually visualized as varying rapidly so that successive
measurements will capture all aspects of the distribution.
Precision, variability, and noise all change quickly.
Repeatability Repeatability is a measure of the ability of a measurement
process to get the same answer when an attempt is made to
keep all factors constant, or at least as stable as possible. The
same test instrument or tool is used sever.31 times in quick
succession without intervening activity, the environment is kept
constant, the same operator is used, etc. The variation in
measurements obtained with one measurement instrument, by
the same operator, measuring the same chalr acteristic on the
same part at or near the same time. Repeatability is very
similar in definition to precision.

Measurement Terms (Continued)

Reproducibility Reproducibility is the measure of the ability of a measurement
process to get the same answer under conditions of all relevant
factors varying normally. The environment is allowed to vary
within the usual limits, the operator may not be the same, even
the test equipment and fixtures can vary if they routinely vary
in practice. Measurements may be made over an extended
period of time. The reproducibility of a single gage is
customarily checked by comparing the results of different
operators taken at different times. Gage reproducibility affects
both accuracy and precision.
Resolution Resolution is a measure of the smallest change in the
measurand that can be represented by the display mechanism
of the measurement system.
Selectivity Selectivity is a measure of the ability of a measurement system
to distinguish between and display the difference in two
measured results when their measurands actually have two
different values.
Sensitivity Sensitivity is a measure of the smallest value of the measured
parameter that can be sensed by a measurement system.
Stability The drift or change in bias obtained with a measurement
system on the same measurement characteristic over an
extended time period.
AIAG (2002)3 defines five sources of measurement variation that can be determined
by gage R&R studies: bias, linearity, repeatability, reproducibility, and stability.
The calibration of measuring instruments is necessary to maintain accuracy (lack

of bias), but does not necessarily increase precision (repeatability). In order to
improve the accuracy and precision of a measurement process, it must have a
defined test method and be statistically stable.

MEASUREMENT SYSTEMS/REPEATABILITY AND REPRODUCIBILITY
Repeatability and Reproducibility

Assuming that gage accuracy and sensitivity are assured, it is often desirable to
determine the variance components of a gage measuring system: repeatability,
reproducibility, and process. There are three widely used m4:!thods to quantify
measurement error: the range method, the average and rangt! method, and the
ANOVA method. A brief description of each follows:
Range Method
Reproducibility is the variability introduced into the measurement system by the bias
differences of different operators. The range method does not quantify repeatability
and reproducibility separately. The range method is a simple way to quantify the
combined repeatability and reproducibility of a measurement system. To separate
repeatability and reproducibility, the average and range method! or the analysis of
variance method must be used.
Average and Range Method
The average and range method computes the total measurement system variability,
and allows the total measurement system variability to be separated into
repeatability, reproducibility, and part variation.
Analysis of Variance Method
The analysis of variance method CANOVA) is the most accurate method for
quantifying repeatability and reproducibility. In addition, the ANCIVA method allows
the variability of the interaction between the appraisers and the parts to be
determined.
Examples of the average and range and ANOVA method will be plresented using the
same data for both examples. The following methodology will be used:
1. Choose five parts at random and select a quality characteristic to measure

2. Identify the parts by numbering them 1 through 5
3. Pick three technicians/inspectors
4. Have them randomly measure the parts using the same mecllsuring instrument
5. Repeat step 4 so that there are two replications for each technician/part
combination
VILLANOVA UNIVERSITY VI-32 LE.~N SIX SIGMA HANDBOOK

Average and Range Method

The average and range method partitions variation into repeatability, reproducibility,
and process variation. The result of this analysis will:
• Determine repeatability by examining the variation between the individual

technicians and within their measurement readings
• Determine reproducibility by examining the variation between the average of

the individual technicians for all parts measured
• Establish process variation by checking the variation between part averages

that are averaged among the technicians
Note that the R&R determination described in the following example is referred to
as the "short method." Ford, Chrysler, and General Motors generally prefer another
format called the "long method."
Technician Part Readings Within Part Within Tech Between Tech
I I 1
I 1st 2nd
Set Set
2.0 1.0
I 1.0
R1 X1
1.5
I R2 X2
I R3
I
2 2.0 3.0 1.0 2.5
A 3 1.5 1.0 0.5 1.25 1.75 2.0
4 3.0 3.0 0.0 3.0
5 2.0 1.5 0.5 1.75
RA = 0.6
1 1.5 1.5 0.0 1.5
2 2.5 2.5 0.0 2.5
B 3 2.0 1.5 0.5 1.75 1.50 1.8
4 2.0 2.5 0.5 2.25
5 1.5 0.5 1.0 1.0
RB = 0.4
1 1.0 1.0 0.0 1.0

2 1.5 2.5 1.0 2.0
C 3 2.0 1.0 1.0 1.5 1.75 1.65
4 2.5 3.0 0.5 2.75
5 1.5 0.5 1.0 1.0
Rc= 0.7
Grand Ranges and Averages 0.567 1.817 1.67 1.817 0.35
-
R1 X 1 Rz X2 R3
Table 6.12 R&R Data for Average and Range Method
VILLANOVA UNiVERSITY VI-33 LEAN SIX SIGMA HANDBOOK

MEASUREMENT SYSTEMS/REPEATABILITY AND REPRODUCIBIILITY
Average and Range Method (Continu~ed)
To proceed further, one must determine several standard deviations using the range
formula:
_ R
(J' =-
d2
The 6' is the predicted value of the standard deviation based on 1the average range.
The factor d 2 depends upon the sample size (n) and the number of samples (K). It
is also convenient to work with values of 1/d 2 so that division can be replaced by
multiplication. As the number of samples, K, approaches infinity, the 1/d2 values
approach those calculated from the d 2 capability factors for control charts given il1
the Appendix. Table 6.13 shows 1/d2 values:
r;zJ 1
I
2
I
3
I
4
I
5
I
10
I
00
I
2 0.709 0.781 0.813 0.826 0.840 0.862 0.885
3 0.524 0.552 0.565 0.571 0.575 0.581 0.592
4 0.446 0.465 0.472 0.474 0.476 0.481 0.485
5 0.403 0.417 0.420 0.422 0.424 0.427 0.429

Where: n =sample size K =number of samples
Table 6.13 R&R Data 1/d2 Values
The calculation for repeatability is:
" Repeat = (d~)( R, ) = (0.885)(0.567) = 0.502
Where 1/d 2 is based on K =15 samples and n =2. From Table 6.13, the 00 column is
used for K and 1/d2 equals 0.885. R is the grand average range within parts.
The calculation for reproducibility is:
" Repro = (:,)(R3 )= (0.524)(0.35) = 0.183
= =
Where 1/d 2 is based on one sample, K 1, and n 3. From Tablle 6.13, 1/d 2 equals
0.524. R3 is the range between the average of all measurements taken by each
technician.

Average and Range Method (Continued)

The total measurement standard deviation is determined by the additive law of
variances according to the following formula:
(J Meas =~«(J Repeatt + «(J Repro)2

(J Meas =~(O.502)2 + (O.183t =0.534
The production process standard deviation is determined by:
(J Process = (d:)(R 2) = (0.420)(1.67) = 0.701
Where 1/d2 is based on three samples, K =

3, and a sample size n 5. From =
Table 6.13, 1/d 2 equals 0.420. R2 equals the average range between technicians of
the (R2 ) ranges of the average individual part readings (X1).
The total observed standard deviation in the example can also be determined by the
additive law of variances according to the following formula:
(J Observed = ~«(J Proct + «(J Meast

(J Observed =~(0.701t + (0.534)2 =0.881
In this example, the measurement error constitutes a substantial portion of total
observed variation (about 37%). Assume that the tolerance of a part is set at 8 sigma
of the process. In this case, the total tolerance would be 5.608 sigma.
The AIAG (2002)3 method of calculating the percentage of tolerance consumed by

the measuring system would be:
(5.15)(OR&R) (100) = (5.15)(0.534) (100) = 49%

Tolerance 5.608
The constant 5.15 comes from the normal curve, representing a 99% confidence
level. Obviously, the 49% value consumed by the measuring system is too large.

MEASUREMENT SYSTEMS/REPEATABILITY AND REPRODUCIBIILITY
Analysis of Variance Method

While the average and range approach is widely used, the ANOVA approach can also
be very useful. Not only can the contribution of repeatability, reproducibility, and
process be partitioned, but possible interactions (between technician and parts) can
be determined.
The following ANOVA examples use the same data as presented in the average and
range methods.
Part(p) Tech(T) Cell Interaction Squares

Number A B C Sum n Avg RowSqs A B C
2 1.5 1
1 8 6 1.333 10.667 4.5 4.5 2
1 1.5 1
2 2.5 1.5
2 14 6 2.333 32.667 12.S 12.5 8
3 2.5 2.5
1.5 2 2
3 9 6 1.5 13.5 3.125 6.125 4.5
1 1.5 1
3 2 2.5
4 16 6 2.667 42.667 18 10.125 15.125
3 2.5 3
2 1.5 1.5
5 7.5 6 1.25 9.375 6.125 2 2
1.5 0.5 0.5
Sum 20 18 16.5 54.5 IPSqs = 108.875 Ic:ell Sqs = 111.125
n 10 10 10 30
Avg 2 1.8 1.65
ColSqs 40 32.4 27.225 IT Sqs = 99.625 I
A ColSq is determined by squaring the Col total and dividil1lg by Col n, e.g.,
202 /10 = 40.
A RowSq is determined by squaring the Row total and dividing by Row n, e.g.,
82 /6 = 10.667.
An interaction CellSq is determined by squaring the Cell total and dividing by cell
sample size n, e.g., (2 + 1)2 / 2 = 4.5.

Analysis of Variance Method (Continued)

IX= 54.5 N = 30
CM = Correction for the mean = (IX)2 / N = (54.5)2 I 30 = 99.008
ITSqs = 99.625 IPSqs = 108.875 ICellSqs = 111.125
TotSS = IX2 - CM = 114.75 - 99.008 = 15.742
TechSS = ITSqs - CM = 99.625 - 99.008 = 0.6167
PartSS = IPSqs - CM = 108.875 - 99.008 = 9.867
InterSS = ICellSqs - CM - TechSS - PartSS = 111.125 - 99.008 - 0.6167 - 9.867 = 1.633
ErrorSS = TotSS - TechSS - PartSS -lnterSS = 15.742 - 0.6167 - 9.867 -1.633 = 3.625
AN OVA TABLE a = 0.05

Source SS OF MS Fcal F(a) Var Adj Var %
Technician 0.6167 2 0.3083 1.28 3.68 0.0067 0.0067 1.08
Part Number 9.867 4 2.467 10.21 3.06 0.3708 0.3708 59.89
Interaction 1.633 8 0.2041 0.84 2.64 -0.0188 0 0
Error 3.625 15 0.2417 0.2417 0.2417 39.03
Total OF 29 SIGe = 0.4916 Totals 0.6192 100.00
SIGtot = 0.7368
Technician OF = Number of technicians - 1 Part OF = Number of parts - 1
Interaction OF = (Technician OF) x (Part Number OF) Total OF = N - 1
Error OF = Total OF - Technician OF - Part Number OF - Interaction OF

F = Effect MS F. = 0.3083 = 1.28
Error MS Tech 0.2417
MS = SS/OF
F. = 2.467 = 10.21 F. = 0.2041 = 0.84
Part 0.2417 Inter 0.2417
Var (Variance) = (Effect MS - Error MS)/(Variance Coefficient)

Analysis of Variance Method (Continued)

The variance coefficient terms come from the original data table, where technician
equals 10, part equals 6 and interaction equals 2. The three calculations follow:
0.3083 - 0.2417
Var(TeCh) = 10
= 0.0067
2.467 - 0.2417
Var(parts) = = 0.3708
6
0.2041 - 0.2417
Var(lnter) = = - 0.0188
2
The adjusted variance (Adj Var) column converts the negative interaction variance
to O. The % column shows the percent contribution of each component based on the
Adj Var column.
SIGe (0.4916) is the square root ofthe Error MS (0.2417) and represents repeatability.
SIGtot (0.7368) is the sigma of total data. The difference between SIGe and SIGtot
is due to the difference among technicians and the difference among parts.
Repeatability is the error variance and contributes 39.03% of the t()tal variation in the
data. Reproducibility is the variation among technicians which contributes 1.08%
of the variation in the data. However, the F ratio test for technicians is 1.28
compared to an F critical value of 3.68 at the 95% confidence level. The null
hypothesis that there is no difference among technicians is not rejected. This
implies that a reduction in measurement variation cannot be achieved by directing
improvement activities at the three technicians. There is no interaction. The
interaction variance is effectively O. This means that each technician measures each
part in the same way.
Because variances are additive, one could say that the total measurement
=
contribution is repeatability variance + technician variance :J9.03% + 1.08% =
40.11 %. If R&R variation is to be reduced, it is the source of repeatability variation
which must be addressed.
Process variation accounts for 59.89% ofthe total variation in the data. Note that the
null hypothesis of no difference between parts would be rejected. Fcal (10.21) is
greater than Fa (3.06). Whether this is too much process variation requires
comparing total data with specifications. The specifications have InO way of knowing
the variance components of product output measurements.

PROCESS CAPABILITY ANALYSIS/HISTOGRAMS
Process Capability Analysis
Process Capability Analysis is presented in the following topic areas:
• Histograms • Process performance

• Normal distribution • Capability indices
• Capability studies • Performance indices
Histograms
Histograms are frequency column graphs that display a static picture of process
behavior. Histograms usually require a minimum of 50-100 data points in order to
adequately capture the measurement or process in question. A histogram is
characterized by the number of data points that fall within a given bar or interval.
This is commonly referred to as "frequency." A stable process is most commonly
characterized by a histogram exhibiting unimodal or bell-shaped curves.
Histogram with Special Causes Bimodal Histogram
LSL USL
Negatively Skewed Distribution Truncated Histogram

(After 100% Inspection)
Figure 6.14 Histogram Examples

PROCESS CAPABILITY ANALYSIS/HISTOGRAMS
Histograms (Continued)
.50 I
.51 III ill
~
.52 UK'I
.53 UK' UIr 28
26
.54 UK' UK' lUI' I 24
.55 UIr UIr UIr UIr I 22
.56 UK' UK' lUI' UK' UK' >- 20
~ 18
.57 UIr UIr UIr UIr UIr II w 16
:::l
.58 UIr UIr IUf UIr UIr III a 14
~ 12
.59 UK' UIr IUf UIr UIr I LL 10
.60 UK' UK' lUI' UK' I 8
.61 UIr UIr IUf 6
4
.62 UK' UIr 2
.63 UK' II
.64 III
.651
Frequency Tally Specification Limits 0.50 - 0.60
Figure 6.15 Histogram Construction Example
• As a rule of thumb, the number of histogram cells should approximate the

square root of the number of observations.
• An unstable normal distribution process is often characteri;~ed by a histogram

that does not exhibit a bell-shaped curve.
• For a normal distribution, variation inside the bell-shaped curve is chance or

natural variation. Other variations are due to special or assignable causes.
• There are many distributions that do not follow the normal curve. Examples
include the Poisson, binomial, exponential, lognormal, rectcmgular, U-shaped,
and triangular distributions.

PROCESS CAPABILITY ANAL YSIS/NORMAL DISTRIBUTION
Normal Distribution
Characteristics of normal distributions include:
• Most of the points (data) are near the centerline, or average

• The centerline divides the curve into two symmetrical halves
• Some of the points approach the minimum and maximum values
• The normal histogram exhibits a bell-shaped distribution
• Very few points are outside the bell-shaped curve
Min Value Max Value
Average
Figure 6.16 Norma"y Distributed Process
When all special causes of variation are eliminated, the process will produce a
product that, when sampled and plotted, has a bell-shaped distribution. If the base
of the histogram is divided into six (6) equal lengths (three on each side of the
average), the amount of data in each interval exhibits the following percentages:
II - 30 II - 20 11 - 0 II II + 0 II + 20 II + 30
1...4 - - - - - - - - - - 99.73% ----------..1
Figure 6.17 The Normal Distribution

PROCESS CAPABILITY ANALYSIS/NORMAL DISTRIBUTION
Normal Distribution (Continued)

Z Value
The Z transformation formula is:

x-J.l
z=--
0'
Where: =
x Data value (The value of concern)
IJ = Mean
a = Standard deviation
This transformation will convert the original values to the number of standard
deviations away from the mean. The result allows one to use a standard normal
table to describe areas under the curve (probability of occurrence). There are
several ways to display the normal (standardized) distribution:
1. As a number under the curve up to the z value:
o 1.0
Figure 6.18 P(z = - 00 to 1) = 0.8413
2. As a number beyond the z value:
o 1.0
Figure 6.19 P(z =1 to +00) = 0.1587


Z Value (Continued)
3. As a number under the curve and at a distance from the mean:
o 1
Figure 6.20 P(z =0 to 1) =0.3413
One must understand the specific normal distribution table that is being used. The
standard normal table in the Appendix uses the second method for calculating the
probability of occurrence.
Z Value Example
To illustrate the Z value, consider the following weights of typical 10th grade
students. The weights are normally distributed with a mean 1..1 = 150 Ib and standard
deviation a = 20.
130 150 170

\..l
Figure 6.21 Student Weight Distribution


Z Value Example (Continued)
Example 6.14: What is the probability of a student weighing between 120 and
1601b?
120 150 160
Figure 6.22 Probability of a Student Weighing Between 120 Ib and 160 Ib
The best technique to solve this problem is using the standard normal table in the
Appendix to determine the tail area values, and to subtract them from the total
probability of 1.
First, determine the Z value and probability below 120 Ib:
P(z = -00 to -1.5) = 0.0668
Second, determine the Z value and probability above 160 Ib:
P(z =0.5 to +00) =0.3085

Third, the probability between 120 and 160 Ib is:
1 - 0.0668 - 0.3085 = 0.6247
Thus, 62.47% of the students will weigh more than 120 Ib, but le~,s than 160 lb.

PROCESS CAPABILITY ANALYSIS/CAPABILITY STUDIES
Process Capability Studies

The determination of process capability requires a predictable pattern of statistically
stable behavior (most frequently a bell-shaped curve) where the chance causes of
variation are compared to the engineering specifications. A capable process is a
process whose spread on the bell-shaped curve is narrower than the tolerance range
or specification limits. Consider 6.23 where USL and LSL are defined as upper
specification limit and lower specification limit, respectively.
(Wortman, 2001 )26
!• LSL USL
···::
1·
·:
:
·:::
:•
·:
Figure 6.23 A Comparison of Process Spread to Tolerance Range
It is often necessary to compare the process variation with the engineering or

specification tolerances to judge the suitability of the process. Process capability
analysis addresses this issue.
A process capability study includes three steps:
• Planning for data collection
• Collecting data
• Plotting and analyzing the results
VILLANOVA UNIVERSITY VI-4S LEAN SIX SIGMA HANDBOOK

Process Capability Studies (Continu.~d)
The objective of a process capability study is to establish a state of control over a

process and then maintain that state of control through time. Actions that change
or adjust the process are frequently the result of some form o,f capability study.
When the natural process limits are compared with the specification range, any of
the following possible courses of action may result:
• Do nothing. If the process limits fall well within the spec:ification limits, no
action may be required.
• Change the specifications. In some cases, specification limits may be set

tighter than necessary. Contact the customer to see if the specifications may
be relaxed or modified.
• Center the process. When the process spread is approximately the same as
the specification spread, an adjustment to the centering olf the process may
bring the bulk of the product within specifications.
• Reduce variability. This is often the most difficult option to achieve. It may
be possible to partition the variation (within piece, batch to batch, etc.) and
work on the largest variation component first. An experirnental design may
be used to identify the leading source of variation.
• Accept the losses. In some cases, management must be content with a high
loss rate. Some centering and reduction in variation may be possible, but the
principal emphasis is on handling the scrap and rework e'fficiently.
Other capability applications:
• Providing a basis for setting up a variables control chart
• Evaluating new equipment
• Reviewing tolerances based on the inherent variability of ia process
• Assigning more capable equipment to tougher jobs
• Performing routine process performance audits
• Determining the effects of adjustments during processing
Modified frlom (Juran, 1999)16

Process Capability Studies (Continued)

Identifying Characteristics
The identification of characteristics to be measured in a process capability study

should meet the following requirements:
• The characteristic should be indicative of a key factor in the quality of the

product or process
• It should be possible to adjust the value of the characteristic
• The operating conditions that affect the measured characteristic should be

defined and controlled
If a part has fourteen different dimensions, process capability would not normally
be performed for all of these dimensions. Selecting one, or possibly two, key
dimensions provides a more manageable method of evaluating the process
capability. For example, in the case of a machined part, the overall length or the
diameter of a hole may be the critical dimension. The characteristic selected may
also be determined by the history of the part and the parameter that has been the
most difficult to control or has created problems in the next higher level of
assembly.
Customer purchase order requirements or industry standards may also determine

the characteristics that are required to be measured. In the automotive industry, the
Production Part Approval Process (PPAP) (AIAG, 1995)1 states: "An acceptable level
of preliminary process capability must be determined prior to submission for all
characteristics deSignated by the customer or supplier as safety, key, critical, or
significant, that can be evaluated using variables (measured) data."
Chrysler, Ford, and General Motors use symbols to designate safety and/or
government regulated characteristics and important performance, fit, or appearance
characteristics. (AIAG, 1995)1

Process Capability Studies (ContinuI9d)

Identifying Specifications or Tolerances
The process specifications or tolerances are determined by custclmer requirements,

industry standards, or the organization's engineering department.
The process capability study is used to demonstrate that the process is centered
within the specification limits and that the process variation predicts the process is
capable of producing parts within the tolerance requirements.
When the process capability study indicates the process is not capable, the
information is used to evaluate and improve the process in order to meet the
tolerance requirements. There may be situations where the specifications or
tolerances are set too tight in relation to the achievable process (:apability. In these
circumstances, the specification must be reevaluated. Ifthe specification cannot be
adjusted, then the action plan is to perform 100% inspection of the process, unless
inspection testing is destructive.
Developing Sampling Plans
The appropriate sampling plan for conducting process capability studies depends
upon the purpose and whether there are customer or standards requirements for the
study. Ford and General Motors specify that process capability studies for PPAP
submissions be based on data taken from a significant production run of a minimum
of 300 consecutive pieces. (AIAG,1995)1
If the process is currently running and is in control, control chart data may be used
to calculate the process capability indices. If the process fits a normal distribution
and is in statistical control, then the standard deviation can be estimated from:
o~~
R d
2
For example, for a project proposal of a new processes, a pilot run may be used to
estimate the process capability. The disadvantage of using a pilot run is that the
estimated process variability is most likely less than the process variability expected
from an ongoing process.
Process capabilities conducted for the purpose of improving the process may be
performed using a design of experiments (DOE) approach. The olbjective of DOE is
to determine the optimum values of the process variables which yield the lowest
process variation.

PROCESS CAPABILITY ANAL YSIS/CAPABILITY STUDIES
Process Capability Studies (Continued)

Verifying Stability and Normality
If only common causes of variation are present in a process, then the output of the
process forms a distribution that is stable over time and is predictable. If special
causes of variation are present, the process output is not stable over time.
(AIAG, 1995)2
The process may also be unstable if either the process average or variation is out-of-
control. Figure 6.24 depicts an unstable process with both process average and
variation out-of-control.
Figure 6.24 Unstable Process with Average and Variation Out-of-control
Common causes of variation refer to the many sources of variation within a process
that has a stable and repeatable distribution over time. This is called a state of
statistical control and the output of the process is predictable. Special causes refer
to any factors causing variation that are not always acting on the process. If special
causes of variation are present, the process distribution changes and the process
output is not stable over time. (AIAG, 1995)2
When plotting a process on a control chart, lack of process stability can be shown
by several types of patterns including: points outside the control limits, trends,
points on one side of the centerline, cycles, etc.
The validity of the normality assumption may be tested using the chi-square
hypothesis test. If the data does not fit a normal distribution, the chi-square
hypothesis test may also be used to test the fit to other distributions such as the
exponential or binomial distributions.

PROCESS CAPABILITY ANALYSIS/PROCESS PERFORMANCE
Process Performance vs. Specificati()ns

In Figure 6.25, one can see that the control limits are determined by process average
values. One can also see the process spread of the individual values. This process
spread can be predicted, and will indicate the range of the individuals being
produced. The process spread of the individuals can then be cClmpared directly to
the specifications to determine if the product is within specification.
UPPER SPECIFICATION LIMIT - - - - - - - - - - - - - - - USL
DISTRIBUTION OF
DISTRIBUTION OF
AVERAGES INDI7~ VALUES
-+-
CONTROL
LIMITS
±3 SX'
/- - - -
PROCESS
SPREAD
± 3 Sp
- UCL
_1_ - - - - - LCL
LOWER SPECIFICATION LIMIT - - - - - - - - - - - - - - - LSL
Figure 6.25 Comparison of Average and Individual Values to Specification Limits
The following calculation will determine the approximate standard deviation of the
process, if the R-bar is known from a control chart.
*d 2 is based on the sample size and comes from a factor table.

Process Capability from Control Charts

The area outside of specification for a normal curve can be determined by a Z value.
x - LSL USL -x
ZLower = S Zupper = S
LSL USL
o 1.0
Figure 6.26 Illustration of Z value Determination
Looking up the value of Z in a standard normal table gives the area outside of
specification. In the example above, Z upper = 1.0. Thus, 15.9% is above the USL.
Process capability (using Z value determinations) can be generated from a control

chart in the following manner:
Example 6.15: Given USL 8.5, LSL 3.5, n= = =3 and the following control chart data,
what are the projected failure rates?
6 4 7 8 1 5 6 4 8 8 6 9 4 6 8 5 6 7 1 6
X 7 6 9 6 7 3 2 6 2 4 6 5 8 6 7 10 6 3 6 1
10 5 5 1 5 2 7 7 9 6 5 2 4 5 8 7 4 4 5 2
X 7.7 5.0 7.0 5.0 4.3 3.3 5.0 5.7 6.3 6.0 5.7 5.3 5.3 5.7 7.7 7.3 5.3 4.7 4.0 3.0
R 4 2 4 7 6 3 5 3 7 4 1 7 4 1 1 5 2 4 5 5
= =
Note that n 3, X 5.465 and R 4.0 = A
0'---
_ R_ 4.0 -- 2 . 363
An estimate of process standard deviation: d2 1.693
Z =USL - X = 8.5 - 5.465 =1 28 Z = X-GLSL = 5.465 - 3.5 =0 83

2.363 .
UPPER GR 2.363 . LOWER
R
From the Z table: 1.28 =10.0% too high, 0.83 =20.3% too low =30.3% total

Process Capability from Control Charts (C()ntinued)
Example 6.15 Continued:
16
15 - LSL = 3.5 USL = 8.5
---
14 20.3% LOW - 10.0'7'0 HIGH
13 or or
12 18.3% LOW 8.3% HIGH
11 - or or
>- -
10 19.5% LOW ~ 9.3% HIGH
U
Z
W 8 -
-
9 ~
5W 67 - -
B: 5 --
~
r--
4 - r--
3 - -
-
2
1
o
-
1 2
n3 4 5 6 7 8 9
n 10
MEASUREMENT
Figure 6.27 Process Capability Histogram for Example 6.15.
=
However, from the histogram: 5/60 8.3% too high, 11/60 = 18.3% too low, 26.6% =
total. The individual data plugged into a calculator yields 9.3% too high, 19.5% too
low and 28.8% total. Given three answers, which is correct? ThE! histogram values
are the short-term truth and the other two are long-term estimat43s.

PROCESS CAPABILITY ANALYSIS/CAPABILITY INDICES
Process Capability Indices

To determine process capability an estimation of sigma is necessary:
R
0'
R
=-
d2
a R is an estimate of process capability sigma and comes from a control chart.
Capability Index
The capability index is defined as:

C = (USL - LSL)
p 60'
R
As a rule of thumb:
1.33 < Cp process is capable

1.00 ~ C p ~ 1.33 process is capable with tight control
C p < 1.00 process is incapable
Capability Index with Process Centering
Cpk is the ratio giving the smallest answer between:
- . [USL -
Cpk -mm X ,X-- -
LSL]
-
30'R 30'R
Capability Ratio
The capability ratio is defined as:

C = 60'R
r (USL - LSL)
As a rule of thumb:
Cr < 0.75 process is capable

0.75 ~ Cr ~ 1.00 process is capable with tight control
1.00 < Cr process is incapable
Note, the rule of thumb logic is somewhat out of step with the six sigma assumption
of ± 1.5 sigma shift. The above formulas only apply if the process is centered, and
remains centered within the specifications, and Cp CPk • =

PROCESS CAPABILITY ANALYSIS/PERFORMANCE INDICES
Process Capability Indices (ContinuE~d)
Cpm Index
The Cpm index is defined as:

C = USL - LSL
pm ~
6 (J.1 - T) 2 + 0'2
Where: USL = upper specification limit

LSL = lower specification limit
IJ = process mean
T = target value
a = process standard deviation
C pm is based on the Taguchi index, which places more emphasis on process

centering on the target. (Breyfogle, 1999)4
Process Capability Exercise
Example 6.16: For a process with X = 12, IJ = 12, a R = 2, USL = 115, LSL = 4, and T =
10, determine C P' C pk , and C pm :
C = (USL-LSL) = (16-4) =1
P 60'R 6(2)
C = min[USL30'R'
pk
- X X - LSL] = min[16 -12 12 - 4J = 0.667
30'R 3(2) , 3(2)
C = USL - LSL = 16 - 4 = 0.707

6~(J.1 - Tl + 0'2 6~(12 _10)2 + 22
pm
Process Performance Indices

To determine process performance an estimation of total data sigma is necessary:
n 2
L(X X) j -
0'.I = =
j 1
(n - 1)

Process Performance Indices (Continued)

Performance Index
The performance index is defined as:

p = (USL - LSL)
p 6al
Performance Index with Process Centering
Ppk is the ratio giving the smallest answer between:
_ . [USL - X ,X-- -
Ppk -mm LSL]
-
3aR 3aR
Performance Ratio
The performance ratio is defined as:
p = 6a j
r (USL - LSL)
Process Performance Exercise
Example 6.17: For a process with X =12, a =2, USL =16, and LSL =4, determine P
j P'
Ppk and Pr :
p = (USL - LSL) = (16 - 4) =1
p 6a j 6(2)
P
pk
=min[USL
3a
-X
R
'
X - LSL]
3aR
=min[163(2)
-12 4] =
12 -
, 3(2)
0.667
P = 6al = 6(2) =1
r (USL - LSL) (16 - 4)
In this example, Pp is equal to Pro

Capability Index Failure Rates

There is a direct link between the calculated Cp (and Pp values) with the standard
normal (Z value) table. A Cp of 1.0 is the loss suffered at a Z value of 3.0 (doubled,
since the table is one sided). Refer to Table 6.28 below.
Cp Z value ppm
0.33 1.00 317,311
0.67 2.00 45,500
1.00 3.00 2,700
1.10 3.30 967
1.20 3.60 318
1.30 3.90 96
1.33 4.00 63
1.40 4.20 27
1.50 4.50 6.8
1.60 4.80 1.6
1.67 5.00 0.57
1.80 5.40 0.067
2.00 6.00 0.002
Table 6.28 Failure Rates for Cp and Z Values
In Table 6.28, ppm equals parts per million of nonconformance (or failure) when the
process:
• Is centered on X
-
• Has a two-tailed specification
• Is normally distributed
• Has no significant shifts in average or dispersion
When the Cp , Cpk , Pp' and Ppk values are 1.0 or less, Z values and the standard normal
table can be used to determine failure rates. With the drivE~ for increasingly
dependable products, there is a need for failure rates in the Cp rcmge of 1.5 to 2.0.

Short-Term vs. Long-Term Capability

Up to this point, process capability has been discussed in terms of stable processes,
with assignable causes removed. In fact, the process average and spread are
dependent upon the number of units measured or the duration over which the
process is measured.
When a process capability is determined using one operator on one shift, with one
piece of equipment, and a homogeneous supply of materials, the process variation
is relatively small. When factors are included for time, multiple operators, various
lots of material, environmental changes, and so on, each contributes to increasing
the process variation. Control limits based on a short-term process evaluation are
closer together than control limits based on the long-term process.
Smith (2001 )21 describes a short run with respect to time and a small run, where
there is a small number of pieces produced. When a small amount of data is
available, there is generally less variation than is found with a larger amount of data.
Control limits based on the smaller number of samples will be narrower than they
should be, and control charts will produce false, out-of-control patterns.
Smith suggests a modified X and R chart for short runs, running an initial 3 to 10
pieces without adjustment. A calculated value is compared with a critical value and
either the process is adjusted or an initial number of subgroups is run. Inflated 0 4
and A2 values are used to establish control limits. Control limits are recalculated
after additional groups are run.
For small runs, with a limited amount of data, Smith recommends the use of the X
and MR chart. The X represents individual data values, not an average, and the MR
is the moving range, a measure of piece-to-piece variability. Process capability or
Cpk values determined from either of these methods must be considered preliminary
information. As the number of data points increases, the calculated process
capability will approach the true capability.
When comparing attribute data and variables data, variables data generally provides
more information about the process, for a given number of data points. Using
variables data, a reasonable estimate of the process mean and variation can be
made with 25 to 30 groups of five samples each. Whereas a comparable estimate
using attribute data may require 25 groups of 50 samples each. Using variables data
is preferable to using attribute data for estimating process capability.

PROCESS CAPABILITY ANAL YSIS/PERFORMANCE INDICES
Process Capability for Attribute Da1ta

The control chart represents the process capability, once special causes have been
identified and removed from the process. For attribute charts, capability is defined
as the average proportion or rate of nonconforming product.
• For p charts, the process capability is the process average nonconforming, p

and is preferably based on 25 or more in-control periods. If desired, the
proportion conforming to specification, 1 - p, may be used.
• For np charts, the process capability is the process avera!~e nonconforming,

p, and is preferably based on 25 or more in-control periods.
• For c charts, the process capability is the average number Clf non conformities,
C, in a sample of fixed size n.
• For u charts, the process capability is the average number e)f nonconformities
per reporting unit, ii.
The average proportion of nonconforming may be reported on a defects per million

opportunities scale by multiplying p times 1,000,000.
(AIAG, 1995)2
VILLANOVA UNIVERSITY VI-58 LE,~N SIX SIGMA HANDBOOK

REFERENCES
References
1. AIAG. (1995). Production Part Approval Process (PPAP). Chrysler Corporation,
Ford Motor Company, and General Motors Corporation.
2. AIAG. (1995). Statistical Process Control (sPC) Reference Manual. Chrysler

Corporation, Ford Motor Company, and General Motors Corporation.
3. AIAG. (2002). Measurement Systems Analysis: Reference Manual, 3rd ed.

Southfield, MI: AIAG.
4. Breyfogle, F.W., III. (1999). Implementing Six Sigma: Smarter Solutions Using
Statistical Methods. New York: John Wiley & Sons.
5. Breyfogle, F.W., III. (2003). Implementing Six Sigma, 2nd ed. New York: John
Wiley & Sons, Inc.
6. Conner, G. (2001). Lean Manufacturing for the Small Shop. Dearborn, MI:
7. Edenborough, N.B. (2002). Quality System Handbook. Terre Haute, IN: Quality
Council of Indiana.
8. Gee, G., McGrath, P., & Izadi, M. (1996, Fall). "A Team Approach to Kaizen."
Journal of Industrial Technology.
9. George, M.L., Rowlands, D., Price, M. & Maxey, J. (2005). The Lean Six Sigma
Pocket Toolbook. McGraw-HiII:New York.
10. Grant, E.L. and Leavenworth, R.S. (1988). Statistical Quality Control, 6th ed.
McGraw - Hill: New York.
11. Gryna, F.M. (2001). Quality Planning and Analysis, 4th ed. New York: McGraw -
Hill.
12. Harry, M. & Schroeder, R. (2000). Six Sigma. New York: Currency, Doubleday.
13. ISO 9001 :2000. Quality management systems - Requirements. Geneva:

International Organization for Standardization.
14. ISO VIM. (1993). International Vocabulary of Basic and General Terms in
Metrology, 2nd ed. Geneve, Switzerland: ISO

REFERENCES
15. ISOITS 16949:2002. Quality Management Systems-Particular Requirements for
the Application of ISO 9001:2000 for Automotive Production and Relevant
Service Part Organizations, 2nd ed. Geneva: International Organization for
Standardization.
16. Juran, J.M. (1999). Juran's Quality Handbook, 5th ed. New York: McGraw-HilI.
17. Productivity Development Team. (1999). Cellular Manufacturing: One-Piece Flow

for Workteams. Portland, OR: Productivity Press.
18. Rother, M., & Shook, J. (1999). Learning to See: Value Stream Mapping to Add
Value and Eliminate Muda. Brookline, MA: The Lean Enterprise Institute.
19. Schmidt, S.R. & Launsby, R.G. (1997). Understanding Industrial Designed
Experiments, 4th ed. Air Academy Press: Colorado Sprin"s, CO.
20. Sharma, A., & Moody, P. (2001). The Perfect Engine: How Iro Win in the New
Demand Economy by Building to Order with Fewer Resources. New York: The
Free Press.
21. Smith, G.M. (2001). Statistical Process Control and Quality Improvement. Upper
Saddle River, NJ: Prentice-Hall.
22. Stein, P.G. (2003). Certified Calibration Technician Primer. Terre Haute, IN: QCI.
23. Triola, M.F. & Franklin, L.A. (1994). Business Statistics, R.!ading: Addison -
Wesley.
24. Wald, A. (1947). Sequential Analysis. New York: John Wiley 4!1t Sons.
25. Wortman, B.L. & Carlson, D.R. (2005). CQE Primer. Terre Itlaute, IN: Quality
Council of Indiana.
26. Wortman, B.L., et al. (2001). CSSBB Primer. Terre Haute, IN: 'Quality Council of
Indiana.

QUESTIONS
6.1. Calculate the takt time for the one hour of 6.7. Why is process capability called a comparison
provided system data? between two independent worlds?
Demand 30 units Ihr A. Because it is based on two independent,
upper and lower, specification limits
B. Because it compares two independent
estimates of the standard deviation
C. Because it compares the world of
specifications with the world of product
spreads
a. 5 minutes c. 3 minutes D. Because it is based on two independent
b. 1.8 minutes d. 2 minutes estimates of normality of the means
6.2. Which of the following statements about check 6.8. The written document that describes the purpose
sheets is NOT true? and scope of an activity at a general level is
called a:
a. Subjective data requires larger forms than
measured data a. Manual c. Work instructions
b. Check sheets can record subjective data b. Procedure d. Check sheet
c. A shopping list is a check sheet
d. Both variables and attribute data can be 6.9. The repeatability of an R&R study can be
recorded on a check sheet determined by examining the variation between:
6.3. Legitimate techniques for ensuring data accuracy a. The individual technicians and within their
and integrity would NOT include which of the measurement readings
options below: b. The average ofthe Individual technicians for
all parts measured
a. Avoid bias relative to targets or tolerances c. Part averages that are averaged among
b. Screen data to remove entry errors technicians
c. Remove outlier data, as they arise d. The individual technicians and comparing it
d. Record data In time sequence, when to the part averages
appropriate
6.10. Which of the following statements describes
6.4. In a normal distribution, what Is the area under attribute data?
the curve between +0.7 and +1.3 standard
deviation units? a. Number of employees wearing green shirts
b. Number of gallons of chemical used in a
a. 0.2903 c. 0.2580 process
b. 0.7580 d. 0.1452 c. Diameter of a hole
d. Miles per gallon of automobile fuel economy
6.5. A histogram is used to plot the number of voids
found versus the weight of a plastic injection 6.11. When conducting a process capability study
molded part. One would expect the shape of the consistent with PPAP requirements, which ofthe
distribution to be: following is mandatory?
a. Normally distributed a. A submission of related control chart data

b. Binomially distributed b. A selected characteristic that is controllable
c. A decreasing slope c. Data collected from a significant production
d. Bimodal run of 300 or more consecutive pieces
d. A demonstrated 5 sigma capability
6.6. The RTY calculation for the following process
steps would be: ~~~~ 6.12. Calculate the performance ratio for a process
with a spread of 20·30 and upper and lower
specification limits of 15 and 35 respectively.
a. 2.0 c. 0.5
b. 1.0 d. 0.4

QUESTIONS
6.13. How is process capability for attribute data 6.18. The calibration of ml!asuring instruments is
calculated? necessary to maintain accuracy. How does
calibration affect preci~;ion?
a. It is not possible to estimate a process
capability since attribute data is abnormal a. The precision increases over the working
b. It is the average proportion or range of the instrument
nonconformities per unit for stable b. The precision decreases over the working
processes range of the instrument
c. It is calculated the same way as process c. Calibration has a minimum impact on
capability for continuous data precision
d. It is the number of conformities per unit for d. The precision will vary over the working
unstable processes range of the instrument
6.14. Select the INCORRECT statement from the 6.19. A process consists of ttlree sequential steps with
following. The IDs of a certain piece of tubing the following yields:
are normally distributed with a mean 1.00". The Y1 = 99.8, Y2 = 97.4, Y3 = 96.4
proportion of tubing with IDs less than 0.90" is: Determine the total defl!cts per unit.
a. Less than the proportion of IDs greater than a. 0.063 c. 0.067

0.90" b. 0.065 d. 0.069
b. Less than 50%
c. Less than the proportion with IDs greater 6.20. A manufacturer of children's wagons would like
than 1.10" to visualize where th~! defects occur on the
d. Less than the proportion with IDs greater wagons. The defects encountered so far include
than 1.00" missing bolts, paint Il,eeling, missing decals,
cracked wood, sharp edges, and bent metal
6.15. The measurement and recording of the length of parts. A useful tool for this application is a:
a product in centimeters would reflect the
application of which measurement scale? a. XYmatrix c. Regression plot
b. Measles chart d. Scatter diagram
a. Nominal c. Interval
b. Ordinal d. Ratio 6.21. A capability study conducted during a pilot run of
100 units indicated the Gpk upper value to be 1.8
6.16. Approximately what percentage of the area under while the Cpk lower value was 0.90. The
the normal curve is included within:!: 3 standard customer requires a Cpl~ minimum value of 1.25.
deviations from the mean? What action should be taken?
a. 95.0% c. 90.0% a. Center the process

b. 99.7% d. 68.3% b. Reduce the variabll ity
c. Renegotiate the cu!;tomer specification
6.17. If the specifications for a process are unknown, d. 100% inspect until new equipment is
but the standard deviation of the process is available
known to be 1.0, which of the following can NOT
be determined? 6.22. With regard to an R&R study, measurement error:
I. Cr a. Is normally the fault: of the Inspector

II. Cp b. Can be determined
III. Pr c. Is usually of little cClnsequence
IV. Pp d. Can be eliminated b~, frequent calibrations of
the measuring devic~e
a. I and II only c. III and IV only
b. I and III only d. I, II, III, and IV 6.23. If a customer needs 2fl0 items per day and a
company operates 7 hou rs per day, what must be
the operational takt timE!?
a. 0.25 minutes/item c. 40 items/hour

b. 3.334 items/minute d. 1.5 minuteslitem

QUESTIONS
6.24. The main purpose of a gage R&R study Is to: 6.30. Which ofthe following is an example of variables
data?
a. Determine the total variation in a part
b. Determine the total variation in a gage a. Paint chips per unit
system b. Tension on fan belts
c. Determine how much of the total variation is c. Percent of defective units in a lot
contributed by measurement d. Audit points
d. Determine the ratio of reproducibility to
repeatability measurements 6.31. If one were to compare short·term capability with
long·term capability for the same process, it
6.25. Random selection of a sample: should not be surprising to flnd:
a. Theoretically means that each item in the lot a. The long·term capability improves
has an equal chance to be selected b. The Cp is better short·term
b. Ensures that the sample average will equal c. The results are very comparable
the population average d. The average drifts but the variation stays
c. Means that a table of random numbers was
used to dictate the selection 6.32. One thousand units of product were examined
d. Is a meaningless theoretical requirement for the possibility of 5 different undesirable
characteristics. A total of 80 defects were found.
6.26. What documents provide details that expand how How many defects would be expected in a million
and when an activity will take place? opportunities?
a. Policy statements a. 16,000 c. 61,458

b. Procedures b. 26,666 d. 80,000
c. Instructions
d. Documents and forms 6.33. Measurement parameters that generally change
slowly over time include all of the following,
6.27. The two most common determinations of EXCEPT:
measurement variation that result from a gage
R&R study are: a. Bias c. Accuracy
b. Precision d. Stability
a. Gage accuracy and precision
b. Gage accuracy and stability 6.34. A process is centered and the Cp is 0.8. This
c. Gage linearity and stability indicates that the specification range Is what
d. Gage reliability and precision percent of the process width?
6.28. The tabulation of the number of times a given a. 60% c. At least 100%
quality characteristic measurement occurs, b. 80% d. 120%
within the product sample being checked, is
called a: 6.35. Use this data for the following question.
a. Histogram Raw data: 307,309,310,315,309,310

b. Normal distribution Coded data: 7,9,10,15,9,10
c. Control chart
d. Random function What is X·bar?
6.29. When would Cp equal Cpk in a capability index a. 10 c. 310

calculation? b. 12 d. 312
a. When the standard deviation is low 6.36. Accuracy is:

b. When the specification limits are wide
c. When CP = PP a. Getting consistent results repeatedly
d. When the process is perfectly centered b. Reading to one decimal greater than the
reported dimension
c. Using the best measuring device available
d. Getting an unbiased true value
VILLANOVA UNIVERSITY VI· 63 LEAN SIX SIGMA HANDBOOK

QUESTIONS
6.37. Reproducibility in an R&R study would be 6.43. Which ofthe following st;:ltements best describes
considered the variability introduced into the a bimodal distribution?
measurement system by:
a. This distribution shows stratified data and
a. The change in instrument differences over has two distinct peaks
the operating range b. This distribution shows a single mode and
b. The total measurement system variation bell shaped distribution
c. The bias differences of different operators c. This distribution is truncated
d. The part variation d. This distribution has several distribution
peaks
6.38. A check sheet is being used to record
Information from several different machining 6.44. The DPMO for a procens is 860. What is the
centers. Whic.h of the following types of approximate 6 sigma level of the process?
information is NOT suitable for inclusion in the
check sheet? a. 4.2 c. 4.6
b. 4.4 d. 4.8
a. Number of rejects c. Machine number
b. Types of defects d. Dimensional size 6.45. The interaction term in an R&R ANOVA indicates
an interaction between:
6.39. The outcome from flipping a coin is considered:
a. The technician and Imeasurement error
a. Discrete data c. Random data b. The technician and tthe part
b. Continuous data d. Probability data c. The part and the totill variation
d. The repeatability and the reproducibility
6.40. Customer satisfaction data is often accumulated
using some form of Likert scale (1-5, 1-7, etc). 6.46. One use of recording check sheets is:
What measurement scale is being used?
a. Automating the chaliing of variables data
a. Nominal c. Interval b. Collecting tally counts of attribute data
b. Ordinal d. Ratio c. Identifying process variables
d. Creating process m;ips
6.41. One of the advantages of using the ANOVA
method versus the average and range method for 6.47. If the value added time for a process is 40
gage R&Ris: minutes per 8-hour shift, what is the process
cycle efficiency expressed as a percentage?
a. Less math
b. Fewer readings a. 5.00% c. 6.24%
c. An ability to measure interactions b. 1.50% d. 8.33%
d. The ability to partition variation into
component parts 6.48. One can say that normailly distributed data will:
6.42. Which two of the following statements are true a. Be wider than the specification limits
for sequential sampling plans? b. Show a larger standard deviation for larger
I. The ASN is relatively low sample sizes
II. These plans are simple to understand c. Be spread approxim;itely equal on either side
III. Samples are taken one item at a time of the average
IV. The administration costs are low d. Be truncated beyond the specification limits
a. I and III only c. I and IV only

b. II and III only d. II and IV only

VII. ANALYSIS TECHNIQUES
DO NOT-PUT YOUR FAITH IN WHAT STATISTICS SAY

UNTIL YOU HAVE CAREFU.LLY CONSIDERED WHAT THEY
DO NOT SA¥.
WILLIAM W. WATT
VILLANOVA UNIVERSITY VII - 1 LEAN SIX SIGMA HANDBOOK

SEVEN CLASSICAL WASTES
Analysis Techniques
Analysis Techniques are described in the following topic areas:
• Seven classical wastes

• Variable relationships
• Hypothesis testing
• Root cause analysis
Seven Classical Wastes

Non-value-added activities are classified as muda. It is another tlerm for waste that
exists in a process. At each step in the process, work is applied to the operation.
The useful activities that the customer will pay for is considered value-added. The
other activities are not important to the customer or contain telements that the
customer will not pay for. Those non-paying activities are mlLlda. Imai (1997)2
provides a list of seven muda categories that have been widely lIsed in industry:
• Overproduction • Processing
• Inventory • Transport
• Repair/rejects • Waiting
• Motion
Overproduction
The muda of overproduction is producing too much at a particlLllar point in time.

Overproduction is characterized by:
• Producing more than is needed by the next process or customer

• Producing earlier than needed by the next process or cust:omer
• Producing faster than is needed by the next process or customer
In the just-in-time environment, producing too early is as bad as producing too late.
Parts need to be available at a certain location at a certain time! according to the
customer's schedule.
VILLANOVA UNIVERSITY VII-2 LEAN SIX SIGMA HANDBOOK

Seven Classical Wastes (Continued)

Overproduction (Continued)
Having the product too early, too late, or in quantities that are too great, will result
in undesirable consequences, such as:
• Extra space used at the customer's plant

• Extra space used at the organization's plant
• Extra raw materials in use
• Extra utilities used
• Extra transportation for the customer and organization
• Extra scheduling costs
Inventory
Parts, raw materials, work-in-process (semi-finished goods), inventory, supplies, and

finished goods are all forms of inventory. Inventory is considered muda since it
does not add value to the product. Inventory will:
• Require space in the shop • Require conveyor systems

• Require transportation • Require additional labor
• Require forklifts • Require interest on materials
In addition, inventory sitting around in various stages can be affected in these ways:
• Gathering dust • Getting wet

• Deterioration • Handling damage
• Becoming obsolete
Repairl Rejects
The repair or rework of defective parts involves a second attempt at producing a
good item. Rejects involving scrapping of the whole part are a definite waste of
resources. Having rejects on a continuous flow line defeats the purpose of
continuous flow. Line operators and maintenance will be used to correct problems,
putting the takt time off course. Repair and rework may require nonconforming
product forms to be filled out by suppliers, generating muda.
Various design changes can be muda also. Design changes are rework or a
development effort causing additional labor.
VILLANOVA UNIVERSITY VII- 3 LEAN SIX SIGMA HANDBOOK


Motion
The efficient use of the human body is critical to the well being of the operator.
Extra unneeded motions are wasteful. Operators should 110t have to walk
excessively, lift heavy loads, bend awkwardly, reach too far, repeat motions, etc.
New tools should be designed to help with strenuous hand or body motions. The
layout of the workplace should be redesigned to take advantage of proper
ergonomics. Each work station should be analyzed for ergonomic and motion
requirements.
Ergonomics can eliminate factors in the workplace that may cause injuries and lost
production. Some guidelines for providing sound ergonomic: principles in the
workplace include:
1. Emphasize safety at all times

2. Fit the employee to the job
3. Change the workplace to fit the employee (not vice versa)
4. Design the workplace so that neutral body positions are maintained
5. Redesign tool handles to reduce stress and injury
6. Vary the tasks through job rotation, every 2 to 4 hours
7. Make the machine serve the human (Sharma, 2001 y
Processing
Processing muda consists of additional steps or activities in the manufacturing
process. This can be described as:
• Having to remove burrs from a manufacturing process

• Reshaping a piece due to poor dies
• Adding an extra handling process due to lack of space
• Performing an inspection step since all inspection is non-value-added
• Repeating product changes that are unneeded
• Maintaining extra copies of information
Transport
All forms of transportation are muda. This describes the use of forklifts, conveyors,
pallet movers, and trucks. This can be caused by poor plant layouts, poor cell
designs, use of batch processing, long lead times, large s.torage areas, or
scheduling problems. The muda of transport is always considered for elimination.
VILLANOVA UNIVERSITY VII- 4 LEJ\N SIX SIGMA HANDBOOK


Waiting
The muda of waiting occurs when an operator is ready for the next operation, but
must remain idle. The operator is idle due to machine downtime, lack of parts,
unwarranted monitoring activities, or line stoppages. A maintenance operator
waiting at a tool bin for a part is muda. The muda of waiting can be characterized
by:
• Idle operators
• Breakdowns in machinery
• Long changeover times
• Uneven scheduling of work
• Batch flow of materials
• Long and unnecessary meetings
(Imai, 1997)2
Additional Classes of Wastes

In addition to the seven classical wastes previously listed, some sources show:
• Misused resources
• Underutilized resources
• Counting
• Looking for tools or parts
• Multiple systems
• Multiple hand-offs
• Unnecessary approvals
• Machine breakdowns
• Waste of sending bad product to customers
• Waste of providing bad service to customers
(Metcalf, 1997)5
The authors have found seven, eight, nine, or ten classes of waste shown in various
sources. The classes of wastes are also categorized for both production and
administrative activities.

VARIABLE RELATIONSHIPS/MULTI-VARI ANALYSIS
Variable Relationships*
Variable relationships is reviewed in the topic areas of:
• Multi-vari analysis
• Linear correlation and regression
Multi-Vari Analysis
In statistical process control, one tracks variables like customer complaints,
pressure, or temperature by taking measurements at certain intervals. The
underlying assumption is that the variables will have approximately one
representative value when measured. Frequently, this is not the case. Temperature
in the cross-section of a furnace will vary and the thickness of a [part may also vary
depending on where each measurement is taken.
Often, the variation is within piece, and the source of this variatioln is different from
piece-to-piece and time-to-time variation. The multi-vari chart is a very useful tool
for analyzing all three types of variation. Multi-vari charts are also used to
investigate the stability or consistency of a process. The chart consists of a series
of vertical lines, or other appropriate schematics, along a time scale. The length of
each line or schematic shape represents the range of values found in each sample
set.
Multi-Vari Sampling Plan Procedure
• Select the process and the characteristic to be investigated

• Select the sample size and time frequency
• Use a tabulation sheet to record the time and values from each sample set
• Plot the chart with time or sequence along the horizontal scale
• Plot the measured values on the vertical scale
• Join the observed values with appropriate lines
• Analyze the chart for within sample, between sample, and time variation
• Conduct additional studies to concentrate on problem are,a(s)
• After improvements, repeat the study to confirm the result:s
* Much of the material in this Section comes from Wortman (2001)9.

Multi-Vari Analysis (Continued)

In Figure 7.1 variation within samples (five locations across the width) is shown by
the line length. Variation from sample to sample is shown by the vertical positions
of the lines.
0.060
I
I I IIII
en
en 0.050
II III !
Cl)
~
c::
(J
.c::
I-
0.040
0.030
0.020
I I
0.010
5 10 15 20
Sample Number
Figure 7.1 An illustrative Multi-Vari Line Chart
To establish a multi-vari chart, a sample set is taken and plotted from the highest to
the lowest value. This variation may be represented by a vertical line or other
rational schematic. Figure 7.2 shows an injection molded plastic part. The
thickness is measured at four points across the width as indicated by arrows.
4
USL = 0.110"
LSL = 0.100"
Figure 7.2 Multi-Vari Measurements on a Plastic Piece
Three hypothetical cases (Figures 7.3,7.4, and 7.5) are presented to help understand
the interpretation of multi-vari charts for the part illustrated in Figure 7.2.

Multi-Vari Analysis (Continued)
USL 0.110" r-----.--t--t-r--.,-..,......-It-t-_+_---,
3
Ai m 0 .105" I-+--+-t-+-f--t-+-t-~_+_+-I_++-It-t-_+_H
LSL 0.100" L -_ _....1..-_ _----1_ _ _....L-._ _ ~
8:00 A.M . 10:00 A.M. Noon 2:00 P.M.
Figure 7.3 Excessive Variability Within Piece (part is t,apered)

USL 0.110"
Aim 0.105" I
1.~
LSL 0.100" L -_ _....1..-_ _ ----1_ _ _- L -_ _- - '
8:00 A.M. 10:00 A.M. Noon 2:00 P.M.
Figure 7.4 Less Variability Within Piece (center is thiicker)
USL 0.110" r - - - - . - - - - - - - r - - - - r - . - - r - + - + _ + _ .
Aim 0 .105"
4
3
2
LSL 0.100" 8;00 A.M. 10:00 A.M . Noon 2:00 P.M.
Figure 7.5 Variability Shift Over Time (part is getting larger)
VILLANOVA UNIVERSITY VII-8 LE.tIN SIX SIGMA HANDBOOK

Mult i-Vari Analysis (Continued)

Interpretation of the chart is apparent once the values are plotted. The advantages
of multi-vari charts are:
• It can dramatize the variation within the piece (positional)
• It can dramatize the variation from piece to piece (cyclical)
• It helps to track any time related changes (temporal)
• It helps minimize variation by identifying areas to look for excessive variation
• It also identifies areas not to look for excessive variation
Table 7.6 identifies the typical areas of time and locational variation.
Product/Process Under Consideration

Areas of Variation Piece Batch
Positional Within Piece Within Batch
Cyclical Piece-to-Piece Batch-to-Batch
Temporal OverTime OverTime
Table 7.6 Areas of Variation
Note that positional variation can often be broken into multiple components:
Cylinder End to end, out of round each end

Batch Top to bottom, side to side
Flat piece Across width, front to back
Table 7.7 Examples of Positional Components

Multi-Vari Case Study

A manufacturer produced flat sheets of aluminum on a hot rolling mill. Although a
finish trimming operation followed, the basic aluminum plcde thickness was
established during the rolling operation.
The thickness specification was 0.245" ±0.005". The operation had been producing
scrap. A process capability study indicated that the process spread was 0.0125" (a
Cp of 0.8) versus the requirement of 0.010".
The operation generated a profit of approximately $200,000 per month even after a
scrap loss of $20,000 per month. Refitting the mill with a more modern design,
featuring automatic gauge control and hydraulic roll bending, would cost $800,000
and result in 6 weeks of downtime for installation.
The department manager requested that a multi-vari study be conducted by a quality

engineer before further consideration of the new mill design or 4lther alternatives.
Four positional measurements were made at the corners of each flat sheet in order
to adequately determine within piece variation. Three flat sheets were measured in
consecutive order to determine piece to piece variation. Additionally, samples were
collected each hour to determine temporal variation.
The pictorial results follow in Figure 7.8.
0.250" r-------;------y------,----~~--=--..,--------,
. . - - - - Left
~~ Right
Front Back
Legend
0.240" L - -_ _- - - - '_ _ _--'----"'-----"_ - - L -_ _ _-'--_ _-----'
BA.M. 9A.M . 10A.M. 11 A.M. 12 A.M.
Time
Figure 7.8 Aluminum Flat Sheet Multi-Vari Chart

Multi-Vari Case Study (Continued)

The results of this short-term study were slightly better than the earlier process
capability study. The maximum detected variation was 0.010". Without
sophisticated analysis, it appeared that the time-to-time variation was the largest
culprit. A gross change was noted after the 10:00 AM break. During this time, the
roll coolant tank was refilled.
Actions taken over the next two weeks included re-Ieveling the bottom back-up roll
(approximately 30% of the total variation), initiating more frequent coolant tank
additions, followed by an automatic coolant make-up modification (SO% of the total
variation).
Additional spray nozzles were added to the roll stripper housings to reduce heat
build up in the work rolls during the rolling process (10% to 1S% of total variation).
The piece-to-piece variation was ignored. This dimensional variation may have
resulted from roll bearing slop or variation in incoming aluminum sheet temperature
(or a number of other sources). The results of this work are reflected in Table 7.9.
Variation % Variation Variation % Variation

Type Variation & Cause Correction Reduction
Time-to-Time SO% low coolant Automatic Nearly
level make-up SO%
Within Piece 30% Non-parallel Re-Ieveled Nearly
settings back-up roll 30%
Within Piece 10%-1S% Inadequate Added more 10%
roll cooling spray nozzles
Piece-to-Piece S%-10% Unknown None None
Table 7.9 Multi-Vari Case Study Results
The results from this single study indicated if all of the modifications were perfect,
the resulting measurement spread would be 0.002" total. In reality, the end result
was:t 0.002" or 0.004" total, under conditions similar to that of the initial study. The
total cash expenditure was $8,000 for the described modifications. All work was
completed in two weeks. The specification of 0.24S" ± O.OOS" was easily met.
Most multi-vari analyses do not yield results that are this spectacular, but the
potential for significant improvement is apparent.

VARIABLE RELATIONSHIPS/LINEAR CORRELATION & REGRESSION
A Simple Linear Model

To introduce the subject of regression analysis, consider the problem of predicting
the test results (Y) for students based upon an input variable (X), the amount of
preparation time in hours.
Hours of study and test results for ten randomly selected students
Student Study Time (Hours) Test Re!sults (%)
1 60 ~57
2 40 ~51
3 50 ~73
4 65 130
5 35 ~50
6 40 !55
7 50 152
8 30 !50
9 45 151
10 55 ~ro
Table 7.10 Study Time Versus Test Results
An initial approach to the analysis of the data in Table 7.10 is to plot the points on
a graph known as a scatter diagram. One will observe that y appears to increase as
x increases. One method of obtaining a prediction equation relating y to x is to place
a ruler on the graph and move it until it seems to pass through the majority of the
points, thus providing what is regarded as the "best fit" line.
81
>
~ 74
~
"'5
~ 67
0::
'ti 60
...
G)
53
30 35 40 45 50 55 60 65
Study Time (Hours), X
Figure 7.11 A Plot of Study Time Versus Test Results
VILLANOVA UNIVERSITY VII ·12 LEJIN SIX SIGMA HANDBOOK

A Simple Linear Model (Continued)

The mathematical equation of a straight line is:
Where Po is the y intercept when x = 0 and P1 is the slope of the line. Please note in
the previous chart that the x-axis does not go to zero so the y intercept appears too
high. The equation for a straight line in this example is too simplistic. There will
actually be a random error which is the difference between an observed value of y
and the mean value of y for a given value of x. For any given value of x, the
observed value of y varies in a random manner and possesses a normal probability
distribution. The concept is illustrated in the following diagram:
Figure 7.12 A Schematic Showing Variation in y as a Function ofx
The probabilistic model for any particular observed value of y is:
mean value of y fOr) ( )

y=( + random error
a given value ofx
VILLANOVA UNIVERSITY VII -13 LEAN SIX SIGMA HANDBOOK

The Method of Least Squares

The statistical procedure of finding the "best fitting" straight line is, in many
respects, a formalization of the procedure used to fit a line by eyl~. One is trying to
minimize the deviations of the points from the prospective line.
If the predicted value of y obtained from the fitted line is denoted as y, the
prediction equation is:
A A A
Y =130 + J31 x
Where: ~o and ~1 represent estimates of the true 130 and 131
81
-tn
~ 60
53
30 35 45 50 55 60 65
Study Time (Hours), X
Figure 7.13 A Plot of Study Time Versus Test Results
Having decided to minimize the deviation of the points in choosing the best fitting
line, one must now define what is meant by "best."

The Method of Least Squares (Continued)

The best fit criterion of goodness known as the principle of least squares is
employed:
Choose, as the best fitting line, the line that minimizes the sum of squares of
the deviations of the observed values of y from those predicted.
Mathematically, one wishes to minimize the sum of squared errors given by:
SSE = ~ Yi - Yi
n ( A )2
I'" 1
Substituting forYi one obtains the following expression:
Sum of squared errors =
The least square estimators of Po and P1 are:

A - A-
and Po = Y- P1 X
where:

Least Squares Example
Once 130 and 131 have been computed, their values are substituted into the equation
of a line to obtain the least squares prediction equation, or regression line.
Example 7.1: Obtain the least squares prediction line for the da"ta below:
Xi Yi x-2I X1Yi y~
60 67 3,600 4,020 4,489
40 61 1,600 2,440 3,721
50 73 2,500 3,650 5,329
65 80 4,225 5,200 6,400
35 60 1,225 2,100 3,600
40 55 1,600 2,200 3,025
50 62 2,500 3,100 3,844
30 50 900 1,500 2,500
45 61 2,025 2,745 3,721
55 70 3,025 3,850 4,900
Sum 470 639 23,200 30,805 41,529
Table 7.14 Data Table for the Study TimelTest Score Example
S
x2
=L.J ~x' _(t,xJ =
i=1
I
n
23200 _ (470)'
, 10
=1110
'
S = ~ x.y.
n
- 1-1
(
1- 1
~XI
n
~ Yi )
)( n
= 30 805 - (
470 639
)( ):: 772
xy ~ II n ' 10
s" =fy;'
i=1
-(t,y,J
n
=41,529 _ (639)' =696.9
10
- 639 - 470
y= - =63.9 x= - =47
10 10
VILLANOVA UNIVERSITY VII - 16 LE.tIN SIX SIGMA HANDBOOK

Least Squares Example (Continued)
Example 7.1 continued:
~1 = Sxy = 772 =0.6955

52
x
1,110
~o=y - ~1i = 63.9 - (0.6955)(47) =31.2115

Y =31.2115 + 0.6955x
A.
One may now predict y for a given value of x by substitution into the prediction
equation. For example, if 60 hours of study time are allocated, the predicted test
score would be:
A
Y = 31.2115 + 0.6955x
y = 31.2115 + (0.6955)(60)
Y= 72.9415 = 73%
Hints on Regression Analysis

• Be careful of rounding errors. The calculator should carry a minimum of six
significant figures in computing sums of squares of deviations. Note, the
prior example consisted of convenient whole numbers which does not occur
often.
• Always plot the data points and graph the least squares line. If the line does
not provide a reasonable fit to the data points, there may be a calculation
error.
• Projecting a regression line outside of the test area can be risky. The above
equation suggests a student without studying, would score 31 % on the test.
The odds favor 25%, if answer a is selected for all questions. The equation
also suggests the student with 100 hours of study, should attain 100% on the
examination - which is highly unlikely.

Calculating &2e , an Estimator of 0 e2

Recall, the model for y assumes that y is related to x by the equation:
If the least squares line is used:
A random error E enters into the calculations of 130 and 131' The random errors affect
the error of prediction. Consequently, the variability ofthe random errors (measured
by a 2e ) plays an important role when predicting by the least squares line.
The first step toward acquiring a boundary on a prediction error requires that one
estimate a 2 • It is reasonable to use SSE (sum of squares for error) based on (n-2)
degrees otfreedom, one for each variable (x and y).
2
n 2 S )
=~(y. - y.) =Sy2 =Sy2
A A
(
SSE ~ I I
- R. S
f-'1 xy
- Sxy
i =1 x2
Sxy and Sx2 were previously defined

-2 SSE
then: a = --
e n-2
Confidence Interval of (1-a) for 131

Formula:
Example 7.2: Substitute previous data into the above formula to obtain the
confidence interval around the slope of the line at a 95% confidence.
Ii
1
=0.6955 ± 2.306 .J1,4.47
110
= 0.6955 ± 0.3094
Intervals constructed by this procedure will enclose the true value of 131 95% of the
time. Hence, for each 10 hours of increased study, the expected increase in test
scores is in the interval of 3.86 to 10.05 percentage points.

Correlation Coefficient (r)

An indicator of the strength of the linear relationship between two variables y and
x is called the Pearson product-moment coefficient of correlation, where:
• r lies in the interval of -1 .:5 r .:5 1.
• The numerator used in calculating r is identical to the numerator of the

formula for the slope ~1' Thus, the coefficient of correlation r will assume
exactly the same sign as ~1 and will equal zero when ~1 = 0.
• A positive value for r implies that the line slopes upward to the right. A
negative value indicates that it slopes downward to the right.
• =
When r 0, the data points are scattered and give no evidence of a linear
relationship. When r = 0, it implies no linear correlation, not simply "no
correlation." A pronounced curvilinear pattern may exist.
• When r =1 or r =-1, all pOints fall on a straight line and SSE equals zero.
• Any other value of r suggests the degree to which the points tend to be
linearly related.
Example 7.3: Using the study time and test score data reviewed earlier in Example
7.1, determine the correlation coefficient.
Given: 5xy =772 x =1,110

52 5y'- = 696.9
772
Solution: r =.j(1,110)(696.9) =0.878

Coefficient of Determination (r2)

The coefficient of determination, also called the square of the- linear correlation
coefficient, is:
S 2 - SSE SSE
R2 r2 = =
y 1 - --
S2y
= S2y
Example 7.4: Using the data from Example 7.3, determine r2.
r2 =(0.878)2 =0.771
One can say that 77% of the variation in test scores can be explained by variation in
study hours.
• r2 lies in the interval of O.s r2 .s 1.
Correlation Example
25
24
23
22
21
MPG
20 . .. . ~? . .. ~ Average
1" 20 MPG
19
18
Dg
17
dg
16 . . . . . . . *""---"''-
~ ,,e...<--r-------,.,-----..,.---
2000 3000 4000
Car Weight
Figure 7.15 Correlation Plot of Car Weight and MPG
VILLANOVA UNIVERSITY VII- 20 LEJIN SIX SIGMA HANDBOOK

Correlation Example (Continued)

SST = L D~ + D~ + ... + D~
SSE = L d~ + d~ + ... + d~
r2 =1 _ SSE = SST - SSE
SST SST
Where SST = total sum of squares (from the experimental average) and SSE = total
sum of squared errors (from the best fit). When SSE is zero, r2 equals one and when
SSE equals SST, then r2 equals zero.
Multiple Linear Regression

Multiple linear regression is an extension of the methodology for linear regression
to more than one independent variable. By including more than one independent
variable, a higher proportion of the variation in y may be explained. A full
explanation of multiple linear regression is left to the resources ofthe readers ofthis
Handbook. A few models and definitions are presented:
First Order Linear Model
Second Order Linear Model (Two Predictor Variables)
S2 = SSE
n- (k + 1)
Source DF SS MS
Regression k SSR MSR =SSRlk
Error n-(k+1) SSE MSE =SSE/[n-(k+1)]
Total n-1 Total SS
Table 7.16 ANOVA Table for Multiple Regression Analysis
Note: k = the number of predictor variables

p-Value
The traditional approach to hypothesis testing compares a test statistic to a pre-
determined critical probability value known as alpha (0). In m()st textbooks, the
statistical tables for critical 0 values are set at 1% or 5% (t distribution, X2
distribution, and F distribution.) The use of table values avoids the necessity of
performing manual calculations to determine the exact probabilities.
Triola (1994)8 provides a definition of the p-value:
"A p-value is the probability of getting a value of the sample b~st statistic that is
at least as extreme as the one found from the sample data (clssuming that the
hypothesized value is correct)."
In other words, a small p-value is an indication that the null hypothesis is false.
Various statistical software tools such as Microsoft Excel (Microsoft Corporation),
MINITAB (Minitab, Inc.), S-PLUS (MathSoft, Inc.), and SPSS (SPSS, Inc.) provide the
exact "probability due to chance" ofthe hypothesis or claim. Table! 7.17 summarizes
the logic behind the difference of either 0 or p-values being due to chance.
o or p-value Remarks
p>5% Significant difference is not provjm
1% < p~5% A statistically significant differenlce
p~ 1% A highly significant statistical differlence
Table 7.17 Significance of 0 or p-Values

(Langley, 1971)\ (Triola, 1994)8
An example of regression analysis for study time and test results was shown earlier
using manual calculations. This same data can be analyzed usin!9 MINITAB, Excel,
or other statistical analysis programs.
Student 1 2 3 4 5 6 7 8 9 10
Study Time (hrs) 60 40 50 65 35 40 50 30 45 55
Test Results (%) 67 61 73 80 60 55 62 50 61 70
Table 7.18 Study Time vs. Test Results
VILLANOVA UNIVERSITY VII - 22 LEJIN SIX SIGMA HANDBOOK

MINITAB Results
A MINITAB output is displayed below. The regression equation is provided. The
coefficients of the line, coefficient of determination, and an ANOVA have also been
calculated. In addition, the probability of significance for the study time and the
ANOVA have been determined.
Regression Analysis: Test Results Versus Study Time

The regression equation is: results =31.2 + 0.695 study time hours
Predictor Coef SE Coef t p
Constant 31.212 6.465 4.83 Q.OO!)
Study time 0.6955 0.1342 5.18 0.001
S =4.472 R-Sq =77.0% R-Sq(adj) =74.2%
Analysis of Variance
Source OF SS MS F P
Regression 1 536.92 536.92 26.85 CO.OO~t
Residual Error 8 159.98 20.00
Total 9 696.90
Table 7.19 MINITAB Regression Analysis Output Example
The study time coefficient is the slope, 1311 with a t statistic of 5.18. The critical value
is not provided, but the p-value is provided. The p-value of 0.001 indicates that the
study time coefficient is a highly statistical significant factor. It is a very extreme
probability number. Therefore, the regression equation is valid and the student is
reminded to study long and hard.
The R-Sq value of 77.0% is the coefficient of determination (0.77). The square root
of the coefficient of determination is the correlation coefficient. That value would
be 0.8775. (Note: some rounding errors occur due to differences between manual
and computerized calculations.) The t test is used to determine if there is a
significant level for the correlation coefficient. In the MINITAB display, the p-value
is also 0.001 for this determination.
Similar results may be obtained using Excel.

HYPOTHESIS TESTING/BASIC CONCEPTS
Hypothesis Testing
Hypothesis Testing is reviewed in the following topic areas:
• Basic concepts • Variance tests

• Point and interval estimation • Analysis of variancf~
• Means tests
Basic Concepts
Before discussing these specific procedures, it is worthwhile to review a number of
commonly used hypothesis test terms.
Null Hypothesis
This is the hypothesis to be tested. The null hypothesis direct:ly stems from the
problem statement and is denoted as Ho. Examples:
• If one is investigating whether a modified seed will result in a different

yield/acre, the null hypothesis (two-tail) would assume the yields to be the
same Ho: Va = Vb.
• If a strong claim is made that the average of process A is greater than the
average of process B, the null hypothesis (one-tail) would state that process
A ~ process B. This is written as Ho: A ~ B.
The procedure employed in testing a hypothesis is strikingly similar to a court trial.

The hypothesis is that the defendant is presumed not guilty until proven guilty.
However, the term innocent does not apply to a null hypothesis. A null hypothesis
can only be rejected. or fail to be rejected. it cannot be accepted because of a lack
of evidence to reject it. If the means of two populations are different, the null
hypothesis of equality can be rejected if enough data is collected. When rejecting
the null hypothesis, the alternate hypothesis must be accepted.
Test Statistic
In order to test a null hypothesis, a test calculation must be lTIade from sample
information. This calculated value is called a test statistic and is compared to an
appropriate critical value. A decision can then be made to reject 40r fail to reject the
null hypothesis.

Basic Concepts (Continued)

Types of Errors
When formulating a conclusion regarding a population based on observations from

a small sample, two types of errors are possible:
• Type I error: This error occurs when the null hypothesis is rejected when it
is, in fact, true. The probability of making a type I error is called a (alpha) and
is commonly referred to as the producer's risk (in sampling). Examples are:
incoming products are good but called bad; a process change is thought to
be different when, in fact, there is no difference.
• Type II error: This error occurs when the null hypothesis is not rejected when
it should be rejected. This error is called the consumer's risk (in sampling)
and is denoted by the symbol 13 (beta). Examples are: incoming products are
bad, but called good; an adverse process change has occurred but is thought
to be no different.
The degree of risk (a) is normally chosen by the concerned parties (a is normally
taken as 5%) in arriving at the critical value of the test statistic. The assumption is
that a small value for a is desirable. Unfortunately, a small a risk increases the 13
risk. For a fixed sample size, a and 13 are inversely related. Increasing the sample
size can reduce both the a and 13 risks.
The types of errors are shown in Figure 7.20 below:
Null Hypothesis
True False
Fail to p=1-a p=13

Reject Correct Type II Error
The Ho Decision
Decision
Made Reject
Ho p=a p=1-13
Type I Correct
Error Decision
Figure 7.20 Error Matrix


One-Tail Test vs. Two-Tail Test
Any test of hypothesis has a risk associated with it and one is generally concerned
with the a risk (a type I error which rejects the null hypothesis when it is true). The
level of this a risk determines the level of confidence (1 - a) that one has in the
conclusion. This risk factor is used to determine the critical value of the test
statistic which is compared to a calculated value.
One-Tail Test
If a null hypothesis is established to test whether a sample value is smaller or larger

than a population value, then the entire a risk is placed on one end of a distribution
curve. This constitutes a one-tail test.
• A study was conducted to determine if the mean battery life produced by a

new method is greater than the present battery life of 35 hours. In this case,
the entire a risk will be placed on the right tail of the existing life distribution
curve.
=
Ho: new < or to present H1 : new> preselnt
.------
ENTIRE a =5%
o
IJo =35 HOURS
Figure 7.21 Determine If the True Mean is Within the a Cri1tical Region
• A chemist is studying the vitamin levels in a brand of cereal to determine ifthe

process level has fallen below 20% of the minimum daily requirement. It is the
manufacturer's intent to never average below the 20% level. A one-tail test
would be applied in this case, with the entire a risk on the left tail.


Ho: level> or =20% H1 : level < 20%
=5%
--------.
ENTIRE a
o
loLo =20%
Figure 7.22 Determine If the True Mean is Within the a Critical Region
Two-Tail Test
If a null hypothesis is established to test whether a population shift has occurred,
in either direction, then a two-tail test is required. The allowable a error is generally
divided into two equal parts. Examples:
• An economist must determine if unemployment levels have changed

significantly over the past year.
• A study is made to determine if the salary levels of company A differ

significantly from those of company B.
Ho: levels are = H1 : levels are *
-1.96 o +1.96
~o
Figure 7.23 Determine If the True Mean is Within

Either the Upper or Lower a Critical Regions

Practical Significance vs. Statistical Significance

The hypothesis is tested to determine if a claim has significant statistical merit.
Traditionally, levels of 5% or 1% are used for the critical significance values. If the
calculated test statistic has a p-value below the critical level then it is deemed to be
statistically significant. More stringent critical values may be required when human
injury or catastrophic loss is involved. Less stringent criti<:al values may be
advantageous when there are no such risks and the potential economic gain is high.
On occasion, an issue of practical versus statistical significance may arise. That is,
some hypothesis or claim is found to be statistically significaillt, but may not be
worth the effort or expense to implement. This could occur if al large sample was
tested to a certain value, such as a diet that results in a net loss of 0.5 pounds for
10,000 people. The result is statistically significant, but a diet los,ing 0.5 pounds per
person would not have any practical significance. (Triola, 1994)8
Huck (1996)1 indicates that issues of practical significance will often occur if the
sample size is not adequate. A power analysis may be needed to aid in the decision-
making process.

Required Sample Size

In the statistical inference discussion thus far, it has been assumed that the sample
size (n) for hypothesis testing has been given and that the critical value of the test
statistic will be determined based on the a error that can be tolerated. The ideal
procedure, however, is to determine the a and 13 error desired and then to calculate
the sample size necessary to obtain the desired decision confidence.
The sample size (n) needed for hypothesis testing depends on:
• The desired type I (a) and type II (13) risk

• The minimum value to be detected between the population means (1-1 - 1-10)
• The variation in the characteristic being measured (S or a)
Variable data sample size, only using a, is illustrated by the following:
Example 7.5: Assume in a pilot process we want to determine whether an

operational adjustment will alter the process hourly mean yield by as much as 4 tons
per hour. What is the minimum sample size which, at the 95% confidence level
(Z=1.96), would confirm the significance of a mean shift greater than 4 tons per
hour? Historic information suggests that the standard deviation of the hourly output
is 20 tons. The general sample size equation for variable data (normal distribution)
is:
Obtain 96 pilot hourly yield values and determine the hourly average. If this mean
deviates by more than 4 tons from the previous hourly average, a significant change
at the 95% confidence level has occurred. If the sample mean deviates by less than
4 tons/hr., the observable mean shift can be explained by chance cause.
For binomial data, use the following formula:
n = Z2 (p) (1 - p)
(Ap)2
Where,
Z = The appropriate Z value Ii = Proportion rate

=
Ap The desired proportion interval n = Sample size

Hypotheses Tests
-
Z= x-~
o/yfi
Z=~-~
0-
x.-x.
-
Noona' ~
~----~ x'~ =(n - 1)S2
r - - - - - - ' ' ' " - - - , ~--.... 02
Small
-
t= x-~
samples _s/..jfi
S2 = (n1 - 1)sf -+ (n 2- 1)51

~ + r12- 2
S-X-x--Sfrr1
1
-2
-n n + .-
1 2
t = X1 -X2
S-x,-X-
2
A= Sfl n 1
1 - - -..... B=S}ln 2
Welch-5atterthwaile
Approximation
~ df= (A+B)2
A2
-+-
~-1
82
n2 -1
Figure 7.24 Schematic of Normal Distribution Hypotheses

Hypotheses Tests (Continued)
-
I I
,"'" ~ p(~-P) ,"'" z = .Q..:...Q.
pvs.1.l.
) , I op = op
I Binomial I ,""
I
, f
= ~ p(1 -p)(
Z=P,- P2
I P1 VS ,P2
~
Op,-p, 1- 1-)
n,
+
n2
,"'" op _p
1 2
-
P=
n, P, + n2P2 ~
n, + n2
cvs.1.l. 0= r=c Z =C - C
.;c
Poisson
c =no. of defects
cvs. c
k =no. samples
Figure 7.25 Schematic of Attribute Distribution Hypotheses
The charts in Figures 7.24 and 7.25 were provided by DuWayne Carlson.

HYPOTHESIS TESTING/POINT & INTERVAL ESTIMATION
Point Estimate for Population Mean

In analyzing sample values to arrive at population probabilities, two major estimators
are used: point estimates and confidence intervals.
A point estimate of the population mean, ~, is the sample mean, X.
Example 7.6: Given the following values for four customer calli wait times: 28.7,
27.9,29.2, and 26.5 seconds, calculate the point estimation of thE! population mean.
n
LXi
28.7 + 27.9 + 29.2 + 26.5 2808 d
J..I.:= X =-
' -=
"=1
n 4
= . secon s
28.08 seconds is the point estimate for the population me,an.
Point Estimate for Population Varianlce

The sample variance, S2, is the best point estimate of the popul,ation variance, a 2 •
The sample standard deviation, s, is the best point estimate of the population
standard deviation, a.
n 2 n
L(~ -X) L(Xi - J..I.)2

s2 = ~i=....:.1_ _ _ (52 = =.....:1_ __
..:....i
n -1 N
s= (5 = =.....:.1_
..:....i _ _
n -1 N

Confidence Interval for the Mean

Continuous Data - a Known
The confidence interval (CI) or interval estimate of the population mean, J,I, when the
population standard deviation, a, is known, is calculated using the sample mean, X,
the population standard deviation, a, the sample size, n, and the normal distribution.
From sample data, one can calculate the interval within which the population mean,
J,I, is predicted to fall. Confidence intervals are always estimated for population
parameters. A confidence interval is a two-tail event and requires critical values
based on an aiphaJ2 risk in each tail. The central limit theorem term, oJ,,", is
necessary because the confidence interval is for a population mean and not
individual values.
Example 7.7: The average of 100 samples is 18 with a population standard deviation
of 6. Calculate the 95% confidence interval for the population mean.
x -Zal2 :fn
6
18 -1.96--
../100
16.82 ~ Jl ~ 19.18
Continuous Data - a Unknown
The confidence interval of the population mean, J,I, when the population standard
deviation, a, is unknown, is calculated using the sample mean, X, the sample
standard deviation, s, the sample size, n, and the t distribution. If a relatively small
sample is used, e.g. n < 30, then the t distribution must be used. When the sample
size is large, e.g. n > 30, the Z distribution may be used in place of the t distribution.

Confidence Interval for the Mean (Continued)

Continuous Data - a Unknown (Continued)
Example 7.8: The average of 25 samples is 18 with a sample standard deviation of

6. Calculate the 95% confidence interval for the population mean.
s s
X - taJ2 n-1 C ~ J.1 ~ X + taJ2 n-1 C
, "n , "n
6 6
18-2.064- ~ J.1 ~ 18 + 2.064 ~
.J25 ,,25
15.52 ~ J.1 ~ 20.48
Confidence Interval for Proportion

For large sample sizes, with np and n(1-p) greater than or equal to 5, the binomial
distribution can be approximated by the normal distributiclIn to calculate a
confidence interval for population proportion. The following formula is used:
Ps
- ZaJ2 1 P, (1 - P.),; p ,; Ps
+z
aJ2
1P, (1 - P.)
n n
Where: Ps =sample proportion, p =population proportion, 111 = sample size
Note that other confidence interval formulas exist. These include percent
nonconforming, Poisson distribution data, and very small sample size data.
Example 7.9: If 16 defectives were found in a sample size of 200 units, calculate the
90% confidence interval for the proportion.
Ps -zaJ2 t·(1 -P.)

n
<
-
<s
p - P +ZaJ2 1P. (1 -P.)
n
Ps = nx = 200
16
= 0.08
r----:----:-
0.08 _1.645,,/°·08(1 - 0.08) ~ p ~ 0.08 + 1.645,,/0.08(1 - (1.08)

v 200 ~ 200
0.048 ~ P ~ 0.112

Confidence Interval for Variance

The confidence interval or interval estimate for the population variance, a 2 , is given
by:
(n - 1) s~
X~2, n - 1
Where: S2 =sample variance

n = sample size
n - 1 = degrees of freedom
X~2, n-1 and X~~2, n-1 = table values for (n - 1) degrees of freedom
The confidence interval for the mean were symmetrical about the average. This is
not true for the variance, since it is based on the chi-square distribution.
Example 7.10: The sample variance for a set of 25 samples was found to be 36.
Calculate the 90% confidence interval for the population variance.
(n - 1) s~ 2 (n - 1) s~
2 ~ (1 ~
X al2, n - 1 x~ -al2, n - 1
(25 -1)36 (25 -1)36

36.42 13.85
23. 72 ~ (12 ~ 62.38
Confidence Interval for Standard Deviation

The confidence interval for the population standard deviation, a, is given by:
(n -1)s~
X21-al2,n-1

HYPOTHESIS TESTING/MEANS TESTS
Hypothesis Tests for Means

Z Test
When the population follows a normal distribution and the pc~pulation standard
deviation, ax, is known, then the hypothesis tests for comparing a population mean,
J,I, with a fixed value, J,lo' are given by the following:
The null hypothesis is denoted by Ho and the alternative hypothesis is denoted by

H1 • The test statistic is given by:
Z = X - J.10 _ X - r-o
II
- Gx/Jn
... where the sample average is X, the number of samples is n and the standard
deviation of means is ax. If n > 30, the sample standard deviation, s, is often used
as an estimate of the population standard deviation, ax. The \test statistic, Z, is
compared with a critical value, Za or Za/2' which is based on a siSlnificance level, a,
for a one-tailed test or a/2 for a two-tailed test. If the H1 sign is :f::, it is a two-tailed
test. If the H1 sign is >, it is a right one-tailed test, and if the H1 sign is <, it is a left,
one-tailed test. (Triola, 1994)8
Example 7.11: Suppose that the average time that a person keeps a bank account
is 5.00 years, with a standard deviation of 0.12 years. A new type of account was
offered and a sample found the following durations for how long people kept this
type of account: 5.10, 4.90, 4.92, 4.87, 5.09, 4.89, 4.95 and 4.88 ytears.
Can one state, with 95% confidence, that the new accounts am being kept for a
shorter period than the original type of accounts? This question involves an
inference about a population mean with a known sigma. The Z test applies. The null
and alternative hypotheses are:
or
Ho: J,I ~ 5.00 years H1 : J,I < 5.00 years

Hypothesis Tests for Means (Continued)

Z Test (Continued)
The sample average is X = 4.95 years with n = 8 and the population standard
=
deviation is ax 0.12 years. The test statistic is:
Z = X - ).10 = 4.95 - 5.00 = -0.05 =-1 18

C:Sx/Jii 0.12/ .J8 0.042 .
Since the H1 sign is <, it is a left one-tailed test and with a 95% confidence, the level
of significance, a = 1 - 0.95 = 0.05. Looking up the critical value in a normal
distribution or Z table, one finds ZO.05 = -1.645. Since the test statistic, -1.18, does not
fall in the reject (or critical) region, the null hypothesis cannot be rejected. There is
insufficient evidence to conclude, with 95% confidence, that the new accounts are
kept for a shorter period than the original accounts.
If the test statistic had been, for example -1.85, one would have rejected the null
hypothesis and concluded, with 95% confidence, that the new accounts were being
kept for a shorter period than the original accounts.
Student's t Test
The student's t distribution applies to samples drawn from a normally distributed

population. It is used for making inferences about a population mean when the
population variance a 2 is unknown and the sample size n is small. The use of the
t distribution is never wrong for any sample size. However, a sample size of 30 is
normally the crossover point between the t and Z tests. The test statistic formula is:
t = X - ).10
s/Jii
Where: X =The sample mean
J,lo =The target value or population mean
=
s The sample standard deviation
n = The number of test samples
The null and alternative hypotheses are the same as were given for the Z test.


Student's t Test (Continued)
The test statistic, t, is compared with a critical value, ta or ta/2' which is based on a
significance level, a, for a one-tailed test or a/2 for a two-tailed test, and the number
of degrees of freedom, d.f. The degrees of freedom is determined by the number of
samples, n, and is simply:
d.f. = n -1
Example 7.12: The average daily revenues of a hospital has been $880,000 (101 = 880).
A new marketing program has been evaluated for 25 days (n = 25) with a yield of
$900,000 (X) and sample standard deviation, s = $20,000. Can one say, with 95%
confidence, that the revenues have changed?
The null and alternative hypotheses are:
or
Ho: 101 =880
The test statistic calculation is:
t = X - J.Lo = 900 - 880 =20 =5

s/.Jn 201J25 4
Since the H1 sign is '1:, it is a two-tailed test and with a 95% confidence, the level of
significance, a = 1 - 0.95 = 0.05. Since it is a two-tail test, a/2 is used to determine
= =
the critical values. The degrees of freedom, dJ. n - 1 24. Looking up the critical
= =
values in a t distribution table, yields t o.025 -2.064 and t o.975 2.064. Since the test
statistic, 5, falls in the right-hand reject (or critical) region, the null hypothesis is
rejected. One concludes, with 95% confidence, that the revenues have changed.
Note that a t distribution table is shown in the Appendix.
This technique was developed byW.S. Gosset and published in 1908 under the pen
name "Student." Gosset referred to the quantity under study ciS t. The test has
since been known as the student's t test.

HYPOTHESIS TESTINGIMEANS TESTS

Student's t Test (Continued)
Example 7.13: A new spark plug design is tested for wear. A sample of six plugs
yielded: 0.0058, 0.0049, 0.0052, 0.0044, 0.0050 and 0.0047 inches of wear. The
current design has historically produced an average wear of 0.0055 11 • With 95%
confidence, is the new design better?
Example 7.14: A very expensive experiment has been conducted to evaluate the
manufacture of synthetic diamonds by a new technique. Five diamonds have been
generated with recorded weights of 0.46, 0.61, 0.52, 0.57 and 0.54 carats. An average
diamond weight greater than 0.50 carats must be realized for the venture to be
profitable, assuming 95% confidence, what is the recommendation?
Conclusions
STEPS EXAMPLE 7.13 EXAMPLE 7.14
Step 1: Establish the null
hypothesis: [there is no Ho: 1-11 ~ 0.0055" Ho: 1-12 ~ 0.50
difference between the target
value and the sample average] H1: 1-11 < 0.0055" H1: 1-12 > 0.50
Step 2: Determine the critical DF = 5 from (n - 1) DF = 4 from (n -1)

value of t for a 95% confidence Left Tail Right Tail
level from the t distribution
-2.015 2.132
Step 3: Calculate the t
0.0050 - 0.0055 0.054 - 0.050
statistic: t= t=
- 0.0004S/J6 0.056/J5
t=X-~o
t = -2.551 t = 1.597
s/.Jn
Step 4: Can one reject the null Since the value of Since the value of
hypothesis? calculated t is to the left of calculated t (1.597) is not
-2.015, the null hypothesis to the right of the critical t
is rejected. The wear is (2.132), the null hypothesis
less for the new plug can't be rejected.
design. Insufficient evidence
exists for the new
technique to be profitable.
Table 7.26 A Matrix Review of Two Student's t Tests

One underlying assumption is that the sampled population has a normal probability
distribution. This is a restrictive assumption since the distribution of the sample is
unknown. The t distribution works well, however, for distributions that are bell
shaped.

Paired-Comparison Hypothesis Tests

Paired t Test
The paired t test is used to test the difference between 2 samplle means. Data is
taken in pairs with the difference calculated for each pair.
Example 7.15: Calculate a paired t test for two operators that measure the same set
of samples.
Sample Operator 1 Operator 2 Difference (d)

1 3.125 3.110 0.01S
2 3.120 3.095 0.02S
3 3.135 3.115 0.020
4 3.130 3.120 0.010
5 3.125 3.125 0
=
d.f. n -1 4 = a =0.014 *
=
n number of sample pairs
a
* The sign of must be maintained Sd =0.0096
t = d = 0.014 = 0.014 =3.261
Sd/Jn 0.0096/-15 0.0043
= =
The critical value for a two-sided test with a 0.05, and d.f. 4 is t o.025,4 2.776. =
Since the calculated test statistic is in the critical area, the null hypothesis Ho is
rejected, and we conclude there is a difference in the population means.
The paired method (dependent samples), compared to treatin~1 the data as two
independent samples, will often show a more significant diffenmce because the
standard deviation of the d's (Sd) includes no sample to samplle variation. This
comparatively more frequent significance occurs despite the fact that "n - 1"
represents fewer degrees of freedom than "n1 + n2 - 2." In geneml, the paired t test
is a more sensitive test than the comparison of two independent samples.

HYPOTHESIS TESTINGIMEANS TESTS
Hypothesis Tests for Proportions

p Test
When testing a claim about a population proportion, and there is a fixed number of
independent trials with constant probabilities, and each trial has two outcome
possibilities (a binomial experiment), one may use a p test. When np < 5 or n(1-
p) < 5, the binomial distribution is used to test hypotheses relating to proportion.
If conditions that np ~ 5 and n(1-p) ~ 5 are met, then the binomial distribution of
sample proportions can be approximated by a normal distribution. The hypothesis
tests for comparing a sample proportion, p, with a fixed value, Po, are given by the
following:
Ho: P =Po
H1 : p < Po

H1• The test statistic is given by:
z= x - npo
~npo (1-po)
Where the number of successes is x and the number of samples is n. The test
statistic, Z, is compared with a critical value, Za or Zal2' which is based on a
significance level, a, for a one-tailed test or a/2 for a two-tailed test. If the H1 sign
is -1=, it is a two-tailed test. If the H1 sign is >, it is a right, one-tailed test, and if the H1
sign is <, it is a left, one-tailed test. (Triola, 1994t

HYPOTHESIS TESTINGNARIANCE TESTS
Chi-Square (X2) Test

Earlier in this Section it was shown that standard deviation (or variance) is
fundamental in making inferences regarding the population mean. In many practical
situations, variance (a 2 ) assumes a position of greater iml)ortance than the
population mean. Consider the following examples:
1. A pharmaceutical manufacturer wishes to develop a new drug with a more

consistent effect on patients. The variation in patient effects in the new drug
must be less than the variation in the existing drugs.
2. An insurance company wishes to compare the risks of having an accident

among several categories of people.
3. Several inspectors examine finished parts at the end of a manufacturing

process. Even when the same lots are examined by differlent inspectors, the
number of defectives varies. Their supervisor wants to know if there is a
significant difference in the knowledge or abilities of the inspectors.
All of the above problems represent a comparison of a target or population variance

with an observed sample variance, a comparison between several sample variances,
or a comparison between frequency proportions. The standardi;~ed test statistic is
called the chi-square (X2) test.
Population variances are distributed according to the chi-square distribution.

Therefore, inferences about a single population variance will be based on chi-
square.
The chi-square test is widely used in two applications.
Case I. Comparing variances when the variance of the poplulation is known.
Case II. Comparing observed and expected frequencies of test outcomes when
there is no defined population variance (attribute data).

Chi-Square (X2) Test (Continued)

When the population follows a normal distribution, the hypothesis tests for
comparing a population variance, 0 2x' with a fixed value, a2 0' are given by the
following:

H1• The test statistic is given by:
Where the number of samples is n and the sample variance is a2. x

The test statistic,
2
x , is compared with a critical value X2a or X2a/2 which is based on a significance level,
a, for a one-tailed test or a/2 for a two-tailed test and the number of degrees of
freedom, dJ. The degrees of freedom is determined by the number of samples, n,
and is simply:
d.f. = n -1
If the H1 sign is '1:, it is a two-tailed test. If the H1 sign is >, it is a right, one-tailed test,
and if the H1 sign is <, it is a left, one-tailed test. {Triola, 1994)8
The X2 distribution looks like:
Left tail Right tail
o 2
X 1-0 o
Figure 7.27 Chi-square Distribution Tail Areas


Unlike the Z and t distributions, the tails of the chi-square distribution are non-
symmetrical. Note that a chi-square distribution table is shown in the Appendix.
Chi-Square Case I. Comparing Variances When

the Variance of the Population is Known.
Example 7.16: The R&D department of a steel plant has tried to develop a new steel
alloy with less tensile variability. The R&D department claims that the new material
will show a four sigma tensile variation less than or equal to 60 psi 95% of the time.
An eight sample test yielded a standard deviation of 8 psi. Can a reduction in tensile
strength variation be validated with 95% confidence?
Solution: The best range of variation expected is 60 psi. This tralnslates to a sigma
of 15 psi (an approximate 4 sigma spread covering 95.44% of occurrences).
The null hypothesis is: or Ho: a x--

2
::. (15)2
2
The alternative hypothesis is: H:a2<a
1 x 0
or
From the chi-square table: Because alternative hypothesis is H1: .,.2x < a 20 , this is a left
tail test with n - 1 = 7. The critical value for 95% confidence is 2.17. That is, the
calculated value will be less than 2.17 5% of the time. Please nc)te that if we were
looking for more variability in the process a right tail rejection region would have
been selected and the critical value would be 14.07.
The calculated statistic is:
Since 1.99 is less than 2.17, the null hypothesis must be rejected. The decreased
variation in the new steel alloy tensile strength supports the R&D claim.


Chi-Square Case II. Test of Independence by Comparing Observed
and Expected Frequencies of Test Outcomes. (Attribute Data)
It is often necessary to compare proportions representing various process

conditions. Machines may be compared as to their ability to produce precise parts.
The ability of inspectors to identify defective products can be evaluated. This
application of chi-square is called the contingency table or row and column analysis.
The procedure is as follows:
1. The null hypothesis of no difference in the proportions among n populations

is:
The alternative hypothesis that not all n populations are equal is:
H1 : not all Pi are equal for i = 1, 2, ... , n
2. Take one subgroup from each of the various processes and determine the
observed frequencies (0) for the various conditions being compared.
3. Calculate for each condition the expected frequencies (E) under the
assumption that no differences exist among the processes.
4. Compare the observed (0) and expected (E) frequencies to obtain "reality."
The chi-square test, which is the most "famous" chi-square statistic, is totaled
over all of the process conditions:
5. A critical value is determined using the chi-square table with the entire level
of significance, 0, in the one-tail, right side, of the distribution. The degrees
of freedom is determined from the calculation (r-1)(c-1), which is (the number
of rows - 1)(times the number of columns - 1).
6. A comparison between the test statistic and the critical value confirms if a
significant difference exists, at a selected confidence level. (NIST, 2007t

Example 7.17: An airport authority wanted to evaluate the ability of three X-ray
inspectors to detect key items. A test was devised whereby prohibited knives were
placed in ninety pieces of luggage. Each inspector was exposed to exactly thirty
pieces of the forbidden luggage, in a random fashion. At a 95% c:onfidence, is there
any significant difference in the abilities of the inspectors?
Inspectors Treatment
Observed Values
1 2 3 Totals
Knives detected 27 25 22 74
Knives undetected 3 5 8 16
Sample total 30 30 30 90
Table 7.28 Observed Values for Luggage Inspection
Null hypothesis:
There is no difference among the three inspectors, Ho: P1 =P:z =P3
Alternative hypothesis:
At least one of the proportions is different, H1: P1 '¢ P2 '¢ P3
Degrees of freedom (OF) =(rows -1)(columns -1) =(2-1)(3-1) = 2

= =
The critical value of X2 for OF 2 and a 0.05 in the one-tail, right side of the
distribution, is 5.99 (refer to the Appendix Section XII). There is only a 5% chance
that the calculated value of X2 will exceed 5.99.
VILLANOVA UNIVERSITY VII - 46 LEj~N SIX SIGMA HANDBOOK

Example 7.17 (Continued):
Inspectors Treatment
Expected Values
1 2 3 Totals
Knives detected 24.67 24.67 24.67 74

Knives undetected 5.33 5.33 5.33 16
Sample total 30 30 30 90
Table 7.29 Expected Values for Luggage Inspection
Expected values (E j ) = (row total)(column total)/(grand total)
X2 = t
1= 1
(OJ - Ej )2
Ej
X2 = (27-24.67Y + (25-24.67t + (22-24.67)2 + (3-5.33)2 + (5-5.33)2 + (8-5.33t
24.67 24.67 24.67 5.33 5.33 5.33
2 (2.33)2 (0.33)2 (_2.67)2 (_2.33)2 (_0.33)2 (2.67)2
X= + + + + +
24.67 24.67 24.67 5.33 5.33 5.33
X2 = 2.89
Since the calculated value of X2 is less than the previously calculated critical value
of 5.99 and this is a right tail test, the null hypothesis cannot be rejected. There is
insufficient evidence to say, with 95% confidence, that the abilities ofthe inspectors
differ.

F Test
The need for a statistical method of comparing two population variiances is apparent.
We may wish to compare the precision of one measuring device to another, or the
relative stability of two manufacturing processes. The F test, named in honor of Sir
Ronald Fisher, is usually employed.
If independent random samples are drawn from two normal populations with equal
variances, the ratio of (s1)2/(s2)2 creates a sampling distribution known as the
F distribution. The hypothesis tests for comparing a population variance, a 2, with
another population variance, a 22 , are given by the following: 1
The null hypothesis is denoted by Ho and the alternative hypothl~sis is denoted by

H1• The shape of the F distribution is non-symmetrical and will depend on the
number of degrees of freedom associated with s/ and s/. Thlese quantities are
represented by V1 and V2 respectively. (V1 is the OF in the numercltor.)
The F statistic is the ratio of two sample variances (two chi-square distributions) and
is given by the formula:
2
Where S1 and S22 are sample variances.
Since the identification of the sample variances is arbitrary, it is customary to

designate the larger sample variance as a 21 and place it in the numerator.
The F cumUlative distribution function is given in the Appendix. Both the lower and
upper tails are given in this table, but most texts only give one talil, and require the
other tail to be computed using the following expression:

F Test (Continued)
f(F)
f (ex)
Figure 7.30 An F Distribution for Two Normal Samples Both with OF = 10
There are numerous F Table formats based on a values of 0.1 0, 0.05, 0.025, 0.01, etc.
Listed below is a partial F Table for a = 0.05. A more complete table is in the
Appendix.
~ v2
1
1
161.4
2
199.5 215.7
3 4
224.6
5
230.2
6
234.0
7
236.8
8
238.9
9
240.5
10
241.9
2 18.51 19.00 19.16 19.25 19.30 19.33 19.35 19.37 19.38 19.40
3 10.13 9.55 9.28 9.12 9.01 8.94 8.89 8.85 8.81 8.79
4 7.71 6.94 6.59 6.39 6.26 6.16 6.09 6.04 6.00 5.96
5 6.61 5.79 5.41 5.19 5.05 4.95 4.88 4.82 4.77 4.74
6 5.99 5.14 4.76 4.53 4.39 4.28 4.21 4.15 4.10 4.06
7 5.59 4.74 4.35 4.12 3.97 3.87 3.79 3.73 3.68 3.64
8 5.32 4.46 4.07 3.84 3.69 3.58 3.50 3.44 3.39 3.35
9 5.12 4.26 3.86 3.63 3.48 3.37 3.29 3.23 3.18 3.14
10 4.96 4.10 3.71 3.48 3.33 3.22 3.14 3.07 3.02 2.98
Table 7.31 F Critical Values (a=0.05)
Note: The above critical values for F may be used for a one-tailed test (a=0.05, 95%
confidence) or a two-tailed test (a/2=0.05, 90% confidence).

F Test (Continued)
Example 7.18: A materials laboratory is studying the effect of aging on a product.

They want to know if there is an improvement in consistency of strength after aging
for one year (assume a 95% confidence level). The data obtained is listed below:
At Start One Year Later

Number of tests 9 7
Product standard deviation (psi) 900 300
Solution: Ho: a/ ~ a/ H1: a/ > a/ and DF1 = 8 DF2 = 6

We are concerned with an improvement in variation; therefore, WEi will use a one-tail
test with the entire 0 risk in the right tail. From the prior F Tabl,e, we see that the
critical value of F is 4.15. The null hypothesis rejection area is E!qual to or greater
than 4.15. S2 (900)2
F- 1 - -9
- s~ - (300)2 -
Conclusion: Since the calculated F value is in the critical region, the null hypothesis
is rejected. There is sufficient evidence to indicate a reduced variance and more
consistency of strength after aging for one year.
Example 7.19: Two samples consisting of 10 and 8 measurements were observed

= =
to have variances of S1 2 7.14 and S22 3.21, respectively. Do the sample variances
present sufficient evidence to indicate that the population varianc:es are different at
a 90% confidence level?
Solution: Ho: a/ = a/ H1 : a/ '4= a/ DF1 = 9 DF2 = 7 0/2 =: 0.05
The test statistic is:

F = s! = 7.14 =2.22
S2 3.21
From Table 7.31, the right tail critical value of F = 3.68. The left tail critical value of
F is found using the equation on page VII-48 where F = 1/(3.29) :: 0.304, where the
degrees of freedom terms are 7 and 9 when finding the table value. Since the
calculated F value of 2.22 is between the critical values of 0.304 and 3.68, we fail to
reject the null hypothesis and cannot conclude that the population variances are
different at a 90% confidence level.

Summary of Inference Tests
TYPE TEST STATISTIC OF APPLICATION

- -
Z= X -Ilo = X-Il0 Single sample mean. Standard
Z
CJ'x CJ'x / ..Jn N.A. deviation of population is known.
-
t=X-llo Single sample mean. Standard
tTest
s/..Jn n-1 deviation of population unknown.
Z= x - npo Population proportion with fixed

pTest N.A. number of independent trials and
~npo (1-po) np ~ 5 and n(1-p) ~ 5
2 Mean
- - 2 sample means. Variances are
Xi - X2 unknown, but considered equal.
Equal t=
Variance SP~ n1 + 1 n 1+n 2-2 ("1-1) s~ +{n2 -1 )s~
tTest 1 n2 sp=~ n 1+n2-2
2 Mean Xi -X2 2 sample means. Variances are

Unequal t= unknown, but considered
S2 S2
Variance _1 +~ * unequal. *DF from the Welch-
tTest ~ n1 "2 Satterthwaite Approximation.
d 2 sample means. Data is taken in

Paired t= pairs. A different d is calculated
tTest Sd/Jn n-1
for each pair.
X2 X2 = (n -1)s!
2 n-1
Tests sample variance against
a2 Known 0x known variance.
X2 X2 = t(O, - EjY (r-1)(c-1)

Compares observed and expected
j=1 Ej frequencies of test outcomes.
F F=~ n 1 -1 Tests if two sample variances are

S22 n 2 -1 equal.
Table 7.32 Comparison Summary of Inference Tests
The student may wish to refer to {Juran, 1999)3, {NIST, 2007)6 or (Triola, 1994)8.

HYPOTHESIS TESTING/ANOVA
Analysis of Variance
Earlier, techniques were presented for estimating and testing thE! values of a single
population mean or the difference between two means (Z test and student's t test).
In many investigations (such as experimental trials), it is necessary to compare three
or more population means simultaneously. There are underlying assumptions in this
analysis of variance of means: the variance is the same for all factor treatments or
levels, the individual measurements within each treatment are ne,rmally distributed
and the error term is considered a normally and independently distributed random
effect.
With analysis of variance, the variations in response measurem4:!nt are partitioned

into components that reflect the effects of one or more independent variables. The
variability of a set of measurements is proportional to the slum of squares of
deviations used to calculate the variance:
Analysis of variance partitions the sum of squares of deviatiions of individual

measurements from the grand mean (called the total sum of squares) into parts: the
sum of squares of treatment means, plus a remainder whilch is termed the
experimental or random error. When an experimental variable is highly related to the
response, its part of the total sum of the squares will be higlhly inflated. This
condition is confirmed by comparing the variable sum of squams with that of the
random error sum of squares using an F test.
A Comparison of Three or More Means

An analysis of variance to detect a difference in three or more population means first
requires obtaining some summary statistics for calculating variance of a set of data
as shown below:
Where:
DC 2 is called the crude sum of squares
(DC)2/N is the CM (correction for the mean), or CF (correction factor)
DC 2. (DC)2/N is termed 55 (total sum of squares, or corrected 55).

2
De (DC)2/N = a 2 (variance) =
• TotalSumofSqualres
N - 1 Total DF(degrees of freedom)

A Comparison of Three or More Means (Continued)

One-Way ANOVA
In the one-way ANOVA, the total variation in the data has two parts: the variation
among treatment means and the variation within treatments.
Let the ANOVA grand average =IXIN =GM. The total SS (Total SS) is then:
Total SS =I(Xj - GM)2 Where Xj is any individual measurement
Total SS =SST + SSE =

Where SST treatment sum of squares and SSE is the
experimental error sum of squares.
SST = Int (Xt - GM)2 = Sum of the squared deviations of each treatment average
from the grand average or grand mean.
SSE = I(Xt - Xt)2 = Sum ofthe squared deviations of each individual observation
within a treatment from the treatment average.
For the ANOVA calculations:

I(TCM) = L Each treatment total squared (Treatment total CM)
No. obs. in that treatment
SST = I(TCM) - CM
SSE =Total SS - SST (Always obtained by difference)
Total DF =N - 1 (Total Degrees of Freedom)
TDF =t-1 (Treatment DF = number of treatments minus 1)
EDF =(N - 1) - (t - 1) =N - t (Error DF, always obtained by difference)
MST = SST = SST (Mean Square Treatments)

TDF t - 1
MSE = SSE = SSE (Mean Square Error)

EDF N- t


One-Way ANOVA (Continued)
To test the null hypothesis:
HO: J,l1 = J,l2 = ... = J,lt H1: At least one mean different
F= MST When F > Fa , reject Ho

MSE
Example 7.20: As an example of a comparison of three means, consider a single

factor experiment: The following coded results were obtained from a single factor
randomized experiment, in which the weights of hamburgers from three different
companies were compared. Determine if there is a significan1t difference in the
results (a=0.05).
Machines Data Sum n Avg TCM = (Sum)2/n EX 2

A 5,7,6,7,6 31 5 6.2 192.2 195
B 2,0, 1, -2, 2 3 5 0.6 1.8 13
C 1,0, -2, -3, 0 -4 5 -0.8 3.2 14
Total 30 15 197.2 222
IX = 30 N = 15Total DF = N -1 = 15 -1 = 14
GM = IXIN = 30/15 = 2.0 IX2 = 222 eM = (D<)2 = (30)2 :: 60

N 1S
Total SS = IX2 - CM = 222 - 60 = 162 I(TCM) = 197.2
SST = I(TCM) - CM = 197.2 - 60 = 137.2
also SST = Int (X t - GM)2 = 5(6.2 - 2)2 + 5(0.6 - 2)2 + 5(-0.8 - 2)2
= 82.2 + 9.8 + 39.2 = 137.2
SSE = Total SS - SST = 162 -137.2 = 24.8


One-Way ANOVA (Continued)
The completed ANOVA table is:
Source
SS OF Mean Square F Fa,v ,v
(of variation) 1 2
Company 137.2 2 68.6 33.2 F0.05,2,12 = 3.89

Error 24.8 12 2.067
Total 162 14
ae =V2.07= 1.44
Since the computed value of F (33.2) exceeds the critical value of F, the null
hypothesis is rejected. Thus, there is evidence that a real difference exists among
the weights of hamburgers from three different companies.
a e is the pooled standard deviation of within treatments variation. It can also be

considered the process capability sigma of individual measurements. It is the
variation within measurements which would still remain if the difference among
treatment means were eliminated.
Two-Way ANOVA
It will be seen that the two-way analysis procedure is an extension of the patterns
described in the one-way analysis. Recall that a one-way ANOVA has two
components of variance: Treatments and experimental error (may be referred to as
columns and error or rows and error). In the two-way ANOVA there are three
components of variance: Factor A treatments, Factor B treatments, and
experimental error (may be referred to as columns, rows and error).

Two-Factor, Two-Way ANOVA Experimlent
Example 7.21: In a hypothetical example, three different study materials were taught
by two different instructors to three different students with the following results.
The responses are examination results shown as a percent.
Study Materials
1 2 3 Sum n Avg RowSq

Instruct
88 68 76
1 84 62 70 690 9 76.667 52900
92 72 78
60 44 60
2 68 50 52 500 9 55.:556 27778
58 52 56
Sum 450 348 392 1190 80678
n 6 6 6 18
Avg 75 58 65.33
ColSq 33750 20184 25610.7 79544.7
IX2 34852 20792 26200 81844
The null hypotheses: Instructor and study material means do not differ.
EX= 1190 N=18 Total OF= 17 GM = EXIN = 1190/18 :: 66.11
eM = (EX)2/N = (1190 2)/18 = 78672.22
Total SS =IX2
- CM =81844 -78672.22 =3171.78
ColSq =column total squared and divided by the no. of observati4)ns in the column.
RowSq =row total squared and divided by the no. of observations in the row.

Two-Factor, Two-Way ANOVA Experiment (Continued)
SSCol = IColSq - CM = 79544.67 - 78672.22 = 872.44
SSRow = IRowSq - CM = 80677.78 - 78672.22 = 2005.56
SSE = Total SS - SSCol- SSRow = 3171.78 - 872.44 - 2005.56 = 293.78
The next step is to construct the ANOVA table.
Source SS OF MS F FO.OS,v ,v.,

i
Columns
872.44 2 436.22 20.8 FO.05,2, 14 = 3.74
(Matis)
Rows
2005.56 1 2005.56 95.6 Fo.o5,1,14 = 4.60
(Instruct)
Error 293.78 14 20.98
17 SIGe = '1'20.98 = 4.58
SIG total =v'Total SSI(N-1) = 13.66
If no interaction: Col OF = Col -1 = 3 -1 = 2 Row OF = Row -1 = 2 -1 = 1

= =
Error OF Total OF - Col OF - Row OF 17 - 2 -1 = 14
= = =
Col F MSColIMSE 436.22/20.98 20.79. This is larger than critical F =3.74.
Therefore, the null hypothesis of equal material means is rejected.
= = =
Row F MSRow/MSE 2005.56120.98 95.59. This is larger than critical F = 4.60.
Therefore, the null hypothesis of equal instructor means is rejected.
The difference between total sigma (13.66) and error sigma (4.58) is due to the
significant difference in instructor means and study material means. Ifthe instructor
difference and study material differences were only due to chance cause, the sigma
variation in the data would be equal to SIGe, the square root of the mean square
error.

ROOT CAUSE ANALYSIS
Root Cause Analysis

An individual or team is given the responsibility of determining the root cause of a
deficiency and correcting it. The solution to some problems mcly be complex and
difficult. In other cases, the solution may be known, but considerable time will be
required to implement it. The proposed action may take several steps. See the
illustration below:
Immediate Intermediate
Situation Ro_o t Cause Action
Action Action
The dam Plug it Patch the dam Find out what caused the leak
leaks so it does not happen again.
Then rebuild the dam.
Parts are 100% Put an oversize Analyze the plrocess and take
oversized Inspection kickout device in action to elimiinate the
line production of oversized parts.
Table 7.33 Short and Long-Term Corrective Acticlns
To help locate the system's true problem, a variety of problem solving tools are
available. Listed below are commonly used techniques.
Subjective Tools Analytical Tools
Ask why, and then ask why again .... Data collection and analysis
Brainstorming Pareto analysis
Process flow analysis Regression analysis
Plan-do-check-act Checksheets
Systematic problem solving Data matrix analysis
Nominal group technique Process capability analysis
Operator observation Partitioning of variatkm
Fishbone diagrams Subgrouping of data
Consensus exercises Simple trials
Six - thinking hats Statistical design of E!xperiments
Use of teams Analytical tests (X2 , F, Z, t)
FMEAlfauit tree analysis Control charting
When permanent corrective action is proposed, management must determine if:
• The root cause analysis has identified the full extent of th~e problem
• The corrective action is satisfactory to eliminate or prevent recurrence
• The corrective action is realistic and maintainable

ROOT CAUSE ANALYSIS
5 Whys
The 5 whys approach to root cause analysis is described as asking the question
"Why?" five times. This technique is generally attributed to a Japanese method of
determining the root cause. The following is an example of 5 whys.
Symptom: The customer shipments were not made on time.
1. Why? We ran out of parts because the die stamping press broke down.
2. Why? The press had not received scheduled maintenance for a period of
three months.
3. Why? The maintenance department staff had been reduced to six people from
eight.
4. Why? The maintenance department was over budget due to high overtime
costs, and the General Manager eliminated all overtime and required a
25% reduction in personnel for all overhead support departments.
5. Why? The company was not reaching profit goals so the CEO had issued an
order to avoid all unnecessary spending. So the root cause was the
CEO was worried about getting fired for poor profit performance.
There is nothing magical about 5 whys. In fact, the root cause may be found after
3 or 4 whys. In other cases, one may need to ask why six or more times. This
method asks why until the root cause is found.
5 Ws and H (or 2 H)
The 5 Ws and H approach to root cause analysis is summarized as asking the

questions Who?, What?, When?, Where?, Why?, and How? The 5 Ws and H
methodology is an old technique, used by newspaper reporters in asking questions
in order to get the full story. In the quality context, responses to these questions can
be organized with each element depicted as a branch on a cause-and-effect
(fishbone) diagram.
In some references, this same basic method is simply referred to as the 5 Ws (Who,
What, Where, When, and Why). Also note that the order of the Ws varies, depending
upon the referenced source. The technique looks at a problem or symptom from
several viewpoints in order to include as much information as is needed to assist in
determining the root cause.

REFERENCES
References
1. Huck, S. & Cormier, W. (1996). Reading Statistics and Rese,arch, 2nd ed. New
York: HarperCollins.

Management. New York: McGraw - Hill.
3. Juran, J.M. (1999). Juran's Quality Handbook, 5th ed. New York: McGraw - Hill.
4. Langley, R. (1971). Practical Statistics: Simply Explained. (Revised edition.).

New York: Dover Publications.
5. Metcalf, M.D. (1997). "Waste Elimination." Sullivan, IN: Metcalf Publishing.
6. NIST, Information Technology Laboratory, Statistical Eng,ineering Division.

(2007). Engineering Statistics Handbook. R~etrieved from
http://www.itl.nist.gov/div898/handbookl January 8,2007.
7. Sharma, A., & Moody, P. (2001). The Perfect Engine: How j~o Win in the New
Free Press.
8. Triola, M.F., & Franklin, L.A. (1994). Business Statistics, RI!ading: Addison -
Wesley.
9. Wortman, B.L., et al. (2001). CSSBB Primer. Terre Haute, IN: Quality Council of
Indiana.

QUESTIONS
7.1. Which of the following tasks is value added? 7.5. A randomly selected sample of bicycle helmets
was tested for impact resistance. Given the data
a. An appropriate product development results below, what is the 95% confidence
process interval for the mean?
b. Reworking parts to meet customer
requirements Test results:
c. MRB meetings Sample size: 100 helmets
d. Efficient material handling of customer Average impact resistance: 276 g
returns Standard deviation ofthe measurements: 15 g
7.2. In the broadest sense, how many of the a. 276 :t 29.4 g

following areas of variation in multi-vari analysis b. 276 :t 2.47 g
can include process time related elements? c. 276:t 2.94 g
d. 276:t 2.17 g
I. Positional
II. Cyclical 7.6. A p-value is used in many statistical
III. Temporal calculations. It can be described as:
a. I and II only a. Being similar to the critical statistical value

b. I and III only as found in the statistical tables
c. II and III only b. Containing an extreme test statistic
d. I, II, and III probability value as obtained from the
sample data
7.3. If the plating time of a tube increases as the c. Having the typical set values of 5% to 1%
inside diameter decreases, then the correlation d. Being reported only when significant
coefficientforthe inside diameter versus plating
time would be: 7.7. An F distribution most closely resembles:
a. Between 0 and +1 a. A t distribution

b. Between 0 and -1 b. A chi-square distribution
c. Greater than 1 c. An exponential distribution
d. Equal to 1 d. A log normal distribution
7.4. How is the level of significance defined? 7.S. From a statistical standpoint, the sample size
for hypothesis testing depends on:
a. The probability of rejecting a null
hypothesis when it Is true I. The required type I and type II risks
b. The probability of not rejecting a null II. The minimum population shift of interest
hypothesis when it is true III. The variations in the population of interest
c. The probability of accepting a null IV. The cost oftaking a sample
hypothesis when it is true
d. The probability of not accepting a null a. I and III only
hypothesis when it is true b. I, II, and III only
c. II and IV only
d. I, II, III, and IV

QUESTIONS
7.9. When a problem solving team applies the 5 why 7.13. In a single factor analysis of variance, the
technique, they are attempting to: assumption of homoge'neity of variances applies
to:
a. Determine if the interviewee is telling the
truth a. The variances within the treatment groups
b. Understand the basics of the problem b. The variances of the treatment groups
c. Eliminate areas not to investigate c. The total variance
d. Determine the root cause of the problem d. All of the above
7.10. Non-value added activities on the factory floor 7.14. A company has just installed a new computer
are most clearly controlled by the elimination of: data entry system and the new input error rates
must be determined. Management requires the
a. Gembo error rate to be within (]1.5%, at a 95% confidence
b. Muda level. What sample size is necessary if the
c. Poka-yoke population standard dl~viation is 1.2%?
d. Kaizen
a. 15
7.11. Which of the following statements concerning b. 16
the coefficient of simple linear correlation, r, is c. 22
NOT true? d. 23
a. r = 0.00 represents the absence of a 7.15. Given that random l;amples of process A
relationship produced 10 defective and 30 good units, while
b. The relationship between the two variables process B produced .25 defectives out of 60
must be nonlinear units. Using the chi-!iquare test what is the
c. =
r 0.76 has the same predictive power as r probability that the obslarved value of chi-square
=-0.76 could result under the hypothesis that both
d. r = 1.00 represents a perfect relationship processes are operating at the same quality
level?
7.12. The determination of temporal variation in multi-
vari charting means: a. Less than 5%
b. Between 5% and 1D%
a. Variation within piece c. Between 10% and 20%
b. Variation over time d. Greater than 20%
c. Variation piece-to-piece
d. Variation within batch 7.16. The correlation coefficient for test #1 equals -
0.89, and the correlaticm coefficient for test #2
equals 0.79. Assuming the tests are in different
subject areas, which ofthe following statements
is FALSE?
a. Both tests show a strong correlation

b. The scatter around the regression line is
greatest for test #1
c. The slope for test ~~1 is negative
d. Y increases as X illcreases in test #2

QUESTIONS
7.17. What inference test compares observed and 7.21. All of the following statements concerning
expected frequencies of test outcomes? statistical inference are true, EXCEPT:
a. F test a. The confidence interval is a range of values

b. Chi-square test which may include the true population
c. t test value
d. Paired t test b. The confidence interval is normally the
statistical tolerance limits ofthe population
7.18. In order to calculate the confidence interval for parameter
the mean for a small sample, what must be c. The point estimate is a single value used to
known? estimate the population parameter
d. A sample result is often analyzed in order to
I. The average predict the value of the population
II. The standard deviation parameter
III. The appropriate t distribution value
IV. The sample size 7.22. What is the principal advantage of multi-vari
charting?
a. II and III only
b. I and II only a. It graphically displays the variation in a
c. I, II, and III only process
d. I, II, III, and IV b. It keeps track of the time that
measurements were made
7.19. An analysis of variance (ANOVA) is used to c. It assists in the breakdown of components
analyze the data gathered in an experiment. of variation
One-way ANOVA implies that this tool deals d. It is much easier to plot than most control
with: charts
I. One level 7.23. In one-way AN OVA, how is the SSE (the

II. Independent samples experimental error sum of squares) usually
III. One dependent variable determined?
IV. One independent variable
a. By dividing SST by the total OF
a. lonly b. By subtracting SST from the total 55
b. II and III only c. By dividing the total 55 by SST
c. II, III, and IV only d. By subtracting the total SS from SST
d. I. II, III, and IV
7.24. A string tension gauge is used to validate
7.20. What forms of transportation and storage are settings on a tennis racket stringing machine.
considered muda? What are the 90% confidence limits for the
estimated true value of the string tension?
I. Conveyer belts
II. Hand trucks String tension measurements: 57 pounds, 56
III. Fork trucks pounds, 60 pounds, 62 pounds, 55 pounds, and
IV. Pallets and bins 59 pounds.
a. II and III only a. 61.79 pounds, 54.54 pounds

b. I and III only b. 60.28 pounds, 56.06 pounds
c. I and IV only c. 60.34 pounds, 56.00 pounds
d. I, II, III, and IV d. 59.94 pounds, 56.39 pounds

QUESTIONS
7.25. Given a coefficient of determination of 0.85, 7.29. What test statistic mu:st be known in order to
which of the following is true? compute a confidence interval for variation?
a. There is positive correlation in the data a. Chi-square value

b. 85% of the variability is explained by the b. zvalue
regression model c. tvalue
c. The sum of squares of X is greater than the d. x value
sum of squares of Y
d. There is a weak data correlation 7.30. Which of the following is NOT considered
muda?
7.26. Ratios of two variances drawn from the same
normal population are described by which one a. Design reviews
of the following distributions? b. Uneven schedulin!l of work
c. Reshaping a die after Installation
a. F d. Breakdowns in machinery
b. Student's t
c. Chi-square 7.31. Identify the subjective root cause analysis tool
d. Normal from the options below:
7.27. The opposite of an alpha risk in hypothesis a. 5whys

testing is called: b. Pareto analysis
c. Control charts
a. Beta risk d. Regression analysis
b. 1-beta risk
c. Level of confidence 7.32. A small change truly exists between a current
d. 1-level of confidence process and an E!xperimental process.
However, an insufficient sample size is collected
7.28. What table value is normally used to place a to determine this differl,mce. What type of error
confidence interval on the slope of a line in would result?
simple linear regression?
a. A type II error
a. t b. A 1-alpha decision
b. F c. A type I error
c. z d. A 1-beta decision
d. Chi-square
VILLANOVA UNIVERSITY VII·64 LEAN SIX SIGMA HANDBOOK

QUESTIONS
7.33. What is the Z value needed to conduct a two-tail 7.37. A correlation of +0.95 for two variables, x and y,
test in a statistical inference problem, specifying means that:
a 90% level of confidence?
I. There is a strong linear relationship
a. 1.96 between x and y
b. 1.28 II. Y may be predicted from x using the
c. 1.65 equation y = 0.95x
d. 2.24
a. I only
7.34. If a multi-vari chart shows that 60% ofvariation b. "only
is within piece, 25% of variation is piece-to- c. Both I and II
piece and 10% of variation occurs over time, d. Neither I nor II
what would be the indicated improvement
action sequence? 7.38. If four inspectors were evaluated, for the
detection or non-detection of a defect in twenty
a. Temporal, positional, cyclical samples, how many degrees of freedom would
b. Positional, temporal, cyclical be used to determine the critical chi-square
c. Cyclical, positional, temporal value?
d. Positional, cyclical, temporal
a. 16
7.35. The current process produces fifty units per b. 8
shift. A new process produced sixty units per c. 6
shift, for ten consecutive shifts. The highest d. 3
shift was sixty-six units and the lowest shift
during the trial was fifty-four units. With what 7.39. A product was yielding 90% recovery before an
level of confidence can one say the process has Improvement was made. To determine if a 2%
changed? change (in either direction) has been made at
the 95% confidence level, what sample size
a. Since the standard deviation is unknown, should be taken?
the test can't be performed
b. 90% confidence a. 468
c. 95% confidence b. 648
d. >99% confidence c. 864
d. 1728
7.36. Refer to the following partial ANOVA table:
7.40. Which ofthe following characterize the muda of
Source SS OF MS waiting?
Materials 900 2
I. Idle operators
Machines 2100 2 II. Machinery breakdown
Errors 300 10 III. Uneven schedules
14 IV. Unnecessary meetings
a. I and II only
What are the three missing MS values?
b. II, III, and IV only
c. I, II, III, and IV
a. 450, 1050, 30
d. I, III, and IV only
b. 60, 140,20
c. 450,190.91,20
d. 45.0,190.91,30

QUESTIONS
7.41. If the 90% confidence for the mean is 181.3 to 7.45. Which two of the following confidence interval
203.8, which of the following statements is calculations require thl! use of Z table values?
correct?
I. Large sample means
a. 90% of all values in the population fall II. Small sample means
between 181.3 and 203.8 III. Variation confiden.:e intervals
b. 90% of all values are either greater than IV. Proportion confidence intervals
203.8 or less than 181.3
c. The probability of randomly selecting a a. I and II only
value between 181.3 and 203.8 is 90% b. II and III only
d. None of the above c. I and III only
d. I and IV only
7.42. The probabilistic regression model for any
particular observed value of Y contains a term 7.46. If the mean for each of the treatment groups in
/30' which represents: an experiment were ide'ntical, the F ratio would
be:
a. The Y-axis intercept, when X=O
b. The Y-axis intercept, when X=1 a. 1
c. The slope of the model b. 0
d. The X-axis intercept, when Y= 0 c. A positive number between 0 and 1.00
d. Infinite
7.43. Three machines are being evaluated in a one-
way ANOVA. A total of sixteen trials have been 7.47. For a linear correlation, the total sum of squares
conducted. How many degrees of freedom are equals 1,600, and the total sum of squared
available to evaluate the error term? errors equals 1,000. What is r2?
a. 12 a. 0.600
b. 13 b. 0.625
c. 14 c. 0.375
d. 15 d. 0.750
7.44. A multi-vari chart indicates: 7.48. Identify the element that is NOT associated with
excess inventory?
15% within piece variation
65% piece-to-piece variation a. Storage space
10% over time variation b. Additional labor
c. Transportation vehicles
What is the recommended action sequence to d. Expensive poka-yoke devices
reduce variation?
a. Cyclical, temporal, positional

b. Positional, cyclical, temporal
c. Cyclical, positional, temporal
d. Temporal, cyclical, positional

, I
VIII. IMPROVEMENT TECHNIQUES
MOST PEOPLE WOULD RATHER UVE WITH A PROBLEM

THEYCAN'T SOLVE, THAN ACCEPT A SOLUTION THEY
CAN'T UNDERSTAND.
R. E. D. WOOLSEY AND H. S. SWANSON
VILLANOVA UNIVERSITY VIII ·1 LEAN SIX SIGMA HANDBOOK

ELIMINATING WASTES/55
Improving the Process
Improving the process is presented in the following topic areas:
• Eliminating wastes • Validating solutions

• Theory of constraints • Measurement system re-analysis
• Design of experiments
Eliminating Wastes
Eliminating wastes is presented in the following topic areas:
• 55 • 5et up reduction
• Kanban (pull) • Flow improvement
• POka-yoke (mistake proofing) • Quick response manufacturing
58, Housekeeping and Workplace Organi~~ation*
Although 5S is considered the simplest of all lean techniques, it should not be taken
lightly. 55 applications can bring excellent benefits in the fiSJht against muda
(waste). 55 starts by providing a formal approach for housekeeping that becomes
systematic. The 55 technique includes the measurement, auditing, and monitoring
of cleanliness, order, and neatness. The 55 approach exemplifies a determination
to organize the workplace, keep it neat and clean, establish standalrdized conditions,
and maintain the discipline that is needed to do the job.
Once fully implemented, the 55 process can increase mOralE!, create positive
customer impressions, and increase efficiencies. Not only will employees feel better
about where they work, the effect on continuous improvement can lead to less
waste, improved quality, and faster lead times. Any of these itl~ms will make an
organization more profitable and competitive in the marketplace.
'" The material on 55 has been modified and combined from the fl3110wing sources:
{Davidson, 2006)4, {George, 2003)5, {Imai, 1997)10, {i5ixSigma, 2006)1\ {Skaggs,
2006)26, {Wortman, 2006)35.
VILLANOVA UNIVERSITY VIII - 2 LEAN SIX SIGMA HANDBOOK

ELIMINATING WASTES/5S
58 (Continued)
There are now very distinctive English counterparts for each original Japanese 55
word. The concept of 55 has been twisted somewhat due to attempts to keep each
element starting with the letter '5' in English, like the real Nippongo words (seiri,
seiton, seiso, seiketsu, and shitsuke). Different authors state slightly different
English 55 equivalents, but their meanings remain similar to the original
fundamental terms. The term 55 comes originally from five Japanese words briefly
described as follows:
• Seiri (sort or organize)

• Seiton (set, straighten, or store)
• Seiso (shine)
• Seiketsu (standardize)
• Shitsuke (sustain or self-discipline)
• Safety (added by some organizations as 55+1)
Step 1. Seiri (also known as sort or organize)
Seiri focuses on eliminating unnecessary items from the workplace. One should
separate the necessary from the unnecessary ones. An effective visual method to
identify these unneeded items is called red tagging. A red tag is placed on all items
not required to complete the normal job. This process requires an evaluation of the
red tag items. Items that are only used occasionally are moved to an organized
outside storage location. Unneeded items are discarded. Sorting is an excellent
way to free up valuable floor space and eliminate such things as broken tools,
obsolete fixtures, scrap, and excess raw material. The sort process also helps
prevent the JIC Oust-in-case) job mentality. A typical sequence follows:
• Set up a schedule to target each area

• Remove unnecessary items in the workplace
• Red tag unneeded items, record everything that is thrown out
• Keep repaired items that will be needed
• Major housekeeping and cleaning is done by area
• Inspect the facility for problems, breakages, rust, scratches, and grime
• List everything which needs repair
• Deal with causes of filth and grime by prioritizing corrections
• Perform management reviews of this and other steps
VILLANOVA UNIVERSITY VIII-3 LEAN SIX SIGMA HANDBOOK

58 (Continued)
Step 2. Seiton (also known as set. set in order, store. straighten. proper
arrangement. or simplify)
Seiton focuses on efficient and effective storage methods. Oncl~ sorted, items are
arranged systematically to ensure their traceability. Things are put in order and
placed in such a way that they can be easily reached whenevelr they are needed.
Items are arranged and identified for ease of use.
One should ask questions like: What do I need to do the job? Where should this
item be located? How many items are really needed?
Strategies for effective seiton activities include: painting floors, outlining work areas
and locations, shadow boards, and modular shelving and cabinets for needed items
(such as trash cans, brooms, mop and bucket). "A place f()r everything and
everything in its place."
o Have a place for everything and put everything in its plaCE!

o Analyze the existing conditions for equipment, inventory, and supplies
• Decide how things should be put away, including the exac:t locations
• Use labels, tool outlines, and color codes
• Determine everyday controls and out-of-stock conditions
• Reduce inventories and define who does the reordering
• Determine who has misSing items or if they are lost
• Use aisle markings, placement for dollies, forklift, boxes, .~tc.
• Establish pallet zones for work in process (WIP)
Step 3. Seiso (also known as shine. scrub. clean. or sweep)
Once the work area clutter and junk have been eliminated and thl~ necessary items
are identified, the next step is to thoroughly clean the work are'l. Daily follow-up
cleaning is necessary to sustain this improvement. Most WOrkE!rS take pride in a
clean and clutter-free work area.
The shine step will help create ownership in the equipment and fac:ility. Workers will
also begin to notice changes in equipment and facilities such as cdr leaks, oil leaks,
coolant leaks, contamination, vibration, breakage, and misalignment. These
changes, if left unattended, could lead to equipment failure and I()ss of production.
All of these issues impact the company's bottom line.

55 (Continued)
Step 3. Seiso (Continued)
Always keep arranged items ready to use and in a tidy status. Look for ways to keep
the work area clean. One must make the workplace spotless.
• This activity includes ways to keep things clean

• Clean everything in the workplace, including equipment
• Perform root cause analysis and remedy machinery and equipment problems
• Complete training on equipment maintenance basics
• Divide each area into zones and assign individual responsibilities
• Rotate difficult or unpleasant jobs
• Implement 3-minute, 5-minute, and 10-minute 5S activities
• Use inspection checklists and perform white glove inspections
Step 4. Seiketsu (also known as standardize or purity)
Once the first three of the 5Ss have been addressed, one should concentrate on
standardizing the best practices in the work area. All employees involved with the
process, should participate in the development of a standard method. Workers are
a valuable, but often overlooked, source of information regarding the best work
methods.
Maintain cleanliness by instituting perpetual cleaning. Create a process for the

above three stages, standardize them, and review them often. Maintain and monitor
the first three Ss. Make cleaning and checking routine.
• Make 5S activities routine so that abnormal conditions show up

• Determine the important points to manage and where to look
• Maintain and monitor facilities to ensure a state of cleanliness
• Make abnormal conditions obvious with visual controls
• Set standards, determine necessary tools, and identify abnormalities
• Determine inspection methods
• Determine short-term countermeasures and long-term remedies
• Use visual tools such as color coding, markings, and labels
• Provide equipment markings, maps, and charts

5S (Continued)
Step 5. Shitsuke (also known as sustain, or self-discipline, or cc)mmitment)
Sustain the previous 4 steps and continually improve on them. 5hitsuke is by far the
most difficult 5 to implement and achieve. Human nature tends, to resist change.
More than a few organizations have found themselves with a dirt)" cluttered shop a
few months following their attempt to implement 55. The tendency is to return to the
status quo and the comfort zone of the "old way" of doing things. Sustain focuses
on defining a new status quo and standard of workplace organization.
5hitsuke is not a part of the previous core 45s, but is an appropriiate approach and
attitude towards any undertaking to inspire pride and adherenCE! to standards (as
established for the other four components). Using self-discipline, each individual
has to commit to the process, stick to the rules, and maintain m()tivation.
• Sustain the 4 previous steps and continually improve on them

• Acquire self-discipline through the habit of repeating the ~~ previous steps
• Establish standards for each of the 55 steps
• Establish and perform evaluations of each step
Step 6. Safety (added by some organizations as 55+1)
Lockheed Martin and other organizations often add one more 5 (hence the notation
55+1). The additional item is safety. Removing hazards and dangers.
What 5S is Not
Due to the straightforwardness of 55, it is easy to mistakenly conflUse some isolated
cleaning activities and label them as a 55 program. 55 is not:
• An isolated spring cleaning event

• Removing only the things that are no longer used
• Rearranging the same items in a neater way
• Taking items from one location to another
• A memo or e-mail stating that a 55 audit will be executed shortly
• A big, one-time event due to the visit of an important client
• A methodology, in itself, leading to world class performan1ce
VILLANOVA UNIVERSITY VIII - 6 LEJlN SIX SIGMA HANDBOOK

Measuring the Impact of 55

5S programs often lack formal metrics to determine the true impact of its
implementation. Management should take a kaizen blitz or kaizen DMAIC approach
including the state of current metrics. Goals are then defined to make 5S
improvements visible. Alukal (2006)1 provides a comprehensive list of metrics to
help estimate the benefits achieved by applying 5S. Some of the measurable gains
are the following:
• Improved production flow

• Improved quality
• Ease of obtaining information
• Increased efficiency
• Increased safety
• Fewer breakdowns
• Less inventory
• Less search time
• More ergonomic conditions
• Point-of-use storage (POUS)
• Space saved
• Visual controls (Alukal, 2006)1
There are also significant intangibles such as the reward of a better place to work,
personal achievement, and sense of ownership resulting from the 5S
implementation. Employees often talk about taking the 5S approach to their own
lives, houses, communities, etc.
55 Implementation Roadmap
The implementation of 5S can be addressed from three different levels:
• Philosophy level (an alignment with company values)

• Operational level (the typical application)
• Process level (a work station approach)
In most cases, companies dedicate most of their efforts to the 5S operational level,
underestimating the importance of incorporating the principles and practices of the
process and philosophy levels.

55 Philosophy Level
Failures in implementing 5S are often related to a disconnect between company
values with 5S practices. Companies typically conduct the sort, straighten, and
shine stages several times, but they do not standardize or sustain the process.
Sometimes this is related to short-term decision making within the business.
A company should not incorporate 5S merely as a housekeeping program.

Housekeeping can be achieved with mandatory and regulatory pOlicies; but a
company culture to increase productivity, based on 5S, is more th;an housekeeping.
5S needs to be perceived as a fundamental activity and be reinforced by
management philosophy and discipline. When 5S is incorporated, as part of the
normal way of work, long-term good improvement habits are developed.
55 Operational Level
The operational level is the typical case for 5S. In this level, a c:ompany creates,
deploys, and implements sorting, organization, cleaning, and similar activities.
Figure 8.1 illustrates the typical steps.
Pre-Launch Launch Deployment Follow-up

Stage Stage Stage Stage
Figure 8.1 5S Implementation Stages
Pre-Launch
In the pre-launch phase of the 5S initiative, top management l;hould be deeply

involved. In this stage, the company dedicates its efforts to:
• Align the 5S project with corporate strategic priorities

• Select a pilot area for launching 55
• Schedule a training program for the pilot area
• Define a grading system to recognize and award teams
• Establish the base line status of the pilot area
• Define communication methods to inform all employees
• Schedule an official announcement of the 5S initiative
5S implementation is an integral project that must be aligned with the strategic

objectives of the company. In a broad sense, 5S will reinforce the fundamental
values and practices that drive a company's performance.

58 Operational Level (Continued)

Pre-Launch (Continued)
A pilot 55 approach will allow a team to understand, correct, and improve the tactical
and practical implementation details. This will prove valuable when the concept is
deployed in other areas.
Launch
Prior to launch, top management has the responsibility of announcing the 55

initiative to the entire company. A clear alignment of 55 with the corporate vision
can help to facilitate the implementation, both at a pilot level, and later at a broader
level. Top management must communicate how 55 is aligned with the corporate:
• Vision
• Values
• Goals
Positive pilot area results can be a motivational factor in encouraging the rest of the
company to embrace 55. Classic 55 steps will be developed first by the pilot area.
Sort
At the operational level, the sort step attempts to remove unnecessary items from
work areas. Any article, piece of equipment, document, or record, that is not
essential for current production or service delivery, should be removed.
Actions taken during the sort step include:
1. Define criteria to classify items as necessary or unnecessary

2. Create a "red tag" to identify needless items (see Figure 8.2)
3. Clarify rules for removing items to the red tag area
4. Mark a physical zone as the red tag area
5. Proceed to search, identify, remove, and dispose of unnecessary items
6. Perform a root-cause-analysis *
7. Create stratification rules or importance principles
8. Implement stratification rules on a regular basis
* In this case, the analysis deals with why the unnecessary items have
accumulated in the work area.


Out of the 8 fundamental actions taken during the sort step, item number 6, root
cause analysis is often neglected. It is important that the 5S implementation team
take this step seriously. During 5S implementation, ifthe root cause ofthe problem
is not identified, analyzed, and eliminated, the problem will emerge again. It is
necessary to conduct a root cause analysis after every stage of 'the 5S program.
5S Red Tag
Filled by: Date:
Quantity: Location:
Item Description (mark one): Reason for disposition:
1. Raw material 1. Defective
2. WIP 2. Obsolete
3. Finished goods 3. Surplus
4. Repair supplies 4. Not needed in short-term
5. Surplus equipment/tools 5. Not needed in long-term
6. Office supplies 6. Needs identification
7. Customer property 7. Other
8. Unknown
9. Files Final disposition:
10. Other
1. Scrap (with paperwork)
2. Scrap (no paperwork)
Disposition by: 3. Return to vendor
Date: 4. Return to customer
Disposal by: 5. Move to red tag area
6. Move to:
7. Store in:
Figure 8.2 5S Red Tag Example
Often problem solving tools, such as brainstorming, affinity di,agrams, Ishikawa

diagrams, 5W+H, tree charts, etc., are necessary to systematiically identify the
problem that originates the disorganization. A team must isolatle the causes, and
eliminate the sources. Attempting to implement 5S with teams that do not have a
solid and practical knowledge of problem solving tools will increase implementation
difficulties, and will not identify root causes.
VILLANOVA UNIVERSITY VIII -10 LE)!,N SIX SIGMA HANDBOOK


Straighten
Straighten is the stage where the team must analyze and understand how things get
organized. Takashi Osada states that straighten (or neatness) involves 4 stages:
• Analyze the status quo. The most important and often overlooked step in the
straighten stage is analyzing how things get messed up, especially when the
volume and variety of items is large and difficult to control. Again, basic
problem solving tools are useful for this analysis. Once the current state is
understood, the team should advance to where things should be placed.
• Decide where things belong. If the sort stage was properly achieved, a
company should experience a reduction in the total quantity of items on the
floor. A 5S purist would say that one of any part on hand, in any given period,
is enough. Deciding where things belong utilizes stratification methods. That
is, the most frequently used items should be easily accessible.
• Decide how things should be put away. Once the stratification of items has
been completed, guidelines should be established for putting away items (as
well as how and when not to put them away).
• Get everybody to follow the put away rules. Visual management is an

effective way to help and prevent people from shifting from the agreed upon
rules of organization. (Osada, 1991)19
In zero quality control, two conditions of human nature are recognized. One, people
forget things and two, people make unintended mistakes. These facts should be
considered to prevent people from forgetting or failing to follow the straighten rules.
Shine (Sweep)
Shine is more than a simple cleanliness step. During this stage, the 5S team has to
develop other activities such as defining specific cleaning schedules for work areas
and organizing the big kick-off-day for companywide or pilot area cleaning. The
team has to organize the shine efforts in order to:
• Institute cleaning activities as a daily habit

• Reinforce the idea of cleaning as a channel to perform work area inspection
• Educate employees about how to prevent grime
• Define the interrelationships between cleaning, quality, safety, equipment
availability, and team morale


Shine (Continued)
It is a good practice to start the cleaning from the roof and dowll1 to the floor. It is
typical to find work areas in which the floor and walls are clean, but the roof and
ceiling are not. Shine emphasizes a completely concentrated effort to prevent dirt
in all areas of the facility.
The use of cleaning checklists is necessary. Checklists for facilitiies and equipment
require specific areas such as: floors, walls, ceilings, roof~;, machines, and
equipment. Ifthe straighten stage was properly accomplished, thEm a particular area
would have been identified to store cleaning supplies.
Standardize
Standardize is a critical step in 5S. If a company fails to provide, a structured plan

to standardize, sort, straighten, and sweep, the long-term ben.afits of 5S will be
impossible to achieve.
Standardize tools include checklists, visual aids, work instructions, and training.
Activities and practices should be scheduled for daily, weekly, mc)nthly, and annual
events.
Standardize activities need to be explained to the employees, with the intention of

describing clear scenarios about how sort, straighten, and shinE! are connected to
the daily operation of the office, stockroom, or manufacturing flloor. Standardize
attempts to calendar activities that will work as improvement ca1talysts in both the
short and long-terms.
Several 5S purists state that a way to measure progress is tOI assess the total
quantity of cleanups, especially cleaning tasks when customers or auditors visit the
facility.
Sustain
Sustain is associated with discipline and habits. This is the sta"e in which the 5S
program attempts to become an integral, inherent lifestyle throughout the company.
During the sustain phase, the team will review and adjust any audit forms they
developed during the pre-launch stage. An annual audit schedlule is necessary.
Every stage of the 5S program should be addressed.
VILLANOVA UNIVERSITY VIII·12 LEt~N SIX SIGMA HANDBOOK


Sustain (Continued)
Keeping audits as simple as possible is recommended. Those new to 55 are prone

to create extremely complex audit forms, in which large quantities of data must be
evaluated. The team should stay focused on the real value of performing 55 audits.
Lean experts state that companies must be humble enough to recognize that if the
55 program cannot be sustained, consequently more elaborated tools, like TPM,
5MED, etc., will be in jeopardy. Mediocre results in 55 implementation will move to
the next experience and will prevent future success. Knowledge, discipline, and
habits are intrinsic to successful 55 implementation.
Deployment
Once a 55 pilot test has been successful, a clear deployment plan for the rest of the
company must be structured. Specific elements covered during the pre-launching
stage need to be revised, including:
• Companywide training
• Incorporation of 55 principles into new employee training
• Adjustment of work instructions, procedures, and policies
• Execution of grading mechanisms
• Activation of evaluation systems
Follow-up
When the standardize and sustain phases are successful, audits and other metrics
(developed to track improvement) can be added to management reviews. Review of
55 performance by top management will be relevant, only if 55 is aligned with the
values and goals of the company.
VILLANOVA UNIVERSITY VIII -13 LEAN SIX SIGMA HANDBOOK

58 Process Level
At a process level, 5S has a more specific approach, oriented to the application of
the 5S principles to the workstation. Inoculating employees with 5S is a process that
may take months or years. Discipline is not easily achieved.
Sort
Employees must develop a proactive attitude to examine the immf!diate workstation

elements, working only with the minimum basic tools, materials, .~quipment, forms,
etc. Visual aids and procedures, among other tools, can help the operator to
remember "how" or "why" he or she is doing a particular task. Specific operator
activities at a workstation level include:
• Reviewing the necessary item to have in the workstation

• Separating out needless items, prior to the execution of the operation
• Reviewing the way in which he or she works to identify wclste
Sort includes identifying and eliminating wastes. The way in whic:h work is done, in
a workstation, can add waste. Tools such as Therbligs, which ar.~ related to micro-
motion studies, can contributed to identifying and eliminating ""aste in a specific
work area. Therbligs are movements or activities that are done by the operators or
machines. Many of these add no value. In order to apply this analysis, an individual
should review internet sources such as The Gilbreth Network, (2006)31.
Several tasks, such as searching, finding, disassembling, positioning, holding, etc.,

are waste (See Figure 8.3 on the following page). During the sort stage, waste and
variability must be identified after a detailed examination. Cctlors can also be
associated with each Therblig. Examples are: brown, black, gralY, light gray, lake
red, gold ochre, green, olive green, carmine red, sky blue, blue, purple, violet or
heavy, violet or light, burnt orange, lemon yellow, yellow ochre, and orange.
Straighten
At a workstation level, there are many tools and concepts that cain be applied, both
from time and movement studies, or from micro-movement studiies. For example,
using labels to identify bins, racks, products, etc., is very useful. Tools including
visual aids, drawings, marks, and colorful images are of great hellP when improving
the straighten phase at a workstation.

58 Process Level (Continued)
Therblig Icon Therblig Icon Therblig Icon
~ If
Transport
Plan
loaded '01 Assemble
Search ,.b Transport

empty '-.../ Disassemble
iT
Find <:0>
-- Release load r
'" Inspect
0
Select . Pre-position
c5
Avoidable
delay I 0
Grasp
n Position
~
Unavoidable
delay rb
~
Rest for
Hold
1L Use
U overcoming
fatigue
Figure 8.3 Therblig Symbols

(The Gilbreth Network, 2006)31
Shine
Typically, implementation teams define specific times during the day to perform
preventive maintenance, housekeeping, and workstation setup. This is necessary,
but the process needs review. Daily practices related to cleaning, sweeping, or
preventive maintenance can become excessive, trivial, or inconsequential to
employees. Cleaning twice a day, for example, is not shine. Remember that shine's
purpose is to prevent grime.
55 needs to be ingrained as a way of life for employees. When employees identify

how their personal cleaning habits can affect quality, then they have assimilated
shine at a personal level.

ELIMINATING WASTES/KANBAN-PULL
Kanban-Pull
T. Ohno of Toyota Motor Company was the originator of the kanban method. This
idea supposedly occurred to Ohno on a visit to the United States when he visited a
supermarket. In the supermarket, product is "pulled" from the shl~lf and the missing
item is replenished. This is the famous "pull" system in action (Shingo, 1989)25.
Liker (1997)14 suggests that the story of Ohno visiting an American supermarket to
develop Kanban is fiction.
Kanban is a method of material control in the factory. It is intended to provide

product to the customer with the shortest possible lead times. Inventory and lead
times are reduced through Heijunka (leveling of production). For example, if the
plant production goal for day 1 is 8 units of A and 16 units of B, and on day 2 it is
20 units of A and 10 units of B, the usual method is to produce all of A, followed by
all of B. This may be the most efficient use of time for the plant machinery, but,
since production will never go according to plan, the customer may change their
mind on day 2 and order less of A. This causes a pileup of inventory and possibly
increases cycle time.
To reduce the WIP and cycle time, the goal is to be able to produce each part, every
day in some order such as, 2 As, 1 B, 2 As, 1 B, etc. The factory must be capable of
producing such an arrangement. It requires control of the machinE!ry and production
schedule, plus coordination of the employees. {Liker, 1997)14
If a kanban system is used, with cards indicating the need to resupply, the method
of feeding an assembly line could be achieved using the following process:
1. Parts are used on the assembly line and a withdrawal kanban is placed in a
designated area.
2. A worker takes the withdrawal kanban to the previous operation to get

additional parts. The WIP kanban is removed from the parts pallet and put in
a specified spot. The original withdrawal kanban goes back to the assembly
line.
3. The WIP kanban card is a work instruction to the WIP op,erator to produce
more parts. This may require a kanban card to pull mate'rial from an even
earlier operation.
4. The next operation will see that it has a kanban card and will have permission
to produce more parts.
5. This sequence can continue further upstream.
VILLANOVA UNIVERSITY VIII -16 LE~.N SIX SIGMA HANDBOOK

ELIMINATING WASTES/KANBAN-PULL
Kanban-Pull (Continued)
The order to produce parts at anyone station is dependent on recelvmg an
instruction, the kanban card. Only upon receiving a kanban card will an operator
produce more goods. This system aims at simplifying paperwork, minimizing WIP
and finished goods inventories. Examples of kanban cards are shown below.
Production Instruction Production Instruction

KANBAN KANBAN
Code
t I Color Code
I I Color
RZC 5 MBT 8
Type Manual Type Automatic
Quantity 1 Set Quantity 5 Each
Style Standard Style Deluxe
Control 4M539ALR Control 1Z2673YQP
Figure 8.4 Examples of Kanban Cards
Due to the critical timing and sequencing of a kanban system, improvements are
continually made. A kanban system can not have production halted by machine
failures or quality problems. Only a specific amount of product is in the system at
any point in time. A stoppage will cause much distress throughout the production
system. Every effort is made to eliminate causes of machine downtime and to
eliminate sources of errors in production.
As the number of kanban cards is decreased, there is a resulting decrease in stock,

and stoppages are again highlighted. All causes of stoppages must be eliminated,
promoting production efficiency and improving quality. (Shingo, 1989)25
Shingo (1989)25 notes that kanban systems are applicable in repetitive production
plants, but not in one-of-a-kind production operations. Kanban is beneficial for
production systems involving parts with common processes.

ELIMINATING WASTES/POKA-YOKE
Poka-Yoke I Mistake Proofing

Shigeo Shingo (1986)24 is widely associated with a Japanese concept called poka-
yoke (pronounced poker-yolk-eh) which means to mistake proof the process. Mr.
Shingo recognized that human error does not necessarily create resulting defects.
The success of poka-yoke is to provide some intervention device or procedure to
catch the mistake before it is translated into nonconforming product.
Shingo (1986)24 lists the following characteristics of poka-yoke devices:
• 100% inspection is possible

• Poka-yoke devices avoid sampling for monitoring and control
• Poka-yoke devices are inexpensive
Poka-yoke devices can be combined with other inspection systems to obtain near
zero defect conditions. Errors can occur in many ways:
• Skipping an operation
• Positioning parts in the wrong direction
• Using wrong parts or materials
• Failing to properly tighten a bolt (Suzaki, 1993)29
There are numerous adaptive approaches. Gadgets or devices can stop machines
from working if a part or operation sequence has been missed by an operator. A
specialized tray or dish can be used prior to assembly to ensurEl that all parts are
present. In this case, the dish acts as a visual checklist. Otheli service oriented
checklists can be used to assist an attendant in case of interruption.
Numerous mechanical screening devices can be utilized. Applications can be based

on length, width, height, and weight. Cash registers at many fast food outlets have
descriptions or schematics of the product purchased. This system and the use of
bar codes at supermarkets eliminates data entry errors and save~i time. Obviously,
mistake proofing is a preventive technique.
Mistake proofing can also be accomplished through control methlods by preventing

human errors or by using a warning mechanism to indicate an error. Some of the
control methods to prevent human errors include:
• Designing a part so it can not be exchanged by mistake

• Using tools and fixtures that will not load a mis-positioned! part
• Having a work procedure controlled by an electric relay
VILLANOVA UNIVERSITY VIII -18 LEJ,N SIX SIGMA HANDBOOK

Poka-Voke I Mistake Proofing (Continued)

A signaling mechanism can alert a worker of possible sources of error. Several
applications include:
• Having the parts color-coded

• Having tool and fixture templates in place to only accept correct parts
• Having mechanisms to detect the insertion of a wrong part
A buzzer or light will signal that an error has occurred, requiring immediate action.
Root cause analysis and corrective action are required before work resumes.
Other than eliminating the opportunity for errors, mistake proofing is relatively
inexpensive to install and engages the operator in a contributing way. Work teams
can often contribute by brainstorming potential ways to thwart error-prone activities.
A disadvantage is, in many cases, that technical or engineering assistance is
required during technique development.
Design improvements to mistake proof products and processes include:
• Elimination of error-prone components

• Amplification of human senses
• Redundancy in design (back-up systems)
• Simplification by using fewer components
• Consideration of functional and physical environmental factors
• Providing fail-safe cut-off mechanisms
• Enhancing product producibility and maintainability
• Selecting components and circuits that are proven
Everyday Examples of Poka-Yoke
• Gas cap attached to a car

• Gas pumps with automatic shut-off nozzles
• 110 V electrical plugs and polarized sockets
• Microwave automatically stops when door is opened
• Seatbelt buzzer to warn drivers and passengers
• Elevator electric eye to prevent door from closing on people
• Lawn mower safety shut-off when bar is released
• Car keys ground symmetrical to allow two-way insertion
• Item drawings on cash registers at fast food restaurants
• Bar codes for product identification during distribution
VILLANOVA UNIVERSITY VIII- 19 LEAN SIX SIGMA HANDBOOK

Poka-Yoke I Mistake Proofing (Continued)
Errors Cause Countermeasures

Forgetful ness Poor Concentration ChecklistsNisual Aidls
Misunderstanding Unfamiliar Training I Checklists
Situations Work Standardization
Visual Aids I Work In:structions
Identification Similar Appearance Training I Visual Aids;
Amateurs Inexperience Training I Skill Building
Beginners Work Standardization
Visual Aids I Work In:structions
Willful Errors Ignoring Rules Training I Work Instructions
Inadvertent Errors Absent Minded Discipline I Skill Building
Work Standardization
Visual Aids I Checklh;ts
Slowness Judgment Delays Work Standardization
Visual Aids I Work In!structions
Lack of Standards Inadequate Work Standardization
Instructions Work Instructions
Surprise Errors Erratic Equipment TPM I Work Standardization
Intentional Errors Crimes Sabotage Education I Disciplim~
Figure 8.5 Countermeasures for Human Errors
Poka-yoke techniques are especially effective when:
• Vigilance is required in manual operations

• Mis-positioning of components can occur
• Attributes not measurements are important
• SPC is difficult to apply
• Turnover and training costs are high
• Special failure causes occur frequently

ELIMINATING WASTES/SET-UP REDUCTION
Setup Reduction
SUR is an acronym for setup reduction. SMED is an acronym for single minute
exchange of dies. In this discussion, the two terms will be used interchangeably.
SUR is one of the most important tools in the lean manufacturing system. The
concept is to take a long setup change of perhaps 4 hours in length and reduce it to
3 minutes. Most people can not believe that this is possible. Shigeo Shingo,
developer of the SMED system, used it quite effectively in the Toyota Production
System for just-in-time production. Single minute exchange of dies does not literally
require only one minute. It merely implies that die changes are to be accomplished
in a single digit of time (nine minutes or less). (Robinson, 1990)22
Long setup changes present a major problem for many companies dealing with low
volume production. American industry has long held the view that the best
production scheme would be long production runs of the same product. Witness
Henry Ford's Model T assembly line of only black cars. In today's world, that may
not be possible. There may be certain industries where the supplier is dominant in
comparison to the buyer. However, if the customer is more dominant, or the
industry is very competitive, being able to switch production very quickly can create
a competitive edge. There are 3 myths regarding setup times:
• The skill for setup changes comes from considerable practice and experience
• Long production runs are more efficient because they save setup times
• Long production runs are economically better
SUR systems reduce the dependence on the long-term experience of operators to

perform an effective changeover. Traditional methods depend on the setup operator
having full knowledge of the machinery and components, and unique skills in
mounting, removing, measuring, centering, adjusting, and calibrating setups. SUR
systems reduce the skill level needed for setup changes. Long runs will reduce
problems with setup changes, but lead to excess inventories, extra handling, extra
storage, etc. The reduction of setup times:
• Expands production capacity

• Reduces inventories and minimize wastes
• Prevents quicker responses to demand changes
• Increases operating flexibility
• Makes more effective use of floor space
• Improves the utilization of capital equipment
• Reduces material handling
• Increases operator efficiency and safety (in some cases)

Setup Reduction (Continued)

To achieve quick changeover one determines operations that must be done while the
machine is stopped, called internal setup (IS), and distinguish these from those
which can be done while machine runs, called external setup (ES). Any useless
steps are removed. The sequence consists of:
• Removing useless operations

• Converting IS to ES
• Simplifying fittings and installations
• Suppressing adjustments and trials
• Working continually on improvement ideas
Quick changeover methods involves the following sequence of actions:
• Document all elements of current setup

• Separate internal from external operations
• Convert internal setups to external setups
• Generate ideas for reducing external setups
• Generate ideas for reducing internal setups
• Evaluate/test new ideas
• Prepare for the next changeover using new ideas
• Standardize by documenting actions and new procedures
• Continuously improve the process
The traditional setup method requires that all machine operations be stopped and
then operators proceed to think about the setup. A better way is t() identify what can
be performed before shutting down the machine (external setup time), and then to
identify what has to be done when the machine is shut down (internal setup time).
A machine shut down means that no parts are being produced. lin planning a SUR
project, the actual conditions and steps of the die changeover must be detailed.
This can be done by:
• Use of a stopwatch for continuous observation
• Use of a work sampling study
• Worker interviews
• Videotaping the entire setup operation

Setup Reduction (Continued)

After the first step has been performed, a breakdown of the operation can begin.
The setup team may be comprised of operators, setup technicians, engineers, and
maintenance staff. This team reviews the setup elements. Every step in the setup
process (from start to finish) is broken down and classified. A major item is to
separate items that can be done when the machine is running (external setup) and
to separate the items that can only be done when the machine is down (internal
setup). External setup operations should include:
• Preparation of parts
• Finding parts
• Measuring parts
• Maintenance of dies and spares
• Cleaning of spares, etc.
The break down of initial elements into internal and external setup operations is just
a start. The existing internal setup elements should be reexamined to convert more
ofthose elements into external setup. The goal is to reduce the time to under 1 digit.
However, it may take a series of SUR projects to lower the time to 1 digit.
The setup team will need to generate some creative options. They should look for
pre-heating of dies, earlier preparation of parts, simplifying holding devices,
standardizing die heights, and using common centering jigs, multipurpose dies,
parallel operations (2 or more people working), functional clamps, one-turn
attachments, U-shaped washers, one-motion methods, interlocking methods,
elimination of adjustments, etc. Brainstorming and problem solving sessions are
needed to continuously improve the setup process.
All elements of internal and external setup must be reviewed in detail and
streamlined in order to attain the single digit goal. Perhaps the goal is unattainable,
but efforts should be made to go as low as possible. Once a SUR procedure is
agreed upon, the setup team should practice the process and critique itself for
additional improvements. (Robinson, 1990)22

Set-Up Reduction (Continued)

Based on specific applications, the reduction of cycle time can helve a considerable
impact on the containment of costs and improvement in productivity. Consider the
following case study, which is not a single minute exchange, but is representative
of the process.
SUR Case Study

A casting facility in Virginia decided to investigate the time nec:essary to replace
sand molds. In past years, the process had required 2 to 3 hours per change out.
Through a gradual evolutionary process, that time had been redUiced to an hour. A
cross functional team was assigned the investigative task. Members included
representatives of production, maintenance, engineering, and supl:!rvision. The team
had previously undergone problem solving and team dynamics training.
Over a period of three months, the change out time was reduced to a firm fifteen
minutes regardless of station and mold type. Some of the key ingredients in this
success included:
• The mold storage method

• The staging of molds
• The timing of the change out
• A redesign of the die hardware
At least two minor engineering modifications were required. Plant communications,

regarding the implementation of the new methods, were also part of the team's
charter. Interestingly, the team discovered that their nearest competitor was still
needing two hours to do the job. In this case, benchmarking the competitor (even
if ethical) would have been discouraging.
VILLANOVA UNIVERSITY VIII - 24 LE.lIN SIX SIGMA HANDBOOK

ELIMINATING WASTES/FLOW IMPROVEMENT
Continuous Flow Manufacturing (CFM)

In the lean thinking environment, continuous flow manufacturing is one of the basic
principles. The main concept is that material should always be moving one piece at
a time, at a rate determined by the needs of the customer. The flow of product must
be smooth and uninterrupted by:
• Quality issues • Handling methods

• Setups • Transportation arrangements
• Machine reliability • Staging areas
• Breakdowns • WIP inventory problems, etc.
• Distance
(Productivity, 1999)21
The mass production, or large lot production, world is a series of operations that
produce goods in large batches. The sequence of operations used in producing
large batch sizes results in waiting time between operations. Large lot production
has these faults:
• Longer lead times for customer orders and delivery

• Additional resources in terms of labor, energy, and space
• Additional product transportation expenses
• Increased product damage or deterioration costs
Continuous flow or one piece flow will:
• Deliver a flow of products to the customer with less delay

• Provide a mechanism to solve other production problems
• Lower the risk of losses through damage, deterioration, or obsolescence
• Require less storage and transport
(Productivity, 1999)21
Ideally, in a continuous flow manufacturing layout, the production steps are

arranged in a tight sequence, such as a straight line or U-shaped cell, without WIP,
using single piece flow. Inside this flow concept, each station and operator (in fact,
the whole system) must operate with complete reliability to achieve continuous flow
and the desired takt time. A system of production is attained with high quality levels
using a variety of defect elimination and detection techniques.

ELIMINATING WASTES/FLOW IMPROVEMENT
Continuous Flow Manufacturing (Continued)

The following techniques are important for continuous flow manufacturing:
• Poka-yoke to prevent defects from proceeding to the next: step
• Source inspection to catch errors and correct the process
• Self-checks to catch defects and to correct the process
• Successive checks by the next process to catch errors
• TPM is used to help achieve high machine capability
(Womack, 1996)33, {Robinson, 1990)22
Takt Time
In the operation of a continuous flow manufacturing line, takt time takes on great
importance. Takt time is a term used (first by Toyota) to define al time element that
equals the demand rate.
In a CFM or one piece flow line, the time allowed for each line operation is limited.
The line is ideally balanced so that each operator can perform their work in the time
allowed. The word, takt, is a German word for baton, used by an orchestra
conductor (Imai, 1997)10. This provides a rhythm to the pr04::ess, similar to a
heartbeat. The work pace is at a certain pace or rhythm. In CFMI, a certain pace is
maintained, and the line must be engineered to do so.
(Conner, 2001)'1, (Sharma, 2001)23

ELIMINATING WASTES/QUICK RESPONSE MANUFACTURING
Quick Response Manufacturing

According to Professor Ryjan Suri, quick response manufacturing is the next step
for the Toyota Production System (TPS). Ryjan Suri is Professor of Industrial
Engineering at the University of Wisconsin - Madison and Director of the Center for
Quick Response Manufacturing. TPS is now 30 to 40 years old and could be
considered an old technology. QRM helps companies use speed and the reduction
of cycle times to deliver products and services faster than their competitors. This
methodology can be applied to both the shop floor and the office.
In many cases, QRM requires that the managerial mind set must change. The
implementation of QRM in a company requires proper training and orientation to
grasp the dynamics ofthe manufacturing system. It is important to understand how
capacity planning, resource utilization, lot-sizing, etc., interact with each other and
impact lead times. This is very important in the relentless pursuit of lead time
reductions. QRM is especially useful for a product line that has a large variety of
highly engineered products with variable demand.
A specialized material planning technique that is a combination of both "push" and

"pull" termed "POLCA", is used for controlling material flow. POLCA stands for
Paired-cell Overlapping Loops of Cards with Authorization. This is a material control
system that operates in conjunction with MRP and a cellular arrangement. Some
examples of the benefits of QRM include:
• Reduction of lead times by 80% to 95%

• Lowered product costs by 15% to 30%
• Increased on time deliveries from 60% to 99%
• Decreased scrap by 80% (Suri, 2001)27, (Suri, 2006)28
The QRM methodology focuses on speed. Suri highlights these principles:
• Change the management mindset

• Find ways to complete a job, focusing on lead time minimization
• Plan to operate critical resources at 70% to 80%, not 100%
• Use reduction of lead time not utilization as the main performance metric
• Do not use equipment efficiency or utilization as the main metrics
• Lead time reduction is more important than on-time delivery
• Install the POLCA material control system
• Move the suppliers to QRM
• Educate customers on QRM in order to enable smaller lot sizes
• Use quick response office cells (teams) for product families
• QRM will lead to a truly lean company
(Center for QRM, 2006)2

ELIMINATING WASTES/KAIZEN
Kaizen
Kaizen is Japanese for continuous improvement (Imai, 1997)10. The word kaizen is
taken from the Japanese word kai meaning "change" and zen meaning "good." This
is usually referred to as incremental improvement, but on a continuous basis and
involving everyone. Western management is enthralled with radical innovations.
They enjoy seeing major breakthroughs, the "home runs" of business. Kaizen is an
umbrella term for:
• Productivity
• Total quality control
• Zero defects
• Suggestion systems
• Just-in-time production (Imai, 1997)10
Kaizen accomplishes improvements at little or no expense, and without the purchase

of sophisticated equipment. The kaizen strategy involves the following
considerations:
• Management maintains and improves operating standards

• Progress improvement is the key to success
• PDCA improvement cycles are used
• The next process is provided with good parts or informaUon
• Quality is of the highest priority
• Problems are solved with hard data (Imai,1997)10
The Kaizen Blitz

While most kaizen activities are considered to be long-term by numerous
individuals, a different type of kaizen strategy can occur. This has been termed a
kaizen event, kaizen workshop, or kaizen blitz, which involves a Jtaizen activity in a
specific area within a short period of time.
The kaizen blitz, using cross functional volunteers in a 3 to 5 day' period, results in
a rapid workplace change on a project basis. The volunteers clome from various
groups. Ifthe work involves a specific department, more team members are selected
from that department. Blitz teams often require a facilitator. Various metrics are
used to measure the outcomes of a kaizen blitz:
• Floor space saved • Increased quality levels

• More line flexibility • Safer working envir(mment
• Improved work flow • Reduced non-value .added time
• Improvement ideas

THEORY OF CONSTRAINTS
Theory of Constraints
The theory of constraints (TOC) is a system developed by E. Goldratt. In 1986,
Goldratt and Cox published a book titled The Goal (Goldratt, 1986f, which
introduced the subject. The Goal describes a process of ongoing continuous
improvement. Additional books have followed on the subject, including Theory of
Constraints (Goldratt, 1990)8.
Goldratt describes the theory of constraints as an intuitive framework for managing

based on the desire to continually improve a company. Using TOC, a definition of
the goals of the company are established along with metrics for critical measures.
(Goetsch, 2000)6
The Goal is a novel written in a story format describing the dual trials of a plant
manager as he struggles to simultaneously manage his plant and his marriage. The
key concept, "theory of constraints" is never mentioned as such, but is fed to the
reader in bits and pieces. Listed below are many of the key elemental pieces:
• Bottlenecks • Return on investment

• Throughput • Cash flow
• Inventory • Local optimums
• Operational expenses • Systems thinking
• Socratic way • Leadtimes
• Jonah • Reduction of batch sizes
• Common sense • Cost accounting
• Delivery of results • Fear of change
• Goals • Resistance
• Assumptions (most are incorrect) • Net profit
The Goal reminds readers that there are three basic measures to be used in the
evaluation of a system.
• Throughput
• Inventory
• Operational expenses
These measures are more reflective of the true system impact than machine
efficiency, equipment utilization, downtime, or balanced plants.

Theory of Constraints (Continued)

A few of the most widely used TOe concepts are detailed below::
• Bottleneck resources are: "Resources whose capacity is e(tual to or less than

the demand placed upon it. A non-bottleneck is any resourlce whose capacity
is greater than the demand placed on it." If a resource presents itself as a
bottleneck, then things must be done to lighten the load. Some of the
appropriate steps might be to offload material to relieve ,a bottleneck or to
make the bottlenecks work only on parts needed now. OnE! should beware of
lost production at a bottleneck, due to poor quality or rejects.
• Balanced plants are not always a good thing. One should not balance
capacity with demand, but "balance the flow of product thmugh the plant with
demand from the market." The plant may be capable of gemerating inventories
and goods at record levels, which jam up the plant's systems. The idea is to
make the flow through the bottleneck equal to market demand. One can do
more with less by just producing what the market requires at the time. It is
possible that the existing plant has more than enough resources to do any
job, but the flow must be controlled.
• Dependent events and statistical fluctuations are important. A subsequent

event depends upon the ones prior to it. The story of He,rbie and the local
scout pack describes how the slowest member of a group will restrain the
pace of the group. A bottleneck will restrain the entire throughput.
• Throughput is: "The rate at which the system generates money through
sales." The finished product must be sold before it can gtmerate money.
• Inventory is: "All the money that the system has inves1ted in purchasing
things that it intends to sell." This can also be defined as sold investments.
• Operational expenses are: "All the money that the system !spends in order to
turn inventory into throughput." This includes depreciatic)n, lubricating oil,
scrap, carrying costs, etc.
• The terms throughput, inventory and operational expenses, define money as:
"Incoming money, money stuck inside, and money going I)Ut."
"Quotes" from (Goldratt, 1986f


Goldratt (1990)8 recommends that the following 5-step method be used for TOC
implementation:
1. Identify the system's constraints. A system constraint limits the firm from
achieving its optimum performance and goals. Thus, constraints must be
identified and prioritized for maximum impact.
2. Decide how to exploit the system's constraints. The non-constraints in the

system should be managed properly so that resources or materials are
provided to feed the constraints.
3. Subordinate everything else to the above decisions. Constraints may have a

limit, so look for ways to reduce the effects of a constraint, or look to expand
the capacities of the constraints.
4. Elevate the system's constraints. Try to eliminate the problems of the

constraint. Strive to keep improving the system.
5. Back to step 1. After the constraint has been broken, go back to step one and
look for new constraints.
Drum - Buffer - Rope
The drum-buffer-rope concept relates to step 3 above: "Subordinate everything else

to the above decisions." Most discussions on bottlenecks or constraints center on
increasing the capacity or removing factors that slow the bottleneck. However, ifthe
capacity of the bottleneck cannot be increased, then one must accept it, and then
work to maximize the bottleneck's output.
One must ensure a smooth source of materials to the bottleneck. Because machines
and equipment have variation in their output, there may on occasion be too much
work-in-process (WIP) for the bottleneck, or not enough WIP for the bottleneck. The
ideal situation is to always have enough WIP for the bottleneck (which controls the
pace of the line) to keep production rates moving. Therefore, a set amount of
inventory (a buffer) is needed ahead of the bottleneck.


Drum - Buffer - Rope (Continued)
To maintain a proper buffer level, a feedback mechanism is necessary to control the

release of raw materials to the downstream equipment. One such 1technique is called
the drum-buffer-rope concept, as described below:
• Drum: This is the constraint that controls the pace of the process. The "beat"
of this operation sets the pace of the line.
• Buffer: This is the work-in-process, or inventory, for the bottleneck. It must

be available to keep the bottleneck operating at full performance.
• Rope: This is the feedback mechanism from the buffer te) the raw material
input point. The dispatching point will release only enough material to keep
the buffer inventory at the proper level.
The drum-buffer-rope technique can be thought of as a "pull-push" system. Refer

to Figure 8.6 below.
Buffer Drum
Shipping
Material
1--.... 1--... Inventory 1 -. . . or next
Release
operation
Feedback loop
Figure 8.6 Drum - Buffer - Rope Example
Pitcher (2003)18 indicates that the DBR model can work very well in a job shop with
its wide variety of products, routings, and process times. In this environment,
bottlenecks can be everywhere. The use of DBR methods ha:s led to excellent
performances in some situations, because WIP is kept low, and I()wer system cycle
times are achieved.
(Pitcher, 200~1)18, (Yang, 2005)32
VILLANOVA UNIVERSITY VIII - 32 LEJ~N SIX SIGMA HANDBOOK


Goldratt (1990)8 and Goetsch (2000)6 describe the TOe approach as the Socratic
method. This approach makes people find their own answers via the artful use of
questioning. Direct answers are not given, but people are guided to draw their own
conclusions and form their own opinions.
There are several other techniques used in the TOe environment:
• Effect-cause-effect: Use brainstorming to determine an intuitive sense of

problems and their causes. That is, for an effect, provide assumptions for the
effect, then speculate on a plausible cause. The cause is investigated to
verify its validity.
• Evaporating clouds: Find the conflicting assumptions in the composition of

a problem. Simple solutions are sometimes available for complex problems.
Find the solution by reexamining the basic foundations of the problem. Once
changes have been made to the system, the problem may no longer exist. It
may evaporate.
• Prerequisite trees: Something must occur before something else can occur.
TOe is a transition tool from an old way of doing things to a new way.

DESIGN OF EXPERIMENTS
DOE Introduction*
Classical experiments focus on OFAT or 1FAT (one factor at a time) at two or three
levels and attempt to hold everything else constant (which is impossible to do in a
complicated process). When DOE is properly constructed, it cain focus on a wide
range of key input factors or variables and will determine the optimum levels of each
of the factors. It should be recognized that the Pareto principle applies to the world
of experimentation. That is, 20% of the potential input factors generally make 80%
of the impact on the result. The classical approach to experimElntation, changing
just one factor at a time, has the following shortcomings:
• Too many experiments are necessary to study the effects of all the input
factors.
o The optimum combination of all the variables may never be revealed.
o The interaction (the behavior of one factor may be dependent on the level of
another factor) between factors cannot be determined.
• Unless carefully planned and the results studied statisticall)', conclusions may
be wrong or misleading.
o Even if the answers are not actually wrong, non-statisticcll experiments are
often inconclusive. Many of the observed effects tend to be mysterious.
• Time and effort may be wasted through studying the wrong variables or
obtaining too much or too little data.
Design of experiments overcomes these problems by careful p,lanning. In short,

DOE is a methodology of varying a number of input factors simultaneously, in a
carefully planned manner, such that their individual and combined effects on the
output can be identified. Advantages of DOE include:
• Many factors can be evaluated simultaneously, making the DOE process

economical and less interruptive to normal operations.
• Sometimes factors having an important influence on the output cannot be

controlled (noise factors), but other input factors can be controlled to make
the output insensitive to noise factors.
* A sUbstantial portion of the material throughout Section X comes from Wortman

(2006)36 and Wortman (2001)34.
VILLANOVA UNIVERSITY VIII - 34 LE,l\N SIX SIGMA HANDBOOK

DOE Introduction (Continued)

• In-depth, statistical knowledge is not always necessary to get a big benefit
from standard planned experimentation.
• One can look at a process with relatively few experiments. The important
factors can be distinguished from the less important ones. Concentrated
effort can then be directed at the important ones.
• Since the designs are balanced, there is confidence in the conclusions drawn.
The factors can usually be set at the optimum levels for verification.
• If important factors are overlooked in an experiment, the results will indicate

that they were overlooked.
• Precise statistical analysis can be run using standard computer programs.
• Frequently, quality and reliability can be improved with minimal trial costs.
In many cases, tremendous cost savings can be achieved.
DOE Applications
Situations, where experimental design can be effectively used, include:
• Choosing between alternatives
• Selecting the key factors affecting a response
• Response surface modeling to:
• Hit a target
• Reduce variability
• Maximize or minimize a response
• Make a process robust (despite uncontrollable factors)
• Seek multiple goals
VILLANOVA UNIVERSITY VIII·35 LEAN SIX SIGMA HANDBOOK

DOE Steps*
Getting good results from a DOE involves a number of steps:
• Set objectives
• Select process variables
• Select an experimental design
• Execute the design
• Ensure the data is consistent with the experimental assumptions
• Analyze and interpret the results
• Use/present the results (may lead to further runs or DOEs)
Important practical considerations in planning and running experiments are:
• Check the performance of gauges/measurement devices first

• Keep the experiment as simple as possible
• Check that all planned runs are feasible
• Watch out for process drifts and shifts during the run
• Avoid unplanned changes (e.g. switching operators at half time)
• Allow some time (and back-up material) for unexpected events
• Obtain buy-in from all parties involved
• Maintain effective ownership of each step in the experimental plan
• Preserve all the raw data-do not keep only summary avera.gesl
• Record everything that happens
• Reset equipment to its original state after the experiment
* Modified from NIST (2001)17 and used with permission.
A Typical DOE Checklist

Every experimental investigation will differ in detail, but the following checklist will
be helpful for many investigations.
• Define the objective of the experiment.
• The principle experimenter should learn as many facts abclut the process as
possible prior to brainstorming.
• Brainstorm a list of the key independent and dependent variiables with people
knowledgeable ofthe process and determine ifthese factors; can be controlled
or measured.
VILLANOVA UNIVERSITY VIII - 36 LE)!~N SIX SIGMA HANDBOOK

A Typical DOE Checklist (Continued)

• Run "dabbling experiments" when necessary to debug equipment or
determine measurement capability and get some preliminary results.
• Assign levels to each independent variable in the light of all available

knowledge.
• Select a standard DOE plan or develop one by consultation. It pays to have

one person outline the DOE and another review it critically.
• Run the experiments in random order and analyze results periodically.
• Draw conclusions. Verify by replicating experiments, if necessary, and

proceed to follow-up with further experimentation if an improvement trend is
indicated in one or more of the factors.
Select and Scale the Process Variables

Process variables include both inputs and outputs - i.e. factors and responses. The
selection of these variables is best done as a team effort. The team should:
• Include all important factors (based on engineering and operator judgments)
• Be bold, but not foolish, in choosing the low and high factor levels
• Avoid factor settings for impractical or impossible combinations
• Include all relevant responses
• Avoid using responses that combine two or more process measurements
When choosing the range of settings for input factors, it is wise to avoid extreme
values. In some cases, extreme values will give runs that are not feasible.
The most popular experimental designs are called two-level designs. They are
simple and economical. They are ideal for screening designs, and give most of the
information required to go to a multi-level response surface experiment, if one is
needed. However, it is often desirable to include some center points (for quantitative
factors) during the experiment (center points are located in the middle of the design
"box").

Experimental Objectives*
Choosing an experimental design depends on the objectives of the experiment and
the number of factors to be investigated. Some objectives are dliscussed below:
1. Comparative objective: If several factors are under investigatkm, but the primary
goal of the experiment is to make a conclusion about whethE~r a factor (in spite
of the existence of the other factors) is "significant," then the-experimenter has
a comparative problem and needs a comparative design solution.
2. Screening objective: The primary purpose of this experimEmt is to select or

screen out the few important main effects from the many lessler important ones.
These designs are also termed main effects or fractional fact,orial designs.
3. Response surface (method) objective: This experiment is clesigned to let an

experimenter estimate interaction effects, and, therefore, giive an idea of the
shape of the response surface under investigation. RSM designs are used to:
• Find improved or optimal process settings

• Troubleshoot process problems and weak points
• Make a product or process more robust against external influences
Number Comparative Screening Objective Respl)nse Surface

of Factors Objective Objec:tive
1 1 factor completely
randomized design
-- --
2-4 Randomized block Full or fractional Centr.al composite or
design factorial Box-Behnken
5 or more Randomized block Fractional factorial Screen first to reduce
design or Plackett-Burman number of factors
Table 8.7 Design Selection Guidelines
The choice of a design depends on the amount of resources availalble and the degree
of control over making wrong decisions that the experimenter desires. It is a good
idea to choose a design that requires somewhat fewer runs than the budget permits,
so that additional runs can be added to check for curvature alnd to correct any
experimental mishaps.
VILLANOVA UNIVERSITY VIII - 38 LE,4!~N SIX SIGMA HANDBOOK

The Iterative Approach to DOE

It is often a mistake to believe that one big experiment will give the answer. It is
more common to perform two or three, or perhaps more, experiments before a
complete answer is attained. In other words, an iterative approach is best and, in the
end, is the most economical. Putting all of one's eggs in one basket is not
advisable. The reason an iterative approach frequently works best is because it is
logical to move through stages of experimentation, each stage supplying a different
kind of answer.
Experimental Assumptions*
In all experimentation, one makes assumptions. Some of the engineering and
mathematical assumptions an experimenter makes include:
• Are the measurement systems capable for all responses?
• Is the process stable?
• Are the residuals (the difference between the model predictions and the actual
observations) well behaved?
Is the Measurement System Capable?
It is not a good idea to find, after finishing an experiment, that the measurement
devices are incapable. This should be confirmed before embarking on the
experiment itself. In addition, it is advisable, especially if the experiment lasts over
a protracted period, that a check be made on all measurement devices from the start
to the conclusion of the experiment. Strange experimental outcomes can often be
traced to "hiccups" in the metrology system.
Is the Process Stable?
Experimental runs should have control runs which are done at the "standard"
process setpoints, or at least at some identifiable operating conditions. The
experiment should start and end with such runs. A plot of the outcomes of these
control runs will indicate if the underlying process itself drifted or shifted during the
experiment. It is desirable to experiment on a stable process. However, if this
cannot be achieved, then the process instability must be accounted for.

Experimental Assumptions (ContinUE!d)

Are the Residuals Well Behaved?
Residuals are estimates of experimental errors obtained by subtra,cting the observed

response from the predicted response. The predicted response is calculated from
the chosen model, after all the unknown model parameters have been estimated
from the experimental data. Residuals can be thought of as elements of variation
unexplained by the fitted model. Since this is a form of error, the same general
assumptions apply to the group of residuals that one typically uses for errors in
general: one expects them to be normally and independently distributed with a mean
of 0 and some constant variance.
These are the assumptions behind ANOVA and classical regression analysis. This
means that an analyst should expect a regression model to err in predicting a
response in a random fashion; the model should predict values higher and lower
than actual with equal probability. In addition, the level of thl~ error should be
independent of when the observation occurred in the study, 4)r the size of the
observation being predicted, or even the factor settings involved in making the
prediction.
The overall pattern of the residuals should be similar to the bE!II-shaped pattern
observed when plotting a histogram of normally distributed data. Graphical
methods are used to examine residuals. Departures from assumptions usually mean
that the residuals contain structure that is not accounted for in the model.
Identifying that structure, and adding a term representing it to the original model,
leads to a better model. Any graph suitable for displaying the distribution of a set
of data is suitable for judging the normality of the distributilon of a group of
residuals. The three most common types are: histograms, normal probability plots,
and dot plots. Shown below are examples of dot plot results.
• ~-
a
.-
-------:
/.~
• • •• •
~ ~
•• a . a a• •
E • • • • • • ••
• •• • •• • ••
• •
Residuals suggest the X1 Residuals suggest that Residuuls suggest the

model is properly the variance increases need for a quadratic term
specified. with X2 • added to X3 •
Figure 8.8 Residual Types

DOE Terms
Balanced A fractional factorial design in which an equal number of trials

Design (at every level state) is conducted for each factor.
Block A subdivision of the experiment into relatively homogenous
experimental units. The term is from agriculture, where a
single field would be divided into blocks for different
treatments.
Blocking When structuring fractional factorial experimental test trials,
blocking is used to account for variables that the experimenter
wishes to avoid. A block may be a dummy factor which
doesn't interact with the real factors.
Box-Behnken When full second-order polynomial models are to be used in
response surface studies of three or more factors, Box-
Behnken designs are often very efficient. They are highly
fractional, three-level factorial designs.
Collinear A collinear condition occurs when two variables are totally
correlated. One variable must be eliminated from the analysis
for valid results.
Confounded When the effects of two factors are not separable. In the
following example, A, B, and C are input factors and columns
AB, AC, & BC represent interactions (multiplication of 2
factors). Confounded columns are identified by arrows,
indicating the setting of one cannot be separated from the
setting of the other. See the examples below:
A 8 C AS AS;.~ A 8 .c AS A.c~
+ + + + - + -
+ - + - + - + + -
+
+ + +
+
+ + +
G + + +
+ + +
11 t j'
JJ 11 t
Figure 8.9 A is confounded with BC, B with AC, and C with AB
l'
JJ

DOE Terms (Continued)
Correlation A number between -1 and 1 that indicates thE~ degree of linear

Coefficient (r) relationship between two sets of numbers. Zero (0) indicates
no linear relationship.
Covariates Things which change during an experiment which had not
been planned to change, such as temperature or humidity. The
test order should be randomized to alleviate this problem.
Curvature Refers to non-straight line behavior between one or more
factors and the response. Curvature is uSUiilly expressed in
mathematical terms involving the square or cube ofthe factor.
For example, in the model:
Y = Bo + B1X1 + B11 (X1 • X1) + E

The term B11 (X1 • X1) describes curvature.
Degrees of The term used is DOF, DF, df or V. The number of
Freedom measurements that are independently available for estimating
a population parameter.
Design of The arrangement in which an experimental Iprogram is to be
Experiments conducted and the selection of the levels of one or more
(DOE) factors or factor combinations to be included in the
experiment. The term SDE (statistical design of experiment) is
also widely used.
EVOP Stands for evolutionary QReration, a term that describes the
way sequential experimental designs can adapted by learning
from present results and predicting future trecltments for better
responses. Often, small response improvemE~nts may be made
via large sample sizes. The experimental risk is quite low
because the trials are conducted in the near vicinity of an
already satisfactory process.
Experimental Variation in response or outcome of virtually identical test
Error conditions. This is also called residual error.
Experiment A test undertaken to make an improvement in a process or to
learn previously unknown information.

First-order Refers to the power to which a factor appears in a model. If

"X1" represents a factor and "8" is its factor effect, then the
model:
is first-order in both X1 and X2 • First-order models cannot

account for curvature or interaction.
Fractional An adjective that means fewer experiments than the full design
calls for. The two, three-factor designs shown below are two-
level, half-fractional designs.
c
+
+ + +
+ + +
+ + + + +
Full Factorial Describes experimental designs which contain all
combinations of all levels of all factors. No possible treatment
combinations are omitted. A two-level, three factor full
factorial design is shown below:
+
+
+ +
+
+ +
+ +
+ + +
Input factor An independent variable which may affect a (dependent)
response variable and is included at different levels in the
experiment.

Inner Array In Taguchi style fractional factorial experimtent, these are the
factors that can be controlled in a process.
Interaction Occurs when the effect of one input factc>r on the output
depends upon the level of another input facltor.
No Interaction Interaction
~Have
Q) Q)
(/) (/)
s:: s::
0 0
c. c.
~ Eaten
(/) (/)
Q) Q)
0:: 0::
Haven't
Eaten Drugs
2 4 6 8 0 1 2 3
# Drinks # Drinks
Figure 8.10 Interaction Illustration
Level A given factor or a specific setting of an input factor. Four

levels of a heat treatment may be 100 of, 1~~0 of, 140 of and
160 of.
Main Effect An estimate of the effect of a factor independent of any other
factors.
Mixture Experiments in which the variables arE! expressed as
Experiments proportions of the whole and sum to 1.0.
Multi- This occurs when two or more input factors. are expected to
Collinearity independently affect the value of an outPUIt factor, but are
found to be highly correlated. For example, an experiment is
being conducted to determine the market value of a house.
The input factors are square feet of living space and number
of bedrooms. In this case, the two input felctors are highly
correlated. Larger residences have more bedrooms.

Nested An experimental design in which all trials are not fully

Experiments randomized. There is generally a logical reason for taking this
action. For example, in an experiment, technicians might be
nested within labs. As long as each technician stays with the
same lab, the techs are nested. It is not often that techs travel
to different labs just to make the design balanced.
Optimization Involves finding the treatment combinations that give the most
desired response. Optimization can be "maximization" (in the
case of product yields) or "minimization" (in the case of
impurities).
Orthogonal A design is orthogonal if the main and interaction effects in a
given design can be estimated without confounding the other
main effects or interactions. A full factorial is said to be
balanced, or orthogonal, because there are an equal number
of data points under each level of each factor.
Outer Array In a Taguchi-style fractional factorial experiment, these are the
factors that cannot be controlled in a process.
Paired The basis of a technique for treating data so as to ignore
Comparison sample-to-sample variability and focus more clearly on
variability caused by a specific factor effect. Only the
differences in response for each sample are tested because
sample-to-sample differences are irrelevant.
Parallel These experiments are done at the same time, not one after
Experiments another. (e.g., agricultural experiments in a big corn field).
Precision The closeness of agreement between test results.
Qualitative This refers to descriptors of category and/or order, but not of
interval or origin. Different machines, operators, materials,
etc. represent qualitative levels or treatments.
Quantitative This refers to descriptors of order and interval (interval scale)
and possibly also of origin (ratio scale). As a quantitative
factor, temperature might describe the interval value 27.32°C.
As a quantitative response, yield might describe the ratio value
87.42%.

Randomized Frees an experiment from the environmenjt and eliminates

Trials biases. This technique avoids the undue influences of
systematic changes that are known or unknc,wn.
Repeated Trials Trials that are conducted to estimate the pure trial-to-trial
experimental error so that lack of fit may be judged. Also
called replications.
Residual Error The difference between the observed and thE! predicted value
(£) or (E) for that result, based on an empirically determined model. It
can be variation in outcomes of virtually identical test
conditions.
Residuals The difference between experimental responsies and predicted
model values.
Resolution I An experiment in which tests are conducted, adjusting one
factor at a time, hoping for the best. This experiment is not
statistically sound.
Resolution II An experiment in which some of the main effects are
confounded. This is very undesirable.
Resol ution III A fractional factorial design in which no main effects are
confounded with each other, but the main effects and two
factor interaction effects are confounded.
Resolution IV A fractional factorial design in which the main effects and two
factor interaction effects are not confounded but the two factor
effects may be confounded with each other.
Resolution V A fractional factorial design in which no con110unding of main
effects and two factor interactions occurs" However, two
factor interactions may be confounded with three factor and
higher interactions.
Response The graph of a system response plotted against one or more
Surface system factors. Response surface methodology employs
Methodology experimental design to discover the "shape'~1 of the response
(RSM) surface and then uses geometric concepts te) take advantage
of the relationships discovered.

Response The variable that shows the observed results of an

Variable experimental treatment. Also known as the output or
dependent variable.
Robust Design A term associated with the application of Taguchi
experimentation in which a response variable is considered
robust or immune to input variables that may be difficult or
impossible to control.
Screening A technique to discover the most (probable) important factors
Experiment in an experimental system. Most screening experiments
employ two-level designs. A word of caution about the results
of screening experiments: if a factor is not highly significant,
it does not necessarily mean that it is insignificant.
Second-order Refers to the power to which one or more factors appear in a
model. If X1 represents a factor and 8 1 is its factor effect, then
the model:
is second-order in X1 but not in X2 • Second-order models can

account for curvature and interaction.
Sequential Experiments are done one after another, not at the same time.
Experiments This is often required by the type of experimental design being
used.
Simplex Design A spatial design used to determine the most desirable variable
combination (proportions) in a mixture.
Test Coverage The percentage of all possible combinations of input factors in
an experimental test.
Treatments In an experiment, the various factor levels that describe how
an experiment is to be carried out. For example, a pH of 3 and
a temperature of 37°C describe an experimental treatment.

A Three Factor, Three Level Experimont

Often a three-factor experiment is required after screening a larger number of
variables. These experiments may be full or fractional factorial. Shown below is a
one-third fractional factorial design. Generally the (-) and (+) levels in two-level
designs are expressed as 0 and 1 in most design catalogues. Three-level designs
are often represented as 0, 1, and 2.
Concentration (:!22 )
( 200 ) ~~-~--'-':::;;;"--"T""I""
('122 )
( 022 )
( 000 )
-- (001 ) (002 )
Temperature
Figure 8.11 A One-third Fractional Factorial Design, 3 Factors, 3 Levels
From a design catalogue test plan*, the selected fractional factoriall experiment looks
like so:
Experiment Concentration Pressure Temperature

1 0 0 0
2 0 1 2
3 0 2 1
4 1 0 1
5 1 1 0
6 1 2 2
7 2 0 2
8 2 1 1
9 2 2 0
* An example is Hahn and Shapiro (1966)9
VILLANOVA UNIVERSITY VIII - 48 LEJ!.N SIX SIGMA HANDBOOK

Randomized Block Plans

In comparing a number of factor treatments, it is desirable that all other conditions
be kept as nearly constant as possible. The required number of tests may be too
large to be carried out under similar conditions. In such cases, one may be able to
divide the experiment into blocks, or planned homogeneous groups. When each
group in the experiment contains exactly one measurement of every treatment, the
experimental plan is called a randomized block plan.
For example, an experimental scheme may take several days to complete. If we

expect some biasing differences among days, we might plan to measure each item
on each day, or to conduct one test, per day on each item. A day would then
represent a block. A randomized incomplete block (tension response) design is
shown below.
Treatment
Block (Days) A 8 C D
1 -5 Omitted -18 -10
2 Omitted -27 -14 -5
3 -4 -14 -23 Omitted
4 -1 -22 Omitted -12
Only treatments A, C, and D are run on the first day. 8, C, and D on the second day,
etc. In the whole experiment, note that each pair of treatments, such as 8C occurs
twice together. The order in which the three treatments are run on a given day
follows a randomized sequence.
Another randomized block design for air permeability response is shown below:
Fabric Types I, II, III & IV

I II III IV
8(15.1) D(11.6) A(15.4) C(9.9)
Chemical C(12.2) C(13.1) 8(16.3) D(9.4)
Applications
A,B,C,D A(19.0) 8(17.6) D(16.0) 8(8.6)
D(11.5) A(13.0) C(10.8) A(11.5)

latin Square Designs

A Latin square plan is often useful when it is desirable to allow for two sources of
non-homogeneity in the conditions affecting test results. Such designs were
originally applied in agriculture when the two sources of non-homogeneity were the
two directions on the field. The square was literally a plot of gr()und.
In Latin square designs, a third variable, the experimental treatment, is then applied
to the source variables in a balanced fashion. The Latin square plan is restricted by
two conditions:
• The number of rows, columns and treatments must be the same.
• There should be no expected interactions between row and column factors,

since these cannot be measured. Ifthere are, the sensitivit~, of the experiment
is reduced.
A Latin square design is essentially a fractional factorial experim,ent which requires

less experimentation to determine the main treatment results.
Consider the following 5 x 5 Latin square:
Carburetor Type
Car I II III IV V
1 A B C D E
2 B C D E A
3 C D E A B
4 D E A B C
5 E A B C D
In the above design, five automobiles and five carburetors are us<ed to evaluate gas
mileage by five drivers (A, B, C, D, and E). Note that only twenty-five of the potential
125 combinations are tested. Thus, the resultant experiment is a one-fifth fractional
factorial.
Similar 3 x 3, 4 x 4, and 6 x 6 designs may be utilized.

Graeco-Latin Designs
Graeco-Latin square designs are sometimes useful to eliminate more than two
sources of variability in an experiment. A Graeco-Latin design is an extension of the
Latin square design, but one extra blocking variable is added for a total of three
blocking variables.
Consider the following 4 X 4 Graeco-Latin design:
Carburetor Type
Car I II III IV Drivers
A,B,C,D
1 Aa B(3 Cy Dl5
2 Bl5 Ay D(3 Ca
3 C(3 Da Al5 By Days
a, (3,y,l5
4 DV Cl5 Ba AI3
The output (response) variable could be gas mileage by 4 drivers (A, B, C, D).
Hyper-Graeco-Latin Designs
A Hyper-Graeco-Latin square design permits the study oftreatments with more than
three blocking variables.
Consider the following 4 x 4 Hyper-Graeco Latin design:
Carburetor Type
Car I II III IV Drivers Tires
A,B,C,D M,N,O,P
1 AaMcp B(3NX CyO\fJ Dl5P!l
2 Bl5N!l AyM\fJ D(3PX CaOcp
3 C(30X DaPcp Al5M!l ByN\fJ Days Speeds
a, (3,y,l5 cp,X,\fJ,!l
4 DyP\fJ Cl50!l BaNcp A(3MX
The output (response) variable could be gas mileage by 4 drivers (A, B, C, D).

Plackett-Burman Designs*
Plackett-Burman (1946)20 designs are used for screening experiments. PB designs
are very economical. The run number is a multiple of four rather than a power of 2.
PB geometric designs are two-level designs with 4, 8, 16, 32, 64 11 and 128 runs and
work best as screening designs. Each interaction effect is confounded with exactly
one main effect.
All other two-level PB designs (12, 20, 24, 28, etc.) are non-geometric designs. In
these designs a two-factor interaction will be partially confounded with each of the
other main effects in the study. Thus, the non-geometric designs are essentially
"main effect designs," when there is reason to believe any interactions are of little
practical importance. A PB design in 12 runs, for example, may be used to conduct
an experiment containing up to 11 factors. See Table 8.12.
Factors
Exp X1 X2 X3 X4 X5 X6 X7 X8 X9 X10 I X11 Results
1 + + + + + + + + + + +
2 - + - + + - - + - - +
3 - - + - ++ - - - +
+
4 + - - + - + + - - -
+
5 - + - - + - + + + - -
6 - - + - - + - + + + -
7 - - - + - - + - + + +
8 + - - - + - - + - + +
9 + + - - - + - - + - +
10 + + + - - - + - - + -
11 - + + + - - - + - - +
12 + - + + + - - - + - -
Table 8.12 Plackett-Burman Non-Geometric Design (12 Runs/11 Factors)
With a 20-run design, an experimenter can do a screening experiiment for up to 19

factors. As many as 27 factors can be evaluated in a 28 run design.
* Source NIST (2001)17. Used with permission.

Two-Level Fractional Factorial Example

The basic steps for a two-level fractional factorial design will be examined via the
following hypothetical example. A seven-step procedure will be followed:
1. Select a process
2. Identify the output factors of concern
3. Identify the input factors and levels to be investigated
4. Select a design (from a catalogue, Taguchi, self-created, etc.)
5. Conduct the experiment under the predetermined conditions
6. Collect the data (relative to the identified outputs)
7. Analyze the data and draw conclusions
Gasoline Mileage Experiment

Step 1 Select a process *
Drive a car a distance of 375 miles, using 15 gallons of gasoline or less.
Step 2 Identify the output factors
Gasoline mileage performance is based on two results (output factors):
1. Will the destination be reached using the allowed amount of gasoline?
2. What is the optimum gasoline mileage possible?
Step 3 Establish the input factors and levels to be investigated
Study the effects of seven variables at two levels, that are suspected of
affecting gasoline mileage performance.
* The student should again note that this is a hypothetical experiment. However,
it is illustrative of how experimentation can be applied to a variety of processes.

Gas Mileage Experiment (Continuedl)

Step 3 Inputs and levels (continued) (7-factors at 2-1evels)
Input Factors Level 1 (-) Level 2 (+)

Factor A Position of windows Open Closed
Factor 8 Driver Sue Steve
Factor C Speed 55 mph 75 mph
Factor D Air conditioner On Off
Factor E Time of day Daytime Night
Factor F Grade of gasoline Regular Super
Factor G Car weight 2,5001b 3,5001b
Note: The above inputs are both variable (quantitative) and attribute (qualitative).
Table 8.13 Design Using 7-factors at 2-levels
Step 4 Select a design
A screening plan is selected from a design catalogue. Only E!ight (8) tests are
needed to evaluate the main effects of all 7-factors at 2-levels. The design is:
Input Factors
Test A 8 C D E F G
#1 - - - - - - -
#2 - - - + + + +
#3 - + + - - + +
#4 - + + + + - -
#5 + - + - + - +
#6 + - + + - + -
#7 + + - - + + -
#8 + + - + - - +
Table 8.14 Selected Design

Gas Mileage Experiment (Continued)

Step 4 Select a design (continued). One test example:
Input Factors
Test A B C D E F G
#3 oo + + oo oo + +
Table 8.15 One Test Example from the Selected Design
Test #3 means:
A (-) '= Windows are open E (oo) = Time of day is daytime
B (+) = Driver is Steve F (+) = Super grade of gasoline
C (+) = Speed is 75 mph G (+) = Car weight is 3,500 Ib
D (-) = Air conditioner is on
Step 5 Conduct the experiment
Step 6 Collect the data
Input Factors Outputs*

Test A B C D E F G Arrive Mpg
#1 oo oo oo oo oo oo oo Yes 26
#2 .. .. .. + + + + No 21
#3 .. + + .. .. + + No 15
#4 .. + + + + .. .. No 20
#5 + .. + .. + .. + No 17
#6 + .. + + .. + .. No 24
#7 + + .. .. + + .. Yes 28
#8 + + .. + .. .. + Yes 25
Average mpg 22
* Arrive "Yes" requires going 375 miles, using 15 gallons of gasoline or less.
=
Mpg miles per gallon of gasoline performance.
Table 8.16 Experiment Output Results
VILLANOVA UNIVERSITY VIII .. 55 LEAN SIX SIGMA HANDBOOK


Step 7 Analyze the data and draw conclusions
Input Factors Outputs*

Test A B C D E F G Arrive mpg
#1 -26 -26 -26 -26 -26 -26 -26 Yes 26
#2 -21 -21 -21 +21 +21 +21 +21 No 21
#3 -15 +15 +15 -15 -15 +15 +15 No 15
#4 -20 +20 +20 +20 +20 -20 -20 INo 20
#5 +17 -17 +17 -17 +17 -17 +17 No 17
#6 +24 -24 +24 +24 -24 +24 -24 INo 24
#7 +28 +28 -28 -28 +28 +28 -28 Yes 28
#8 +25 +25 -25 +25 -25 -25 +25 Yes 25
Average mpg 22
Difference /!. 12 0 -24 4 -4 0 -20 Effect

Effect /!'+4 3 0 -6 1 -1 0 -5 Gross Net
Abs(/!'+4) 3 0 6 1 1 0 5 16 8
Maximum mpg 30
Table 8.17 Experiment Data Analysis
(+) Means changing from level 1 to level 2 has a positive effect

(-) Means changing from level 1 to level 2 has a negative effect
o Means changing from level 1 to level 2 has no effect
1. The arrive pattern of "Yes" and "No" does not track with any single input factor.
2. The difference (/!.) values are the sum of values for each input factor. The effects
are the differences divided by 4 since four tests were at both high and low levels.
The absolute values are then taken for the effect values. Sum the absolute values
to give the gross effect and divide this by two for the net effect of improvement.
The gross effect is divided by 2 because the experiment is conducted in the
middle ofthe high and low levels and only halfthe difference (~.) can be achieved.
VILlANOVA UNIVERSITY VIII-56 LEJ~N SIX SIGMA HANDBOOK


Step 7 (continued)
3. The previous analysis leads to the following conclusions:
• Factor A Position of windows Level 2 (+) Closed improves by 3 mpg

• Factor B Driver Sue or Steve has no effect on mpg
• Factor C Speed Level 1 (-) 55 mph improves by 6 mpg
• Factor D Air conditioner Level 2 (+) Off improves by 1 mpg
• Factor E Time of day Level 1 (-) Daytime improves by 1 mpg
• Factor F Grade of gasoline Regular or Super has no effect on mpg
• Factor G Car weight Level 1 (-) 2,500 Ib improves by 5 mpg
4. To calculate the optimum gasoline mileage:
The net effect of 8 mpg is added to the average of 22 mpg to determine the
maximum of 30 mpg. The optimum gasoline mileage would be obtained by
running the following trial:
Factor A Factor B Factor C Factor D Factor E Factor F Factor G

Level 2 (+) 0 Level 1 (-) Level 2 (+) Level 1 (-) 0 Level 1 (-)
Where: Level 2 ::: (+) Level 1 ::: (-) o ::: Does not matter
Table 8.18 Optimum Factor Levels
The above trial was one of the 120 tests not performed out of 128 possible choices.
DOE is almost magical. The predicted gasoline mileage can be confirmed by
additional experimentation.
28
22
16
-A +A -8 +8 -C +C -0 +0 -E +E -F +F -G +G
Figure 8.19 Miles per Gallon Gasoline Mileage vs. Factor Levels

VALIDATING SOLUTIONS
Implementing and Validating Solutions

There are a number of tools and techniques that can be used to assist in the
implementation and validation of solutions. The following post improvement
considerations will be discussed here.
• Brainstorming
• FMEA
• Multi-vari reanalysis
• Post-improvement capability analysis
• DOE improvement analysis (manual, MINITAB, F test, t test, etc.)
• Measurement system reanalysis
Brainstorming
Quite often, brainstorming is useful in generating ideas for both simple problem
solving approaches (basic quality tools) or advanced problem solving approaches
(DOE or ANOVA).
In a perfect world, a problem will be satisfactorily resolved, yield will be increased,

or the customer will stop complaining, etc. In the real world, there may be a need
to revisit the brainstorming process. That is, the result was no~ quite as good as
anticipated or the DOE failed to find the critical input factors. In these cases, the
original brainstormed list can be revisited or the improvement team can simply start
over with fresh ideas.
FMEA
Consider that a design improvement team has worked diligently on a power lock
assembly with a potential failure mode of clamping force loss. The risk assessment
(RPN) before improvement was an arbitrary value of 32. After improvement, the RPN
is a value of 8.
Obviously, there has been a dramatic improvement. Whether the team has
completed their task, at this point, depends upon the hazard cc)nsequences of a
failure, as well as the RPNs for other potential failure modes within the same power
lock assembly.

Multi-Vari Re-analysis
Working on a process using multi-vari analysis is somewhat like draining a swamp.
The large stumps are uncovered first. In Figure 8.20 it is noted in the left drawing
that a substantial amount of variation occurred over time. This has now been
corrected by an improvement team, as reflected in the current drawing on the right.
Now the improvement team must decide if reducing the remaining within-piece
variation is worthy of additional effort (since piece-to-piece variability is of much
lower concern)
USL 0.110"
4
!3
Aim 0.105" 2
4 1
3
2
1
LSL 0.100"
Time ~ Time ---.
Figure 8.20 Multi-Vari Charts Reflecting Before and After Improvement
Post-Improvement Capability Analysis

Assume that an improvement team has focused their attention on reducing the
variability in a process. As shown in the left portion of Figure 8.21 below, the Cp was
1.0 before improvement. The right portion of Figure 8.21 indicates that a Cp of 1.67
has now been attained. The only decision facing the team at this point is whether
the new Cp is satisfactory.
USL LSL USL LSL
AIM AIM
Figure 8.21 Capability Studies Before and After Improvement

Response Surfaces
An experimental equation represents a response line, plane, or surface for the
factors being evaluated. Refer to Figure 8.22 in which the S, C, and T values depend
on the size of slopes, curves, and twists, respectively.
Y = 8 0 + S1X1 + S2X2 ..• ",S)(j

+ C1X 12 + C2X22 .•. ",CjXj2
+ T1,2X 1X2 ••• ...Tj,jXIX j
<-------2 D Picture---------><--3 D Picture -----><----More than a 3 D Picture------->
Figure 8.22 Description of Equation Componen1ts
3D Response Surface Matching Dome Contour
Additive Additive
Figure 8.23 Comparison of 3D and 2D Response Surfaces
Rising Ridge Stationary Ridge Saddle Minimax
Figure 8.24 Various Contour Examples

Response Surfaces (Continued)

The improvement team or principal experimenter should not assume that just
because yield or some other output result has improved from 56% to 64% that the
optimum result has been attained. As demonstrated with some of the contours on
the prior page, yields of 80% or 90% are possible from the two input variables
X1 and X2 • They simply haven't been discovered yet.
Certain surface response experiments are helpful in reaching the optimum results
in a relatively few experiments. However, there are experimental risks involving
potential product losses using these techniques.
EVOP Evolutionary Operations

Certain response surface designs emphasized a bold strategy for improvement. By
contrast, EVOP emphasizes a conservative experimental strategy for continuous
process improvement. Refer to Figure 8.25 below:
91% 94%
E E
69% 83% 96%

B B D
71%
pH A
E
92%
69% B D
A 87% 88%
A A C
63% 70% 84%
Concentration
Figure 8.25 Illustration of EVOP Experimentation
Tests are carried out in phase A until a response pattern is established. Then phase
B is centered on the best conditions from phase A. This procedure is repeated until
the best result is determined. When nearing a peak, switching to smaller step sizes
or examining different variables are helpful approaches. EVOP usually involves
small incremental changes so that little or no process scrap is generated. Large
sample sizes may be required to determine the appropriate direction of
improvement. The method can be extended to more than two variables.

DOE Improvement Analysis

DOE analysis using both manual and computer techniques will be illustrated with the
following example.
Suppose that pressure, temperature, and concentration are three suspected key
variables affecting the yield of a chemical process which is currently running at 64%.
An experimenter may choose to fix these variables at two levels (high and low) to
see how they influence yield. In order to find out the effect of all three factors and
= =
their interactions, a total of 2 x 2 x 2 23 8 experiments* must be conducted. This
is called a full factorial experiment. The low and high levels of input factors are
noted below by (-) and (+).
Experiment No. Temperature Pressure Concentration Yield %

1 - - - 55
2 + - . 77
3 - + - 47
4 + + - 73
5 - .. + 56
6 + - + 80
7 - + + 51
8 + + + 73
Average 64
Temperature: (-) =120 DC (+) =150 DC

Concentration: (-) = 10 N (+) = 12 N
Pressure: (-) = 10 psi (+) = 14 psi
Table 8.26 Full Factorial Experiment Using 3 Factors at 2 Levels
* A rapid mnemonic memory aid to determine the number of test trials is the
following: levels are level to the ground while factors fly. That iis, 3 factors at two
= =
levels is 23 8 trials and 2 factors at three levels is 23 9 trial~:;.

DOE Improvement Analysis (Continued)

To find the effect of temperature, one sums the yield values when the temperature
is high and subtracts the sum of yields when the temperature is low, dividing the
results by four.
(77+73+80+73) - (55+47+56+51) = 23.5
The temperature effect:
4
When the temperature is set at a high level rather than at a low level, there is a gain
of 23.5% yield. All ofthis yield improvement can be attributable to temperature alone
since, during the four high temperature experiments, the other two variables were
twice low and twice high.
The pressure effect: (47+73+51+73)-(55+77+56+80) _

-'-------'---'------~ - -6
4
The effect of changing pressure from low level to high level is a loss of 6% yield.
Th t f ff t (56+80+51+73) - (55+77+47+73) 2
e concen ra Ion e ec: 4- =
Higher concentration levels result in a relatively minor 2% improvement in yield.
The interaction effects between the factors can be checked by using the T, P, and C
columns to generate the interaction columns by the multiplication of signs:
Interactions
EXP. T P C TXP PXC TXC TXPXC Y1ELD
1 - - - + + + - 55
2 + - - - + - + 77
3 - + - - - + + 47
4 + + - + - - - 73
5 - - + + - - + 56
6 + - + - - + - 80
7 - + + - + - - 51
8 + + + + + + + 73
Table 8.27 Full Factorial Experiment Interactions

DOE Improvement Analysis (Continuled)

Following the same principles as used for the main effects analysis:
T x P interaction*: (55+73+56+73) - (77+47+80+51) = 0.5

4
*Interaction means the change in yield when the pressure and b~mperature values
are both low or both high as opposed to when one is high and the other is low.
The T x P interaction shows a marginal gain in yield when the temperature and
pressure are both at the same level.
P x C interaction:
(55+77+51+73) - (47+73+56+80) =0
4
T x C interaction: (55+47+80+73) - (77+73+56+51) =-0.5

4
T x P x C interaction: (77+47+56+73) - (55+73+80+51) = -1.5

4
In this example, the interactions have either zero or minimal negative yield effects.
If the interactions are significant compared to the main effects, they must be
considered before choosing the final level combinations.
The best combination of factors here is: high temperature, low pressure, and high
concentration (even though the true concentration contribution is probably minimal).
MINITAB Results
Most people don't analyze experimental results using manual techniques. On the
following page is a synopsis of the effects of temperatur1e, pressure, and
concentration on yield results using MINITAB.


MINITAB Results (Continued)
Source OF 55 M5 F P
T 1 1104.50 1104.50 803.27 0.000
P 1 72.00 72.00 52.36 0.002
C 1 8.00 8.00 5.82 0.073
Error 4 5.50 1.37
Total 7 1190.00
Table 8.28 Analysis of Variance for Yield versus T, P, C
The F values and corresponding p-values indicate that temperature and pressure are
significant to greater than 99% certainty. Concentration might also be important but
more replications would be necessary to see if the 93% certainty can be improved
to something greater than 95%.
Predictor Coef 5E Coef t p

Constant -34.7500 6.4090 -5.42 0.006
T 0.7833 0.0276 28.34 0.000
P -1.5000 0.2073 -7.24 0.002
C 1.0000 0.4146 2.41 0.073
5 =1.173 R2 =99.5% R2 (adj) =99.2%
The regression equation is:
Yield =-
34.75 + 0.783 T -1.50 P + 1.00 C
Table 8.29 Regression Analysis for Yield versus T, P, C
The regression equation will yield results similar to those for the previous manual
calculations. Again, the p-values for temperature and pressure reflect high degrees
of certainty.

DOE Improvement Analysis {Continu~ed}
Fractional Factorial Design Comparison
Frequently, one can derive the same conclusions by conducting fewer experiments.
Suppose the experiments cost $10,000 each, one might then dncide to conduct a
one-half, fractional factorial experiment. Assume the following balanced design was
chosen for the original experiment:
Experiment T P C Yield %
2 + - - 77
3 - + - 47
5 - - + 56
8 + + + 73
Average yield .;3.25
Table 8.30 One-half Fractional Factorial Experiment Using 3 Factors at 2 Levels
Since a fractional factorial experiment is being conducted, only the main effects of
factors can be determined. Please note that experiments 1, 4, 6, and 7 would have
been equally valid.
The temperature effect: (77+73) - (47+56) = 23.5

2
The pressure effect: (47+73) - (77+56) = -6.5

2
The concentration effect: (56+73) - (47+77) = 2.5
2
The results are not exactly identical to what was obtained by conducting the eight
full factorial experiment. However, the same relative conclusion!:; can be drawn as
to the effects of temperature, pressure, and concentration on thE~ final yield.
Note that the average yield is 63.25%. If the temperature is high, an 11.75% increase
is expected, plus 3.25% for low pressure, plus 1.25% for high cOl1lcentration equals
an anticipated maximum yield of 79.5% even though this experiment was not
conducted. This yield is in line with the actual results from exp,eriment number 6
from the full factorial.


Fractional Factorial Design Comparison (Continued)
Predictor Coef SE Coef t p

Constant -36.7500 0.0000 * *
T 0.7833 0.0000 * *
P -1.6250 0.0000 * *
C 1.2500 0.0000 * *
5 =* R2 = * R2 (adj) = *
The regression equation is:
Yield = -36.75 + 0.783 T -1.625 P + 1.250 C
* indicates value cannot be calculated.
Table 8.31 Regression Analysis for Yield versus T, P, C

MINITAB Results for the Fractional Factorial Example
Using either the manual or MINITAB recaps, would the experimenter stop at this
point? Might a follow-up experiment (perhaps at three levels) looking at higher
temperatures and lower pressures payoff? After all, the yield has improved by 16%
since experimentation started.

MEASUREMENT SYSTEM RE-ANALYSIS
Measurement System Re-analysis

Note that Measurement Systems are described in Section VI of this Handbook.
Process improvements often result in reduced variation. With an increased

understanding of a process, variation that was previously considered to be random
or chance cause can be reduced. Continuous improvement efforts resulting in
reduced process variation, may require a re-evaluation of the meclsurement system.
For example, micrometers that measured to 0.001" were used for several years. As
machining capability improved and the specification requirements became tighter,
micrometers with the ability to measure to 0.0001" became necessary. Although
both of these micrometers still have many applications todalY, more accurate
instruments such as "supermicrometers" are now required for some length
measurements. Using specialized equipment in temperature and humidity controlled
rooms, measurements to 1 millionth or 0.1 millionth of an inch for length are
possible today, and are routinely done.
An old rule of thumb was that the measurement instrument needled an accuracy of
one-tenth of the specification tolerance. This was also stated as the "10:1" rule. If
the specification required 0.345" ± 0.005", the proper instrument had an accuracy of
0.0001" because the specification required measurement to tlhe nearest 0.001"
(Some interpreted the requirement as 0.0005" or 0.001" based on 0.1" of the
tolerance). Other sources used a "4:1" ratio, or the equivalent 25%, as a guide for
the measurement uncertainty as compared with the specificati()n tolerance. The
disadvantage of either of these rules is that in some application:s achieving a 10:1
ratio is easily done, while in other applications it would require expensive equipment
to attain the suggested uncertainty.
The new quality system standards require " ... an estimation of the measurement
uncertainty as well as statistical techniques for analysis of test andlor calibration
data." (ISO/IEC 17025, 1999)12. Actually this is not a new requirement because
similar wording was in MIL-STD-45662A (1988)15 " ... the collective uncertainty of the
measurement standards shall not exceed 25% of the acceptable 1tolerance for each
characteristic being calibrated."
For automotive suppliers, Quality Management Systems (ISOrTS 16949, 2002)13

states "Statistical studies shall be conducted to analyze the varialtion present in the
results of each type of measuring and test equipment system .... The analytical
methods and acceptance criteria used shall conform to those in customer reference
manuals on measurement systems analysis" (i.e., bias, linearity, stability,
repeatability and reproducibility studies).
VILLANOVA UNIVERSITY VIII - 68 LEJ\N SIX SIGMA HANDBOOK

MEASUREMENT SYSTEM RE-ANAL YSIS
Measurement System Re-analysis (Continued)

The Measurement Systems Analysis Reference Manual (MSA, 1998)16 does not state
specific requirements for the measurement system, but does provide guidance for
calculation of the bias, linearity, stability, repeatability and reproducibility. For
example, guidelines for acceptance of gage repeatability and reproducibility (% R&R)
are:
• < 10% error The measurement system is acceptable.
• 10% to 30% error May be acceptable based upon importance of application,

cost of gage, cost of repairs, etc.
• > 30% error Measurement system needs improvement. Make every

effort to identify the problems and have them corrected.
(MSA, 1998)16
Note that although the factors are different, the term <10% looks similar to the 10:1
ratio and 10% to 30% looks similar to the 4:1 ratio or 25% from earlier methods.
The expression of measurement uncertainty includes both a range and the level of
confidence at which the statement is made. For example, measurement uncertainty
of a 100 gram mass could be stated as:
"ms = (100.02147 ± 0.00070) g, where the number following the symbol ± is the
numerical value of an expanded uncertainty U = kuc' with U determined from a
combined standard uncertainty (i.e., estimated standard deviation) Uc = 0.35 mg
and a coverage factor k= 2. Since it can be assumed that the possible estimated
values of the standard are approximately normally distributed with approximate
standard deviation uc ' the unknown value of the standard is believed to lie in the
interval defined by Uwith a level of confidence of approximately 95%."
(Taylor, 1994)30
Each organization needs to determine the measurement system requirements and

evaluate if the measurement system capability meets those requirements. The
requirements are dependent upon customer specifications, the intended use, and
costs of implementation. As the process capability improves, the measurement
system needs to be re-evaluated to determine if requirements or measurement
capability also need to be improved to achieve the desired level of control.

REFERENCES
References
1. Alukal, G. & Manos A. (2006). Lean Kaizen. A Simplified Apl,roach to Process
Improvements. Milwaukee, WI: ASQ Quality Press.
2. Center for Quick Response Manufacturing, University of Wisconsin - Madison.

(2006). Retrieved November 22,2006 from:
http://www.engr.wisc.edu/centers/cqrm/pub/qrm21st.htm.
3. Conner, G. (2001). Lean Manufacturing for the Small Sh"p. Dearborn, MI:
4. Davidson, P. (2006). "Kaizen Events -55's," Lean Six Sigma. Biisk Education, Inc.
and Villanova University.
5. George, M.L. (2003). Lean Six Sigma for Service. New York: McGraw-Hili.
6. Goetsch, D.L. & Davis, S.B. (2000). Quality Management, Introduction to Total
Quality Management for Production, Processing, and Servi'ces, 3rd ed. Upper
Saddle River, NJ: Prentice Hall.
7. Goldratt, E., & Cox, J. (1986). The Goal: A Process of Ongo'ing Improvement,
Revised Edition. Croton-on-Hudson, NY: North River Press.
8. Goldratt, E. (1990). Theory of Constraints. Great Barrington, MA: North River

Press.
9. Hahn, G.J. & Shapiro, 8.5. (1966, May). A Catalog and Comput,er Program for the
Design and Analysis of Orthogonal Symmetric and Asymmetric Fractional
Factorial Experiments. Schenectady, NY: General Electric, Research and
Development Center.
10. Imai, M. (1997). Gemba Kaizen: A Commonsense, Low-Cost Aproach to

11. iSixSigma (2006). "55". Downloaded November 14, 2006 from

http://www.isixsigma.com/dictionary/5S-486.htm
12. ISO/IEC 17025 (1999). General Requirements for the Competence of Testing and
Calibration Laboratories, 1st ed. Geneva: International Organization for
Standardization.

REFERENCES
References (Contin ued)

13. ISOITS 16949 (2002-03-01 ). Quality management system - Particular
requirements for the application of ISO 9001 :2000 for automotive production
and relevant service part organizations, 2nd ed. Automotive Industry Action
Group (AIAG).
14. Liker, J. (editor). (1997). Becoming Lean: Inside Stories of U.S. Manufacturers.
Portland, OR: Productivity Press.
15. MIL-STD-45662A (1 August 1988). Military Standard, Calibration Systems

Requirements. Department of Defense.
16. MSA (1998). Measurement Systems Analysis Reference Manual, 2nd ed.
Automotive Industry Action Group (AIAG).
17. NIST, Information Technology Laboratory, Statistical Engineering Division.

(2001). Engineering Statistics Handbook. Retrieved September 21, 2001 from
http://www.itl.nist.gov/div898/handbookl
18. Pitcher, M. (March 13, 2003). "Examining Just-in-Time and Theory of

Constraints." Lean Directions. [Electronic version]. Retrieved November 29,
2006 from: http://www.sme.org/cgi-bin/get-newsletter.pl?lean&2003013&1.
19. Osada, T. (1991). The5S's: five keys toa total quality environment. Tokyo: Asian
Productivity Organization.
20. Plackett, R.L. & Burman, J. P. (1946). "The Design of Optimal Multifactorial
Experiments." Biometrika (Vol.33).
21. Productivity Development Team. (1999). Cellular Manufacturing: One-Piece Flow

for Workteams. Portland, OR: Productivity Press.
22. Robinson, A. (1990). Modern Approaches to Manufacturing Improvement: The

Shingo System. Portland, OR: Productivity Press.
23. Sharma, A., & Moody, P. (2001). The Perfect Engine: How to Win in the New
Free Press.
24. Shingo, S. (1986). Zero Quality Control: Source Inspection and the Poka-yoke
System. Stamford, CT: Productivity Press.

REFERENCES
25. Shingo, S. (1989). A Study of the Toyota Production System /From an Industrial
Engineering Viewpoint. Cambridge, MA: Productivity Press.
26. Skaggs, T. (2006) "The 5S Philosophy." Metaltek Mfg. Inc. Papa Kaizen.
Downloaded November 1, 2006 from:
http://www.tpmonline.com/papakaizen/articls_on_lean_mclnufacturing_strat
egies/5s.htm
27. Suri, R. (May 2001). Quick Response Manufacturing. Advancl~d Manufacturing.

Retrieved November 23,2006 from:
http://www.advancedmanufacturing.com/May01/quick.htm .
28. Suri, R. (2006). Quick Response Manufacturing: A Competitive Strategy for the
21 st Century. Retrieved November 23,2006 from:
http://www.advancedmanufacturing.com/may01/qrm.pdf.
29. Suzaki, K. (1993). The New Shop Floor Management: Empowering People for
Continuous Improvement. New York: The Free Press.
30. Taylor, B.N. & Kuyatt, C.E. (1994). NIST Technical Note 12~J7. Guidelines for
Evaluating and Expressing the Uncertainty of NIST Mea~;urement Results.
United States Department of Commerce, Technology Administration, National
Institute of Standards and Technology.
31. The Gilbreth Network. (2006). "Therbligs: The Keys to Siimplifying Work."
Downloaded November 30, 2006 from
http://gilbrethnetwork.tripod.com/therbligs.html
32. Yang, K. (2005). Design for Six Sigma for Service. New York: McGraw-Hili.
33. Womack, J., & Jones, D. (1996). Lean Thinking: Banish Wastei~ndCreate Wealth
in Your Corporation. New York: Simon & Schuster.
34. Wortman, B.L., et al. (2001). CSSBB Primer. Terre Haute, IN: IJuality Council of
Indiana.
35. Wortman, B.L., et al. (2006). CSSGB Primer. Terre Haute, IN: IJuality Council of
Indiana.
36. Wortman, B.L., Carlson, D.R. & Richardson, W.R. (2006). CQE Primer. Terre

QUESTIONS
8.1. Goldratt's theory of constraints deals with 8.4. The floor of a small shop looks dirty,
money In relationship to three fundamental disorganized, and messy. The manager tells you
measurements for evaluating a system. Align that this is ok, they will perform their annual 55
the following measures with their monetary day as soon as they finish a current large order.
equivalents. You tell the manager this is not really 55
because:
I. Throughput
II. Inventory a. 55 is systematic and formal
III. Operational expenses b. 55 requires a coordinator
c. There will probably be another big order after
A. Money stuck inside the current one
B. Incoming money d. 55 is usually performed more often than
C. Money going out annually
a. I - C, II - B, III - A 8.5. One would say that continuous flow

b. I - B, II - A, III - C manufacturing:
c. I - A, II - B, III - C
d. I - B, II - C, III - A I. Increases product damage
II. Requires less storage and transport
8.2. Consider the following data: III. Reduces cycle time
IV. Provides mechanisms to solve other
Temp Time Oil Response production problems
-1 -1 -1 60
1 -1 -1 72 a. I and II only
b. I and III only
-1 1 -1 54
c. II and IV only
1 1 -1 68 d. II, III, and IV only
-1 -1 1 52
1 ·1 1 83 8.6. In conducting a fractional factorial it is known
-1 1 1 45 that factors A, B, C, D, and E are independent of
1 1 1 60 each other, but factors F and G may not be
independent. A small sub-experiment revealed
What is the main effect of temperature in this a high correlation between factors F and G.
experiment? What is the name of this condition?
a. 18.00 a. Collinearity
b. 70.75 b. Confounded
c. 61.75 c. Coplanarity
d. 15.43 d. Covariates
8.3. Certain six sigma improvement efforts have 8.7. How does poka-yoke respond to human error?
resulted In the need to replace the existing
measurement system and others have not. What a. By eliminating human error
could be the reasons for staying with an existing b. By punishing human error
system? c. By rewarding defect detection due to human
error
I. It's precise enough d. By catching human error before it becomes a
II. Better equipment does not exist defect
III. The current measurement is based on
counted data 8.8. A Latin square design is an experimental design
IV. The improvement has not been successful which:
a. I only a. Cannot be used when estimation of the

b. II and III only interaction effects is desired
c. I, II, and IV only b. Affords a good estimate of interaction effects
d. I, II, III, and IV c. May not permit all treatments in every block
d. May require the need to estimate the
parameters during the experimentation
VILLANOVA UNIVERSITY VIII·73 LEAN SIX SIGMA HANDBOOK

QUESTIONS
8.9. Which of the following is considered the sixth 8.15. What are the major reasc)n(s) the Japanese place
"S"? such a high emphasis 0111 housekeeping items in
the 5S approach?
a. Sanitation
b. Safety I. They figure that workplace organization and
c. Simplification manufacturing are inseparable
d. Setup reduction II. They recognize that workplace organization
must precede 01ther high levels of
8.10. Identify the lean enterprise technique in which performance
the videotaping of a segment of the operation is III. The Japanese tend to be a nation of tidy
helpful: people
a. SMED a. I only
b. TPM b. I and II only
c. Takt c. II and III only
d. FIFO d. I, II, and III
8.11. The theory of constraints concentrates mainly 8.16. Identify the most difficult limitation in achieving
on: continuous flow.
a. Understanding customer needs a. Untrained employees

b. Developing a value stream map b. Existing equipment
c. Achieving on-time goals c. Employee culture
d. Removing process bottlenecks d. Middle management involvement
8.12. Identify the improvement methodology which 8.17. Ideally, what sequenc:e of experimentation
would be LEAST effective in validating the should one use to optimize the response of a
outcome of an improvement activity when only process?
quantitative data is involved.
a. Use response surfalce methodologies at all
a. Multi-vari reanalysis stages
b. Post-improvement capability analysis b. Use screening firs,t and then response
c. Pareto diagram reanalysis surface techniques
d. Response surface plotting c. Use charting techlliques first and then
ANOVA
8.13. If an experiment has an alias, one would say that d. Use experimental designs first and then
the two factor effects are: ANOVA
a. Confounded 8.18. SMED helps a compal1lY create a competitive

b. Blocked edge by allowing the company to:
c. Misnamed
d. Mixtures a. Complete production setups quickly
b. Produce long prodUl:tion runs
8.14. A poka-yoke device detects an error and stops c. Understand the voicl9 of the customer
the process. What should the operator do next? d. Limit the number of product selections
a. Find the root cause of the problem and carry 8.19. The smallest run number possible in order to
out corrective action examine the main effect. of 22 factors at 2 levels
b. Throw out the defective piece and continue is:
with normal operations
c. Stop the process, assemble a team, and plan a. 23
a kaizen event b. 24
d. If the problem persists, implement process c. 44
inspection as a precaution d. 56

QUESTIONS
8.20. Which of the following are Included in the 8.26. Red tagging is used during which 5S stage?
principles of 5S?
a. Standardize
I. Sorting out b. Sustain
II. Systematic arrangement c. Straighten
III. Self-discipline d. Sort
IV. Systems management
8.27. The main objection to designed experimentation
a. I only in an industrial environment is:
b. I, II, and III only
c. II, III, and IV only a. Obtaining more information for less cost than
d. I, II, III, and IV can be obtained by traditional
experimentation
8.21. When performing "one experiment with five b. Getting excessive scrap as a result of
repetitions, " what are the six experiments choosing factor levels that are too extreme
called? c. Verifying that one factor at a time is the most
economical way to proceed
a. Randomization d. Obtaining data and then deciding what to do
b. Replications with it
c. Planned grouping
d. Sequential 8.28. When comparing breakthrough achievement with
kaizen techniques, which of the following
8.22. Which of the following techniques does NOT statements is true?
necessarily compliment the visual factory
concept? a. Kaizen techniques provide more rapid
Improvement
a. Kanban b. Breakthrough achievement is generally less
b. Tool boards expensive
c. 55 c. Breakthrough achievement would be used for
d. Poka-yoke low tech products
d. Kaizen techniques are more easily applied at
8.23. Which of the following is NOT a defect the floor level
elimination and detection technique?
8.29. Which of the following is NOT an external setup
a. Vendor certificate of compliance operation?
b. Poka-yoke
c. Source inspection a. Preparing parts
d. Self checks by operators b. Changing dies
c. Cleaning spares
8.24. A2-1evel, 5-factor experiment is being conducted d. Finding parts
to optimize the reliability of an electronic control
module. A half replicate of the standard full 8.30. The main difference between traditional kaizen
factorial experiment is proposed. The number of and the kaizen blitz approach is:
treatment combinations will be:
a. The number of people involved
a. 10 b. The pace of the change effort
b. 16 c. The amount of floor space saved
c. 25 d. The commitment level of management
d. 32
8.31. What is considered an ideal batch size in a
8.25. What does the Japanese concept of poka-yoke continuous flow operation?
mean?
a. Large batches are considered ideal
a. Root cause analysis b. It depends on the bin size pre-determination
b. Standardizing corrective actions c. One piece is considered ideal
c. Mistake proofing d. Takt time batch size is considered ideal
d. Reengineering

QUESTIONS
8.32. An experiment yielding the following equation: 8.37. In CFM, the time per op1aration is based on:
y =14.0+12.7 ~ +16.1 ~ +9.8 ~ +6.4~ ~ a. The takt time

b. Line balancing
c. The existing constriilint
What CANNOT be said about the response surface? d. Machine capacity
a. It contains more than a 3-D picture 8.38. Just-in-time (JIT) is collectively characterized by:
b. It contains a curve component
c. It contains two slopes I. Producing the right items
d. It contains a twist component
II. Delivering the right items
III. In the right amounts
8.33. In the SMED system what is the real meaning of
IV. At the right time
the single minute concept?
a. I and III only
a. That dies will be exchanged in 1 minute or
b. II and III only
less c. I, II and IV only
b. That dies will be exchanged in less than 5
d. I, II, III and IV
minutes
c. That dies will be exchanged in less than 10 8.39. Which of the following tools would be of LEAST
minutes value when assessing the results of an
d. That dies will be exchanged in less than 15 improvement team's activities?
minutes
a. A post redesign FMI:A
8.34. Why is deep cleaning (shine) performed until the b. A follow-up capabililty analysis
third step of 5S and not before? c. Multi-vari studies
d. A brainstorming session
a. Because only specialized personnel should
take care of cleaning activities 8.40. Kanban concepts include all of the following,
b. Cleaning is ranked third in importance after
EXCEPT:
sorting and straightening
c. It makes sense to perform cleaning after a. Pull instead of push
sorting and straightening the work place
b. JIT
d. Cleaning is often not necessary after sorting
c. One piece flow
and straightening d. Storage in the warl3house but not on the
production floor
8.35. EVOP should be used:
8.41. In a production facto~" that utilizes the pull
a. When there is a manufacturing problem system, which of the following may NOT be
b. When a process is not in statistical control
achieved?
c. When an experimenter first begins working
on a new product a. Reduced raw materi,al inventory
d. When a process is producing satisfactory
b. Reduced in-process inventory
material
c. Increased prices
d. Reduced finished gClods inventory
8.36. Identify the tool that would generate an
experimental equation after a DOE has been
8.42. What is the resolution of a Plackett-Burman
completed.
design with n runs and n-1 factors?
a. Multi-vari reanalysis
a. Resolution V
b. Capability analysis
b. Resolution VI
c. Computer software
c. Resolution II
d. FMEA studies
d. Resolution III

QUESTIONS
8.43. A concept that includes lean enterprise but goes 8.49. When studying a combination of qualitative and
a step further by integrating all aspects of quantitative factors in a 25 full factorial design, a
product cycle time is called: member of a lean six sigma group recommends
removing the qualitative factors in order to study
a. TOC only numerical combinations of effects. What is
b. JIT your reaction to this request?
c. TPS
d. QRM a. You must remove the qualitative factors in
order to calculate quantitative effects
8.44. In every experiment there is experimental error. b. You keep the qualitative factors even though
Which of the following statements is true? no effects can be calculated
c. You remove the qualitative factors and make
a. It is due to a lack of material uniformity and an attributes chart with them
to inherent experimental variability d. You keep both factors and calculate the
b. This error can be changed statistically by effects of both factors
increasing the degrees of freedom
c. The error can be reduced only by improving 8.50. One would say that the kanban method would be
the material most closely associated with:
d. In a well-designed experiment there is no
interaction effect a. The elimination of non-value-added activities
in the process
8.45. Which of the following 5S stages is primarily the b. The development of a future state process
responsibility of top management? stream map
c. Making problems visible in a process, thus
a. Shine clarifying targets for improvement
b. Sustain d. The control of material flow
c. Sort
d. Straighten 8.51. Which of the following is NOT a TOC principle?
8.46. Kaizen techniques: a. Maximizing throughput

b. The system optimization is the sum of the
a. Are initiated by manufacturing workers optimum individual process steps
b. Should be planned properly c. System thinking is essential in managing
c. Can be expensive and are not worth doing in change and solving problems
many cases d. Continuous improvement is required
d. Are important only when producing products
8.52. What is the technique used to control the release
8.47. Experiments can have many different objectives. of raw material from a constraint to the
Which of the following would be included in the downstream equipment?
options?
a. Kanban cards
I. Comparative objective b. Drum - buffer - rope
II. Screening objective c. Production leveling
III. Response surface determination d. Throughput
a. I and II only 8.53. An incomplete experimental block design may be

b. II and III only especially suitable when:
c. I and 1111 only
d. I, II, and III a. There is missing data
b. There is need for fractional replication
8.48. The term kaizen blitz entails: c. It may not be possible to apply all treatments
in every block
a. Long-term projects d. There is a need to estimate the parameters
b. Employee suggestion systems during the experimentation
c. Undocumented quick improvements
d. Rapid workplace change on a project basis

QUESTIONS
8.54. Which of the following operations would benefit 8.S8. The reduction of lead times is one of the main
most from a kanban system application? objectives of kanban. Which of the following is
the other main objectiv<e?
a. One-of-a-kind production operations
b. Service operations a. To optimize floor sp1ace
c. Repetitive production plants b. To increase "heijunlka"
d. Hospital emergency centers c. To control the factory
d. To reduce WIP
8.SS. Ergonomics reduce injuries and lost production.
Which of the following are sound ergonomic 8.S9. There is strong int'3raction between two
principles In the wOrkplace? variables. This means lthat:
I. Safety programs a. An interaction occurs when the effect of one

II. Changing employees to fit the workplace input factor on the Cl1utput depends upon the
III. Redesigning tools to reduce stress level of another input factor
IV. Job rotations b. An interaction occurs when there is a strong
correlation between two factors
a. I only c. An interaction occurs when the effect of one
b. II and'" only input factor on the Cl<utput depends upon the
c. I, III, and IV only level of the same input factor
d. I, II, III, and IV d. An interaction occurs when two or more
factors are confounded with each other
8.S6. Which is the main activity necessary to keep a SS
program working? 8.60. Which of the following maintains and improves
working standards by small gradual evolutionary
a. SS periodic cleaning improvements.
b. SS refreshers
c. SS audits a. PDCA
d. SS sorting b. Taguchi's loss function
c. Poka-yoke
8.S7. Why would lean six sigma activities prompt a d. Kaizen
reanalysis of measuring systems?
a. It's mandated by the MSA Reference Manual

b. Variability may be reduced, requiring more
precise measurements
c. Lean six sigma is a more quality conscious
control technology
d. Most lean six sigma projects include
statistical software requiring frequent
reanalysis
VILLANOVA UNIVERSITY VIII -78 LE.lIN SIX SIGMA HANDBOOK

I
IX. CONTROL CONCEPTS
THE CONTROL CHARJ WILL TELL YOU WHETHER YOUR

PROCESS IS IN STATISTICAL CONTROL.
W. EDWARDS DEMING
VILLANOVA UNIVERSITY IX -1 LEAN SIX SIGMA HANDBOOK

QUALITY CONTROLS
Control Concepts
Control Concepts are presented in the following topic areas:
• Quality controls • Visual systems

• Control plans • Standard work
• Control charts • Training requirements
• TPM
Written Procedures
A procedure is a document that specifies the way to perform an activity. For most
operations, a procedure can be created in advance by the appropriate individual(s).
Consider the situation where a process exists, but has not been documented. The
procedure should be developed by those responsible for the process. Some
procedures may be developed by the quality department for use by other operating
departments. Generally, the operating departments provide input.
Work Instructions
Procedures describe the process at a general level, while work instructions provide
details and a step-by-step sequence of activities. Flow charts may be used with
work instructions to show relationships of process steps. Controned copies of work
instructions are kept in the area where the activities are performed.
Some discretion is required in writing work instructions. The levEl1 of detail must be
appropriate for the background, experience, and skills of the personnel using the
instructions. Typically, the people that perform the activities described in the work
instruction should be involved in its creation. The wording and terminology used
should also match that used by the personnel performing the tasks.
Quality Controls
Production operations, which directly affect quality, are identifiEld and planned to
ensure that they are carried out under controlled conditions. Controlled conditions
include the following:
• The prior approval of processes and equipment, as appropriate
VILLANOVA UNIVERSITY IX -2 LEJI,N SIX SIGMA HANDBOOK

QUALITY CONTROLS
Quality Controls (Continued)

• Documented procedures defining the manner of production where the
absence of instructions would adversely affect quality
• The use of suitable production, installation, and servicing equipment, in an

appropriate working environment
• Compliance with reference standards, codes, and quality plans, and/or

documented procedures
• The monitoring and control of suitable process and product characteristics
• Workmanship criteria, stated in the clearest practical manner (written

standards, representative samples, illustrations, etc.) must be provided
• The suitable maintenance of equipment to ensure continued process

capability
Processes, which cannot be fully verified by subsequent inspection and testing,

must be carried out by qualified operators and/or continuously monitored to ensure
that the specified requirements are met. Records are maintained for all qualified
process equipment and personnel.
Quality Control Methods

Determining product and process control methods is often called the development
of a quality plan. Quality planning employs the coordination of company resources
to meet customer requirements.
It is necessity to first identify all of the key internal and external customer
requirements. One should remember to include all of the critical product and
process characteristics uncovered throughout the complete design process.
Examples of customer requirements are listed in the first column in Table 9.1.
The second step is to identify the manufacturing process flow and the
manufacturing support processes. Examples of these are listed in the second
column in Table 9.1.
The third step is to identify the quality tools that a company will use to control the
processes. Examples of these are listed in the third column in Table 9.1.

QUALITY CONTROLS
Quality Control Methods (Continuedl)
Requirements Support Processes Quali~y Tools

Outside diameter Receiving inspection Part work insltructions
Finishing Subassembly Part inspecticln instructions
Packing System assemble Process work instruction
Length Final test Quality procedures
Material Shipping Purchase order data
Safety tests Purchasing SPC charts
Labeling Outside processing Test fixtures
Identification Training Routing sheefts
Table 9.1 Examples of Product & Process Control Oletail
After the customer requirements, processes, and quality tools have been identified,
a control plan can be detailed. First, one takes a customer requirement, then
decides which process step and which quality tool should be used to satisfy the
requirement. Some quality requirements need to be addressed in two or more
process steps with different quality tools. Some examples are:
• The outside diameter is checked at subassembly using a part work

instruction.
• The packing is completed at shipping using a process work instruction.
• The material is controlled by purchasing using purchase order data, requiring

a certificate of analysis, and at receiving inspection w:;ing a part work
instruction.
• The length is checked at final inspection using an inspection report.
• A safety test is performed by an outside processor usin!~ the appropriate

purchase order data.
When the quality plan is developed, each line item must be implemented and
verified. The quality planning process generates a quality plan. The part of the
quality plan that focuses on production is often called a control plan.

CONTROL PLANS
Control Plans
A control plan is a document describing the critical characteristics (key input or
output variables) of the part or process. Through this system of monitoring and
control, customer requirements will be met and the product or process variation will
be reduced. However, the control plan should not be a replacement for detailed
operator instructions in the form of work instructions or standard operating
procedures. Each part or process must have a control plan. A group of common
parts using a common process can be covered by a single control plan.
Types of Control Plans

For the automotive sector, ISOIlS 16949:20029 and the Advanced Product Quality
Planning APQP (2000)1 identify three control plan phases:
• Prototype
• Pre-launch
• Production
A prototype control plan is used in the early development stages when the part or
process is being defined or configured. This control plan will list the controls for the
necessary dimensional measurements, types of materials, and required performance
tests.
A pre-launch control plan is used after the prototype phase is completed and before
full production is approved. It lists the controls for the necessary dimensional
measurements, types of materials, and performance tests. This plan will have more
frequent inspections, more in-process and final check points, some statistical data
collection and analysis, and more audits. This stage will be discontinued once the
pre-launch part or process has been validated and approved for production.
A production control plan is used for the full production of a part. It contains all of
the line items for a full control plan: part or product characteristics, process
controls, tests, measurement system analysis, and reaction plans. A more detailed
list of input factors is provided later in this discussion.
(APQP, 2000) 1 , (MSA, 2002)12

CONTROL PLANS
Control Plan Organization

Rybarczyk (2005)17 points out that the control phase is the forgotten "C" in DMAIC.
The project control phase is necessary in order to sustain the project gains. The
control plan must truly be a "living document" (APQP, 2000)1 '[for it to remain an
effective mechanism to monitor and control the process.
A responsible person must be placed in charge of the control plan. This ensures
successful monitoring and updating. A green belt or black belt mayor may not be
a suitable person for the role, as he/she may be replaced or transferred to a different
position. A better selection would be the process owner.
The current process owner can be listed on the control plan, but, in reality, it is a
functional role that is to be passed on to the next individual in that same
organizational position. If the control plan is not maintained, t:he benefits of the
project could slowly be lost. The frequent changing of process 40wners, combined
with large numbers of process projects, can easily result in neglected or lost control
plans.
Some considerations in the closing phase of a LSS project incllJlde:
• Identify the process owner

• Involve the team in the control plan
• Create new or updated work instructions and procedures
• Notify the affected personnel
• Train the affected personnel
• Ensure that the control plan training is effective
• Place the control plan in the proper quality system document
• Attain agreement between the team members and process owner
(Rybarczyk, 2005)17, (Six Sigma Academy, 2002)22
VILLANOVA UNIVERSITY IX·6 LE~IN SIX SIGMA HANDBOOK

CONTROL PLANS
Control Plan Input Sources

A number of inputs or sources contribute to understanding, manufacturing, and
controlling the part or process. Many of the following are included:
• Process flow diagrams

• System FMEAs, DFMEAs, and PFMEAs
• Cause-and-effect analysis
• Special customer characteristics
• Historical data
• Lessonslearned
• Team process knowledge
• Design reviews
• Quality function deployment
• Designed experiments
• Statistical applications
• Multi-vari studies
• Regression analysis (Breyfogle, 2003)5, (APQP, 2000)1
The student is provided an illustrative blank control plan, a description of the line
items in the control plan, and a filled in example control plan. Customer
requirements may dictate the exact form of the control plan. (APQP, 2000)1. Often,
there is some flexibility in the construction of the forms.
Control Plan (Sample)

Control Plan for:
Control number: Team members: Page:
Original date:
Contact person (typically process owner): Revision date:
>a
-
G)
c. u
G) :E C ...
~
-e .
cu G) "C
III .;::
- .
G)
III _::::J .2-C 0 C
-
III C ::::J J:
III III cu 0
CO"
G).- :ccu G) 0" cu
G) III > -> .~
G) G)
G) 0.
::::J_ E.EU ~ :EE
G)
u u cu C. ~
E
...0 ...C.0 ::::J C. G) cu III c. ·iii G) C
-
C. U 0
Q. -::::J-
G) (ij ~
...
::::J_
G)
U G) G) 0 C C ...
1:cu •
C o.Q
>a.!!!
·u ·u III G) G)
m
0. 0. (ij o 0
III C. ~
::::J
U
.Q
::::J
>a
G)- G) ...
G)
c.
G)
c.
cum
G) cu cu E
cu
E
cu E ... III G) C
0
cu
Q. CJ) ~~ ~~ CJ) CJ) :Em C!) CJ) CJ) C G) G)
Q. ...
E 0
G)
0::
Figure 11.2 Illustrative Blank Control Plan

CONTROL PLANS
Description of Control Plan Line Itenls

1. Control Plan: Provide a title for the control plan. The control plan will often be
placed into another document, such as an operating instrw:tion or six sigma
database. If necessary, indicate if this is a prototype, pre-launch, or production
plan.
2. Control number: Provide a reference number. This number nlay be supplied by

the responsible department.
3. Team members: If a cross-functional team is involved, provide the member's

names.
4. Contact person: This could be the green belt in charge of the project, however,
the name and function of the process owner are more important.
5. Page: Provide page numbers if the control plan exceed~; one page. The
examples shown in this description are brief. Control plans may run up to 20
pages.
6. Original date: Indicate the original date of issue of the control plan.
7. Revision date: Provide the latest revision date of the control plan.
8. Part! process: List the part number or the process flow being charted.
9. Sub-process step: Indicate the sub-process step being descrilbed (if applicable).
10. Key input variable (X): Note the key input variable, when appropriate. On any
line item, only the X or V variable is filled out, not both. This is to clearly indicate
which item is being monitored and controlled.
11. Key output variable (V): Note the key output variable, when Blppropriate.
12. Special characteristics note: Indicate if a special charalcteristic is to be

monitored and controlled.
13. Specifications: For manufacturing applications the engineering specifications

of the part should be monitored and controlled. For other applications, one
would provide upper and lower specification limits, as well a:s the target value.
VILLANOVA UNIVERSITY IX - 8 LEJ~N SIX SIGMA HANDBOOK

CONTROL PLANS
Description of Control Plan Line Items (Continued)

14. Measurement gage technique: The gage or measurement technique should be
described. The gage, tool, fixture, or test device used for data collection must
be in conformance with the needed measurement system analysis (MSA)
requirements. When necessary, this would include: linearity, stability, accuracy,
reproducibility, repeatability, and uncertainty analysis. This can be more difficult
when attribute data is of concern. The AIAG MSA manual is a good guide.
15. Gage capability: Provide the current capability ofthe measurement system. The
AIAG MSA manual lists:
• Under 10% error as acceptable

• 10% to 30% error may be acceptable, depending on the cost and the situation
• Over 30% error is not acceptable
Measurement devices may need uncertainty determinations.
16. Sample size: Provide the sample size for each subgroup.
17. Sample frequency: List how often the inspection or monitoring of the part or
process is required.
18. Initial Cpk : This provides an indication of process capability.
19. Person responsible for measurement: Indicate who will make and record the
measurement.
20. Control method: Note how this X or Y variable will be controlled. Examples
include control charts, checklists, visual inspections, automated measurements,
etc.
21. Reaction plan: Describe what will happen if the variable goes out of control.
How should the responsible person respond to the situation?
(Six Sigma Academy, 2002)22, (ISOITS 16949:2002)9

CONTROL PLANS
Illustrative Control Plan

Control Plan (Example)
Control Plan for: LSS Primer
Control number: LSS-001 Team members: Glenn, Wes, Page: 10f1
Tim, Bob, Odls, Bill
Original date: February 16, 2007
Contact person (typically Process Owner): Bill Revision date: March 2, 2007
..
g E
GI
::ctV
..
u
fI/
>.
.E
GI
::c
-
!l. GI "t:
GI ::ctV .! u ·iii
... ..
C
~
.~ GI "0
.. Sc
U GI
fI/ .~ tV tV fI/ ;:':::1
...
0
.r. C
-
fI/ :::I
fI/ fI/
fI/
CQ
>
> CQ 0
C CC"
CD- ::ctV CD C" !l.GI
CQ
CD
:a E GI C.
..
N
~
..
.r.
U GI
..9-
U
...
:::I
U CQ E.E !l. ·iii .¥
E
-..
0 :::I !l. GI U CQ !l. .. GI C
0 !l. U .. GI GI C .. 0
n. ::::I iii I;:: :::1- U GI U "0
.E 0 ·uGI_GI ·uGI C. C. 0:::1
~
(II GI GI (II (II U
t!tV .a >. >. CQC)
GI CQ
C)
tV
E E ..
GI GI
CQ C
0
tV
GI
~
::::I GI !l.O !l. tV CQ
n. m ~ mc m :::iC) CI m m :E n.E u a::
Receive ring Heavy Notify Bill

Primer binders metal duty slant visual NA 5 lot NA clerk checklist contact
10320 Notify Bill
binder Gunny Arrestoxx contact
fabric Sak Master 10% 5 lot 1.5 clerk checklist supplier
Notify Bill
binder PMS 293 Pantone contact
print (Blue) color 10% 5 lot 1.5 clerk checklist supplier
Notify Bill
binder 2.10" steel contact
width +/- 0.03" ruler 6% 5 lot 1.7 clerk checklist supplier
Notify Bill
binder 11.63" steel contact
height +/- 0.03" ruler 6% 5 lot 1.7 clerk checklist supplier
Figure 11.3 Example Control Plan for Receiving Binders
In the example above, note that only the key input column is controlled.
Control Plan Summary

Control plan construction is often led by the green or black belt in charge of the lean
six sigma project. The team is usually cross·functional with individuals from
different areas, including the process owner. The team will ensure that the control
plan contains the critical variables, the Xs and the Ys, of the product or process.
The control plan must show compliance and control before plr oject closure. A
successful control plan will remain a living document to ensure that the benefits of
the project will be fully realized.
VILLANOVA UNIVERSITY IX ·10 LEJ,N SIX SIGMA HANDBOOK

CONTROL CHARTS
Control Charts *
Control charts are the most powerful tools to analyze variation in most processes -
either manufacturing or administrative. Control charts were originated by Walter
Shewhart in the mid 1920s. The publication of Economic Control of Quality of
Manufactured Product {Shewhart, 1931)20 made the concepts more widely known.
Upper control limit (UCL)
Process average
Lower control limit (LCL)
Figure 9.4 An Illustrative Control Chart
Control charts using variables data are line graphs that display a dynamic picture
of process behavior. Control charts require approximately 25 subgroups of size 4
or larger to calculate upper and lower control limits, but require only periodic small
subgroups or Xs to continue to monitor the process. Control charts for attributes
data require 25 or more subgroups to calculate the control limits.
A process which is in statistical control has most of the plot points randomly
distributed around the mean with fewer plot points as one approaches the control
limits. Points exceeding the controls are very rare. When a process is in control, it
is predictable.
Control charts may be used by lean six sigma improvement teams to control critical
products or processes after the appropriate variable or attribute levels have been
attained. These control charts can be maintained manually, electronically, computer
based or some combination of these techniques.
* Much of this material has been adapted from the CQE Primer {Wortman, 2006)27.
Excellent control chart references include: Grant {1988f, Western Electric (1956)25
and Besterfield (1993)4. See the references at the end of this Section.

CONTROL CHARTS
Types of Charts
There are many variations of possible control charts. The two primary types are:
Control Charts for Variables
Plots specific measurements of a process characteristic (temperature, size, weight,

sales volume, shipments, etc.).
Types: X - R charts (when data is readily available)

Run charts (limited single-point data)
MX - MR charts (limited data - moving average/moving range)
X - MR charts (limited data)
X - S charts (when sigma is readily available)
Median charts
Short run charts
There are other more advanced variable charts like CuSum (cumulative sum) and
EWMA (exponentially weighted moving average) charts.
Control Charts for Attributes
Plots general measurement of the total process (the number of complaints per order,
number of orders on time, absenteeism frequency, number of ermrs per letter, etc.).
Types: p charts
np charts
c charts
u charts
Short run varieties of the above four charts
Charts for variables are generally more costly since each separatEl variable (thought
to be important) must have data gathered and analyzed. In some cases, the
relatively larger sample sizes associated with attribute charts can prove to be more
expensive.
Variable Charts
Often, variable charts are the most valuable and useful because the specific
measurement values are known. There is a large number of variable chart options.
Variable charts are reviewed first in the following text. Five of these will be
discussed and three of these will be plotted.
VILLANOVA UNIVERSITY IX -12 LEJ!.N SIX SIGMA HANDBOOK

CONTROL CHARTS
X and R Chart Terms

n Sample size (subgroup size).
X A reading (the data).
X Average of readings in a sample.
X Average of all the Xs. It is the value of the central line on the X chart.
R The range. The difference between the largest and smallest value in
each sample.
R Average of all the Rs. It is the value of the central line on the R chart.
UCULCL Upper and Lower Control Limits. The control boundaries for
99.73% of the population. They are not specification limits.
20.5
..... . _ .... . . . .... .... ....................... .................. .................._ . ........._ .. - UCLx = 20.0
20.0
19.5
X 19.0
Average 18.5
18.0
17.5
5 10 15 20 25 30
4.5
4.0 ................ _ •••••-" .. _ .. - .............................. .... __ •.••:•••.•• ".- ••• - . ..... ... ............ UCLR = 4.0
3.5
R
3.0
Range 2.5
2.0 R= 1.9
1.5
1.0
0.5 . . ._ _ _ _ _ _ _ _ _ _ _ _ _ _ _ .LCL.=O
0
5 10 15 20 25 30
Figure 9.5 Typical X - R Control Chart

CONTROL CHARTS
Steps for Constructing X- R Charts

1. Determine the sample size (n = 3, 4, or 5) and the frequenc:y of sampling.
2. Collect 20 to 25 sets of time-sequenced samples (60 to 100 data points).
3. Calculate the average for each set of samples, equals X.
4. Calculate the range for each set of samples, equals R.
5. Calculate X (the average of all the X's). This is the center line of theXchart.
6. Calculate R (the average of all the R's). This is the center Iline of the R chart.
7. Calculate the control limits:
X Chart: UCLx =X + A2R*

LCLx = X - A2R*
R Chart: UCLR = 0 4 R
LCLR =D3 R
8. Plot the data and interpret the chart for special or assignable causes.
I n I A2 I Da I D~ I d2 I
2 1.88 0 3.27 1.13
3 1.02 0 2.57 1.69
4 0.73 0 2.28 2.06
5 0.58 0 2.11 2.33

6 0.48 0 2.00 2.53
Table 9.6 Common Factors used forX- R Control Limits
* Note: A2 R can be shown to be identical in value to 3S x
VILLANOVA UNIVERSITY IX -14 LE.t\N SIX SIGMA HANDBOOK

CONTROL CHARTS
X-R Control Chart Data

Closure Removal Torques (In Ibs)
Sample Measurements Sample Sample
1 2 3 4 5 Avg.X Range R
1 12 12 13 15 12 12.8 3
2 15 17 16 17 18 16.6 3
3 13 18 14 14 15 14.8 5
4 10 12 11 10 11 10.8 2
5 13 16 14 15 14 14.4 3
6 15 12 13 15 11 13.2 4
7 15 16 15 16 15 15.4 1
8 15 17 16 14 12 14.8 5
9 22 17 15 17 14 17.0 8
10 16 15 17 15 18 16.2 3
11 16 18 16 16 16 16.4 2
12 15 16 17 17 14 15.8 3
13 17 15 16 15 16 15.8 2
14 16 15 18 18 16 16.6 3
15 17 19 17 15 17 17.0 4
16 19 17 15 15 17 16.6 4
17 16 19 16 15 14 16.0 5
18 16 15 17 16 18 16.4 3
19 17 13 17 15 14 15.2 4
20 19 18 17 15 16 17.0 4
21 14 17 16 14 13 14.8 4
Totals 323.6 75
X 15.4
UCLx =-X + A2R- =15.4 + (0.58 x 3.6) =17.5 R 3.6
LCLx =X - A2R =15.4 - 2.1 =13.3
UCL =D4R =2.11 x 3.6 =7.6
R
LCL =D3R =0 x 3.6 =0

R

CONTROL CHARTS
X· R Control Chart
Chart No. 1
Product Name: Tablets Process Closure Department Operator Bill
Variable: Removal Torques SDecification Limits: LSL =10 LBS USL =22 LBS IUnits of Me,asure: LBS
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 H 20 21 22 23 24 25
- - -. - - .- ..
18
- -~ .~ -. ~ - - .- ~. ~ ~
U( 4: 17.5
,.
17
16
15
- " /
["10.
" " I \
X 1 .4
1/1
C\)
..
-.- .
CI I I'\. I
14
IV
C\)
13
~
,.,
I
... '- - - .. .... - - - LC -= ~3.
~
12 Is
)eC
11
.. ..,,:; . :al
8
7
.. - 1-
:i; - .1-
UC 7.E
1/1 I
6
C\)
CI 5
C 4
IV
a: 3
R= 3.6
2 !.' \.
""'~
1
1
- ... .....
2 3 4 5 -6 7 8
I_~ P-
-- -
9 10 11 12 13 14 15- 16'- 17 18 -
- . 19 -20 [Lei "11= ~I!
21 22 23 24 25
I--f-
Date Start 1/12198 1112 ~ 1/13
Time 7:05 7:10 7:35 8:10 8:15 9:10 9:12 9:33

"
bill ~1 :43 12:05 13.:05 13:45 3:55 ~4:20 ~4:55 7:00 7:55 9:01) 9:12 9:32
1 12 15 13 10 13 15 15 15 22 16 16 15 17 16 17 19 16 16 17' 19 14
2 12 17 18 12 16 12 16 17 17 15 18 16 15 15 19 17 19 15 13 18 17
3 13 16 14 11 14 13 15 16 15 17 16 17 16 18 17 15 16 17 17 17 16
Sample 4 15 17 14 10 15 15 16 14 17 15 16 17 15 18 15 15 15 16 1~, 15 14
Measurement 5 12 18 15 11 14 11
15 12 14 18 16 14 16 16 17 17 14 18 1~, 16 13
Total 64 83 74 54 72 66 77 74 85 81 82 79 79 83 85 83 80 82 76 85 74
Averagei 12.8 16.6 14.8 10.8 14.4 ~ 3.2 ~5.4 14.8 17 16.2 1~A 15.8 15.8 16.6 17 rl6.6 16 ~6.4 1U 17 14.8
Range R 3 3 5 2 3 4 1 5 8 3 2 3 2 3 4 4 5 3 4 4 4
Figure 9.7 An x- R Chart for Closure Removal Torques

This was a start up process. The process started smoothing lout from data set
number 10 on. If continued, the chart would need new control limits from that point.

CONTROL CHARTS
X-Bar and Sigma Charts
X-bar (X) and sigma (S) charts are often used for increased sensitivity to variation
(especially when larger sample sizes are used). These charts may be more difficult
to work with manually than the X - R charts due to the tedious calculation of the
sample standard deviation (S). Often, S comes from automated process equipment
so the charting process is much easier. The formula is:
S= ~L(X-X)'
n-1
Where: I means the sum
X the individual measurements
X-bar the average
n the sample size
The X chart is constructed in the same way as described earlier, except that sigma ( S )
is used for the control limit calculations via the following formulas:
= - = -
UCLx = X + A3 S LCLx = X - A3 S
Where: X is the grand average

A3 is a factor based on sample size (see Table 9.8 below)
S is the average sample standard deviation
The control limits for the sigma (S) chart are calculated using the following formulas
and table:
S is the average sample standard deviation and is the centerline of the sigma chart.
I n
I 2
I 3
I 4
I 5
I 6
I 7
I 8
I 9
I 10
I 25 I
84 3.27 2.57 2.27 2.09 1.97 1.88 1.82 1.76 1.72 1.44
83 * * * * 0.03 0.12 0.18 0.24 0.28 0.56
A3 2.66 1.95 1.63 1.43 1.29 1.18 1.10 1.03 0.98 0.61
*The lower control limit for a sigma chart when (n) is less than 6 is zero.
Table 9.8 Sigma Chart Factors

CONTROL CHARTS
Median Control Charts

There are several varieties of median control charts. One type plots only the
individual measured data on a single chart. The middle value is circled. Median
charts may use an odd number of readings to make the median vallue more obvious.
Another variety records the data and plots the median value and range on two
separate charts. Minimal calculations are needed for each subgroup. The control
limits for the median chart are calculated using the same formulas as the X - R
chart:
UCL-x = X + A2R
The A2 values are somewhat different than the A 2 values for thEl X - R chart since
the median is less efficient and, therefore, exhibits more variatio,n.
n 2 3 4 5
1.88 1.19 0.80 0.69
Table 9.9 A2 Factors for Median Control Charts

The range factors (0 3 and 0 4 ) and process standard deviation factclr (d 2) are the same
as used for the X -R chart. The specific advantages of a median chart are:
• It is easy to use and requires fewer calculations
• It shows the process variation
• It shows both the median and the spread

CONTROL CHARTS
X-MR Charts
Control charts plotting individual readings and a moving range may be used for
short runs and for destructive testing. X-MR charts are also known as I - MR,
individual moving range charts. The control limits for individual charts are
calculated using the formulas and factors as illustrated in Table 9.10 below:
I n
I 2
I 3
I 4
I 5
I
04 3.27 2.57 2.28 2.11
03 0 0 0 0
E2 * 2.66 1.77 1.46 1.29
'n
*E 2 =A2"n
Table 9.10 X-MR Control Chart Factors
The control limits for the range chart are calculated exactly as for the X-R chart.
The X-MR chart (for individuals and moving ranges) is the only control chart which
may have specification limits shown.
There are drawbacks in the interpretation and use of X-MR charts:
• All interpretation is faulty if the data distribution is not normal.
• Individual charts do not separate piece-to-piece repeatability of the process.
• Variability patterns cannot be assured until 80-100 readings are taken.
• Individual charts are not as sensitive to changes in the process as the X-R
chart (or MX-MR, when n = 3).
MX-MR charts with n = 3 are recommended by the authors when information is

limited.

CONTROL CHARTS
x - MR Chart Example
Product Name ADDie Strudel Process LineA Charlt No. 7 IOperat or you
Vari able Stick Weights I Specification Limit T85, High 88, Low 82 IUnits of Measure Grams
1 12 3 4 5 6 7 8 9 1 10 11 112 13 14 15 , 16 17 18 19 20 21 22 23 24 25
,
95 : , ,
~ (jet=-: 9 .
ca I I
::l 90
-
.-"t3:> 85
~
I
I
-s:: 80
,
,
I ,
•
75 . I
•
,
15
. ,
CI) ,
Q.) I , ,
tn 10
s:: ,
ca ,
a::: 5
. . . -
, I
1 2 , 3 14 5 6 7 8 I 9 10 11 1 12 13 14 15 16 17 18 1!1 20 21 22 23 24 125
Date 4116/ 98 I I
Time I I I I I
1 I I I I
2 I I
3 I I 1
Sample 4 I I
Measurements 5 1 I
Individuals. X 85 87 86 86 77 83 184 87 90 184 89 182 84 86 88 85 90 83 84 87 87 y= tS.4
Range,R
Notes
2 1
I I
0/
6 11 3 3 6 5
I
7 2 2 2 3 5 7 1 3 0 ~R d,.,?+_
Figure 9.11 X - MR Chart Example
X = 1794 =85.4 - 68
21
MR = -20 =3.4
Note k =21 for X J k =20 for MR. n =2 for the MR chart above
UCL XJ LCL x =X ± E2MR UCLMR =D4MR =(~L27)(3.4) =11.1
UCLXJ LCL x =85.4 ± (2.66)(3.4) LCL MR =0
UCLx = 94.4 LCL x =76.4

CONTROL CHARTS
MX-MR Chart Example Data
Sample Measurements Sample Sample

Avg. Range
1 2 3 R
X
1 2.83 2.80 2.95 2.86 0.15
2 2.80 2.95 2.92 2.89 0.15
3 2.95 2.92 2.93 2.93 0.03
4 2.92 2.93 2.97 2.94 0.05
5 2.93 2.97 2.95 2.95 0.04
6 2.97 2.95 2.86 2.93 0.11
7 2.95 2.86 2.89 2.90 0.09
8 2.86 2.89 2.78 2.84 0.11
9 2.89 2.78 2.96 2.88 0.18
10 2.78 2.96 3.02 2.92 0.24
11 2.96 3.02 2.93 2.97 0.09
12 3.02 2.93 2.88 2.94 0.14
13 2.93 2.88 2.97 2.93 0.09
14 2.88 2.97 3.04 2.96 0.16
15 2.97 3.04 3.05 3.02 0.08
16 3.04 3.05 3.03 3.04 0.02
17 3.05 3.03 3.04 3.04 0.02
18 3.03 3.04 3.04 3.04 0.01
19 3.04 3.04 3.01 3.03 0.03
20 3.04 3.01 3.02 3.02 0.03
21 3.01 3.02 3.00 3.01 0.02
22 3.02 3.00 3.00 3.01 0.02
23 3.00 3.00 3.03 3.01 0.03
24 3.00 3.03 3.07 3.03 0.07
25 3.03 3.07 2.92 3.01 0.15
Note how data is used three times. Totals DC = 74.10 ~R = 2.11
MX = 2.96 MR= 0.0844
UCLMX =2.96 + (1.02)(0.0844) =2.96 + 0.086 =3.05

LCL MX =2.96 - (1.02)(0.0844) =2.96 - 0.086 =2.88
UCL =(2.57)(0.0844) =0.217
MR LCL =0
MR

CONTROL CHARTS
MX-MR Charts
MX-MR (moving average-moving range) charts are a variation of X-R charts where
data is less readily available. There are several construction techniques. An
example for n =3 for the data on the prior page is shown below. Control limits are
calculated using the X-R formulas and factors.
PRCXl..CT NNIE L1DUID I PROCESS ~ILLING OPERATION CH,'RT NO. 2 OPERATOR BILL
VARIABLE VEIGHT ISPECI~ICATION LIHITS USL·3. I LSL-3.0 l UNlTS O~ HE.ASURE GRAHS
2 3 4 5 6 7 6 9 10 II 12 13 14 15 16 17 16 19 20 21 22 23 24 25
LCL X·3.0S
3.05
3.00
(.I)
W ~·2.96
l:l 2.95
<J:
Ck:
101
>
<J:
2.B5
.3
(.I)
~ LCL· .217
Z .2
<J:
Ck:
I 2 3 4 5 6 7 B 9 10 II 12 13 14 15 16 17 1B 19 20 21 22 23 24 25
DATE
TIME '7: IB 7:2E7 ; 33 7:4~ B : a, 8:S!: 1a:2R 8:31 8:45 9:1i!lIi 9: 15 9:3111 9:45 9:59
1 2 . B3i? .81112.952.9 2 . 9 2 . 912 . 9~ 12 . 8E 2.8!3 2. 7S 2 . 9£ a.11I 2 . ~ 2 . 81 ~.~ 3 . 1114 3 . 111! 3.~ 3.~~1114 3. I1IJ 13. 1112 3.1111113.1111113.111
2 2 . BR ~ . 95 2 . 922 . 9 2.9 2.952 .Be 2 .8~ 2. 7e 2.96 3.111. 2.9~ 2 .8e 2.9 3.11143.11153.111 3.11143 . 1114 3.1111 3 . 11J.'2 3 . 0U.11I11I3 . 11I3 S. 0'1
3 2.952.922.932.9 2.9~2 .86 2 . 8~ 2. 7E 2 . 96 3 . 02fz . 9 2 .BE 2 . 9il 3 . 111 3 . @!i 3.~ 3 . 111 3.04 3.013 . 111 3.l1Ie~ . l1Ie 3 . 11133.11172.92
SAMPLE
I1EASUREMENTS 4
5
AVERAGE. x 2 . Be 2 . 892.932 . 9 2.952.9 2 . 9Q 2 .B4 2. BS 2 . 9 2 . 9? 2 . 9412 . 9 2.9E 3 . 1112 3 . 111 3 . 111 3 . 11I43 . 11I~ 3 . 111 3.111 3 .B I 3 . 1111 3 . 111~ 3 . 111/
RANGLR .15 . 15 . 1113 . 85 . B4 . 11 .1119 . 11 . 18 . 24 . 1119 . 14 . 1119 .16 .1118 . 02 . 1112 .1111 .1113 . 1113 .1112 . 1112 .1113 . 1117 . 15
NOTES
Figure 9.12 MX-MR Chart Example
The student would naturally be unfamiliar with the data in the ablDve example. It is
actual data for a filling operation. According to chart interpretation rules, did a
change occur between plot points 14 and 24? If so, referring to, the specification
limits, was it a good change?
VILLANOVA UNIVERSITY IX -22 LE.tIN SIX SIGMA HANDBOOK

CONTROL CHARTS
Attribute Charts
An attribute chart plots characteristics such as short or tall, fat or thin, blue or
brown, pass or fail, okay or not okay, good or bad, etc. Attributes are discrete,
counted data. Unlike variables charts, only one chart is plotted for attributes. There
are four types of attribute charts, as summarized below:
Chart Records Subgroup size

p Fraction defective Varies
np Number of defectives Constant
c Number of defects Constant
u Number of defects per unit Varies
I 100p*
I Percent defectives
I Varies
I
Table 9.13 Types of Attributes Charts
* The p chart reflected in percent.
The best uses of an attribute chart are to:
• Follow trends and cycles

• Evaluate any change in the process
Key points to consider when using attribute charts:
• Normally the subgroup size is greater than 50 (for p charts).
• The average number of defects/defectives is equal to or greater than 4 or 5.
• If the actual p chart subgroup size varies by more than ± 20% from the average
subgroup size, the data point must either be discarded or the control limits
calculated for the individual point.
• The most sensitive attribute chart is the p chart. The most sensitive and
expensive chart is the X - R chart.
• The defects and defectives plotted in attribute charts are often categorized in
Pareto fashion to determine the vital few. To actually reduce the defect or
defective level, a fundamental change in the system is often necessary.

CONTROL CHARTS
Attribute Chart Formulas
Defectives (Binomial Distribution)

p Chart np Ch;art
0/0 Defectives Defectives
p = np for fraction nli = Lnp

n k
p = np x 100 for % UCLnp = nli + 3 Jnli( 1-li)
n
UCL = Ii + 3l'(1.:p)
LCL np = nli - 3/nli{1-p)
p n
k = number of samples
LCL = Ii - 3l'(1-li)
p n
- L
n=-
n -
p=
Lnp
k Ln
Defects (POisson Distribution)
u Chart c Chart
Average Number of Defects Number of IDefects
c
u=-
n
-c =LC
--
k
- ~
UCLu = u + 3 ~ UCLc = C + 3~:
LCLc = C -3~
- ~
LCL = u - 3 - k = number of ~iamples
u ~
- L
n
n=-
-
u=-
LC
k Ln
Sample Size Varies Sample SiZI~ Fixed
Figure 9.14 Attribute Chart Formulas
VILLANOVA UNIVERSITY IX ·24 LEJ~N SIX SIGMA HANDBOOK

CONTROL CHARTS
p Chart Example Data

Lot Sample Size Number Defective Percentage Bad
Identification (Varies) In the Sample (np) in each Sample
1 1250 8 0.64%
2 1250 0 0.00%
3 1350 12 0.89%
4 1200 3 0.25 %
5 1150 5 0.43 %
6 1100 0 0.00%
7 1100 2 0.18%
8 1350 2 0.15%
9 1250 1 0.08%
10 600 3 0.50%
11 1150 0 0.00 %
12 1100 5 0.45%
13 1050 10 0.95%
14 1050 0 0.00%
15 1100 0 0.00%
16 1000 0 0.00%
17 1200 0 0.00 %
18 1050 0 0.00%
19 1150 1 0.09%
20 1050 0 0.00%
k = no. of lots In = 22,500 Inp = 52
k= 20
From the Example:
n= 22,500 = 1125 p=
total defective
=
52
= 0.23 %
20 total inspected 22,500
The control limits are determined by the formulas:
UCLp, LCLp = p
_
± 3
~P(100n- Ii)
r-------;-------:-
UCL LCL = 0.23% ± 3 0.23%(100% - 0.23%)

p' p 1125
UCLp, LCLp = 0.23% ± (3)(0.143%)
UCLp = 0.66% LCLp = -0.20% = 0%

CONTROL CHARTS
p Chart Example
Attributes Control Chart Form p~ n POCO IJ 0
PARTR: OESCRIPTI~: UNCOATED TABLETS CHARACTERISTIC: 7. DEFECTI'IES OllTE: 6-I-U
SlUCE: TABLET DEPMTt£NT IJI[RIITOR: YMIIlJS INSPECTlIl;...·---=B=IL_L_ _ __
III : .66 lCl: 0 II¥EM : ----::.=23:....-_ _'"'"i"i:::::;:::::;::::::;::;:-_ _
1-0
UCL I) =.
P+3VO( 100-0 I ii
.9
='123 3'V11 r.II 0-.2: I
II 15
.8 = ,23 '3( . 14 3)
= ,66
--
.7
.6
- - - I- '- - '- - ~
- -
~ uc = 6E
.5
.4
.3
p= 23
.2
.1
0
U ..... :0
Sonp le
] 2\/3 4 5 6 7\ 8 9 A~A~A~ ;~A4 ;~ ]6 ;~A~ ;~h~ktk~Ak~ k~
{n I 2SO 2SO 350 I~ II SO 1100 1100 1350 2SO 600 ISO 100 050 050 100 000 I~ 1050 IISQ 10SO ~ =2 .. 5 )0 ~=l 125
~~er
( . c) S 0 12 3 5 0 2 2 I 3 0 5 10 0 0 0 0 0 I 0 ~ np 52
Froct i on
®ul .64 0 .89 .25 .43 0 . IS . 15 .08 .SO 0 .45 .95 0 0 0 0 0 .09 0 p= 23
Dotel
~ %~ ~ %%% ~I ~3 ~4 ~4 ~5 ~6 ~S ~9 ~g
""
Tine
Notes
'" '"
Figure 9.15 P Chart Example
Note that a system change of major consequence occurred at picot point 14. If this
chart were to be continued, the sample size would have to be increased
substantially.
VILLANOVA UNIVERSITY IX -26 LEJIN SIX SIGMA HANDBOOK

CONTROL CHARTS
c Chart Example Data

Lot Sample Size Number of Defects
Identification (Constant) (In the Sample)
1 100 5
2 100 8
3 100 7
4 100 5
5 100 7
6 100 3
7 100 3
8 100 4
9 100 2
10 100 2
11 100 3
12 100 3
13 100 2
14 100 3
15 100 1
16 100 9
17 100 6
18 100 7
19 100 7
20 100 4
21 100 7
22 100 1
23 100 6
24 100 5
25 100 _4_
k= 25 =
n 100 I,c = 114
From the Example:
k=25 c = total defectives = 114 = 4.6

no. of lots 25
The control limits are determined by the formulas:
UCLc =c+ 3~ =4.6 + 3.J4.6 =11

LCLc =c -3~ =4.6 - 3.../4.6 =-1.8 =0

CONTROL CHARTS
c Chart Example
Attributes Control Chart Form pO np 0 c~ uD
PART "* : Encyclopedia DESCRIPTION: SPC Checklist CHARACTERISTIC: ...::;:D.::..;ef=ec=tse-_ _ DATE: . . :. ;10::.. ;/1,--_
SOURCE: Bindina Deoartment OPERATOR: INSPECTOR: -'-Y.::..:ou=---_ _ __
UCL LCL: 0 AVERAGE:
I ,
14
I
I
12
10
4 I
2 ,
,
0
/ 1\ 2 13 14\ 5 16 7 Is\MAo 11 1~ 13A~A5 A~AAA~A~ho h1/~\ h~h~h~
Sample
(n)
N u m ~er
(np,c)
/
5 8 7 5 7 3
0 eF xed Sta nda dS ~m~ Ie
3 4 2 2 3 3 2 3 1 9 6 7 7 14 7 1 6 5
"4
Fr~ction
%p,u) c= 4.6
Datel
Time 1~ Shift
Cha age
Notes
Figure 9.16 c Chart Example
Note that a shift change occurred between plot points 15 and 16"
Is this significant?
VILLANOVA UNIVERSITY IX -28 LEJ\N SIX SIGMA HANDBOOK

CONTROL CHARTS
np Chart Example Data

Lot Sample Size Number Defective
Identification (Constant) (In the Sample)
1 100 9
2 100 6
3 100 7
4 100 3
5 100 4
6 100 6
7 100 3
8 100 3
9 100 3
10 100 4
11 100 4
12 100 6
13 100 4
14 100 5
15 100 2
16 100 3
17 100 2
18 100 8
19 100 4
20 100 3
21 100 2
22 100 3
23 100 4
24 100 8
25 100 6
k= 25 n = 100 I,np = 112
From the Example:
k=25 np = total defective = 112 = 4.5

no. of lots 25
The control limits are determined by the formula:
UCLnp ' LCLnp =np- ± 3 n

-( 1 - npJ
np =4.5 ± 3.j(4.5)(0.955)
UCLnp =10.7 LCL np = -1.72 = 0

CONTROL CHARTS
np Chart Example
Attributes Control Chart Form p D np [g1 CD U D
PART # : Encyclopedia DESCRIPTION: SPC Checklist CHARACTERISTIC: Any De fective DATE: 9/25 - 10/1
SOURCE: Binding Department OPERATOR: INSPECTOR: ~Yo=u~_ _ __
OCl ~~ o AVERAGE:
I
I
I
,,
,
I
14 ,
12
10 I
4
~
- I
2
,
0 •
Sample
1 2 / 3 / 4\ 5\ 6 Ji\Js\ g\Ao /11 1~A3 A4 1~AM7 A~ 1~ :~~h~h2 /23 /24 2~
/ "-
(n) I
101 Un ts F ixec I
Number
(np,c) 9 6 7 3 4 6 3 3 3 4 4 6 4 5 2 3 2 8 4 3 2 3 4 8 6
Fraction
%(p,u)
Datel
Time
Notes
l%5 9'1
Figure 9.17 An np Chart Example
UCL"p = np + 3~ LCL"p = np - 3~
UCLnp = 4.5 + 3..j(4.5)(0.955) LCL np =4.5 - 3..j(4.5)(0.955)
UCLnp = 10.7 LCLnp = -1.72 = 0
VILLANOVA UNIVERSITY IX -30 LEj~N SIX SIGMA HANDBOOK

CONTROL CHARTS
Control Chart Analysis

If a process is out-of-control, then special causes of variation are present in either
the average chart or range chart, or both. These special causes must be found and
eliminated in order to achieve an in control process. A process out-of-control is
detected on a control chart by either having any point outside the control limits or
by unnatural patterns of variation.
~ _ _ _ _ Upper Control Limit
± 1S = 68.27%
±2S = 95.45% Grand Average
±3S = 99.73%
f--- - Lower Control Limit
Figure 9.18 Control Limits Illustration
Control limits are the boundaries set by the process which alert us to process
stability and variability. Remember, our control limits are 3 standard deviations
above and below the grand average. If the process is in control, 99.73% of the
averages will fall inside these limits. The same is true for the range control limits.
Because there are two components to every control chart -- the average and the
range -- there are four possible conditions could occur in the process.
1. Average Out-of-Control
Range In Control
2. Average In Control Process

Range Out-of-Control Out-of-Control
3. Average Out-of Control

Range Out-of-Control
4. Average In Control Process

Range In Control In Control
Each of these conditions is illustrated on the following pages.
VILLANOVA UNIVERSITY IX - 31 LEAN SIX SIGMA HANDBOOK

CONTROL CHARTS
1. Process Average Out-ot-control

Average Shifting
+ 15 /[)<~
w + ~ 0 _.,' ...... _...•- ...- -_. '7P>-- ~ ':::::~- UCL
Variation Stable
~ 5 ~ I " =
0:: 0 X
~ 1-"- .•- ......:); t<. .. -_. _ .. --- -- -... ~- LCL
< -5
-10~~~~~~-+-+-+-+~
-15~~~~~~~~~~~
25~~~~~~~~~~~
20~~~~~~-+-+-+-+~
w I5 1--' ...... ...... ... _... "-' ................. "'" 1-.. UCL
~
~ 10~~--+,/.....,..-t""'""",,-+--+-+/"'-+~-+-+~ R
5~~~/~~~~~~~
o ~~~~~---'---L.----'-----L..----'-~ LCL
2. Process Variation Out-ot-control

Average Stable
+ 10 ..... ,........... -- --... --........... ". -- .... ... UCL Variation Changing
~ +5 l--f----<~-/-,!"""~~~--+-+-/*""--I X
ffi
>
0
..... ,.......... -- ............ -.... ............... .... LCL
< -5
- 10l--!---i~---+~~~--+-+-+--I
- I5 L---l-...JL-.J-1------'-------'---L- - L - - ' -- - ' -- - '
~
25
20
1~
....... .......... ........ ....~ ~.. ........... ........ ........ ..1. .\- ......... UCL
~ 15
~ 10 I
j \ r---..
f-" ~II
I \\ -
0:: R
5 ...........
o LCL
VILLANOVA UNIVERSITY IX -32 LEJ~N SIX SIGMA HANDBOOK

CONTROL CHARTS
3. Process Average & Variation Out-of-control

+ 15
".-... Average Shifting
,...6«
-... . . ._.- .._- 1-' r1 1>"'- ~-\ Variation Changing
~ +5
+ 10 -- ..... ........ ....
.......
UCL
< ./ ........... """'V /'
X
-~ ~ - r"- ._- -- ....,.. _...
f5
;;: -5
0
-- . . ..
..- ~- ..
'I-'
LCL
-10
-15
25 ""
~r
...~ x.. .-.. ..•
20
~ 15
..... ..". _4H'" .... -~ .f- l . H ••• UCL
~ 10 / t". f-'" II \
a:: I 1"'-., R
5 ........
o LCL
4. Process In Control
+15~~-'~~~-'-'-'-'~ Average Stable
+ 10 -- _.- --- - -- .__• '-1--' -- ..,- .•... UCL Variation Stable
w +5~~~~-;~-+-+-+~~
~ /"""' =
c:t:: 0 X
~ -_. -" i--- ._- ...- •__ • - - -_. -'" .... ..... LCL
< -5
- 10 ~~I--l~-;~-+-+-+-I--i
- I5 '--'----'L........I----'---'--..L-I...-I...---1...----'---'
25
20 _. - _.
~ 15
~- ...... ... ..- ..... UCL
~ 10
c:t::
5
I' " V
v ." .......... /' --
-
R
LCL
o

CONTROL CHARTS
Control Chart Interpretation
Five Common Rules
(Rule 2) 4 out of 5
points in zone B (Rule 4) 8 or more consecutive
points on one side of center line UCL
Zone A
Zone C
ZoneC
ZoneB
Zone A
LCL
(Rule 3) 2 out of 3 Rule 5 A trend is
(Rule 1) A point beyond points in zone A 6 or more consecutive points
the contrallimit increasing or decreasing
Comment: Some authorities say 7 or more consecutive points for both Rules 4 and 5.
Other Unusual Patterns

UCL
------------...------------
Zone A
ZoneB
Zone C
Zone C
ZoneB
Zone A
(Rule 6) Stratification. LCL
15 or more points in zone C (Rule 7) Mixture or
systematic variation

CONTROL CHARTS
Process In Control with Chance Variation
This is an example of a process which is in control. Notice that it looks good, but
not too good.
Trends
- R CHART CAUSES
X CHART CAUSES
• Deterioration of machine • Improvement or deterioration of operator skill

• Tired operator • Tired operator
• Tool wear • Change in incoming material quality
CORRECTIVE ACTION
• Repair or use alternate machine if available

• Discuss operation with operator to find cause
• Rotate operator
• Change, repair, or sharpen tool
• Investigate material

CONTROL CHARTS
Jumps in Process Level
- R CHART CAUSES
X CHART CAUSES
• Changes in proportions of materials coming from • Change in material

different sources • Change in method
• New operator or machine • Change in operator
• Modification of production method or process • Change in inspe1ction
• Change in inspection device or method
CORRECTIVE ACTION
• Keep material supply consistent

• Investigate source of material
• Check out machine capability
• Examine operator methods and instruction
• Check calibration of measurement device
Recurring Cycles
- R CHART CAUSES
X CHART CAUSES
• Physical environment • Scheduled maint,enance

• Temperature • Tired operator
• Humidity • Tool wear
• Tired operator
• Regular rotation of machine or operator
CORRECTIVE ACTION
• If environment is controllable, adjust it

• Service equipment
• Rotate operators
• Evaluate machine maintenance
• Replace, sharpen, or repair tool
VILLANOVA UNIVERSITY IX ·36 LEAN SIX SIGMA HANDBOOK

CONTROL CHARTS
Points Near or Outside Limits
- R CHART CAUSES
X CHART CAUSES
• Over control • Mixture of material of distinctly

• Large systematic differences in material quality different quality
• Large systematic differences in test methods or
equipment
CORRECTIVE ACTION
• Check control limits

• Investigate material variation
• Evaluate test procedures
• Evaluate inspection frequency or methods
• Eliminate operator over adjustment of the process
Lack of Variability
- R CHART CAUSES
X CHART CAUSES
• Incorrect calculation of control limits • Collecting, in each sample, a

• Improvement in process since limits were number of measurements from
calculated widely differing lots
• Employee may not be making checks • Improvement in process since
limits were calculated
CORRECTIVE ACTION
• Check control limits

• Validate rational sample subgroupings
• Verify checking procedure, gages, etc.
• Verify proper employee measurement
• Congratulate someone for improvement

CONTROL CHARTS
The Pre-control Technique

An easy method of controlling the process average is known as "pre-control." Pre-
control was developed in 1954 by a group of consultants (includiing Dorin Shainin)
in an attempt to replace the control chart. Pre-control is most successful with
processes which are inherently stable and not subject to rapid plrocess drifts once
they are set up. Pre-control can act both as a guide in setting process aim and
monitoring the continuing process.
The idea behind pre-control is to divide the total tolerance intc» zones. The two
boundaries within the tolerance are called pre-control (P-C) line:s. The location of
these lines is halfway between the center of the specification and specification
limits. It can be shown that 86% of the parts will be inside the P-C lines with 7% in
each of the outer sections, if the process is normally distributed ailld Cpk 1. Usually =
the process will occupy much less of the tolerance range, so this extreme case will
not apply.
P-C P-C
LINE LINE
7% 7%
RED YELLOW YELLOW RED
+-- TARGET----.
SPECIFICATION
Figure 9.19 A Pre-Control Schematic
The chance that two parts in a row will fall outside either P-C line iis 1n times 1n, or
1/49. This means that only once in every 49 pieces can one expec1t to get two pieces
in a row outside the P-C lines due to chance. There is a much greater chance (48/49)
that the process has changed. Therefore, it is advisable to reset the process to the
center. It is equally unlikely that one piece will be outside one P-G line and the next
outside the other P-C line. This is a definite indication that a !»pecial factor has
widened the variation and action must be taken to find that spe!cial cause before
continuing.
VILLANOVA UNIVERSITY IX - 38 LE.t~N SIX SIGMA HANDBOOK

CONTROL CHARTS
The Pre-control Technique (Continued)

Pre-control rules:
• Setup: The job is OK to run if five pieces in a row are inside the target
• Running: Sample two consecutive pieces
• If the first piece is within target, run (don't measure the second piece)
• If the first piece is not within target, check the second piece
• If the second piece is within target, continue to run
• If both pieces are out of target, adjust the process, go back to setup
• Any time a reading is out-of-specification, stop and adjust
The ideal frequency of sampling is 25 checks until a reset is required. Sampling can
be relaxed if the process does not need adjustment in greater than 25 checks.
Sampling must be increased if the opposite is true. To make pre-control even easier
to use, gauges for the target area may be painted green. Yellow is used for the outer
zones and red for out-of-specification.
The advantages of pre-control include:
• Shifts in process centering or increases in process spread can be detected

• The percentage of non-conforming product will not exceed a pre-determined
level
• No recording, calculating, or plotting is required
• Attribute or visual characteristics can be used
• Can serve as a setup plan for short production runs, often found in job shops
• The specification tolerance is used directly
• Very simple instructions are needed for operators
The disadvantages of pre-control include:
• There is no permanent paper record of adjustments

• Subtle changes in process capability cannot be calculated
• It will not work for an unstable process
• It will not work effectively if the process spread is greater than the tolerance

TPM
Total Productive Maintenance (TPM)

Total productive maintenance (TPM) is an activity that promotes c:oordinated group
activities for greater equipment effectiveness and requires olperators to share
responsibility for routine machine inspection, cleaning, maintenance, and minor
repairs. A professional maintenance staff retains respom;ibility for major
maintenance activities and acts as coaches for the routine and minor items.
Productive maintenance combines preventive, predictive, maintainability

improvement techniques, and equipment life cycle costs of equipment to increase
reliability and ease of maintenance. There are "six big losses" that contribute
negatively to equipment effectiveness:
• Equipment failure: from breakdowns

• Setup and adjustment: from setup changes
• Idling and minor stoppages: defective sensors, parts caught on conveyor, etc.
• Reduced speed: the loss between designed and actual op1erating speeds
• Process defects: scrap and quality defects
• Reduced yields: loss of product from machine startups and shutdowns
A further description of the six big losses follows:
1. Equipment failure: Breakdowns cause time losses and lower productivity, as

well as quality losses due to defective products. One must change the mind
set to have zero breakdowns.
2. Setup and adjustment: These are losses from setup changes. One should
reduce the setup times and have better adjustment periods.
3. Idling and minor stoppages: Defective sensors, parts caught on conveyor,

etc., will cause slowdowns and losses. Zero minor stoppages are essential
for unmanned production.
4. Reduced speed: This is the loss resulting from the differences between
designed and actual operating speeds.
5. Process defects: Malfunctioning equipment will produce scrap and quality

defects.
6. Reduced yield: Product losses often result from machinle shutdowns and
startups.
VILLANOVA UNIVERSITY IX -40 LEJ~N SIX SIGMA HANDBOOK

TPM
Total Productive Maintenance (Continued)

Table 9.20 shows illustrative goals for each of the six big losses.
Six big losses Goal

Breakdown losses 0
Setup and adjustment losses <10 minutes
Speed losses 0
Idling and minor stoppage losses 0
Quality defect losses 0
Yield losses minimize
(Nakajima, 1989)15
Table 9.20 Illustrative Goals for the Six Big Losses
Elimination ofthe six big losses, will lead to dramatically improved plant conditions.
The lean manufacturing system could not exist without TPM.
(Nakajima, 1988, 1989)14,15
The most important features of total productive maintenance (TPM) are:
1. Efforts to maximize equipment effectiveness
2. A system of productive maintenance for a machine's life span
3. Implementation by departments such as engineering, maintenance, and

production
4. Involvement of every employee, from top management to the floor employees
5. Autonomous maintenance by operators
6. Company led small group activities

TPM
Total Productive Maintenance (Continued)

The "total" in total productive maintenance (TPM) has the following meanings:
1. Total effectiveness in the pursuit of economic efficiency and profitability.
2. Total maintenance includes maintenance prevention, m,aintainability, and

preventive maintenance.
3. Total participation of all employees includes autonomous maintenance by

operators and small group activities. (Nakajima, 1988)14
The goal of maximizing equipment effectiveness requires the complete elimination

of failures, defects, wastes and losses due to equipment related operations. The
objectives of TPM are zero breakdowns and zero defects. There are reports of firms
achieving just 2% of the original number of breakdowns and a 90% reduction in
process defects through the use of TPM.
Lohr & Bromkamp GmbH (Lobro) in Offebach, Germany chose to place TPM at the
center of its quality initiative. The TPM effort began with an overall cleaning of the
equipment, involving all 1,800 employees. The program had 6 elements:
1. Autonomous maintenance
2. Eliminating the six big losses
3. 100% production quality
4. A planning system for new machines
5. Training for all operators
6. Increased office efficiency (Imai, 1997)8
As a result of the TPM efforts, and other quality improvements, Lobro reported the
following results for 1990 to 1995:
1. Absenteeism was reduced by one-half

2. Suggestions per employee increased from 0.15 to 8 per year
3. Scrap was reduced by one-half
4. Customer rejects were reduced by 90%
5. Training days, per employee, increased from 0.8 to 5 days
6. Setup times were reduced by one-half
7. Throughput time was reduced by 30%
8. Inventory was reduced by 40%

TPM
TPM Metrics
Overall equipment effectiveness is the prime measure used to evaluate TPM. There
are several formula variations. Plants can adjust the factors according to their
needs.
The formulas are:
Overall.Equipment = Availability x PerfC?rmance x Rate of

Effectiveness Efficiency Quality Products
Availability = Operation Time = Loading Time - Downtime

Loading Time Loading Time
Loading time is the available time per shift or per unit minus planned downtime.
Planned downtime includes scheduled maintenance and meetings. Operation time
is loading time minus unscheduled downtime.
Example 9.1: If there are 480 minutes per shift (available time), 15 minutes of setup
time, 10 minutes of planned downtime, 30 minutes of unscheduled equipment
failures, what is the loading time and the availability?
Solution:
Loading Time = Available Time per Shift - Planned Downtime
Loading Time = 480 - 10 = 470 minutes
Availability = 470 - (30 + 15) = 425 =0.904 = 90.4%

470 470
Performance efficiency is defined as the operating speed rate multiplied by the net
operating rate. The operating speed rate is the ratio of the theoretical cycle time to
its actual operating cycle time.
Operating Speed Rate = Theoretical Cycle Time

Actual Cycle Time

TPM
TPM Metrics (Continued)
Example 9.2. If the theoretical cycle time for an operation is 1 miinute per unit, and
the actual cycle time is 1.5 minutes per unit, what is the operating speed rate?
Solution:
Operating Speed Rate = 1.51 minute

minutes
= 0.667 = 66.7%
The net operating rate measures the stability of the equipmen1t, the losses from
minor stoppages, small problems, and adjustment losses.
Net Operating Rate = ActualProcessing Time

Operating Time
Net Operating Rate = Processed Amount x Actual Cycle Time

Operating Time
Example 9.3. If the processed amount is 185 units, the actual cycle time is 1.5
minutes per unit and the operation time is 425 minutes, what is the net operating
rate?
Solution: Net Operating Rate = 185 x 1.5 = 65.3%

425
Now, the performance efficiency can be calculated:
Performance Efficiency = Operating Speed Rate x Net Opelrating Rate
Example 9.4. Using the data from examples 9.2 and 9.3, determin4e the performance
efficiency.
Performance Efficiency = 0.667 x 0.653 = 43.6%
VILLANOVA UNIVERSITY IX-44 LEJ~N SIX SIGMA HANDBOOK

TPM
TPM Metrics (Continued)

Overall equipment effectiveness (OEE) is equal to availability x performance
efficiency x rate of quality products.
Example 9.5. Using the information from Examples 9.1 and 9.4, if the percentage of
good products is 95%, what is the overall equipment effectiveness?
Overall Equipment Effectiveness = 0.904 x 0.436 x 0.95 = 0.374 = 37.4%
The overall equipment effectiveness (OEE) for this example is a poor 37.4%. TPM
prize winning companies have OEE's above 85%. The ideal conditions are:
• Availability greater than 90%

• Performance efficiency greater than 95%
• Rate of quality greater than 99%
{Nakajima, 1988)14
For additional TPM metrics, refer to Nakajima (1989)15.
Steps to Implement TPM

The Japan Institute of Plant Maintenance (JIPM) annually awards a preventative
maintenance (PM) prize. The criteria is based on improvements from TPM, such as
increased productivity and quality. Other factors include the following:
1. Reduced costs
2. Reduced inventory
3. Accident reduction/elimination
4. Pollution control
5. Work environment
The proper integration ofthe philosophy ofTPM within the company will bring about
improved worker and equipment utilization. These changes are aided by improving
employee attitudes, increasing their skills, and providing a supporting work
environment.

TPM
Steps to Implement TPM (Continued)

Nakajima (1988)14 recommends twelve necessary steps to achieve prize winning TPM
results within a 3 year time table.
Step 1: Announce management's commitment to TPM (newsletters, bulletins)

Step 2: Introduce TPM with company communication programs
Step 3: Organize every functional level to promote TPM (in small groups)
Step 4: Establish TPM policies and goals
Step 5: Prepare a detailed master plan for TPM
Step 6: Hold the TPM kick-off (the start of serious work)
Step 7: Form project teams to improve equipment effectiveness
Step 8: Develop the autonomous maintenance program by building skills
Step 9: Develop a scheduled maintenance program
Step 10: Conduct training for operator and maintenance skills
Step 11: Develop early equipment management programs
Step 12: Implement TPM and aim for perfection
Autonomous TPM Small Group Activities

Small group activities, involving overlapping functions, are the heart of TPM. The
department supervisor will normally serve as the team leader. The activities of
equipment cleaning, lubrication, bolting, inspection, etc., are porformed as small
group activities. The activities and time spent on TPM activities are carefully
monitored. The teams will progress through 4 stages:
1. Self-development: learning by members
2. Improvement activities: group improvement activities are fulfilled
3. Problem solving: the group starts to actively select problems to solve
4. Autonomous management: the group selects high-level goals and manages

their work independently
I[Nakajima, 1988)14

VISUAL SYSTEMS
Visual Systems
Visual control systems can be described as the use of production boards, schedule
boards, tool boards, jidohka devices, and kanban cards on the factory floor. The
intent of these techniques is to provide management and workers with a visible
display of what is happening at any moment. In an ideal situation, anyone can
glance across the room and be able to assess or "feel" the condition of the shop
floor, including whether it is operating properly. Visual controls assist the operators
by providing them with information that they would not normally receive. The
operators typically do not have enough information regarding order status, product
quality, lead times, customer demands and costs. (Imai, 1997)8, (Liker, 1997)11
Imai (1997)8 provides three reasons for using visual management tools:
• To make problems visible
• To help workers and management stay in contact with the workplace
• To clarify targets for improvement
Production boards and schedule boards are examples of visual control. These
generally include the posting of daily production, maintenance items, or quality
problems for everyone to see and understand. Before the start of a shift, the
department supervisor uses the boards for a short discussion on planned activities
for the day and any specific problems. The boards are also used by managers and
guests, walking through the plant, to gauge the progress ofthe various departments.
Among the items that can be displayed on the boards are:
• Quality problems • Training records

• Trend charts • Cost reductions
• Checklists • Total productive maintenance
• Safety trends • Machine downtime
• Productivity • 5S activities
• Delivery information • Standardized work instructions

VISUAL SYSTEMS
Visual Systems (Continued)

Visual aids should be:
• Visible: easy to see and understand
• Informative: conveying useful information
• Useful: making a job easier, faster, safer, and rewarding
• Attractive: interesting and enjoyable to view
• Stimulating: motivating to the worker
• Lit: for greatest visual impact they should flash or change
The basic objectives of management are to either influence or diirect behavior.
Visual displays that influence behavior include:
• Safety boards
• Daily meeting boards
• Improvement boards
• Status boards
• Layout maps
• Goal and mission statements
Visual controls that direct behavior include:
• Production/schedule boards
• Defect boards or bins
• Maintenance boards
• Production status boards
• Standard size containers
• Poka-yoke techniques
• Andon or trouble lights
• Labels and color coding
• Tool boards
VILLANOVA UNIVERSITY IX -48 LEJ\N SIX SIGMA HANDBOOK

VISUAL SYSTEMS
Visual Systems (Continued)

Jidohka is defined as a device that stops a machine whenever a defective product
is produced. It is autonomation, that is, a form of automation with human elements
attached. When an equipment malfunction occurs, a light turns red or a signal
comes on to indicate a problem. The operator or maintenance personnel must
respond to find the source of the problem and to resolve it. A machine that is
stopped is an indication of a problem that must be fixed. {Miller, 1993)13
The kanban system provides material control for the factory floor. These cards are
a form of visual control for the flow of production and inventory.
The tool board is a display designed for the tools needed at a work station. This
method is a part of 55 activities. The board is constructed to hold or mark the place
for the tools and includes only the tools required for that work station. When the
tool is not being used, if it is not in its spot, then it is missing or lost and another
tool must be obtained for effective operation of the process. {Liker, 1997)11
The visual factory places an emphasis on setting and displaying targets for
improvement. The concept is that various operations have a target or goal to
achieve. The standard time is initially set higher than the target. However, as the
operation is performed, the operator tries to beat the old time, until the goal is met.
Imai (1997)8 illustrates this concept using an example of a plant goal to reduce the
setup time of machines. A display board was placed next to a machine and the
current setup time was displayed on a graph. Each time a setup was performed, the
setup time was marked. Over time, the operator reduced the setup time by avoiding
past errors and improving methods. Eventually, the targeted setup time was
reached.
In summary, visual systems enable management and employees to see the status
of the factory floor at a glance. The current conditions and progress are evident and
any problems can be seen by everyone.

STANDARD WORK
Standard Work
The operation of a plant depends on the use of policies, procedures, and work
instructions. These could be referred to as standards. Maintainiing and improving
standards leads to improvement of the processes and plant effelctiveness.
If things go wrong in gemba, the workplace, such as defects or unhappy customers,

efforts are made to seek out the root causes, implement corn!ctive action, and
change work procedures. If no problems are encountered in routine daily work
(called maintenance), the process is under control. The first requirement of
management is to maintain the standards. With a system under control, an
improvement stage can be started. That is, management must nClt be satisfied with
the status quo.
If there is a need to increase production, management must find a way to do so. One
of the ways is to have operators change the way they do their jobs. The use of
kaizen activities, kaizen blitz, etc., can be used to improve the process. Once the
changes have been made, efforts should commence to standardize the new
procedures. (Imai, 1997)8
Imai (1997)8 provides a discussion of the word "standards." It seems that the word
standards has a very bad connotation in the western world versu:s that in Japan. In
Japan, standards are used to control the process, not the wor.cers. In the west,
standards imply the use of unfair conditions on workers, such CiS working harder
under extreme conditions, etc. In Japan, standards are used to describe a process
that is the safest and easiest for the workers, and is the most t::ost-effective and
productive way for the company. It is a balance between the twe. parties.
The following are some examples of standards that go beyond procedures and work
instructions:
• Yellow lines on the floor

• Color coding
• Production control board
• Level indicators for minimum and maximum inventory
• Cross-training matrices
• Trouble lights (Suzaki, 1993)24
VILLANOVA UNIVERSITY IX - 50 LEt~N SIX SIGMA HANDBOOK

STANDARD WORK
Standard Work (Continued)

Standards have the following features:
1. Standards are the best, easiest, and safest way to do a job.
2. They preserve know-how and expertise. Years of experience and knowledge

can be lost from the loss of employees.
3. They provide a way to measure performance.
4. Correct standards show the relationship between cause-and-effect, leading

to desired effects.
5. Standards provide a basis for maintenance and improvement.
6. They provide a set of visual signs on how to do the job.
7. Standards are a basis for training.
8. They are a basis for auditing.
9. They are a means to prevent recurrence of errors.
10. Standards minimize variability.

(Imai, 1997)8
Standard work instructions are the documentation of each action required to

complete a specific task. Standard work instructions should always be displayed
at the workplace. (This is similar to the ISO 9001 standards which mention that work
instructions should be accessible to the operator.) If abnormalities appear in the
system, those items can be spotted and eliminated. (Sharma, 2001)19

STANDARD WORK

Standard work is regarded to be one of the most important techniques for achieving
a perfect process. This approach provides the discipline for attclining perfect flow
in a process. Under normal work conditions, with no abnormalilties in the system,
the flow is perfect. The standard work conditions are determined for:
• Takt time
• Ergonomics
• Parts flow
• Maintenance procedures
• Routines
Shingo (1989)21 wrote that Toyota's Ohno mentioned that for standard operations,
much ofthe information will be contained in standard work sheetsi. A standard work
sheet combines the 3 elements of materials, workers, and mcllchines in a work
environment. Toyota refers to it as a work combination.
Standard work sheets, that operators will have confidence in, should consider the
following areas:
• Resource availability • Optimized inventory

• Machine arrangements • Defective prevention
• Process improvements • Operational mistakes deterred
• Worker input • Safe workplace conf::epts
• Tooling improvements • Autonomous systems installed
• Minimized transport
In order to have standard work sheets, waste elimination, problem solving, and
quality methods must be accomplished. The elements that comprise the standard
work operations are:
• Cycle time: The time allowed to make a piece of producUon. This will be
based on the takt time. The actual time will be compared to the required takt
time to see if improvements are needed.
• Work sequence: The order of operations that the worker must use to produce
a part: grasp, move, hold, remove, delay, etc. The same order of work must
be done every time. A standard time is provided for each eIE~ment. A standard
work combination sheet (providing element times), standard work layout
sheet (workplace layout), or planning capacity table (machiine capacity data),
or all 3, are provided to the operator for use.

STANDARD WORK

• Standard inventory: This is the minimum allowable in-process material in the
work area, including the amount of material on the machinery, needed to
maintain a smooth flow. One piece in the machine and one piece handoff are
optimal.
(Shingo, 1989)2\ (Sharma, 2001)19
Shingo (1989)21 and Sharma (2001)19 indicate that standard charts will also include:
• Capacity charts by part: A chart on order of processes, process names,

numbers, basic times, tooling needs, etc.
• Standard task combination sheets: Order of operations for a job.
• Task manuals: Detailed instructions on tool changes, setup changes, or parts

assembly, etc.
• Task instruction manuals: A training manual fortraining workers. These have

equipment layouts, quality check methods, operational procedures, standard
stock required, etc.
• Standard operating sheets: Details from the task instruction manual with
information on equipment layouts, cycle times, order of operations, standard
on hand stock, net work times, safety checks, and quality checks.
Shingo (1989)21 wrote that standard operating sheets should be improved

continuously. Sharma (2001)19 provides a monthly time length for review of work
sheets. The proper implementation of standard work is not as exciting as
performing "kaizen events." It is a time-consuming process. The supervisors and
workers must always be looking to update and maintain the standards.
Sharma (2001 )19 provides a final definition of standard work:
"The best combination of machines and people working together to produce

a product or provide a service at a particular point in time."

TRAINING REQUIREMENTS
Training Requirements
Obviously training is a preventive technique. Effective employee training is one of
the first considerations that an organization must undertake. Trclining is presented
at this point because it is also an effective control technique.
ANSIiISO/ASQ Q 9004 - 2000 2 stipulates that an organization sholuld encourage the

involvement and development of its people by providing ongoing training. The
objective of this standard element is to provide people with the education and
training necessary to improve the effectiveness and efficiency of the organization.
The criteria for the Malcomb Baldrige award (2006)3 has a 25-poilnt category which
emphasizes an organization's education, training, and career dev4elopment methods
to build employee knowledge, skills, and capabilities. This requirement focuses on:
• Tying organizational needs to education and training

• Addressing organizational needs with appropriate training
• Seeking employee input to training and development
• Developing effective education and training
• Reinforcing new knowledge and skills
• Evaluating the effectiveness of education and training basE!d on performance
Kirkpatrick (1994)10 presents 10 factors for the planning and implementation of an

effective training program:
1. Determining needs
2. Setting objectives
3. Determining subject content
4. Electing participants
5. Determining the best schedule
6. Selecting appropriate facilities
7. Selecting appropriate instructors
8. Preparing audiovisual aids
9. Coordinating the program
10. Evaluating the program

Training Requirements (Continued)

Training must be incorporated as part of the strategic business plan and directed to
the achievement of business objectives. Objectives of training include:
• New employee orientation

• Communication of skills unique to the organization
• Team building and productivity improvement
• Imparting new skills, to place responsibilities at lower organizational levels
• Disseminating information on new technology, regulations, or processes
• Development of employees for new positions and promotions
• Corrective action training to prevent recurrence of a mistake
• Retraining employees whose positions have been eliminated
Training Performance Gaps

Wile (1996)28 has developed a cause analysis tool that isolates the specific causes
of performance gaps for the individuals or groups. This tool divides performance
gaps into two major causes (external or internal to the performer), leading to eight
root cause categories. The categories are:
External Performer Factors
1. Organizational systems: clear goals, job design, policies

2. Incentives: compensation, reinforcement, meaningful work
3. Coaching & reinforcement: coach, encourage, provide feedback, acknowledge
4. Cognitive support: job aids, electronic systems documentation
5. Tools: computers, software, instruments
6. Physical environment: noise, light, temperature, layout
Internal Performer Factors
1. Skills and knowledge: training, on-the-job training, self-study

2. Inherent ability: intelligence, emotional attributes, education, motivation
Both Robinson (2000)16 and Wile (1996)28 present the case that the six external
elements are controlled by management, whereas, the performer controls the two
internal elements.
VILLANOVA UNIVERSITY IX- 55 LEAN SIX SIGMA HANDBOOK

Training Needs Analysis

Training needs analysis is a critical step in the training or performance improvement
process. It is usually used to uncover the gaps between adequate and inadequate
job performance. If it is performed at the beginning stages olf a training effort,
proper content direction should be indicated. Smith (1987)23 states that the results
of a training needs analysis are twofold:
1. What is being done now

2. What should be done
Conducting a training needs analysis requires diligent work and effort on behalf of
the analyst. Many or all of the details will lead to the formati(:>n of the training
solution. Smith (1987)23 provides some guidelines for conducting a training needs
analysis in three steps: surveillance, investigation, and analysis"
Surveillance
An organizational trainer should gather information in regards to many key company

needs. These facts, figures, documents, data, etc., should include: policy
statements, news releases, newsletters, policies, procedures, attitudes and feelings
of people, and contacts with people. These and other factors cain provide sensory
data for training efforts.
Investi gation
In this stage, a possible performance gap may be suspected. The next step would
be to gather more details in that specific area. Among the data gathering techniques
are:
1. Individual observation 5. Work sampling

2. Interviews with groups 6. Performance '3ppraisals
3. Questionnaires or checklists 7. Work studies
4. Records of activities 8. Personnel testing
Analysis
At the analysis stage, the gathered information is sorted and examined for validity.
It is then summarized with conclusions drawn and recommendations provided.
Smith (1987)23 divides analysis preparation into the groupings of:
• Organization wide: a focus on performance gaps at the or!ganizational level

• Operations (job) level: overall performance of the job
• Performer level: the individual job holder's performance

Training and Development

Training and development is the process aimed at improving the skills and
expanding the knowledge of employees. Training provides very specific employee
development intended to close the gap between current and desired abilities.
Development is the learning of basic principles, which can be applied to future
conditions. Education teaches a general understanding in selected subject areas.
Training and development should be a continuous process for all employees,
regardless of their position or age. Some of the possible training methods include:
• On-the-job training (done in a job setting)

• Coaching (an experienced person provides specific advice)
• Apprenticeship (an employee serves as an understudy)
• Modeling (job skills are demonstrated)
• Off-the-job training (professional off work site training)
• Management development (this can include executive coaching)
• Mentoring (the pairing of a seasoned professional and a younger one)
• Management simulation games ("real-life"problem solving situations)
(Schermerhorn, 1993)18
Importance of Top-down Support

While most managers agree that training and development of employees is vital to
their company's future, not all managers are willing to allocate the time, money, or
resources. Training is expensive. Training for most organizations is not the product
or service offered for sale. The benefits of training are not always immediately
visible, but the costs are. Upper management must decide the extent of training that
an organization will provide. The desire for training cannot be fulfilled at the lower
levels in the organization, without management top-down support to provide the
necessary funding.
New and rapidly growing companies are often very supportive of training programs.
As the company expands, new skills are needed, and the company fills this need by
hiring new people and by training current personnel.
Training programs need upper, middle and lower management support, if they are
to be successful. The desire for training by a subordinate cannot possibly overcome
resistance by higher management for that training. It is management's authority to
allocate the time and resources to permit training to take place.
VILLANOVA UNIVERSITY IX- 57 LEAN SIX SIGMA HANDBOOK

Resources for Training

The objectives of training are specific new skills and knowledge. The training
department must understand the needs of the employees, their background, and
motivation for learning.
The training materials should be presented at the level appropriate to the target
training group. Teaching statistical process control (SPC), m;ing conventional
statistical formulas, to workers with an elementary grade education will be of little
value. However, showing the same workers how to apply SPC to the products they
manufacture will give these people useful information.
The time and location that training takes place also has an effect on the quality of
training. Training classes at 8:00 AM when people are fresh and alert may be fine for
first shift workers, but it is probably the worst time for third shift: workers.
Environmental and location factors to consider (which enhance the quality of

training) include:
• Good lighting - Lighting should be sufficient for reading, without shadows and
without flickering.
• Room temperature and ventilation - A slightly cooler than normal room

temperature helps keep people awake, but a cold room is undesirable.
Smoking in class is prohibited in most situations.
• Acoustics - Everyone should be able to hear the instructor. Disruptive noises

and vibrations should be minimal.
• Line of sight - Each person should be able to see the black board, overheads,
and other visual displays.
• Furniture - Seating must be comfortable, with back support, and with a writing
surface large enough for a book and paper.
• Accommodations - Convenient access to rest rooms, drinking fountains, and

food services (where applicable) should be provided.
• Timing - Training should be held at a time corresponding to the normal

working hours of the employee, if possible.
• Physical location - The training should be centrally locate!d, with adequate,

safe parking.

Learning Principles
Goetsch (2000)6 summarizes the following principles of learning:
• People learn best when they are ready to learn

• People learn more easily when a new topic relates to a known topic
• People learn best in a step-by-step manner
• People learn by doing (very important)
• Retention is best when people use the new knowledge frequently
• Success in learning tends to stimulate additional learning
• People need both immediate and continual progress feedback
Smith (1987)23 provides the following learning principles:
1. Whole or part learning: Material can be provided in separate or whole

segments. Small or simple segments can be as bad as those that are too
large.
2. Spaced learning: Learning is best at regular intervals (versus cramming). The

availability of the trainer or the trainees may dictate the spacing of learning.
The "incubation" time to absorb new material must be considered.
3. Active learning: People learn by doing. This requires an active role by the
student in the learning process. In the area of statistical training, numerous
articles point to the benefits of active learning.
4. Feedback: There is a need for students (learners) and trainers to know how
the effort is progressing. The learner must know if progress is being made,
and the trainer must know if the message is being transmitted.
5. Overlearning: Forgetting learned material invariably occurs after the class

ends. To reduce the amount of loss in knowledge or skill, the objective is to
learn the skill or knowledge until it can be recalled perfectly. That is, the
subject is overlearned. This requires active participation by the learner.
6. Reinforcement: Providing an incentive or reward for learning is beneficial.

This type of reinforcement is positive. Punishment for a wrong effort is not
considered reinforcement.
7. Primacy and Recency: It appears that information presented at the beginning

and at the end of a learning session are more likely to be remembered than
the middle segments. The trainer may want to develop a strategy to re-
emphasize the middle portions of a session.

Learning Principles (Continued)

8. Meaningful material: During the presentation of new materiial, efforts must be
made to relate the material to what is currently known and to provide
motivation on the usefulness of the new information.
9. Multiple-sense learning: The instructor should plan to use more than one
sense to help the student learn. Students on the average take in information
as follows:
• 80% if information is by sight

• 11 % if information is by hearing
• 9% if information is by other senses
Another source, Goetsch (2000t, provides the following percenta~les regarding what
learners retain from instruction they receive:
• 10% of what is read

• 20% of what is heard
• 30% of what is seen
• 50% of what is seen and heard
o 70% of what is seen and spoken
o 90% of what is said while doing what is talked about
10. Transfer of learning: Transferring learning to the workplace will depend on the
closeness of training to the workplace and to the workplat:e design.

REFERENCES
References
1. Advanced Product Quality Planning (APQP) and Control Plan Reference
Manual, 3rd ed. (2000). Southfield, MI: AIAG.
2. ANSIiISO/ASQ Q9004-2000 Quality management systems - Guidelines for

performance improvements.
3. Baldrige National Quality Program, Criteria for Performance Excellence. (2006)

NIST: Gaithersburg, MD. Available from NIST, http://www.quality.NIST.gov
4. Besterfield, D.H. (1993). Quality Control, 3rd ed. Englewood Cliffs: Prentice
Hall.
5. Breyfogle, F., III. (2003). Implementing Six Sigma, 2nd ed. New York: John
Wiley & Sons.
6. Goetsch, D.L., & Davis, S.B. (2000). Quality Management, Introduction to Total
Quality Management for Production, Processing, and Services, 3rd ed.
Upper Saddle River, NJ: Prentice Hall.
7. Grant, E.L. & Leavenworth, R. (1988). Statistical Quality Control, 6th ed. New
York: McGraw-Hili.

9. Isons 16949:2002 Quality Management Systems, (Corrected version).

Geneva: ISO.
10. Kirkpatrick, D. (1994). Evaluating Training Programs: The Four Levels. San
Francisco: Berrett-Koehler.
11. Liker, J. (editor). (1997). Becoming Lean: Inside Stories of U.S. Manufacturers.
Portland, OR: Productivity Press.
12. Measurement Systems Analysis: MSA, 3rd ed. (2002). Southfield, MI: AIAG.
13. Miller, W.B., & Schenk, V.L. (1993). Alii Need to Know About Manufacturing
I Learned in Joe's Garage. Walnut Creek, CA: Bayrock Press.
14. Nakajima, S. (1988). Introduction to TPM: Total Productive Maintenance.

Cambridge, MA: Productivity Press.

REFERENCES
15. Nakajima, S. (editor). (1989). TPMDevelopmentProgram: Implementing Total
Productive Maintenance. Cambridge, MA: Productivity Press.
16. Robinson, J. (2000, March). "Engaging Line Managers, in Performance

Improvement." Paper presented at the meeting of the Central Illinois
Chapter of ASTD, Champaign, IL.
17. Rybarczyk, P. (November 2005). "The Forgotten C in DMAIC." Quality

Progress, 38 (11) Milwaukee: ASQ.
18. Schermerhorn, J. (1993). Management for Productivity, 4th ed. New York:
John Wiley & Sons.
19. Sharma, A., & Moody, P. (2001). The Perfect Engine: How to Win in the New
Demand Economy by Building to Order with Fewer Resources. New York:
The Free Press.
20. Shewhart, W.A. (1931 reprinted 1980). Economic Control of Quality of

Manufactured Products. Milwaukee: ASQ Quality Press.
21. Shingo, S. (1989). A Study of the Toyota Production System From an

Industrial Engineering Viewpoint. Cambridge, MA: Productivity Press.
22. Six Sigma Academy. (2002). The Black Belt Memory Jogge'r: A Pocket Guide
for Six Sigma Success. Salem, NH: GoaI/QPC.
23. Smith, B., & Delahaye, B. (1987). How to be an Effective Trainer, 2nd ed. New
York: Wiley Professional Development Programs.
24. Suzaki, K. (1993). The New Shop Floor Management: Empclwering People for
Continuous Improvement. New York: The Free Press.
25. Western Electric. (1956). Statistical Quality Control Handbook. Indianapolis,

IN: AT&T.
26. Wile, D. (1996, February). "Why Doers Do." P & I, 35 (2).
27. Wortman, B.L.; Carlson, D.R.; & Richardson, W.R. (2006). CQE Primer. Terre

QUESTIONS
9.1. When investigating TPM, which of the following 9.6. A particular department is using an X-bar chart
would NOT be considered one of the six big with specification limits instead of control limits
negative contributors to equipment for monitoring a key characteristic. The
effectiveness? department supervisor states that the
specifications have more real meaning than the
a. Setup and adjustment control limits. The best immediate reaction
b. Reduced speed should be to:
c. Reduced yield
d. Work cell arrangement a. Congratulate the supervisor for an
outstanding application of control chart
9.2. Identify the LEAST likely result of adopting theory
standardized work procedures: b. Stop the process and check if all readings are
within control
a. They tend to minimize variability c. Stop the process immediately and replace the
b. They are a basis for training specification limits with control limits
c. They ensure marketplace success d. Do nothing; the supervisor should be allowed
d. They preserve know-how and expertise to use the chosen criteria
9.3. There are several types of control plans. Identify 9.7. A lean six sigma project has progressed to the
the appropriate control plan to be used for part point that a control plan is required. Control plan
mock-Ups. activities can be considered closed after which of
the following?
a. Prototype
b. Pre-launch a. A process owner is named for the control
c. Production plan
d. Generic b. A responsible engineer is designated
c. The cross functional team signs off on the
9.4. A P chart: control plan
d. The control plan is a "living document" and is
a. Can be used for only one type of defect per rarely closed
chart
b. Plots the number of defects in a sample =
9.8. A process is in control with p-bar 0.10 and n =
c. Plots either the fraction or percent defective 50. The three sigma control limits for the np
in order of time control chart are which of the following?
d. Plots variations in dimensions
a. 1.37,12.91
9.5. What is a jidhoka? b. 0,11.36
c. -2.01,12.01
a. A totally automated process to reduce d. 3.05, 8.05
defectives
b. A device that identifies good parts
c. A technique which separates defectives from
good parts
d. A device that stops the machine whenever a
defective is produced

QUESTIONS
9.9. After the successful completion and 9.13. If one were to summarizf! the results of a training
implementation of the first four phases of DMAIC, needs analysis into a few words, what would be
the control phase should be completed by which the best selection from the choices presented
of the following? below?
a. The black belt alone; the project has been a. Providing incentives and meaningful work
successfully completed b. Giving cognitive support
b. The process owner alone; it is hislher c. Identifying performa nce gaps
responsibility to take the project from here d. Developing skills anl:l knowledge
c. The process owner and black belt together
d. The complete team is the most desirable 9.14. An X-bar and R chart was prepared for an
option operation using twenty !lamples with five pieces
in each sample; X-double bar was found to be
9.10. Standard work sheets are required for standard 33.6 and R-bar was 6.20. During production, a
operations. Which element is NOT included on sample of five was taken and the pieces
the sheets? measured 36, 43, 37, 25, and 38. At the time this
sample was taken:
a. Cycle time based on takt time
b. Work sequence - the order of operations a. Both the average Elnd range were within
c. Standard inventory on hand control limits
d. The annual or semi-annual review date b. Neither the average nor range were within
control limits
9.11 Many training instructors have developed c. Only the average was outside control limits
approaches to emphasize multiple sense d. Only the range was c)utside control limits
learning. Which of the following options would
be generally recognized to best foster student 9.15. The visual factory would be supported to the
retention? greatest extent by:
a. Reading and hearing I. Muda

b. Hearing and seeing II. Kanban
c. Seeing and speaking III. Gemba
d. Reading and seeing IV.5S
9.12. Considering that all of the following terms have a. I and II

benefits, which would LEAST likely affect b. II and III
product quality? c. II and IV
d. III and IV
a. Muda
b. 5S 9.16. Which of the following is NOT a normally
c. Lean manufacturing accepted control chElrt interpretation rule
d. Total productive maintenance violation?
a. Seven or more consecutive points above or

below the centerline in 20 plot points
b. Six or more conseclutive points, upward or
downward, in more than 20 plot points
c. Two of three points in zone 3
d. Ten points in a row in zone 1
VILLANOVA UNIVERSITY IX-64 LE,'N SIX SIGMA HANDBOOK

QUESTIONS
9.17. Process A consists of several machines that 9.21. Advantages of control charts do NOT include:
combine their output into a common stream.
Once combined, it is impossible to trace single a. They can detect trends of statistical
pieces to specific machines. Process B receives significance
the mixed pieces. A corrective action requires b. They provide straightforward, easily
finding the root cause of a defect found in some interpreted information
of these pieces. A team assigned to this problem c. They provide an ongoing measure of process
is thinking of using SQC to detect the source of capability
the problem. Where should SQC be d. They can detect special causes of variation
implemented?
9.22. After a six sigma team reached the improve
a. At the beginning of Process B phase of the DMAIC cycle, it was clear that a full
b. At each machine in Process A redesign of the part was the right path to take.
c. At the end of Process B From then on, the DMAIC project became a DFSS
d. At the beginning of Process A project. Which control phase should be used for
the newly redesigned part?
9.18 Pre-control starts a process specifically centered
between: a. FMEA
b. Pre-launch
a. Process limits c. Production
b. Specification limits d. Prototype
c. Normal distribution limits
d. Three sigma control limits 9.23. Control limits are set at the three sigma level
because:
9.19. A P chart has exhibited statistical control over a
period of time. However, the average fraction a. This level makes it difficult for the output to
defective is too high to be satisfactory. Internal get out of control
improvement can be obtained by: b. This level establishes tight limits for the
production process
I. A change in the basic design of the product c. This level reduces the probability of looking
II. Instituting 100% inspection for trouble in the process when none exists
/II. A change in the production process through d. This level assures a very small type II error
substitution of new tooling or machinery
9.24. Training needs for an individual are based on all
a. lonly of the following, EXCEPT:
b. I and /II only
c. II and /II only a. The employee's performance
d. I, II, and /II b. The company's objectives
c. The gap between current and desired skills
9.20. Which of the following would be a device d. Availability of qualified instructors
associated with the visual factory?
a. Standard work
b. Andon board
c. Queue time
d. Work cell

QUESTIONS
9.25. Standard work as visualized by the Japanese 9.29. Which of the following tools would NOT be part
means: of the control plan devtllopment?
I. Working harder a. Cause-and-effect aria lysis

II. Working safer b. QFD
III. Working productively c. Regression analysi!.
IV. Working under extreme constraints d. Measurement scalels
a. I and II only 9.30. The Japanese technique that does NOT

b. III and IV only compliment the visual factory concept is:
c. II and III only
d. I and IV only a. Kanban
b. Poka-yoke
9.26. A process is checked by inspection of random c. 5 -whys
samples of four shafts after a polishing d. 5Ss
operation, and X-bar and R charts are
maintained. A person making a spot check picks 9.31. Which of the following indicates an out of control
out two shafts, measures them accurately, and condition?
plots the average value on the X-bar chart. The
point falls Just outside the control limits. The I. A trend of 9 points in an upward direction
department foreman is advised to stop the II. 10 consecutive poilnts above or below the
process. This decision indicates that: center
III. 2 of 3 consecutive points outside the two
a. The process level is out of control sigma warning limitls
b. Both the level and dispersion are out of
control a. I and II only
c. The process level is out of control but not the b. II and III only
dispersion c. I and III only
d. The person Is not using the chart correctly d. I, II, and III
9.27. Ultimate responsibility for training is vested with: 9.32. What is the importance of the reaction plan in a
control plan?
a. The department supervisor
b. Co-workers a. It describes what will happen if a key variable
c. The human resource department goes out of control
d. The employee b. It indicates that a nE!W team must be formed
to react to a problem
9.28. The factor D" In X-bar and R control charts is c. It lists how often lthe process should be
used to: monitored
d. It defines the special characteristics to be
a. Determine the upper control limit of a range monitored
chart
b. Establish the control limits for the average
chart
c. Correct the bias in estimating the population
d. Determine the lower control limit of a range
chart
VILLANOVA UNIVERSITY IX·66 LEAN SIX SIGMA HANDBOOK

QUESTIONS
9.33 The ideal setting for training: 9.37. During variable control charting, a trend of four
consecutive points is noted on both the average
a. Is at the employee's normal work station and range charts. The average chart is
b. Is at the end of the employee's regular work increasing and the range chart is decreasing.
shift One may make which of the following
c. Will have environmental conditions go conclusions?
unnoticed by the student
d. Should match the conditions in an outdoor a. The nominal measurement is increasing
park in Spring time b. The variability is decreasing
c. The process is improving
9.34. The control chart that is most sensitive to d. No conclusions may be made yet
variations in a measurement is:
9.38. After the successful implementation of an
a. p chart "Improve" stage, a DMAIC team is considering
b. np chart control charts to monitor the new gains for a
c. c chart process. Which approach would provide the
d. X-bar and R chart best results?
9.35. How many of the following Japanese techniques a. Chart as many characteristics as possible
are supportive of operational control? b. Chart the most important input variables
associated to the CTQs
I. Poka -yoke c. Use control charts for variables only
II. 5S d. Use numerous control charts for attributes
III. TPM and variables
IV. Kanban
9.39 What % of product should fall between the P - C
a. II and IV only lines (green zone) on a pre-control chart,
b. I, II, and IV only assuming that the process is stable?
c. II, III, and IV only
d. I, II, III, and IV a. 60.0%
b. 68.4%
9.36. Given that resistors are produced in lots of 1,000, c. 86.0%
and the average number of defective resistors d. 95.4%
per lot is 12.7, what are the upper and lower
limits for the control chart appropriate for this 9.40. Which control chart pattern best represents an in
process? control process?
a. =
LCL 2.0 UCL 23.4 = a. A consecutive run of seven or more points on
b. =
LCL 3.8 UCL 20.2 = one side of the centerline
c. LCL = 0.031 UCL 0.131 b. A random distribution of points with one
d. LCL = 1.5 UCL = 26.7 point outside the control limits
c. A random distribution of points on both sides
of the centerline
d. A steady trend of points toward either control
limit

QUESTIONS
9.41. An R chart is generally used to: 9.45. Standard work does NOT rely on which of the
following tools?
a. Determine if the process is in control
b. Determine if the process mean is in control a. Visual factory
c. Determine if the process variance is in b. 5S
control c. Kanban
d. Determine the variance of the process d. EWMA
9.42. Which of the following elements is NOT part of a 9.46 The single most important factor in establishing
control plan form? an effective company trclining effort is:
a. Specifications a. Top management SUIPport

b. Potential causes of failure b. Willingness of an employee to learn
c. Key input variables c. Immediate supervisor's support
d. Key output variables d. Good instruction and materials
9.43. Which of the following charts have upper control 9.47. An R chart is most closely related to which ofthe
limits, but frequently have lower control limits of following?
zero?
a. c chart
a. X-bar and individual charts b. S chart
b. c charts and u charts c. u chart
c. p charts and np charts d. X-bar chart
d. R and sigma charts
9.48. What lean technique is most widely used to make
9.44. The design of a control plan for a particular part problems visible?
incorporates information from a variety of
sources such as flow charts, QFD, FMEAs, a. JIT
designed experiments, and statistical studies. It b. 5S
is a tool to monitor and control the part or c. Kaizen
process. If used properly, the control plan d. SMED
avoids which of the following problems?
a. Becoming a substitute for written operator

instructions
b. Having a listing of the critical Xs and Ys of
the process
c. Error proofing the process through various
control plans
d. Being used as evidence of installed controls

I
x. DESIGN IMPROVEMENT
THE ONLY THING WE CAN'T MAKE IS SOMETHING WE

CAN'T THINK ABOUT.
HENRY J. FORD
VILLANOVA UNIVERSITY X ·1 LEAN SIX SIGMA HANDBOOK

X. DESIGN IMPROVEMENT
DFSS, IDOV, DMADV
Design Improvement
Design Improvement is presented in the following topic areas:
• DFSS (IDOV & DMADV) • DFX

• QFD • TRIZ
• Robust Design • Systematic Design
• FMECA (FMEA) • Creative Design Ideas
Design for Six Sigma

Design for six sigma is the suggested method to bring order to product design.
Hockman (2001)19 indicates that 70% - 80% of all quality problems i:lre design related.
Emphasis on the manufacturing side alone will concentrate at the tail end of the
problem solving process. The emphasis should be at the frolnt end. Problem
solving downstream is more costly and time-consuming than fixing it at the source.
In 1999, NIST reported that the automotive supply chain lost at lealst a billion dollars
a year due to poor interoperability of digitally designed product data.
There has been considerable emphasis in recent years by American industry in

downsizing, restructuring, process redesigning, and instituting ICOSt containment,
etc. These methods are directed at holding the line on costs. Hamel (1994)16
describes this as denominator management.
In the business world, the equation for return on investment, or return on net
operating assets, has both a numerator - net income, and a denominator -
investment. Managers have found that by cutting the denominator, investments in
people, resources, materials, or other assets is an easy way to make the desired
return on investment rise (at least short-term).
To grow the numerator of the equation requires a different way of thinking. That
thinking must include ways to increase sales or revenues. One of the ways to
increase revenues must include introducing more new products"
Cooper (1993)9 states that new products account for a large percentage of company
sales (40%), and profits (46%). Of course, not every new product will survive. Two
studies listed in Table 10.1 provide some statistics.
VILLANOVA UNIVERSITY X-2 LEAN SIX SIGMA HANDBOOK

DFSS, IDOV, DMADV
Design for Six Sigma (Continued)
Development Items Study A Study B

New product ideas 7 11
Products entering development 4 3
Products launched 1.5 1.3
Successful products 1 1
Table 10.1 Progression of New Products Through Development
Table 10.1 indicates that a large amount of ideas are needed. These ideas are
sorted, screened, and evaluated in order to obtain feasible ideas, which enter the
development stage, pass into the launch stage, and become successful products.
Cooper (1996)10 provides more details of how winning products are obtained:
1. A unique, superior product: This is a product with benefits and value for the
customer.
2. A strong market orientation: An understanding of customer needs and wants

exists.
3. Predevelopment work: Up front activities such as screening, market analysis,

technical assessment, market research, and business analysis are vital.
4. Good product definition: A company must undertake good product and

project definition before development begins.
5. Quality of execution: The development process has many steps. A company

must execute these steps with the proper amount of detail and correctness.
6. Team effort: Product development is a team effort that includes research &
development, marketing, sales, and operations.
7. Proper project selection: Poor projects must be killed atthe proper time. This
provides adequate resources for the good projects.
8. Prepare for the launch: A good product launch is important and resources
must be available for future launches.

DFSS, IDOV, DMADV

9. Top management leadership: Management must provide gluidance, strategy,
resources, and leadership.
10. Speed to market: Product development speed is the wealPon of choice, but
sound management practices should be maintained.
11. A new product process: This is a screening (stage gate!) process for new
products.
12. An attractive market: An attractive market makes ilt easier to have a

successful product.
13. Strength of company abilities: The new product provides cl synergy between
the company and internal abilities.
There are many product development processes to choose from. Rosenau (1996)30
suggests that the former "relay race" process (one function pa:ssing the product
from marketing to engineering to manufacturing and back thr10ugh the loop) is
obsolete. Multi-functional team activities involving all departments are necessary
for effectiveness and speed to market. The process is comprised of 2 parts: a
"fuzzy front end" (idea generation and sorting) and new product dE!velopment (NPD).
The complete NPD process includes 5 activities:
• Concept study: A study is needed to uncover the unknowns; about the market,
technology, and/or the manufacturing process.
• Feasibility investigations: There is a need to determine thE! limitations of the

concept. Find out if the unknowns are resolvable, or if new research improves
the project.
• Development of the new product: This is the start of the NPD process. This
includes the specifications, needs of the customer, target markets,
establishment of multi-functional teams, and determination 4:>f key stage gates.
• Maintenance: These are the post delivery activities associated with product
development.
• Continuous learning: Project status reports and evaluaticms are needed to

permit learning.
(Rosenau, 1996)30

DFSS, IDOV, DMADV

Stage Gate Process Clarification
A stage gate process is used by many companies to screen and pass projects as
they progress through development stages. Each stage of a project has
requirements that must be fulfilled. The gate is a management review of the
particular stage in question. It is at the various gates that management should make
the "kill" decision. Too many projects are allowed to live beyond their useful lives
and clog the system. This dilutes the efforts of project teams and overloads the
company resources. Table 10.2 illustrates some sample stages from Rosenau
(1996)30.
Stage Xerox Exxon CalComp

Get an idea Pre-concept Idea
Prove it works Concept Preliminary
assessment
Financial Assessment Market
assessment specifications
Develop Design Development Design engineering
and demonstration model prototype
test
System verification
Mfg verification
Scale up Production Validation Production
Launch Launch Commercial
launch
Post delivery Maintenance End of life
support (obsolete)
Continuous Review
learning
Table 10.2 Product Development Stages for Various Companies
The above Table presents several examples of new product development processes.
The individual organization should customize their process and allow a suitable time
period for it to stabilize.

DFSS, IDOV, DMADV

In the area of new product management, Crawford (1997)11 and Cooper (1993)9
describe some commonly accepted new product terms:
1. New-to-the-world products: These are inventions and discoveries such as

Polaroid cameras, laser printers, in-line skates, etc.
2. New category entries: These are company products that are not new to the
world, but new to the company. A "me too" type product.
3. Additions to product lines: These products are extensions of the

organization's existing product line. Examples are Diet Coke, caffeine free
Diet Coke.
4. Product improvements: Current products made better.
5. Repositionings: Products that are retargeted for a new use. The original
purpose was not broad enough. Arm & Hammer baking soda has been
repositioned as a drain deodorant, refrigerator deodorant, etc.
6. Cost reductions: New products which are designed tel replace existing
products, but at a lower cost.
GE Plastics (Mt. Vernon, IN) has formalized their product design development
process (Harold, 1999)17. It is described as designing for six sigma using the
product development process. The methodology is used to pn:>duce engineered
plastics through a series of tollgates that describe the elements needed for
completion of a stage. Best practices are used in each stage, including:
• Understanding critical to quality characteristics for extE!rnal and internal

customers
• Conducting failure mode and effects analysis (FMEA)
• Performing design of experiments to identify key variables
• Benchmarking other facilities, using competitive analysis, surveys, etc.
Phadnis (2001)26 in an article "Design for Six Sigma Roadmap" provides various
checklists to guide the project team toward completion of a project.

DFSS, IDOV, DMADV

IDOV
With DMAIC, a lean six sigma team takes an existing process, defines a problem
within that process, and follows a series of steps to improve the current state. IDOV
(identify, design, optimize, and validate) takes this approach one step further. The
central concept of variation reduction provides new leverage when it starts at the
design phase of products and processes.
DMAIC attains improvement in existing products and processes while IDOV, and
other DFSS methodologies, quantifies the steps necessary to achieve six sigma
quality in new products and processes. (Shina, 2001)31
Treffs (2001 )37 and Simon (2000)32 provide additional insight on the development of
other six sigma design methods. A standardized approach has not yet been
established, but most authors recommend a framework that tries to remove "gut
feel" and substitutes more control.
Treffs (2001)37 presents the following four-step IDOV model:
• Identify: Use a team charter, QFD, FMEA, and benchmarking.
• Design: Emphasize CTQs, identify functional requirements, develop

alternatives, evaluate, and select.
• Optimize: Use process capability information, a statistical tolerancing

approach, robust design, and other various six sigma tools.
• Validate: Test and validate the design.
Vandervort (2001 )38 defines each IDOV step as follows:
• Identify: Overall product requirements are defined. Marketing information and

customer feedback are quantified and ranked. Quantified customer
information becomes the first set for critical to quality (CTQ) features.
Technical requirements, performance targets, and specifications are
identified.

DFSS, IDOV, DMADV

IDOV (Continued)
• Design: Transfer functions are applied to develop the ()verall layout and
geometry of the product. A statistical quality approach is used in this step.
Product quality scorecards (PQS) are defined at this point. Some of the tools
common to this phase are design of experiments (DOE), finite element
analysis (FEA), and statistical inference.
• Optimize: The objective of the optimize step is to minimize:! the sensitivity of

the new design to eTQ design features. Capability analysis is performed for
each component and sUb-component according to the product quality
scorecard. Tolerances are determined from the PQS and the resulting
capability assessment.
• Validate: PQSs are used as input to show the required IE!Ve I of quality has
been achieved. Data from quality control and field studies confirm that the
product satisfies the required specifications.
DMADV
Simon (2000)32 provides a five-step DMADV process for six sigma design. The
DMADV method for the creation of a new product consists of the following steps:
• Define: Define the project goals and customer deliverable:,
• Measure: Measure and determine customer needs and spt!cifications
• Analyze: Analyze the process options to meet the customers' needs
• Design: Develop the process details to meet the customers' needs
• Verify: Verify the design performance and ability to meet customer needs
The DMADV methodology should be used when:
• A product or process is not in existence and one needs to be developed
• The product or process exists and has been optimized, but still doesn't meet
the customers' specification or six sigma level
(Simon, 2000)32
VILLANOVA UNIVERSITY x-a LEAN SIX SIGMA HANDBOOK

DFSS, IDOV, DMADV

Planning for DFSS
As six sigma implementations progress, new projects become more complex. The
development phases will typically follow the sequence below:
• Training and launch: Teams learn the DMAIC methodology and projects
selected represent "low-hanging fruit" which yield gains from obvious
improvement opportunities with easy solutions.
• Implementation: Teams work on improvement projects which address core

business objectives and use statistical tools to enhance process efficiency
and predictability.
• Design for six sigma (DFSS): Teams attack the most complex projects, with
the greatest potential gains. Solutions require more than improvement and
become a matter of new process and product designs.
{ASQ Staff, 2003)5
DMAIC projects tend to be easier and faster than design projects, hence they provide
early gains. DFSS projects tend to take longer to implement, but res~lt in solutions
to complex design projects. Organizations with well-developed six sigma programs
run DMAIC and DFSS projects concurrently. The following steps help determine an
organization's readiness to deploy DFSS:
1. Monitor sigma levels: Steadily rising sigma levels are an indication of an

effective program. When increases begin to slow, existing processes may
have been improved as far as possible, and redesign may be necessary.
2. Written schedule of prioritized project ideas: Plan projects well in advance

and prioritize them by complexity. When simpler improvements run out, make
sure the organizational staff is ready to apply DFSS.
3. Remain aware of marketplace changes that could cause products or

processes to become obsolete: Changes in customer requirements, market
demand, or technology may suggest a need to apply DFSS to develop new
products or processes.
4. Gauge the organization's capability for success with DFSS: Are improvement
projects being completed within projected timelines? Should the
management team be more involved? Are adequate resources available?
{ASQ Staff, 2003)5

DFSS, IDOV, DMADV

When presented with a new or modified product or process, the design engineerwill
select a design process. A typical design process (The French model) is depicted
from Cross (1994)12:
Analysis of Statement of Conl;eptual

Need
--+ Problem L-+ Problem r+ DI~sign
I
+
Selected Embodiment Working
Schemes --+ of Schemes --+ Detailing
r--+ Drawings , etc
Figure 10.3 The French Design Model
The designer (and design team) will capture the needs, provide analysis, and
produce a statement of the problem. The conceptual design will generate a variety
of solutions to the problem. This brings together the elements of engineering,
science, practical knowledge, production methods, and practices. The embodiment
of schemes step produces a concrete, working drawing (or item) from the abstract
concept. The detailing step consolidates and coordinates the fine points of
producing a product.
The designer of a new product is responsible for taking the initial concept to final
launch. In this effort, the designer will be part of a team. The project manager,
product manager, or general manager for a new product or new design team (which
includes marketing, sales, operations, design, and finance) will n,eed to manage the
process.
The use of six sigma tools and techniques must be introduced in a well thought out
manner at various phases of the project. The remainder of this Section details some
of the available tools. The student should be aware that there are numerous other
DFSS techniques that are not presented.
VILLANOVA UNIVERSITY X -10 LEAN SIX SIGMA HANDBOOK

QFD
Design Using Quality Function Deployment

Quality function deployment is a tool that is sometimes referred to as the "voice of
the customer," or as the "house of quality." Hauser (1988)18 illustrates how the
house of quality can be used to change the product. By describing the product in
the language of the engineer, along the top of the house of quality, the design team
lists those engineering characteristics that are likely to affect one or more of the
customer attributes. "The interfunctional team fills in the body of the house, the
'relationship matrix,' indicating how much each engineering characteristic affects
each customer attribute." The strength of these relationships is indicated by
numbers or symbols.
"By comparing weighted characteristics to actual component costs, creative design

teams set priorities for improving components." (Hauser,1988)18. It is important to
focus on customer satisfaction values when considering engineering
characteristics. Increasing one engineering characteristic may have a negative
impact on another engineering characteristic. It may also be an indication that a
different solution is required in the design. "Engineering is creative solutions and
a balancing of objectives. The house of quality's distinctive roof matrix helps
engineers specify various engineering features that have to be improved
collaterally." (Hauser, 1988)18
The foundation of the house contains the benchmarking or target values. The values
indicate "how much" for each of the measures. For the example given in Figure
10.4, the "how much" target values include the temperature range, book weight,
maximum errors, cost, etc. This area may also be used to indicate objective
measures for competitive products and technical importance of each factor. There
is a tendency to give a tolerance for the target values, but it is better to set the target
values as single objectives, and then rate the engineering characteristics in terms
of the ability of achieving the target values.
The right-hand wall of the house in Figure 10.4 indicates the customer competitive
assessment and other factors affecting the customer. The competition comparison
shows graphically the relative weights between this product and the competition.
The elements which are included in the house are customized to the particular
product or service being described. When reviewing a completed house, the easiest
method is to look at each area separately (walls, ceiling, and foundation), to
understand the factors, relative strengths, and interactions.

QFD
Design Using Quality Function Deployment (Continued)
-
~ DESIGN FEATURES
P = POSITIVE INTERACTIONS
N = NEGATIVE INTERACTIONS
I--
Z (J) COMPETITIVE
W Z
RANKINGS I--
0 COMPARISON
Z
5 = MOST IMPORTANT 0 i=
0= NO IMPORTANCE u W
N (J) U x:
W 0
W > in 0:: W (J) III
..J I--
I-- 0 COMPETITIVE
(!) III ::::i u 0:: U W
..J
I--
Z ASSESSMENT
Z <C iii 0-<C w0:: ii: III W a::
I-- 0- 5 = BEST w
52 0- <C <C 0:: (/)
w w
CUSTOMER Z <C 0:: :i!i ;: ;: > 0:: 3 = AVERAGE a:: :!!;
I-
I-
NEEDS <C a => 0 W 0 0 => 1 = WORST 0 « w
- 0:: <C a u u. ..J :i!i u :s: (/) al
COMPREHENSIVE 5 1 2 1 1 3 5 5 4 3 4 4 3
LOW COST
4
-
2 0 4 3 1 5 3 0 4 3 4 3 3 3 /
-5
UP·TO-DATE
I
EASILY AVAILABLE
TEST QUESTIONS
RATINGS
-4
-5
'--
4
0
5
1
1
0
0
3
2
3
0
5
165 121 136 [46 135 153 170

0
4
1
2
0
5
5
0
5
I
5
4
5
..J
<C
=>
3
4
4
3
4
2
4
5
5
I--
0::
3
3
4
e>
a
3
3
4
(J)
<:
..J Z
Z <C 0 (J) 0
TARGET VALUES 0:: u. (J) U It) (J) 0:: <C x: => 0- ::c
0 0 III W
(J)
..... W W :i!i 0 Z 0- i= I--
I-- ....""+ ..J
CD U5
~
v ..J ;: 0:: 0
III
<C =>
(J)
0::
W
W
:i!i
U v 0 W => :i!i
>
=> i:i: 0:: ::c :; 0 >< > ..J
<C a (J)
0:: 0 M W <C W
I-- 0 0:: U ..J
(J) 0:: 0:: <C
f1 Z (J)
~ W ::c
M U
W
wI-- I--
OUR MANUAL 89 98 5.2 2 65 3 91
OBJECTIVE o::>Z
-w
MEASURES X BOOK 77 95 3.9 5 65 0 79 w!::::i!i
:i!il--(J)
Y MANUAL 81 76 4.5 4 65 3 83 OW(J)
I--O-W
ABSOLUTE 83 27 44 91 89 15 72 (J):i!i(J)
TECHNICAL =>O(J)
MPORTANCE RANKING 3 6 5 1 2 7 4 uu<c
Figure 10.4 A Hypothetical House of Quality Example
The "how" portion ofthe above Figure is identified as design featu res. Various other
texts have used "engineering characteristics," "design requirements," "technical
descriptors" and "technical details" as descriptions. The "how mLich" floor area and
the "comparison" right wall area of the house in Figure 10.4 are included to show
their impact on the design of the product.

QFD
Design Using Quality Function Deployment (Continued)

After setting the primary design characteristics, Hauser (1988)18 suggests using the
"hows" from the house of quality as the "whats" of another house that depicts
detailed product design. This process is repeated with a process planning house
and then production planning house. In this way, the voice of the customer is
carried through from design to manufacturing. This is illustrated in Figure 10.5.
DIU
o
0-"0-
c::';:
0
..
o
o
U
00
00
GI"
GIGI
c ....
0_ U
Clco
0
t:~
0
co ....
o.U
I!!
-
.. ....
Glc
Uo
0 0
o.l!!
-
c .. >.GI
w~ co GlQ.
.c ~o
U U
House of Parts Process Production

quality deployment planning planning
(Hauser, 1988)18
Figure 10.5 Linked House of Quality Example
While it is easy to get caught up in the process of constructing the house(s) and
completing entry ofthe data, one should not lose sight ofthe objectives ofthe house
of quality methodology. Hauser (1988)18 states, "The house of quality is a kind of
conceptual map that provides the means for interfunctional planning and
communications." "The principal benefit of the house of quality is quality in-house.
It gets people thinking in the right direction and thinking together."
The voice of the customer, both external and internal, is quantified and presented
in the format of a house of quality. The different organizational functional groups,
engineering, marketing, manufacturing and so on, are able to see the effect of design
and planning changes in order to balance customer needs, costs, and engineering
characteristics in the development of new or improved products and services.
Various authors use the basic theme of the house of quality presented by Hauser
and include additional factors and symbols to show other data relationships.

ROBUST DESIGN
Robust Design Introduction

Dr. G. Taguchi wrote that the United States has coined the term "Taguchi Methods"
to describe his system of robustness for the evaluation and improvement of the
product development processes. He has stated that he preferred the term "quality
engineering" to describe the process. (Ashley, 1992)\ (Taguchi, 1993)35. Other
authors have used robust design or robust engineering (Wilkins, 2000)40 to describe
the process. Any of the above mentioned terms can be used. Quality engineering
will be the term used when material is referenced from Taguchi's texts or articles.
Otherwise in this Section, robust design will be the main term 01f choice.
Robust Design Approach

Robust design processes are one of the more important developments in design
processes in recent years. The use of robust approaches for dl~sign is a process
that, when used, can produce extremely reliable designs both during manufacture
and in use. Robust design uses the concept of parameter control to place the design
in a position where random "noise" does not cause failure. The diagram below
describes the process of robust design.
Noise Factors
,r
Signal Factor ""

rr Products/Procedures .. Response
..III-
Control Factors
Figure 10.6 A Robust Design Schematic

ROBUST DESIGN
Robust Design Approach (Continued)

The concept is that a product or process is controlled by a number of factors to
produce the desired response. The signal factor is the signal used for the intended
response. That is, the actions taken (signal) to start the lawn mower (response) or
the dial setting (signal) to obtain a furnace temperature (response). The success of
obtaining the response is dependent on control factors and noise factors.
Control factors are those parameters that are controllable by the designer. These
factors are the items in the product or process that operate to produce a response
when triggered by a signal. For instance, in the case of the furnace, the control
factors might be the design of the thermocouple and heat controller. Control factors
are sometimes separated into those which add no cost to the product or process
and those that do add cost. Since factors that add cost are frequently associated
with selection of the tolerance ofthe components, these are called tolerance factors.
Factors that don't add cost are simply control factors. Noise factors are parameters
or events that are not controllable by the designer. These are generally random, in
that only the mean and variance can be predicted.
Noise factors in furnace design might be:
• Line voltage variations

• Outside temperature
• Parallax errors in dial setting
These noise factors have the ability to produce an error in the desired response. The
function of the designer is to select control factors so that the impact of noise
factors on the response is minimized while maximizing the response to signal
factors. This adjustment of factors is best done using statistical design of
experiments or SDE.
A Robust Design Example

Experimental design techniques were developed in England and in the United States
many years ago. They were primarily developed by and for statisticians. Engineers
were not widely exposed to the techniques in ways which seemed relevant to them.
Meanwhile, Genichi Taguchi, a Japanese engineer, researched the literature and
found the techniques very useful. He concluded, however, that Japanese engineers
probably would not use the techniques unless they could be presented
pragmatically. He decided to do that, and by the early 1980s, Americans visiting
Japan noticed a "new" powerful technique being used. SDE is another Japanese
technique invented in America and England.

ROBUST DESIGN
A Robust Design Example (Continued)

The most celebrated case of design of experiments was probably that of a parameter
design experiment at a tile manufacturing company in Japan. In simple terms, some
factors (which were less expensive to control) were fixed at requisite levels such that
the output (variation in tile dimension), was made insensitive to a noise factor
(temperature variation). See the Genichi Taguchi (1987)34 reference at the end ofthis
Section.
In 1953, a mid-size tile manufacturing company in Japan was having a serious

problem with their $2 m kiln purchased from West Germany. The problem was
extreme variation in the dimensions of the tile produced. The stacked tiles were
baked inside a tunnel kiln as shown below. Tiles toward the outside of the stack
tended to have a different average dimension and exhibited more variation than
those toward the inside of the stack.
Burners
Inside Outside
Tiles Tiles
Figure 10.7 A Schematic of a Tile Tunnel Kiln
The cause of variation was readily understandable. There was an uneven

temperature profile inside the kiln. To correct the cause, the company would have
to redesign the kiln, which was a very expensive proposition. This company's
budget didn't allow such costly action, but the kiln was creating a tremendous
financial loss for the company, so something had to be done.
Although temperature was an important factor, it was treated as a noise factor. This
meant that temperature was a necessary evil and all other factors would be varied
to see if the dimensional variation could be made insensitive to tE~mperature. In Dr.
Taguchi's words, whether the "robustness of the tile design" cc,uld be improved.
People (the engineers, chemists, etc.) having knowledge about the process were
brought together. They brainstormed and identified seven major cc>ntrollable factors
which they thought could affect the tile dimension.

ROBUST DESIGN
A Robust Design Example (Continued)

These were: (1) limestone content in the raw mix; (2) fineness of the additives; (3)
amalgamate content; (4) type of amalgamate; (5) raw material quantity; (6) waste
return content; and (7) type of feldspar. After testing these factors over specified
levels using an orthogonal design, the experimenters discovered that factor #1
(limestone content) was the most significant factor, although other factors had
smaller effects. It was found that by increasing the limestone content from 1% to
2%, and by choosing a slightly better level for other factors, the percent warpage
could be reduced from 30% to less than 1%. Fortunately, limestone was the
cheapest material in the tile mix. Moreover, they found through the experimentation
that they could use a smaller amount of amalgamate without adversely affecting the
tile dimension. Amalgamate was the most expensive material in the tile.
This is a classic example of improving quality (reducing the impact of a noise factor),
reducing costs (using less amalgamate) and drastically reducing the number of
defectives at the same time.
Some of the key robust design steps discussed by Phadke, (1989)25 are concept
design, parameter design, and tolerance design.
Concept Design
Concept design is the selection of the process or product architecture based on

technology, cost, customer, or other important considerations. This step depends
heavily on the abilities and creativity of the designer.
Parameter Design
During the parameter design stage, the design is established using the lowest cost
components and manufacturing techniques. The response is then optimized for
control and minimized for noise. If the design meets the requirements, the designer
has achieved an acceptable design at the lowest cost.
Tolerance Design
If the design doesn't meet requirements, the designer then gives consideration to
more expensive components or processes that reduce the tolerances. The
tolerances are reduced until the design requirements are met. With robust design
approaches, the designer has the ability to produce a design with either the lowest
cost, the highest reliability, or an optimized combination of cost and reliability.

FMECAlFMEA
Failure Mode and Effects Analysis (FMEA)(IFMECA)

FMEAs are used by companies and institutions to manage risks. The risks involve
potential failures and their effects. A FMEA should be an up-front event; certainly
not an "after-the-fact" exercise. The costs associated with prevl~nting failures are
certainly less than the costs of correcting a failure after it has oc:curred.
AIAG (2001)1 defines FMEA as:
" ... a systematic group of activities intended to: a) recognize, and evaluate the
potential failure of a product/process and the effects of that failure, b) identify
actions that could eliminate or reduce the chance of the! potential failure
occurring, and c) document the entire process."
A FMECA provides the design engineer, reliability engineer, and others a systematic
technique to analyze a system, subsystem, or item for all potential or possible failure
modes. This method then places a probability that the failure mode will actually
occur and what effect this failure has on the rest of the system.
A FMEA or FMECA (in some cases there is little difference) is a dletailed analysis of
a system, down to the component level. Once all items are clas.sified as to the 1)
failure mode, 2) effect of failure, and 3) probability failure will occ:ur, they are rated
as to their severity via an index called a risk priority number (RPN).
(Dovich, 2002)13
FMEA Steps and Benefits

FMEAs are used for new designs and technologies, the modifica1tion of an existing
design/process, in a new environment or in a new application. ThE! purpose of using
FMEAs is to reduce or eliminate failures or errors using the following steps:
• Decide on the process

• Study the process
• Identify potential failure modes
• Rate the severity, occurrence, and detection levels
• Compute the risk priority number (RPN) and prioritize
• Implement the corrective actions and re-evaluate the RPN~;
• Update the table (an often neglected step)
VILLANOVA UNIVERSITY X -18 LEJ\N SIX SIGMA HANDBOOK

FMECAlFMEA
FMEA Steps and Benefits (Continued)

The benefits to a properly performed FMEA include:
• Improved product/service functions

• Lowered warranty costs
• Reduced manufacturing problems
• Increased safety and reliability of products and processes
• Improved customer satisfaction measures
• Decreased business problems
In the DFSS realm, FMEAs will be used in the design phase of product development
(concept, design, optimize, and verify), while FMEAs are used in the development
phase of technology development (invent/innovate, develop, optimize, and verify).
(AIAG, 2001) 1 , (Stamatis, 1995)33
FMEA Types
There are a variety of types of FMEAs. However, Stamatis (1995)33 has classified
them into four types: system, design, process, and service. Other writers provide
more detailed breakdowns, but these four types are primarily used. The FMEA form
is not standardized, so a company can customize it for their use. Most of the forms
have common elements. The form to use may be mandated from the customer. The
big three American automotive manufacturers request the use of forms consistent
with the AIAG FMEA manual.
System FMEAs
System FMEAs (SFMEA) deal with systems, subsystems, and components, along
with the interaction between systems and elements of the system. There is to be a
balanced approach to effectiveness, performance, and costs. The inputs can be
derived from a QFD (quality function deployment - voice of the customer). The
potential causes of failures can be gleamed from warranty claims, customer
complaints, field service data, reliability data, and feasibility studies.

FMECAlFMEA
Design FMEAs
An objective design FMEA (DFMEA) will reduce the risk of failUl~es. This includes
functional requirements and design alternatives. It should analyze products prior
to the release of production drawings for tooling and to manufacturing (before-the-
fact). The focus is on failure modes caused by design deficienci4es. Manufacturing
issues are not in the DFMEA. Those should be included in process FMEAs.
Design FMEAs do not rely on process controls to correct potential design flaws.
One should acknowledge the manufacturing process capabilitiies with regard to
specifications, tolerances, finishes, tooling, or other sources. The team should
involve individuals from design, testing, reliability, materi;:lls, service, and
manufacturing. DFMEAs are conducted on the superior concepts that emerge from
a selection process. 76% of engineering changes are due to pOlor design and the
remaining 24% are due to improvements.
The potential causes of failure modes may be related to materia ~s, processes, and
costs. For detection methods, one would use simulation, mathematical modeling,
prototype testing, design of experiments, verification testing, pmduct testing, and
tolerance stack-up studies. A design FMEA will be presented la1ter in this Section.
Process FMEAs
Process FMEAs (PFMEA) study the manufacturing and assemlbly process. The
purpose is to focus on the failure modes caused by process or assembly operations
and to minimize those deficiencies. The PFMEA is initiated before or after the
feasibility stage, prior to tooling for production, and accounts for all manufacturing
operations. A flow chart of the process must be made available, along with the
DFMEA document for reference. A QFD can provide further customer input. The
improvement team should involve individuals from design, assembly, operations,
materials, quality, service, and suppliers.
The PFMEA should not rely on product design changes to correct process
weaknesses. Some of the potential issues include too much, tOCI little, missing, or
wrong elements pertaining to labor, machine, methods, materials, measurement, and
environment. The characteristics of a product can be classifiE!d as: 1. critical -
shall/must comply with safety/government regulations, or service requirements; 2.
significant - customer and supplier quality features; 3. key - characteristics that
provide prompt feedback on the product aiding in the corrective action process.

FMECAlFMEA
Service FMEAs
Service FMEAs investigate services before they reach the customer. The focus is
on failure modes (tasks, errors, mistakes) caused by system or process deficiencies
before the first service. The service FMEA will cover the non-manufacturing aspects
and includes maintenance contractors, financial services, legal, hospitality services,
government and educational institutions, and healthcare. The inputs to the FMEA
may be derived from QFD, benchmarking, market research, and/or focus groups.
The potential causes of failures may be from: labor, machine, methods, materials,
measurement and environment. The service FMEA can be termed healthcare FMEAs
for the healthcare industry.
FMEA Follow-up
The team leader (and eventually the process owner) should follow-up on the FMEA
to verify that the implementation actions were undertaken and completed. Some of
the follow-up steps include:
1. Reviewing the FMEA for function, purpose, and objectives

2. Noting the high risk, high severity, high RPN values
3. Identifying the critical, significant, and major characteristics of the process
4. Seeing that the control plan is in place
5. Conducting capability studies, if needed
6. Working on processes below Cpk of 1.33
7. Working on improving other processes with Cpk of 1.33 (AIAG,2001)1
VILLANOVA UNIVERSITY x - 21 LEAN SIX SIGMA HANDBOOK

FMECAlFMEA
Design FMEA (FMECA) Process Steps

The following steps are taken to prepare a design FMECA. An example of a
completed FMEA is shown in Table 10.8.
1. FMEA number: This should be a log controlled assigned number

2. The part number, name, or other appropriate description
3. The design responsibility: Which department or group is responsible?
4. The person responsible for FMEA preparation
5. The date the FMEA was prepared and any necessary revision level
6. The subsystem or component part number getting detailed analysis
7. The component function
8. The potential failure mode
9. The potential effect of failure
10. The potential cause of failure
11. What are the current controls in place to prevent the cause from occurring?
The next major step is to weigh the risks associated with the current component,
effect, and cause with the controls that are currently in place.
12. P is the probability this failure mode will occur. Values for this index
generally range from 1 to 10, with 10 being near certainty !o f occurrence.
13. S is the severity of the effect of the failure on the rest of: the system if the
failure occurs. These values are often indexed from 1 to 10, with a 10 meaning
that the safety of the user is in jeopardy.
14. D is a measure ofthe effectiveness ofthe current controls (in place) to identify
the potential weakness or failure prior to its release to production. This index
may also range from 1 to 10. A value of 10 indicates the design weakness
would most certainly make it to final production without detection.
15. RPN. The risk priority number is the product of the indices from the previous
three columns. This RPN is dimensionless since there is no real meaning to
a value of say 600 versus 450 except in the difference in lTIagnitude.
RPN =P·S·D
16. The actions then are based upon what items either have, the highest RPN
and/or where the major safety issues are.
17. There is a column for actions to reduce the risk, a column for responsibility,
and a column for the revised RPN after corrective action.
VILLANOVA UNIVERSITY X -22 LE.t~N SIX SIGMA HANDBOOK

FMECAlFMEA
Design FMECA Example

Part No.lName: 37XT11-Lock Mech. P = Probability FMEANo.43
Project: Re-design S = Seriousness Final Design Deadline: July 1, 2007
Other Departments : Shop Service, etc. o = Likelihood Prepared By: RCD
Subsystem Name: Quill Clamping Mechanisms RPN = Risk Priority Reviewed By: BLW
Suppliers Involved: Wilton and others Number FMEA Date: 12-20-06 Rev.
Design Responsibility: Bob Dovich
PART FUNCTION POTENTIAL POTENTIAl. CURRENT RISK RECO"MMENDED A9TlON(S) REVISED ~ISK RESPONSIBU
NUMBER FAlL.URE ~FFECT(S) POTENTIAL CONTROLS ASSESSMENT CORRECTIVE TAKa! ASSESSMENT DEPT OR
NAME MODE(S) FFAILURE CAUSE(S) OF ACTION(S) INDIVIDUAL
FAILURE
'p S D RPN P S D RPN
WILTON CLAMP LEAK HOUSE- ACCEPT DISCUSS WITH

POWER KEEPING WEAR SUPPLIER'S 243 24 SUPPLIER
LOCK INFO
LOSES MACHINING INADEQUATE PERFORM LOAD

ENG.
CLAMPING PARTS SIZE POWER 244 32 TESTS
STANDARD
FORCE OVERSIZE LOCK
POOR WORK & NONE

Q. C.STD. 142 8
MATERIALS
REVIEW NEED
OVER
NONE 242 16 FOR PREVENTION
PRESSURE
SYSTEM
REVIEW PRESSURE
PUMP ENG. 1 4 2 8 PELIVERED IN FIELD
SIZING STANDARD
AND ACTUAL NEED
Table 10.8 An Illustrative FMECA
Risk Assessment
Risk assessment is the combination of the probability of an event or failure, and the
consequence(s) of that event. The analysis of risk of failure normally considers the
severity and probability of failure. The severity of failure is generally defined by the
categories from MIL-STD-1629 (1980)22. These are shown in Table 10.9.
Classification Description
I Catastrophic A failure that may cause death or mission loss
II Critical A failure that may cause severe injury or major system damage
III Marginal A failure that may cause minor injury or degradation in mission
performance
IV Minor A failure that does not cause injury or system damage but may
result in system failure and unscheduled maintenance
Table 10.9 Hazard Severity Categories

FMECAlFMEA
Risk Assessment (Continued)

The categories in Table 10.9 were built with military missions in mind. Commercial
companies should modify the definitions. An example, from Ireson (1996)21 for
commercial applications uses a severity index based on a scale from 1 to 10.
Rank Criteria
1 It is unreasonable to expect that the minor nature of this failure will
degrade the performance of t he system.
2-3 Minor nature of failure will cause slight annoyance to the customer.
Customer may notice a slight deterioration of the system performance.
4-6 Moderate failure will cause customer dissatisfaction. Customer will
notice some system performance deterioration.
7-8 High degree of customer dissatisfaction and inoperation of the system.
Does not involve safety or noncompliance to government regulations.
9 -10 Very high severity ranking in terms of safety-related failures and
nonconformance to regulations and standards.
Table 10.10 Commercial Severity Index (Scale 1 - 10)
The hazard classification or severity index is generated for each component or

subsystem by the reliability analyst. This classification is based on the expected
results of the failure of the component or subsystem. The probability of failure may
also be ranked. A common ranking of failure probabilities is shown in Table 10.11.
Faillire Probability Level Description Probability

A High likelihood of occurrence >10-1
B Probable occurrence 10-1 to 10-2
C Occasionally occurs 10-2 to 10-3
D Remote probability 10-3 to 10-6
E Highly unlikely <10-6
Table 10.11 A Common Failure Probability Ranking
A number of systems are used to combine the probability of failure and the hazard
category. These systems are based on accepting a degree of risk (~f occurrence with
respect to the severity of the hazard.

DFX
Design for X (DFX)

Design for X (DFX) is defined as a knowledge-based approach for designing
products to have as many desirable characteristics as possible. The desirable
characteristics include: quality, reliability, serviceability, safety, user friendliness,
etc. This approach goes beyond the traditional quality aspects of function, features,
and appearance of the item.
AT&T Bell Laboratories coined the term DFX to describe the process of designing
a product to meet the above characteristics. In doing so, the life cycle cost of a
product and the lowering of down-stream manufacturing costs would be addressed.
The DFX toolbox has continued to grow in number from its inception 15 years ago
to include hundreds of tools today (Huang, 1997)20. The user can be overwhelmed
by the choices available. Some researchers in DFX technology have developed
sophisticated models and algorithms. The usual practice is to apply one DFX tool
at a time. Multiple applications of DFX tools can be costly. The authors note that a
systematic framework is not yet available for use with DFX methodology. A set
methodology would aid in the following ways:
• Understanding how DFX works
• Aiding in the selection of a tool
• Faster learning of DFX tools
• Providing a platform for multiple DFX tools
Usage of DFX Techniques and Tools

The following material is based principally on the work of Watson (1998)39.
1. Design guidelines:
DFX methods are usually presented as rules ofthumb (design guidelines). These
guidelines provide broad design rules and strategies. The design rule to increase
assembly efficiency requires a reduction in the part count and part types. The
strategy would be to verify that each part is needed.

DFX
Usage of DFX Techniques and Tools (Continued)

2. DFX analysis tools:
Each DFX tool involves some analytical procedure that measures the
effectiveness ofthe selected tool. Boothroyd (1994f provides; a DFA (design for
assembly) procedure that addresses the handling time, insertion time, total
assembly time, number of parts, and the assembly efficiency. Each tool should
have some method of verifying its effectiveness.
3. Determine DFX tool structure:
A technique may require other calculations before the technique can be

considered complete. An independent tool will not dependl on the output of
another tool. The handling analysis, insertion analysis, and nlUmber of parts are
all capable of being calculated, but the total assembly time requires subsystem
times for each component.
4. Tool effectiveness and context:
Each tool can be evaluated for usefulness by the user. The tooll may be evaluated
based on accuracy of analysis, reliability characteristics, and/or integrity of the
information generated.
5. The focus of activity and the product development process:
Use of the DFX tools will be of benefit if the product development process is
understood by the design team. Understanding the process ,activities will help
determine when a particular tool can be used.
6. Mapping tool focus by level:
The mapping of a tool by level implies that DFX analysis can be complex. Several
levels of analysis may be involved with one individual tool. lrhe structure may
dictate the feasibility of tool use. For routine product redesigns, the amount of
information needed may already be available. For original de~signs, the amount
of interdependence of tools can make it difficult to coordinate all of the changes
downstream.

DFX
DFX Characteristics
The following characteristics and attributes should be considered by DFX projects.
(Bralla, 1999)8
Function and performance:
These factors are vital for the product.
Safety:
Design for safety requires the elimination of potential failure prone elements that
could occur in the operation and use ofthe product. The design should make the
product safe for: manufacture, sale, use by the consumer, and disposal.
Quality:
The three characteristics of quality, reliability, and durability are required and are
often grouped together in this category.
Reliability:
A reliable design has already anticipated all that can go wrong with the product,
using the laws of probability to predict product failure. Techniques are employed
to reduce failure rates in design testing. FMEA techniques consider how
alternative designs can fail. Derating of parts is considered. Redundancy
through parallel, critical component systems may be used.
Testability:
The performance attributes must be easily measured.
Ma n ufactu rability:
The concept of design for manufacturability (DFM), includes testability, and

shipability. Producibility and manufacturability are terms used since the 1960s.
Design for manufacturability (DFM) has been the dominant term used since 1985.
A design must simplify the manufacture of a product through a reduced number
of parts, and a reduced number of manufacturing operations. (Gryna, 1999)15

DFX
DFX Characteristics (Continued)

Assembly (Design for Assembly, DFA):
DFA means simplifying the product so that fewer parts are in"olved, making the
product easier to assemble. This portion of DFX can often provide the most
significant benefit. A product designed for ease of assembly can: reduce
service, improve recycling, reduce repair times, and ensure faster time to market.
This is accomplished by using fewer parts, reducing engineering documents,
lowering inventory levels, reducing inspections, minimizing setups, minimizing
material handling, etc.
Environment:
The objective is minimal pollution during manufacture, use, cmd disposal. This
could be defined as design for the environment (DFE). The concept is to increase
growth without increasing the amount of consumable resources. Some
categories of environmental design practices include: rec:overy and reuse,
disassembly, waste minimization, energy conservation, maiE!rial conservation,
chronic risk reduction, and accident prevention.
Serviceability (Maintainability and Repairability):
A product should be returned to operation and use easily afte!r a failure. This is
sometimes directly linked to maintainability.
Maintainability:
The product must perform satisfactorily throughout its intended life with minimal
expenses. The best approach is to ensure the reliability of components. There
should be: reduced down time for maintenance activities; reduced user and
technician time for maintenance tasks; reduced requirements for parts; and
lower costs of maintenance. Endres (1999)14 provides some specific methods for
increasing maintainability (decreasing diagnosis and repair times): use modular
construction in systems, use throwaway parts (instead of parts requiring repair),
use built-in testing, have parts operate in a constant failure rate mode, etc.

DFX
DFX Characteristics (Continued)

User Friendliness, or Ergonomics:
Human factors engineering must fit the product to the human user. Some
guidelines to consider are: fitting the product to the user's attributes, simplifying
the user's tasks, making controls and functions obvious, anticipating human
error, providing constraints to prevent incorrect use, properly positioning
locating surfaces, improving component accessibility, and identify components.
Appearance (Aesthetics):
Attractiveness is especially necessary for consumer products. These

characteristics include: special requirements of the user, relevancy of the style,
compatibility of materials and form, proportional shapes, or protection from
damage in service.
Packaging:
The best package for the product must be considered. The size and physical
characteristics of the product are important, as is the economics of the package
use. The method of packaging must be determined. Automated packaging
methods are desirable.
Features:
Features are the accessories, options, and attachments available for a product.
Time to Market:
The ability to have shorter cycle times in the launch design of a product is
desirable. The ability to make the product either on time, or faster than the
competition is of tremendous advantage.

TRIZ
TRIZ
TRIZ is a Russian abbreviation for "the theory of inventive problem solving."
(Altshuller, 1996)2, (Altshuller, 1998)3. TRIZ in Russian is TeoplUI pememHI
H306peTaTeJIbCKHX 3a;a;a-q (TPH3). (The School of Triz, 2ooofl6. L. Shulyak is the
translator of a series of books originally written by Genrich Altshuiller, on the subject
of inventive problem solving. Provost (1998)27 also provides a description of the
TRIZ methodology in Quality Progress. Altshuller states that inventiveness can be
taught. Creativity can be learned, it is not innate. One does not h.:lve to be born with
creativity.
Altshuller asserts that traditional inventing is "trial and error" resulting in much
wasted time, effort and resources. Through his years ()f education and
imprisonment, he solidified a theory that one solves problems through a collection
of assembled techniques. Technical evolution and invention have certain patterns.
One should be knowledgeable with them to solve technical problems. There is some
common sense, logic and use of physics in problem solving. (Altshuller, 1996)2
There are three groups of methods to solve technical problems:
1. Various tricks (a reference to a technique)
2. Methods based on utilizing physical effects and phenomlEma (changing the

state of the physical properties of substances)
3. Complex methods (combination of tricks and physics)
Altshuller (1998)3 provides an introduction to ARIZ (algorithm to solve an inventive

problem). This is a sequence of 9 action steps in the use of TRIZ. The steps are:
1. Analysis of the problem
2. Analysis of the problem's model: Use of a block dia~lram defining the

"operating zone"
3. Formulation of the ideal final result (IFR): Providing a desc:ription of the final
result, which will provide more details
4. Utilization of outside substances and field resources
5. Utilization of an informational data bank: Determinin!;;1 the physical or

chemical constraints (standards) on the problem
VILLANOVA UNIVERSITY x- 30 LElIN SIX SIGMA HANDBOOK

TRIZ
TRIZ (Continued)
6. Change or reformulate the problem
7. Analysis of the method that removed the physical contradiction: Is a quality

solution provided?
8. Utilization of the found solution: Seeking side effects of the solution on the
system or other processes.
9. Analysis of the steps that lead to the solution: An analysis may prove useful
later.
Initially there were 27 TRIZ tools (Altshuller, 1996)2, (Provost, 1998)27, which were
later expanded to 40 innovative, technical tools (Altshuller, 1998)3. This short
section on TRIZ can not fully describe the methodology that has proven to be very
successful. Users of TRIZ report great results.
The 40 TRIZ tools are:
• Segmentation • Partial or excessive action

• Extraction • Transition into a new dimension
• Local quality • Mechanical vibration
• Asymmetry • Periodic action
• Consolidation • Continuity of useful action
• Universality • Rushing through
• Nesting • Convert harm into benefit
• Counterweight • Replacement of mechanical systems
• Prior counteraction • Pneumatic or hydraulic construction
• Prior action • Flexible membranes or thin films
• Cushion in advance • Porous material
• Equipotentiality • Changing the color
• Do it in reverse • Homogeneity
• Feedback • Rejecting or regenerating parts
• Mediator • Transformation of properties
• Self-service • Phase transition
• Copying • Thermal expansion
• Dispose • Accelerated oxidation
• Spheroidality • Inert environment
• Dynamicity • Composite materials

SYSTEMATIC DESIGN
Systematic Design
Systematic design is a step-by-step approach to design. It provides a structure to
the design process using a German methodology. It is stated that: systematic design
is a very rational approach and will produce valid solutions. This approach is similar
to the German design standard: Guideline VOl 2221 "Systematic Approach to the
Design of Technical Systems and Products."
Pahl (1988)24 presents four main phases in the design process:
• Task clarification: collect information, formulate concepts, identify needs

• Conceptual design: identify essential problems and sub-functions
• Embodiment design: develop concepts, layouts, refineme!nts
• Detail design: finalize drawings, concepts and generate documentation
An abstract concept is developed into a concrete item, represented by a drawing.

Synthesis involves search and discovery, and the act of combining parts or elements
to produce a new form.
Modern German design thinking uses the following structure:
• The requirements of the design are determined

• The appropriate process elements are selected
• A step-by-step method transforms qualitative items to quantitative items
• A deliberate combination of elements of differing comple)~ities is used
The main steps in the conceptual phase:
o Clarify the task

• Identify essential problems
• Establish function structures
• Search for solutions using intuition and brainstorming
• Combine solution principles and select qualitatively
• Firm up concept variants: preliminary calculations and la~'outs
• Evaluate concept variants
Systematic design includes tools and methods suggested for valrious steps along
the design process. The creativity of the designer is encouraged in this method, but
on a more structured basis. Any and all design methods must employ the designer's
creativity to find new innovative solutions. . (Pahl, 1988)24
VILLANOVA UNIVERSITY x - 32 LEJ\N SIX SIGMA HANDBOOK

CREATIVE DESIGN IDEAS
Creative Design Ideas

A.F. Osborn is credited with creating the concept of brainstorming in the 1940s. The
following list is adapted from Osborn (1963)23. These are creative ways to examine
and improve a part or process.
Elimination: What will happen if something is eliminated?
Reversal: What will happen if something is reversed?
Normal or abnormal: Is something an abnormality or does it occur all the

time?
Constant or variable: What will happen if only the changing item is treated
as an exception?
Enlargement or What will happen if something is enlarged or reduced?

reduction:
Linking or separating: What will happen if some things are joined or taken
apart?
Concentration or What will happen if something is concentrated or

dispersion: dispersed?
Addition or removal: What will happen if something is added or removed?
Replacement or Can a replacement or substitution be made? Can

sUbstitution: waiting time be used for another purpose?
Change the sequence: What will happen if a different procedure is used?
Parallel or serial: Can two or more things be done at the same time?
Can they be done sequentially?

Creative Design Ideas (Continued)
Similarities: Is there a method to group similar t:hings?
Use it another way: Is there another way to use it while keeping the
current setup? Can anything else be produced?
Borrow an idea Form ideas by benchmarking or t::ombination. Can

elsewhere: ideas used elsewhere be adapted?
Change or replace it: Change the shape, color, sound, mc)vement, location,
orientation, power source, or rotation.
Expand it: Add something; spend more time; increase the

repetition; make it stronger, longer, or thicker; add
some other value; double it, duplicate it; increase it;
exaggerate it.
Reduce it: Remove something; make it smaller or stronger;

divide it; simplify it; reduce it; lightcm it; express it in
a more subtle way.
Use alternatives: Use someone or something else. Us;e other elements,

ingredients, materials, meth()ds, locations,
approaches, or voice tone.
Replace it: Use different elements, ingredients, dies, layouts,

sequences, or arrangements. Clhange the pace,
speed, or schedule.
Reverse it: Turn it upside down; invert it; revelrse the positions,
front and back. Change the roles: r orientations, or
setups.
Recombine it: Mix or assemble it differently.

Process Design Example

A large manufacturer in the entertainment industry had an existing process cycle
time that averaged 10 -14 days. An extremely valuable master was shipped from a
customer or a customer's outside provider. Because of the "risks" involved, a
series of management paperwork sign-offs had been established between receiving,
planning, the vault, and production. At least two management supervisors were
required to participate at each transfer point.
What occurred in practice was that the master (which was subject to damage) sat
around for many days in a safe (or in various other locations because the system
didn't make sense) waiting for the appropriate authority figures to come back from
vacation or attend to paperwork. Refer to Figure 10.12.
The process problem was attacked by a trained team who began by flow charting the
current process. They asked "Why do we do it this way" questions at each key step.
As it turned out, this expensive master had a very low likelihood of theft. Other
producers, who could even use the master, were few indeed and were known to have
integrity. Hourly employees at the firm also considered themselves to have as much
honesty and integrity as management. To them, the established system was a joke.
The cycle time was reduced to 2 days (the majority of which was a scheduling
decision). See Figure 10.13.
This example makes an interesting case study illustrating how teamwork and the
development of a process flow chart can be used to simplify an existing process.
However, the question arises: Why is this example included in design improvement?
The logical explanation is that many processes happen because someone in

management developed it in a vacuum or with limited input from others. The same
effort should be directed at creating an administrative process as correcting one
later. All service organizations (hospitals, banks, non-profit agencies, the military,
educational sites, government agencies, etc.) have a multitude of processes. All
manufacturing companies have similar information flows. The point is that
considerable effort should be directed at the simplified design of these processes.

Process Design Example (Continued)

MASTER
ARRIVES
C MASTER
ARRIVES
PUT IN DOC~~
"
SUPERVISOR
DOCK OR REPLACEMENT
SAFE RECEIVES MASTER
YES
DELIVER MASTER
TO PLANNING
+
ADVISE PLANNING
.
DELIVER MASTER
TO VAULT
OF AVAILABILlTf
AUTHORITY
H
NO--.-t
PUT MASTER
IN VAULT
YES
DELIVER MASTER
TO VAULT
."
[RODUCTION IS
SCHEDULED
>---NO
H
( PRODUCT IS )
PRODUCED
YES
Figure 10.13 The New System

PUT ADVISE
MASTER PLANNING
IN VAULT OF AVAILIABLE
PRODUCTION
IS SCHEDULED
Figure 10.12 The Old System

REFERENCES
References
1. AIAG. (2001). Potential Failure Modes and Effects Analysis (FMEA) Reference
Manual, 3rd ed. Southfield, MI: AIAG.
2. Altshuller, G. (1996). And Suddenly the Inventor Appeared: TRIZ, the Theory of
Inventive Problem Solving. (L. Shulyak, Trans.). Worcester, MA: Technical
Innovation Center.
3. Altshuller, G. (1998). 40 Principles: TRIZ Keys to Technical Innovation. (L.

Shulyak, Trans.). Worcester, MA: Technical Innovation Center.
4. Ashley, S. (July, 1992). "Applying Taguchi's Quality Engineering to Technology

Development." Mechanical Engineering.
5. ASQ Staff Writer. (2003, Nov. 26). "Planning for DFSS." Downloaded Jan. 4, 2006
from: http://www.asq.org/forums/sixsigma/articles/champion/pdf/
chp_planning_for_DFSS.pdf
6. Bongiorno, J. (2000, October). "Improving FMEAs: FMEAs can Transform

Compliance into Competitive Advantage." (Electronic version). Quality Digest,
37-40.
7. Boothroyd, G.; Dewhurst, P.; & Knight, W. (1994). Product Design for
Manufacture and Assembly. New York: Marcel Dekker.
8. Bralla, J. (1999). Design for Manufacturability Handbook. 2nd ed., New York:
McGraw-Hili.
9. Cooper, R. (1993). Winning at New Products, 2nd ed. Reading, MA: Perseus
Books.
10. Cooper, R. (1996). "New Products: What Separates the Winners from the
Losers." In Rosenau, et. al., The PDMA Handbook of New Product
Development. New York: John Wiley & Sons.
11. Crawford, C. (1997). New Products Management, 5th ed. Chicago: Irwin.
12. Cross, N. (1994). Engineering Design Methods: Strategies for Product Designs,
2nd ed., Chichester: John Wiley & Sons.
13. Dovich, R.A., & Wortman, B.L. (2002). eRE Primer. Terre Haute, IN: Quality
Council of Indiana.

REFERENCES
14. Endres, A. (1999). "Quality in Research and Development." In Juran's Quality
Handbook, 5th ed., Juran, J., et al., New York: McGraw-Hili.
15. Gryna, F. (1999). "Operations." In Juran's Quality Handboolc~ 5th ed., Juran, J.,
et al., New York: McGraw-Hili.
16. Hamel, G., & Prahalad, C. (1994). Competing for the FuturE!. Boston: Harvard
Business School Press.
17. Harold, D. (1999). "Designing for Six Sigma Capability." Retrieved

September 15, 2001 from:
http://www.controleng.com/archieves/1999/ctI0101.99/01a103.htm
18. Hauser, J.R., & Clausing, D. (1988). "The House of Quality." Harvard Business
Review, May-June, p. 63-73.
19. Hockman, K. (2001). "Why is a Design for Six Sigma Methodc,logy Necessary?"
Retrieved September 10, 2001, from:
http://www.sixsigmaforum.com/protected/articles/ds_dfss•. shtml
20. Huang, G., & Mak, K. (1997, Sept). "Developing a Generic DI~sign for X Shell."
Journal of Engineering Design.
21. Ireson, W.G.; Coombs, C.F.; & Moss, R., eds. (1996). Handbook of Reliability
Engineering and Management. New York: McGraw-Hili.
22. MIL-STD-1629. Procedures for Performing a Failure Mode, Effects and Criticality
Analysis.
23. Osborn, A. F. (1963). Applied Imagination: Principles and Procedures of Creative

Thinking. New York: C. Scribner's Sons.
24. Pahl, G., & Beitz, W. (1988). Engineering Design: A Systlematic Approach.
London: The Design Council, and Berlin: Spring-Verlag.
25. Phadke, M.S. (1989). Quality Engineering Using Robust Dc~sign. Englewood
Cliffs. NJ: Prentice Hall.
26. Phadnis, S. (2001). "Design for Six Sigma Roadmap." Refer to

http://www.isixsigma.comllibrary/content/c01091 Oc.asp
VILLANOVA UNIVERSITY X- 38 LEJ~N SIX SIGMA HANDBOOK

REFERENCES
27. Provost, L. & Langley, G. (1998, March). "The Importance of Concepts in
Creativity into Quality Management." Quality Progress. Milwaukee: ASQ.
28. Reid, R. (May, 2005). "FMEA- Something Old, Something New." Quality
Progress. 38(5), 90-93.
29. Reiling, J.; Knutzen, B.; & Stoecklein, M. (August 2003). "FMEA - The Cure For
Medical Errors." [Electronic version]. Quality Progress, 36(8), p. 67-71.
30. Rosenau, M. (1996). "Choosing a Development Process That's Right for Your
Company." In Rosenau, et. aI., The PDMA Handbook of New Product
Development. New York: John Wiley & Sons.
31. Shina, S. (2001). Six Sigma for Electronics Design and Manufacturing. New York:
McGraw-Hili.
32. Simon, K. (December, 2000). "DMAIC Versus DMADV." Retrieved August 17,
2001 from:
http://www.isixsigma.comllibrary/content/c001211a.asp
33. Stamatis, D. (1995). Failure Mode and Effect Analysis: Failure From Theory to
Execution. Milwaukee: ASQC Quality Press.
34. Taguchi, G. (1987). System of Experimental Design, 2 Volumes, Dearborn MI:

Unipub/Kraus/American Supplier Institute.
35. Taguchi, G. (1993). Taguchi on Robust Technology Development. New York:

ASME Press.
36. The School of TRIZ.. Retrieved on October 17, 2000 from:

http://www.triz.minsk.by/
37. Trefts, D.; Rabeneck, S.; & Rabeneck, L. (2001). "An Approach to Achieving Six
Sigma Design and Optimize (Improve) Phase:" Retrieved September 10,2001,
from:
http://www.sixsigmaforum.com/protected/articles/ds_approach1.shtml

REFERENCES
38. Vandervort, C., & Kudlacik, E. (2001). "GE Generator Technology Update."
Retrieved January 8,2006 from:
http://mjharden.com/prod_serv/products/tech_docs/en/downloads/ger4203.pdf
39. Watson, B., & Radcliffe, D. (1998, Sept). "Structuring Design for X Tool Use for
Improved Utilization." Journal of Engineering Design.
40. Wilkins, Jr., J. (May, 2000). "Putting Taguchi Methods to Work to Solve Design
Flaws." Quality Progress.

x. IMPROVEMENT TECHNIQUES
QUESTIONS
10.1. FMEAs satisfy which of the following basic 10.4. The most effective and efficient method of
organizational needs? solving quality problems for a product is to
concentrate efforts in which of the following
a. FMEAs are of no value since they are very areas?
tedious
b. Auditors will not hand out a finding if FMEAs a. Design
are in compliance b. Production
c. The bottom line is automatically improved c. Quality improvement
d. The risk of quality defects is greatly reduced d. Systematic troubleshooting
10.2. An improvement team has completed several 10.5. Which of the following organizations is ready
RPN values and is now ready to institute for a DFSS approach?
improvements. There are three failure modes
with the same RPN score in the table below. a. The organization has recently become ISO
The first line item which should be reduced 9001 certified
would be: b. The first full-time black belt was recently
hired and oriented within the organization
Failure Mode Severity Occur. Detection RPN c. After many successful LSS projects the rate
of improvement has slowed
Bad battery 9 7 5 315 d. The new CEO has a strong engineering
Cracked welds 7 9 5 315 background and wants change
Missing bolts 5 7 9 315 10.6. The design steps in Taguchi's robust design
sequence are:
a. The bad battery because it has the highest
severity ranking I. Concept design
b. The cracked welds because of the high II. Parameter design
occurrence III. Tolerance design
c. The missing bolts because it has the highest
detection level a. I, II, III
d. The rankings are the same, so choose any b. I, III, II
one to start c. II, I, III
d. III, I, II
10.3. What is the main difference between DMAIC and
DFSS? 10.7. Which of the following is NOT a widely
recognized topic area for DFX?
a. DFSS does not have a control phase
b. DMAIC attacks existing problems while a. Design for profit
DFSS provides new product and process b. Design for assembly
design c. Design for reliability
c. DFSS works only for engineers and DMAIC d. Design for appearance
can be used by everybody in an organization
d. DMAIC is a methodology while DFSS is a 10.8. Which of the following statements is an
tool INCORRECT description of QFD?
a. It translates fuzzy customer requirements

into design specifications
b. It is an iterative process, similar to FMEA
c. It is similar to project management
d. It identifies risk areas

X. IMPROVEMENT TECHNIQUES
QUESTIONS
10.9. In the design of a product, design for X (DFX) 10.13.ln robust design. a factor that can cause
can be described as: unknown variability, or an error in the response
factor, is considered a:
a. Design for anything
b. A very effective method for manufacturability a. Signal factor
of parts b. Control factor
c. A new technique encompassing six sigma c. Noise factor
d. A knowledge-based design approach d. Response factor
10.10. Arrange the following design steps into a logical 10.14. The term "severity" in ia FMEA describes the:
sequence from start to finish.
a. Difficulty of completing the FMEA form
I. Measure and determine customer needs and b. Possible impact to a system user of a low-
specifications level failure
II. Define the project goals and customer needs c. Likelihood of a failure
specifications d. Time for which the system is expected to be
III. Analyze the process options down
IV. Verify and validate the design
V. Develop the design details for producing the 10.15. When faced with a complex problem which
customer needs requires an inventive solution, the method
which produces the results with the least
a. I, II, III, IV, V wasted time, effort. and resources is:
b. II, I, III, V, IV
c. II, I, IV, III, V a. Trial and error
d. I, II, V, III, IV b. Innate inventitivene:ss
c. Using ARIZ steps in the TRIZ method
10.11.ln Taguchi methods, a factor that results In a d. Plan-do-check-act
response variable and can be equated to an
Independent variable is called a: 10.16. The analysis of risk involves two measures of
failure. These meaSUrE!S are:
a. Control factor
b. Signal factor a. Failure analysis and failure effects
c. Noise factor b. Failure mode and failure method
d. Dependent variable c. Failure severity and failure probability
d. Failure mechanism ::md failure mode
10.12. A number of authors have recommended
sequences by which the HOQ (QFD) can capture 10.17. Cooper stresses that new products will have a
customer needs in the design. Please arrange greater chance of succoss If they have all of the
the following design details in appropriate following characteristic;s EXCEPT:
sequence from start to finish.
A. An attractive market
I. Production requirements B. A unique and superior product
II. Key process operations C. Being first to market
III. Parts characteristics D. A good product laurlch
IV. Engineering characteristics
a. I, II, III, IV
b. II, I, IV. III
c. IV, II. III, I
d. IV, III, II, I
VILLANOVA UNIVERSITY X -42 LEjI,N SIX SIGMA HANDBOOK

QUESTIONS
10.18. Identify the design acronym(s) that would be 10.23. TRIZ is a methodology for problem solving and
considered (a) subset(s) of DFX: Is quite useful in the design phase of a product.
Which of the following methods are employed in
I. DFSS TRIZ?
II. DFA
III.DFM I. Trial and error
II. Reference to a trick
a. lonly III. Use of physical effects (physics)
b. II and III only IV. Combination oftrlcks and physics
c. I and III only
d. I, II, and III a. I and II only
b. II and III only
10.19.ln robust design technology, which of the c. I, II, and III only
following factors would NOT be considered a d. II, III, and IV only
noise factor?
10.24. The stage gate process is used by many
a. Ambient temperature companies to screen and pass projects. Many
b. Carbon content in cast iron companies may fail to use the process properly.
c. Weather conditions A common problem in the stage gate process
d. Sun spots that could negate its benefits would be:
10.20. The product development process should a. Failure to kill a project

consist of: b. Too many kills in the system
c. Teams are overloaded with projects
a. An individual product developer d. Requirements are not established
b. A relay team of function-to-function
specialists 10.25. DFX has many Internal and external customer-
c. Internal quality teams based objectives. Which of the following would
d. Multi-functional teams be considered an external target?
10.21. DFM (or design for manufacturability) attempts a. Design for assembly
to accomplish all of the following objectives, b. Design for features
EXCEPT: c. Design for testability
d. Design for manufacture
a. Testability
b. Environmental control 10.26. New products will account for a large
c. Producibillty percentage of company sales. The typical ratio
d. Reduced number of manufacturing of new product ideas to successful products is
operations in the range of:
10.22. Robust design means that the designer has a. 1:1

attempted to: b. 7:1
c. 20:1
a. Build the design with high durability d. 50:1
components
b. Generated designs using computer models 10.27. Robust design methods can be used In product
c. Minimized random noise effects design. The various factors to consider in
d. Used high strength materials robust design include all the following,
EXCEPT:
a. Tolerance parameter factors

b. Noise factors
c. Signal factor
d. Control factors

QUESTIONS
10.28. Review the following DFX statements and 10.31. A new product such as ;:1 Pepsi with lemon twist
identify the one true description: would be considered:
a. DFSS is a subset of DFX a. New to the world

b. The selection of DFX tools is relatively b. An addition to the product line
simple c. A repositioned prod uct
c. DFX is a targeted development approach d. Product improvement
d. DFX was first created in the 1990s
10.32. Arrange the following basic design process
10.29. Quality function deployment, also known as the steps into an appropriate sequence from start to
house of quality, is used in the design process. finish.
It has various benefits, the most critical being:
I. Establish the desigl1l objectives
a. People are aligned and think together II. Analyze the possiblo solutions for best fit
b. Designers have access to the right technical III. Generate ideas to create solutions
specs IV. Implement the selected design
c. Customers put their wants down on paper
d. Competitive analysis is developed a. I, II, III, IV
objectively b. I, III, II, IV
c. III, I, II, IV
10.30. The initial construction of several matrices to d. II, I, III, IV
display a QFD involves the building of a "house
of quality." The side walls normally depict:
I. Customer needs
II. Customer competitive assessment
III. Design features
IV. Technical competitive assessment
A. I and II only
B. II and III only
C. III and IV only
D. I, II, III, and IV

XI. CASE STUDIES
I HEAR, I FORGET.
I SEE, I REMEMBER.
I DO, I UNDERSTAND.
CHINESE PROVERB
A number of case studies are presented in this Section:
• QCI Spaghetti Studies
• Software License Improvement at Exactus
• LSS Implementation at Ludovico
• Insurance Claim Flow Improvement
• Medical Center Improvement using A3 Problem Solving
VILLANOVA UNIVERSITY XI-1 LEAN SIX SIGMA HANDBOOK

XI. CASE STUDIES
QCI SPAGHETTI STUDIES
QCI Spaghetti Studies

Quality Council of Indiana makes technical training manuals, CDs, and solution
texts, principally for quality professionals. The storage, production, and shipping
activities at QCI are restricted to a 40' x 40' floor space. There are some legacy
structural issues (five functional doorways, a staircase, and a support beam in the
center that helps hold the ceiling up). These problems could be overcome, in the
name of efficiency, for a price. Figure 11.1 is a diagram of the layout and material
flow in 1996.
TABS
BINDER STORAGE
CLOSET I
SHIPPING
DOOR
STORAGE I
. - - - - - - , INACTIVE
DOOR
ACTIVE
DOOR
MEDIA
CENTER
Figure 11.1 1996 QCI Copy Area Material Flows
Three employees A, B, and C worked in the general locations as shown.
Reflected in Figure 11.1 is the normal production of training binde!rs, solution texts,
and question disks (at the time). Also included are items like replenishing paper,
retrieving tabs, and moving packaging materials.
VILLANOVA UNIVERSITY XI- 2 LE~~N SIX SIGMA HANDBOOK

XI. CASE STUDIES
QCI Spaghetti Studies (Continued)

The situation in Figure 11.1 depicts some obvious problem areas:
• Paper and tab storage are poorly located

• The media center is outside the copy room
• There are a lot of copy machines
• Shipping is done manually
• Employee A travels long distances
Not apparent in Figure 11.1 were the following issues:
• Paper copy masters were necessary

• Copy quality was marginal
• Employees Band C worked at a frantic pace
• 3 of 6 copy machines were inoperative at any given time
• Copy errors were difficult to catch
• When volumes increased, the problems got worse
Since copy machines (or printers) are normally leased or purchased on a 5 year
basis, there were some limitations on rapid movement in the equipment area without
serious financial consequences.
Over an 8 to 10 year period, a number of improvement teams were formed.

Membership consisted of production and shipping employees as well as members
of management. The average team membership consisted of five people.
Improvement tools and techniques included the following:
• Brainstorming sessions
• Benchmarking similar operations
• Investigations and trials of new equipment
• The use of spaghetti flow diagrams
• Simulated layouts (both computer generated and graphical cut-outs)
• Prototype equipment trials
• The installation of simple visual control systems
After the above activities were concluded, only two operators were necessary. The
resulting spaghetti chart for material flows is shown in Figure 11.2.

XI. CASE STUDIES
QCI Spaghetti Studies (Continued)

Two operators now function from starting points A and B. The o'perators fill either
a shipping (A) or production (B) role. For variety and currency IPurposes, the two
roles are rotated on a daily basis. To increase copy capacity, tine work shifts are
slightly staggered, providing 9.5 hours of daily coverage, Monday through Friday.
- ACTIVE
~ ]
...J
~ W
~
~
'" i2
'"
SERVER
~
0:
0
8
SHIPPING
DOOR
PRINTER S
1
PAPER
STORAGE
I NiICn VE
, - -- - - - , ~
ACTIVE ACTIVE
DOOR DOOR
Figure 11.2 Current QCI Copy Area Material Flows
So with two employees now working in the area, what happl~ned to the third
employee? What happens when one of the two employees is sick or on vacation for
a week?
Normal attrition was factored into the short-term manpower equation. However, one
employee was soon moved into a technical writing position, so the over-all company
head count remained the same. Fortunately, QCI busy periods are not in desirable
vacation periods, and in most situations, one operator is sufficient for a week.
Furthermore, vacations are restricted to permit only one planned cllbsence per week.
VILLANOVA UNIVERSITY XI- 4 LEP.N SIX SIGMA HANDBOOK

XI. CASE STUDIES
act Spaghetti Studies {Continued}

Between the two positions, A and B, the shipping responsibilities of position A takes
precedent because of the small buffer stock. It should be noted that a considerable
amount of product can be produced at position B without operator intervention or
quality issues. If all else fails, other full-time employees have been trained in the
area such that a triaging decision can be made quickly.
There are standard QCI procedures for making manuals, comb-bound books, and
CDs. However, specific customer orders drive the entire process and they vary
considerably. In an 80 customer order day, at least 75 orders will be different.
Specific material flow sequences under these conditions do not follow the same
step-by-step routines. Think about how order fulfillment works at a fast food
restaurant. There are detailed instructions for making french fries and hamburgers
but specific material flows depend upon the immediate desires of the customer.
Compared to Lands End, QCI has a relatively uncomplicated product base. However,
QCI has increased the number of available products from 12 in 1996 to 64 today (with
3 more in the works). Potential product combinations run to the millions.
In a few areas, QCI has been successful in reducing binder, tab, and cover inventory
levels. However, all binder and tab manufacturers offer a tremendous discount on
larger quantities. Visual systems are used to control inventories and work in
progress.
It should be noted that printer change over times are one minute maximum. Paper
scrap is now virtually non-existent. Sales have also increased 40% between Figures
11.1 and 11.2. Significant advances over the years have also included:
• Improved quality and productivity from the printers

• Paper storage moved closer to the machines
• The media center moved into the copy area
• Storage of all product information on a digital server
• An on-line shipping manifest system
Progress has been remarkable when viewed long-term. However, improvement on

a quarterly basis has been somewhere between evolutionary and glacial. QCI views
itself as an educational company that happens to manufacture training products.
QCI routinely reviews and benchmarks the administrative, design, and marketing
areas of the company.
VILLANOVA UNIVERSITY XI- 5 LEAN SIX SIGMA HANDBOOK

XI. CASE STUDIES
EXACTUS SOFTWARE LICENSE IMPROVEMENT
Software License Improvement at Exa(:tus

Introduction
Exactus Corporation is one of the largest business software d4:!velopers in Latin

America. The company has offices in most countries in the area iand serves mainly
large corporations, both local and multi-national. Quality and continuous
improvement are strategic issues for Exactus. The company is liSa 9001 certified,
and has achieved CMM level 3 status, making Exactus one of the 1~ew software firms
with this achievement in Latin America. The following project was developed by
Gloriana Jimenez, and a team from Exactus, as a green belt certification project.
Edwin Garro acted as black belt advisor for the project.
Defi ne Stage
Project Title
Issue and renew software licenses correctly and within the requiired time period.
Business Case
When a software license expires, the software stops working until a new license
number is entered in the system. If a customer has requested alnd completed the
license renewal prior to the expiration date, then the softwarE! should continue
working without any impediment. New software installations also require a proper
license number request. Two problems can occur after the customer has completed
his/her part, either the wrong license is issued or the license is not issued prior to
the expiration date. Both problems create a series of internal and E!xternal dilemmas.
Internally, a series of costly problems are generated and externally the customer is
stressed and dissatisfied.
Problem
Incorrect and late licenses were a chronic problem. During the month of November
2005, 5 ERP licenses were either wrong or late. In total, 46 out olf 150 licenses had
some type of problem (this includes November and December, ~!005 and January,
2006).
Goal/Scope
The goal was to completely eliminate problems with software licenses. This
includes both new and renewed licenses.

XI. CASE STUDIES
Software License Improvement at Exactus (Continued)

Measure Stage
Data from the time window of November, 2005, December, 2005, and January, 2006,
showed the following results for 150 issued software licenses:
Wrong Late Wrong

licenses licenses and late
14 32 46
Percentage of total 9.33% 21.33% 30.66%
ppm 93,333 213,333 306,666
Sigma level 2.82 2.30 1.99
Table 11.3 Results for November, 2005 through January, 2006
The causes of both wrong and late licenses were separated as follows:
Cause Frequency
Not knowing or not following procedures 8
Error from third party 3
Error in data entry 1
Lack of data 1
Wrong information 1
Table 11.4 Causes of Issuing a Wrong License
Cause Frequency
Internal software error 29
Not knowing or not following procedures 3
Table 11.5 Causes of Late Licenses
VILLANOV A UNIVERSITY XI-7 LEAN SIX SIGMA HANDBOOK

XI. CASE STUDIES
8 100
7 1
... -.. -_. -r-I 80
6 ~--~~---------------------------,
5 ......... . .... _. . _ ...... _. . .. _........... - _. _.. _.•. _... _. - - 60
4
" . &-
. . _......... - - - - -- - - _. - _. _.- - - - -- - - - -- - - _. - - _. -- - - - - - - - - -- -_. 40
3
2
f- 20
1
0 ~~~~----~~1~----~~1~L---~~=L--~==r=L--L 0
Procedures Third party Data entry Lack of data Wrong info
Figure 11.6 Pareto Chart of Causes of Issuing a Wrong License
The two main causes of wrong and late software licenses were:
1. Internal software errors
2. Not knowing or not following procedures
Analyze Stage
A series of brainstorming sessions and current state process mapping revealed the
following root causes of problems:
1. The current process was so complicated, that employees did not follow it.
2. Each employee created his/her own procedure.
3. The internal software created to issue software licenses did not work properly.
4. There was improper registration of new licenses.

XI. CASE STUDIES
J
Improve Stage
Three actions were implemented to eliminate the problem:
1. Create a new and simplified process.
2. Delineate responsibilities and accountabilities for issuing licenses.
3. Replace the use of the current software with a customer relationship

management (CRM) software, which was already available in the company.
An implementation schedule was developed. The following are the main elements
of the implementation stage.
• Prove the CRM software to all involved in the licensing process

• Train all personnel who received the new CRM software
• Create a procedure for third-party installations
• Train appropriate personnel in the third-party installations procedure
• Make one person responsible for follow-up on the licensing process
• Give the responsible person sufficient authority and resources
• Train the sales force in the new procedure
• Train implementation consultants
• Conduct a pilot test in Costa Rica
• Communicate the implementation results to all offices in all countries
• Expand implementation to all countries
Control Stage
• All licenses are now documented and reviewed

• The licensing process has been included in the internal audit schedule
• Customer service contracts the customer for closure
• Satisfied customers end the new and renewed license process

XI. CASE STUDIES

Project Results
A "before and after" project analysis is presented in Table 11.7 below.
Before After
There was not a defined procedure for issuing The new procE!dure covers all
new licenses. Two departments had different situations: new licenses, renewed
and inconsistent procedures. Third-party licenses, and third-party
implementations had no procedure at all. implementations;.
Information in the old software had to be The CRM system was substituted
included in the CRM system. No license could for the old licensing software. All
be generated until all information was in both information is c:ompleted in one
systems. system.
The one person who generated the license Licenses are generated by a sales
information had no relation to the sales assistant.
department. Sales representatives were
constantly called to get missing information.
Some countries were not considered in the All countries arE! part of the new
old process. process.
Five employees from three departments took Two employees are in charge of
part in the old licensing process. the whole proce!;s.
The old procedure took 19 steps. The new procedure takes 8 steps.
Third-party implementations required double Third-party implementations are
the effort to get a license. now consideredl a regular sale
from the beginning.
The old procedure was complex and The new system has an easier
bureaucratic. procedure.
The old failure rate was 300,000 ppm. So far, the new failure rate is 0
ppm.
Table 11.7 Before and After Exactus License Anal)'sis
Exactus now consistently uses DMAIC and lean six sigma practices for all
continuous improvement projects.
VILLANOVA UNIVERSITY XI-10 LEJi~N SIX SIGMA HANDBOOK

XI. CASE STUDIES
LSS AT LUDOVICO
}
Lean Six Sigma Implementation at Ludovico
Introduction
Ludovico Producci6n Grafica (Graphic Production) is a small printing shop in San
Jose, Costa Rica. The eight year old company principally serves international
corporations with operations in Costa Rica. Approximately 60% of Ludovico's
production involves manuals and inserts for electronic and medical device
companies. International customers such as Panduit, Arthrocare, Suttle, and Cytyc
require a commitment to quality and speed.
Often, a company's whole purpose is explained by its name "Ludovico. Graphic

Production" indicates that the company's main purpose is to manufacture high
quality printed products. A Graphic Arts Shop, on the other hand, is where the
customer goes for art designing. The customer wants, and pays for, something
pleasing to their taste. In Graphic Production the objective is to reproduce the
original idea as many times (sometimes millions) as requested. Ludovico is paid for
consistently reproducing an image. Nice and beautiful were previously agreed upon.
Terms like "identical", "equal", "the same" define Ludovico's customer's needs.
The purpose of this business case is to show a successful utilization of very simple
lean and six sigma tools in a small business environment. The very same tools that
apply to giants such as GE or Toyota, apply to small companies all over the world.
Ludovico's Six Core Principles

• No job is more important that any other. Each job is the most important.
• No job position is more important that any other job position.
• The customer determines the date of delivery. It must be Ludovico's

obsession to fulfill the delivery date requested by the customer.
• The quality of the job goes beyond what is seen. Ludovico will only accept a
job if the specifications are completely understood.
• The quality of a job begins with the quality of the shop. The shop must always
be clean and orderly.
• Each job begins with cleaning and finishes with cleaning.
The above six core principles pave the way for other lean and six sigma applications.

XI. CASE STUDIES
LSS AT LUDOVICO
Lean Six Sigma Implementation at Ludovico (Continued)

Values and principles are the first step in creating a world class company.
Techniques and methodologies such as lean and six sigma maY' come and go, but
the need for continuous improvement remains.
The Need for Lean Six Sigma Tools
Figure 11.8 shows monthly sales from December, 2004 to Novennber, 2006.
80,000
70,000
60,000
~
II)
50,000
:::l
II) 40,000
w
...J
c(
II) 30,000
20,000
10,000
o:t It) It) It) CD CD
-
It) It) It) CD CD CD
~ ~ c~
0
U
Ql
0
lJ ...C.
~ 0
C:
::J
~
en U
0
U
Ql
0
lJ
0
.:.C. c:
0
C, ....
t,)
Ql ::J Ql ::J ::J
C II. c( "") c( 0 C II. c( "") c( ()
MONTH
Figure 11.8 Ludovico Monthly Sales
The upward trend that started in 2004 was marked by the arrival ,of a second multi-
national customer with manufacturing operations in Costa Rica. The previous
market was composed mainly of advertisement agencies, and marketing
departments. The ad and marketing business is not very demanding. Quantities can
drift as much as ±5%. An occasional faulty product is allowed and even expected.
Manufacturing companies, on the other hand, demand exact quanitities, zero defects
and no delivery problems.
VILLANOVA UNIVERSITY XI-12 LE~\N SIX SIGMA HANDBOOK

XI. CASE STUDIES
LSS AT LUDOVICO
i
V.E.R© The Foundation of Speed, Delivery and Quality
Along with the six core principles above, Ludovico centers its quality in a simple but
powerful system called V.E.R. ("to see" in Spanish). V.E.R. is an acronym for
VERIFICAR (verify), EJECUTAR (execute) and REVISAR (review or check). Simply
stated, the system verifies or confirms that all the elements, prior to printing, comply
with customer specifications (samples, engineering drawings, color proofs,
purchase orders, digital images, customer signed approvals, etc.); executes the
work order according to approved first piece and process controls; and verifies
quantities, final process inspections, special packaging instructions, etc.
Minimal documentation is maintained so operators are not overwhelmed by paper

work. A dual document, with a work order on one side and quality instructions on
the other side, accompanies the job at all times. Each operator reads and signs this
document as the job passes through his/her workstation. Two main procedures
govern daily operations, one for production and one for quality. Visual management
is promoted and encouraged whenever possible.
For internal purposes and identification, the V.E.R. system is represented by the
popular Egyptian eye of Horus (or Ra). See Figure 11.9 below.
LUDOVICO®
produccion grafica
V.E.R.
VERIFICAR, EJECUTAR, REVISAR
Figure 11.9 The Ludovico Logo and Supporting Principles

XI. CASE STUDIES
LSS AT LUDOVICO
The First Kaizen Event-Flow Follows VallJle
When demand was softer and the print shop had extra capac:ity, there was an
opportunity to teach, for the first time, the concepts of ValUE! and flow to the
Ludovico troops.
A regular customer entered a routine monthly order of 10,000 ma~lazines. It took the
shop about 2 days to print the 10,000 magazines and then 4 more! days to complete
the binding and finishing portions. Figure 11.10 illustrates what went on in the
binding and finishing process.
ffi ffi
~"~ ~ 1; ~' D:q ;,
MAGAZINE MANUAL MAGAZINE STAPLING MAGAZINE FOLDING
~,
~
IfDt I"""TR....
,~, ~,
IMM....
IN....
G & -- . . ,
CT= 4 sec. CT= 2.5 sec. CT=4sec. CT=4sec.
C/O= a sec. C/O= a sec C/O= 7 min. C/O=5 min.
Batch= 10,000.u. Batch= 10,000.u. Batch= 10,000.u. Batch= 10,000.u.
1 DAY 1 DAY 1 DAY 1 CAY 4 DAYS
I 4 SEC. 4 SEC. 4 SEC.

<--_4S_EC
_. _ - - - ' rC.
Figure 11.10 Magazine Work Flow Before Improvement
The four employees in the finishing department would receive the 10,000 printed but
unfinished magazines from the cutting machine. They would thlen spend one day
manually collecting the 10,000 magazines (Step 1 in the value strt!am map). During
the second day the same 4 operators would staple the magazine (Step 2 in the value
stream map). During a third day they would fold the 10,000 maga2:ines (Step 3 in the
value stream map) and finally during the fourth day they would finish, trim, and
package the batch for delivery (Step 4 in the value stream map).
It is easy to calculate, from the cycle times, that less than a full d~IY was required to
complete each step. However, the same 4 operators were not necE!ssarily idle. They
often completed other small work orders.
VILLANOVA UNIVERSITY XI-14 LEtlN SIX SIGMA HANDBOOK

XI. CASE STUDIES
LSS AT LUDOVICO

The First Kaizen Event-Flow Follows Value (Continued)
Each day the crew would create a beautiful 10,000 magazine stack of WIP. Not until
4 days (and 4 seconds) was an actual single magazine ready. Operators were asked
what would happen if the customer entered the shop and requested a copy of their
magazine? About 16 seconds after the question was asked, a full magazine emerged
from the pile of WIP and was proudly given to the inquiring party.
The basic concept of flow, value, and WIP were introduced to the crew at that point.
At first they did not accept the new concepts at all. Some of them had worshiped
WIP for years and were afraid to try either a one piece flow or small batch approach.
After considerable discussion, a new method was created (see Figure 11.11). The
operators learned to act as a manufacturing cell, pulling one magazine at a time from
the previous operation. A fifth person (a handler) was introduced to the line. This
new task was undertaken to balance workload, move material, tally the finished
product, and set a pace for the group. The same batch that required 4 days for
completion was finished in 3 hours.
......------. r-D
Magazine finishing
ro =5
3 HOURS 3 HOURS
CT= 2.5 sec.
C/O=7 min.
Batch= 2 units
Figure 11.11 Magazine Work Flow After Improvement
VILLANOVA UNIVERSITY XI-1S LEAN SIX SIGMA HANDBOOK

XI. CASE STUDIES
LSS AT LUDOVICO
\ .
Lean Six Sigma Implementation at Ludovico {Continued}

The First Kaizen Event-Flow Follows Value (Continued)
A few changes suggested by the operators resulted in even faster cycle times than
anticipated. In addition, due to a coincidental design feature, tVifO magazines (not
one) could be finished every 2.5 seconds.
Notice that no sophisticated kanban or takt time calculations wer'e actually needed.
For a "make to order" company like Ludovico, dramatically redulcing times means
extra capacity to deliver more product to more customers.
From the day of this experiment, operators from the finishing department have
learned to balance their workload and achieve optimum work order flow. Thus, Work
Principle 3 "The customer determines the date of delivery" could be achieved and
even more customers could be accommodated.
The Second Kaizen Event-5S Implementation

The print shop demand kept growing. Ludovico's commitment t,o on-time delivery
was well appreciated by all customers. However, shop neatness and cleanliness
paid the price for speed and stress. All employees, 12 at that time, were trained in
5S. Some of the results from the first S (sort) were the following:
• 2 truckloads of materials, machine parts, and garbage were discarded

• Enough paper stock was discovered to last for 3 weeks
• Old "pantone" color inks (not used for a long time) were discarded
• Informal ink inventories ("just in case") were found at each press machine
• The first realistic tool inventory was conducted
• 2 broken industrial staplers became one good one
• 1 unused printing machine was sold to a smaller shop
• 1 unused plate burner was sold to a smaller shop
• 400 old plates were sold for recycling
• 1,000 pieces of film were sold for recycling
Application of the second S (set in order or straighten) resulted in the creation of a

tool board, a carton box supermarket, a printed product supermar~~et, and a reusable
plate storage area. More accomplishments will be presented in the visual
management discussion that follows.
VILLANOVA UNIVERSITY XI-16 LE)I,N SIX SIGMA HANDBOOK

XI. CASE STUDIES
LSS AT LUDOVICO

The Second Kaizen Event - 5S Implementation (Continued)
Application of the third S (shine) was immediately aligned with work principles 5 and
6 "The quality of a job begins with the quality of the shop. The shop must always be
clean and orderly" and "Each job begins with and finishes with cleaning." Clean
floors, clean workstations, clean machines, and clean offices, become the norm.
Occasional touch-up cleaning might be done for important visitors and customer
audits. At least once a month there is a deep cleaning, sorting and setting in order
event, to complement the daily cleaning activities.
There are no formal 5S audits yet. A second wave of 5S implementation is planned

for 2007. Lessons learned from the first implementation will be put into action in this
evolutionary form of 5S.
Winning the War Against Physical Space Restrictions
When the shop was acquired in 1999, space was immediately a concern. The old
warehouse that houses Ludovico is about 4,300 square feet. Recent flow projects,
cubic space utilization, and other kaizen projects have resulted in 210 square feet
rescued for productive use.
Standarization Through Visual Management
Visual management has been fundamental to sustain the 200% plus growth
experienced during the last two years. Examples of visual management include a
planning board, a training board, a tool board, and a floor area layout. A first piece
control board is also very important. Figure 11.13 is a representation of the training
board which shows 13 floor operators versus all machines and manual processes.

XI. CASE STUDIES
LSS AT LUDOVICO

Standardization Through Visual Management (C()ntinued)
Figure 11.13 Training Board - Individual Competencies
Due to visual management techniques, rework, the main quality issue at Ludovico,
has been reduced from 57,142 ppm (3.08 sigma level) to 6,250 ppm (3.99 sigma
level), through November, 2006.
All operators at Ludovico have 100% authorization to stop a procE~ss whenever they
have any doubts about their performance. Employees are encouraged to ask
questions and find answers from their superior (or support staff). Both the V.E.R.
logo and visual management systems support this initiative.

XI. CASE STUDIES
LSS AT LUDOVICO

Lessons Learned from Lean and Six Sigma Implementation
• No implementation has been perfect, but all have achieved improvements.
• Some ideas failed from day one, but that did not discourage the teams from
trying new approaches.
• Ludovico could not have sustained its growth without lean and six sigma
tools.
• The use of tools is backed up by a higher set of work principles and work
culture. Without that higher purpose, the tools would have been marginally
effective.
• Ludovico needs to be more formal and consistent in its continuous

improvement effort.
Summary of Results
Sales have grown more than 200% during the last 2 years. Lean and six sigma tools
have been essential to sustain this dramatic growth. Repeat customers are the main
source of income for Ludovico.
A key process (magazine finishing) went from 4 days to 3 hours through the use of
lean techniques. Operators consistently use flow and flexible lines to comply with
customer delivery dates.
A first wave of 5S implementation provided both tangible and intangible results.

Tangible results include more than $7,000 of savings, including the recycling and
sale of disposed items. Intangible results include the creation of a culture of order
and cleanliness consistent with company values.
Almost 5% of unproductive floor space has been converted into productive floor
space.
Visual management and other techniques have decreased rework from 57,142 ppm
(3.08 sigma level) to 6,250 ppm (3.99 sigma level).

XI. CASE STUDIES
INSURANCE CLAIM FLOW IMPROVEMENT
Insurance Claim Flow Improvement

As part of an overall continuous improvement initiative, an insurance services fund
chartered a series of kaizen teams. One project team was challenged to develop an
improved work flow and workplace organization for the mail rClom and microfilm
operations. This group handled paper-based claims coming in 1~rom the USPS.
The reason for selecting this area for a kaizen event revolved around the need to
introduce new document scanning technology to replace the antiquated microfilming
system. This unionized organization did not have a very go()d track record in
making major work changes. Typically, changes faced difficulties, many false starts,
and a number of heated arguments with union representatives.
The company steering committee, which was composed of eqlUal representation

from management and union leadership, took great care to charter a cross-
functional team that would include representatives from a nLimber of involved
functions. This team was led by a highly experienced facilitator .:md included team
members made up of mail room and microfilm employees, insurance claim
adjudication techs, information technology professionals, a union steward, and a
supervisor from the mail room.
A five day kaizen event was planned to allow the team adequate tiime to learn and to
develop the best possible plan for later implementation by the team and others over
a three month period. Unlike traditional kaizens, where a 'team attempts to
implement changes in a single week, this was a special situatio,n for a number of
reasons.
The workforce involved with the current state process worked in lthree departments
on two different floors of the office building. Secondly, there was Ito be considerable
expense putting in the necessary cabling and power to support the new scanner
technology, so they wanted to make sure the location was ri!~ht the first time.
Thirdly, the optical recognition scanning technology was new to the company and
it was expected that there would be a significant learning cur'Ve. Lastly, there
needed to be adequate time to work out programming issues to integrate data from
the new scanning results with the existing IT infrastructure.
The steering committee provided a charter for the team: Find wa~,s to take time out
of the process while at the same time working to improve prclductivity and the
quality of results. The first step in the process was to give this team a solid
foundation by providing adequate training. The training content included waste and
waste elimination, value stream mapping, one piece flow, and significant team
building activities. Working in teams was a new concept for the individuals involved.
VILLANOVA UNIVERSITY XI-20 LE.tIN SIX SIGMA HANDBOOK

XI. CASE STUDIES
Insurance Claim Flow Improvement (Continued)

The actual focus for this team was to:
• Establish a clear picture of the current state

• Identify and eliminate waste
• Address error-proofing opportunities
• Implement safety improvements
• Create work flow improvements
• Develop standard visual work
• Recommend workplace design changes
• Develop implementation action plans
• Measure and compare the results to base lines
• Prepare and deliver a final report
The team was challenged to create a value stream map (VSM) and spaghetti diagram
of the current state processes, including an estimate of physical distances traveled
for in-bound paper claims.
The team created a spaghetti diagram which indicated the following travel distances
for the paper claims:
• Mail room team deliverables - 255 feet on the 1st floor.
• Mail room team deliverables - 461 feet on the 3rd floor.
• A total of 716 feet of movement by people with ZERO value added.
The team brainstormed and agreed upon the information required to give them an
accurate picture of the current state of the process. The VSM databox is shown in
Table 11.14.
The team broke into sub-teams and began gathering observations and data to
populate the data boxes with pertinent information. They also captured as many
ideas and recommendations for potential improvements as possible. Information
was recorded at each "touch" occurrence for the maiVdocuments passing through
the value stream.

XI. CASE STUDIES
Step name:
Place it occurs:
Who does this?

How many people?
Total cycle time of task C~,cles per min.
Average daily quantity Qty: I
Non-value-added time comments
Average delay (hours, minutes, or

days)
Number of units waiting Qty: I
Comments
Top 3 quality or rework issues 1.

2.
3.
Artifacts at this step?
Comments and observations:
Table 11.14 Value Stream Map Databox
VILLANOVA UNIVERSITY XI- 22 LEJIN SIX SIGMA HANDBOOK

XI. CASE STUDIES

The finished VSM was constructed using paper and pencil on wall charts. Each
data box was printed on a piece of copy paper and the data was filled-in with pencils
(to allow for easy corrections). A photo of the completed initial current state map is
shown in Figure 11.15.
Figure 11.15 Current State Value Stream Map
What the team found was that each and every piece of claim mail encountered at
least 10 significant hand-offs (touches), if all went well the first time! There were
many "artifacts" collected and displayed. Included were forms, screen shots of
necessary entry data, and tracking documents. The team went on to summarize the
VSM statistics and learned the following:
• The average paper claim cycle time (touch time) was 49 seconds.
• The paper claim through time was four 8 hour days (32 hours).
• The value added time was 0.04% - 49 seconds divided by 32 hours.
Needless to say, the team was highly motivated and interested in seeing what they
could do to eliminate all the delays and wasted efforts in the current state of the
value stream. Everyone agreed that making customers wait 4 days to start the
claims process was unacceptable!

XI. CASE STUDIES

On the third day of the kaizen event, the team moved on to the critical next step -
brainstorming value stream improvement ideas. Initially, over 100 improvement
ideas were generated. The team went on to use affinity diagramming techniques to
group the ideas into logical categories. These, in turn, lead to targeted
improvements as summarized below:
• Reduced movement of people • 5S activities

• Improved ergonomics and safety • Improved work fllow
• Required materials at point of use • Standardized visiual work
• A production communication board • Reduced handling
The team agreed that there had to be a way to get all the processe:s into a single area
on the first floor of the building. This meant that many paradigmn about the current
work had to be broken to condense the desired workflow into the smallest possible
space. One big shift was to move away from mail sorting and pr1eparation at desks
and, instead, perform this work at a central pass-through mail siorting table. This
change would facilitate a smooth transition to the scanners.
Creative thinking was required to conceptualize a new layout. The team was
challenged: "Come up with a workflow that allows a piece of mail to continuously
move, without stopping, from the time it is opened to the time it is scanned." The
team accepted the challenge and developed the following concE!ptual work flow:
slots
Rx .,
Rx
•
dust/noise
----~
barrier
X-ray
dental •
non-scan
scan kick-out
30 day storage 2
3rd floor
electronic transfer
Figure 11.16 Utopian Conceptual Work Flow

XI. CASE STUDIES

In the conceptual design, claims would move from envelopes to their final resting
place, after scanning, utilizing a tiny fraction of the space that the current state
process required.
Next, the team set the conceptual layout aside and began working with the actual
building footprint itself. They accepted the built-in limitations of the current building
and derived a layout that would come as close as possible to their "utopian" plan.
Three layout scenarios were developed. One option was selected. The team then
created an implementation plan to make the physical transition a reality.
Goals and Objectives
The team then established their objectives for the project:
• Improve productivity by 20% by July 31. This means reducing the cycle time
per piece through workflow improvements.
• Decrease the through-time for mail to the DC room by 50% by July 31. This
means all received mail will be delivered in no more than two days.
• Reduce the number of sorts in the mail room by 50% by June 30. The current
19 sorts would be reduced to 9 or 10.
• Reduce the physical movement of mail by 50%. That is, cut the travel distance
of 716 feet by half.
Implementation
Over several months the team continued to implement their previously developed
ideas to reduce movement, improve ergonomics and safety, provide materials at the
point of use, reduce handling, institute better flow, develop standard work, institute
58 activities, and utilize visual communication boards.
One significant innovation was the development and implementation of a visual

"pass-through" work table to facilitate their goals. Many excellent improvements
were incorporated as seen in the photos in Figures 11.19 and 11.20 on the next page.

XI. CASE STUDIES

Implementation (Continued)
Figure 11.19 The Pass Through Figure 11.20 The Co-Located

Sorting Operation Scanning Operation
The Rest of the Story
At the completion of the kaizen project, a circle team was chartered to sustain the
gains and continue the improvement process. This was accomplished without any
issues with the union. All changes were accomplished without the typical problems
encountered when management imposed workforce changes.
The achieved results included a 25% improvement in productivit~', a 50% reduction

in space, and a 75% reduction in travel distance. There were across-the-board
improvements in the quality of the work itself.
Today the average in-bound piece of mail is handled in under two (eight-hour) days
and the team is far from satisfied. The mail room team is also actively working with
other departments to identify activities they can perform to he~p eliminate other
wastes and speed information through the system. This depalrtment has made
continuous improvement part of their culture.

XI. CASE STUDIES
A3 MEDICAL CENTER IMPROVEMENT
Medical Center Improvement using A3 Problem Solving*

Long Patient Transportation Times
The Context
At Community Medical Center (CMC), two types of patients are sent to the diagnostic
departments for procedures: outpatients and inpatients. The outpatients come to
the hospital, register, complete a procedure, and leave on the same day. The
inpatients reside in the clinical departments overnight and are sent to the diagnostic
departments for various procedures depending on the medical necessity. Once the
procedure is complete, the patient is returned to the clinical department.
Some of the outpatients who come for the procedures are frail, and therefore, unable
to walk to the diagnostic department. It is the responsibility of the transportation
department to move the patient to the appropriate department. Similarly, it is the
responsibility of the transporter to move the inpatients to the diagnostic department
from the clinical departments when they are scheduled for procedures.
The diagnostic departments (Operating Room, Radiology, Nuclear Medicine,

Cardiology, Endoscopy, and Emergency Room) in CMC regularly complained that
patient transporters took an excessive time, causing delays in treatment and patient
waits. Many thought that the transporters were having long coffee breaks and did
not respond. The transportation manager became exasperated with the influx of
complaints and decided to address the issue with an A3 improvement team.
The A3 Process
A group of individuals representing the diagnostic departments (radiology,
endoscopy, special procedures, cardiology), nursing, transportation, and quality risk
management met to discuss the issue and initiate an A3 problem solving initiative.
These individuals formed the core A3 problem solving team.
To understand the problem first hand, the transportation manager and four
transporters observed the current process. They observed the request for patient
transportation process as it unfolded for 10 days. The manager also contacted and
interviewed different individuals in the diagnostic and clinical departments to get
first hand information about the process.
* Slightly modified from Sobek (2007f and used with permission.

XI. CASE STUDIES
A3 Medical Center Improvement (Continued)

The transportation manager said, "We observed all nursing static)ns and procedure
areas, and noted a process full of miscommunications. For example, I observed the
ward secretary who said she would call [a transporter] right away. She actually
made the call 37 minutes later, 3 minutes before the procedurte. This happened
frequently. The procedure department and the nursing station ne'lfer communicated
as to the expected patient transport or procedure times. The patient's nurse often
did not even know the patient was going to a procedure so the patient medications
were not always met for the procedure."
I n addition, the manager of transportation completed a patient transportation survey.

In this survey, he measured the time difference between the transporters receiving
a beep (a request from the diagnostic department) to the time the patient was
delivered to the diagnostic or procedure areas. The results of 23 patients surveyed
over a three-day period (1/15/03 to 1117/03) showed an average request to delivery
time of 56 minutes. The actual patient transport time was only 5 minutes and the
rest was preparation time and delays in communication. The communication delays
caused delays in timely procedures, resulting in unhappy patient~i, clinical workers,
and physicians.
In the current state, somebody from the diagnostic department, usually a technician,
called or paged the transporter. At other times, somebody fr()m the diagnostic
department called the ward secretary on the floors who then called the RN and the
transporter. The transporter did not know who was paging. Sometimes the
message the transporter received said, "Bring down John Doe to Badiology." There
was no information on the room number, bed number, floor, or area. The transporter
did not know from where the person was paging and did not alwclYs know whom to
call to clarify. He/she only knew a patient needed to be transported. A great deal of
time was, thus, expended by the transporter on patient search.
If the information was complete and the patient was ready for the procedure, the
transporter reached the patient and transported him or her to the diagnostic
department. However, in many situations when the transporter n~ached the patient
(usually inpatients), the patient was not ready and was in need of medications, MRI
screening, bathroom break, IV changes, etc. In these situations, ,as the patient was
not ready for transport, the transporter contacted the nurse. The transporter left the
room and waited for the call from the nurse when the patient Wall) ready.

XI. CASE STUDIES

The transportation manager drew the current state (patient ready for transport)
drawing on the A3 Report with appropriate icons and arrows to indicate the flow of
information and patients through the system, as shown in Figure 11.21. On the
current state drawing, he recorded the shortest (9 minutes), longest (177 minutes),
and average transportation time (56 minutes) from the data collected earlier. The
identified problems were no written message to request a transporter, and late
arrival by the patients at the diagnostic departments. These are depicted as storm
clouds on the current state diagram.
Charge
1\ ·fi
Diagnostic
department
Ward
secretary
RN
h Patient
----EJ
Transporter
Figure 11.21 Patient Transport Current State Map
The A3 problem solving team brainstormed the root causes of the problems using
the "S-Whys" approach. The analysis of the first storm cloud revealed that the staff
members calling from the diagnostic department were often too busy to send written
messages to the transporter or to the floors, and therefore, the message lacked
complete information causing delays. The analysis of the second storm cloud
revealed that as the RNs or ward secretaries were sometimes not aware that a
patient needed a procedure, they failed to prepare the patient on time, which
eventually led to late arrival of the patient at the diagnostic department.
Based on the understanding of the current state and the associated root causes, the
team embarked on devising the target state. The transportation manager termed the
problem solving as "Road to Recovery."

XI. CASE STUDIES

In the target state, the staff in the diagnostic department (usually a technician in
radiology or endoscopy, or ward secretary in surgery) will page both the charge RN
and the transporter at the same time. The information includled in the page is
complete information, for effective transport of the patient to the diagnostic
department (i.e. patients first and last name, medical record number, room number,
destination, etc.). The charge RN will attend to the nursing care needs of the patient
and the transporter will attend to the comforts during transport such as shoes,
blankets, chairs/stretcher, etc. If everything is found in order, the patient will be
transported to the diagnostic department for the procedure. Upon completion of the
procedure, the diagnostic department will page the transporter \litho will return the
patient back to the patient's room.
The transportation manager drew the target state drawing on the A3 report as
illustrated below in Figure 11.22.
f\ .1\
Charge RN Technician
Diagnostic
Department
Should be able to do in
Less than 30 minutes
Transporter
Figure 11.22 Patient Transport Target State Map
The specific countermeasures to achieve the target state were:
• The diagnostic departments will beep the charge RN and thE~ transporter at the
same time
• The page will include specific information and a reference card

XI. CASE STUDIES

• The charge RN (or designated person) and the transporter will attend to the
patient, with specific responsibilities
• The patient will be made aware of the ensuing procedure
As part of the implementation plan, the team created a specific action plan. First, a
designated transporter and a staff member responsible for communication in CMC
developed a "group page" whereby two or more people could be paged
simultaneously by the diagnostic departments. Second, the transportation manager
and the charge RNs met and developed a patient tracking sheet (a log sheet for the
floor staff to sign off when the patient is transported). Third, the transportation
manager and the designated transporter developed a reference card that contained
the pager numbers of the charge RNs of each clinical department and the transport
pager numbers that the diagnostic departments should page. It contains the
information that needs to be paged by the diagnostic department when asking for
a patient transport. This information included:
• Name of the department from where the message is paged
• First and last name of the patient
• Patient's medical record number
• Room number
• Patient's destination
• Preferred mode of transport (chair, stretcher)
The reference card also contained a step-by-step procedure for requesting a

transporter by the diagnostic departments. The transportation manager sent copies
of this information to every department to ensure safe, accurate, and efficient
transport of patients.
To ensure smooth implementation of the improved process, the transportation

manager met with key individuals in all clinical departments on a one-on-one basis.
He explained the necessity of the new process and got their feedback on the new
process and how it could be improved. He also had meetings with the house
supervisors to get them on board with the new process.

XI. CONTROL CONCEPTS
REFERENCES

The transportation manager mentioned that the ward secretaries were initially tough
to deal with. However, the house supervisors were very supportive. "They made the
ward secretaries conform to the policies," said the transportation manager. The
procedural departments were very supportive as well.
The transportation manager set the target time from request to delivery at 30
minutes. When asked about the rationale for setting such a high time, he responded,
"The procedural departments were happy with 30 minutes. They were tickled to
death. Moreover, most procedure departments schedule in 30 minutes increments."
He carried out follow-up surveys at regular intervals to assess transport time. The
following table presents the collected data:
Survey Date Time (In Mins)

March, 2003 14.7
April, 2004 11.0
September, 2004 11.5
May, 2005 9.15
Table 11.23 Follow-up Survey Results (Sobek, 2007)1
Reactions to the A3 Process
The transportation manager felt the A3 process was every effE!ctive for problem
solving in healthcare. He wrote, "I find the A3 process a very important tool for
evaluating problems and/or processes. It allows a person or team to look at how a
process flows and where the problem or work around area may be. It promotes team
work on solving problems by giving a global and unbiased look into procedures. It
involves a positive thought process and invigorates the mind to think in alternative
ways of problem solving by including all aspects of a process. It gives all parties
involved a way to express and present their perceptions or data. IOn a scale of 1-10,
ten being the highest, I would rate the A3 process at a 10."
Reference
1. Sobek, O.K. (2007). "A Case Study in A3 Problem Solvin!)." Montana State
University. Downloaded January 25, 2007 from
http://www.coe.montana.eduIlE/faculty/sobeklA3

XII. APPENDIX
INDEX LEARNING TURNS NO STUDENT

PALE, YET HOLDS THE EEL OF SCIENCE
BY THE TAIL.
ALEXANDER POPE
VILLANOVA UNIVERSITY XII -1 LEAN SIX SIGMA HANDBOOK

XII. APPENDIX - TABLES
Table I
Standard Normal Table
o z
z X.XO X.X1 X.X2 X.X3 X.X4 X.X5 X.X6 XJ(7 X.X8 X.X9
0.0 0.5000 0.4960 0.4920 0.4880 0.4840 0.4801 0.4761 0.4n1 0.4681 0.4641
0.1 0.4602 0.4562 0.4522 0.4483 0.4443 0.4404 0.4364 0.4~125 0.4286 0.4247
0.2 0.4207 0.4168 0.4129 0.4090 0.4052 0.4013 0.3974 0.3936 0.3897 0.3859
0.3 0.3821 0.3783 0.3745 0.3707 0.3669 0.3632 0.3594 0.3~j57 0.3520 0.3483
0.4 0.3446 0.3409 0.3372 0.3336 0.3300 0.3264 0.3228 0.31192 0.3156 0.3121
0.5 0.3085 0.3050 0.3015 0.2981 0.2946 0.2912 0.2877 0.2f143 0.2810 0.2776
0.6 0.2743 0.2709 0.2676 0.2643 0.2611 0.2578 0.2546 O.2~i14 0.2483 0.2451
0.7 0.2420 0.2389 0.2358 0.2327 0.2297 0.2266 0.2236 0.2206 0.2177 0.2148
0.8 0.2119 0.2090 0.2061 0.2033 0.2005 0.1977 0.1949 0.1~122 0.1894 0.1867
0.9 0.1841 0.1814 0.1788 0.1762 0.1736 0.1711 0.1685 0.1660 0.1635 0.1611
1.0 0.1587 0.1562 0.1539 0.1515 0.1492 0.1469 0.1446 0.1423 0.1401 0.1379
1.1 0.1357 0.1335 0.1314 0.1292 0.1271 0.1251 0.1230 0.12:10 0.1190 0.1170
1.2 0.1151 0.1131 0.1112 0.1093 0.1075 0.1056 0.1038 0.10'20 0.1003 0.0985
1.3 0.0968 0.0951 0.0934 0.0918 0.0901 0.0885 0.0869 0.0853 0.0838 0.0823
1.4 0.0808 0.0793 0.0778 0.0764 0.0749 0.0735 0.0721 0.0708 0.0694 0.0681
1.5 0.0668 0.0655 0.0643 0.0630 0.0618 0.0606 0.0594 0.0582 0.0571 0.0559
1.6 0.0548 0.0537 0.0526 0.0516 0.0505 0.0495 0.0485 0.0475 0.0465 0.0455
1.7 0.0446 0.0436 0.0427 0.0418 0.0409 0.0401 0.0392 0.0384 0.0375 0.0367
1.8 0.0359 0.0351 0.0344 0.0336 0.0329 0.0322 0.0314 0.0307 0.0301 0.0294
1.9 0.0287 0.0281 0.0274 0.0268 0.0262 0.0256 0.0250 0.0244 0.0239 0.0233
2.0 0.0228 0.0222 0.0217 0.0212 0.0207 0.0202 0.0197 0.0192 0.0188 0.0183
2.1 0.0179 0.0174 0.0170 0.0166 0.0162 0.0158 0.0154 0.0150 0.0146 0.0143
2.2 0.0139 0.0136 0.0132 0.0129 0.0125 0.0122 0.0119 0.0116 0.0113 0.0110
2.3 0.0107 0.0104 0.0102 0.0099 0.0096 0.0094 0.0091 0.00B9 0.0087 0.0084
2.4 0.0082 0.0080 0.0078 0.0075 0.0073 0.0071 0.0069 0.0068 0.0066 0.0064
2.5 0.0062 0.0060 0.0059 0.0057 0.0055 0.0054 0.0052 0.0051 0.0049 0.0048
2.6 0.0047 0.0045 0.0044 0.0043 0.0041 0.0040 0.0039 0.00:38 0.0037 0.0036
2.7 0.0035 0.0034 0.0033 0.0032 0.0031 0.0030 0.0029 0.0028 0.0027 0.0026
2.8 0.0026 0.0025 0.0024 0.0023 0.0023 0.0022 0.0021 0.00:21 0.0020 0.0019
2.9 0.0019 0.0018 0.0018 0.0017 0.0016 0.0016 0.0015 0.00'15 0.0014 0.0014
3.0 0.00135
How to read "z" value:

1. Find the first two digits of Z in the far left-hand column and use that row.
2. Find the third digit in the top row and read down that column until you
reach the row with the first two digits.
3. Multiply that number by 100 to get a percentage.
A standard normal table gives the percentage under one-half of the normal curve
at different Z values. Look at the values of 0, 1.0, 2.0 and 3.0 to better understand
th is concept.
VILLANOVA UNIVERSITY XII - 2 LEA.N SIX SIGMA HANDBOOK

Table II - Six Sigma Failure Rates
With a 1.5 a Process Shift With No Process Shift

Z ppm Z ppm Z ppm Z ppm
1.0 697,672.15 3.6 17,864.53 1.0 317,310.52 3.6 318.29
1.1 660,082.92 3.7 13,903.50 1.1 271,332.20 3.7 215.66
1.2 621,378.38 3.8 10,724.14 1.2 230,139.46 3.8 144.74
1.3 581,814.88 3.9 8,197.56 1.3 193,601.10 3.9 96.23
1.4 541,693.78 4.0 6,209.70 1.4 161,513.42 4.0 63.37
1.5 501,349.97 4.1 4,661.23 1.5 133,614.46 4.1 41.34
1.6 461,139.78 4.2 3,467.03 1.6 109,598.58 4.2 26.71
1.7 421,427.51 4.3 2,555.19 1.7 89,130.86 4.3 17.09
1.8 382,572.13 4.4 1,865.88 1.8 71,860.53 4.4 10.83
1.9 344,915.28 4.5 1,349.97 1.9 57,432.99 4.5 6.80
2.0 308,770.21 4.6 967.67 2.0 45,500.12 4.6 4.23
2.1 274,412.21 4.7 687.20 2.1 35,728.71 4.7 2.60
2.2 242,071.41 4.8 483.48 2.2 27,806.80 4.8 1.59
2.3 211,927.71 4.9 336.98 2.3 21,448.16 4.9 0.960
2.4 184,108.21 5.0 232.67 2.4 16,395.06 5.0 0.574
2.5 158,686.95 5.1 159.15 2.5 12,419.36 5.1 0.340
2.6 135,686.77 5.2 107.83 2.6 9,322.44 5.2 0.200
2.7 115,083.09 5.3 72.37 2.7 6,934.05 5.3 0.116
2.8 96,809.10 5.4 48.12 2.8 5,110.38 5.4 0.067
2.9 80,762.13 5.5 31.69 2.9 3,731.76 5.5 0.038
3.0 66,810.63 5.6 20.67 3.0 2,699.93 5.6 0.021
3.1 54,801.40 5.7 13.35 3.1 1,935.34 5.7 0.012
3.2 44,566.73 5.8 8.55 3.2 1,374.40 5.8 0.007
3.3 35,931.06 5.9 5.42 3.3 966.97 5.9 0.004
3.4 28,716.97 6.0 3.40 3.4 673.96 6.0 0.002
3.5 22,750.35 6.1 2.11 3.5 465.35 6.1 0.001
VILLANOVA UNIVERSITY XII - 3 LEAN SIX SIGMA HANDBOOK

XII. APPENDIX· TABLES
Table III
Poisson Distribution
Probability of r or fewer occurrences of an event that has an average number of
occurrences equal to np.
~ 0.02
0
0.980
1
1.000
2 3 4 5 6 7 8 9
0.04 0.961 0.999 1.000

0.06 0.942 0.998 1.000
0.08 0.923 0.997 1.000
0.10 0.905 0.995 1.000
0.15 0.861 0.990 0.999 1.000

0.20 0.819 0.982 0.999 1.000
0.25 0.779 0.974 0.998 1.000
0.30 0.741 0.963 0.996 1.000
0.35 0.705 0.951 0.994 1.000

0.40 0.670 0.938 0.992 0.999 1.000
0.45 0.638 0.925 0.989 0.999 1.000
0.50 0.607 0.910 0.986 0.998 1.000
0.55 0.577 0.894 0.982 0.998 1.000

0.60 0.549 0.878 0.977 0.997 1.000
0.65 0.522 0.861 0.972 0.996 0.999 1.000
0.70 0.497 0.844 0.966 0.994 0.999 1.000
0.75 0.472 0.827 0.959 0.993 0.999 1.000
0.80 0.449 0.809 0.953 0.991 0.999 1.000

0.85 0.427 0.791 0.945 0.989 0.998 1.000
0.90 0.407 0.772 0.937 0.987 0.998 1.000
0.95 0.387 0.754 0.929 0.984 0.997 1.000
1.00 0.368 0.736 0.920 0.981 0.996 0.999 1.000
1.1 0.333 0.699 0.900 0.974 0.995 0.999 1.000

1.2 0.301 0.663 0.879 0.966 0.992 0.998 1.000
1.3 0.273 0.627 0.857 0.957 0.989 0.998 1.000
1.4 0.247 0.592 0.833 0.946 0.986 0.997 0.999 1.000
1.5 0.223 0.558 0.809 0.934 0.981 0.996 0.999 1.0010
1.6 0.202 0.525 0.783 0.921 0.976 0.994 0.999 1.0010

1.7 0.183 0.493 0.757 0.907 0.970 0.992 0.998 1.0010
1.8 0.165 0.463 0.731 0.891 0.964 0.990 0.997 0.9919 1.000
1.9 0.150 0.434 0.704 0.875 0.956 0.987 0.997 0.999 1.000
2.0 0.135 0.406 0.677 0.857 0.947 0.983 0.995 0.999 1.000
VILLANOVA UNIVERSITY XII·4 LEAN SIX SIGMA HANDBOOK

Table III
Poisson Distribution (Continued)
~ 2.2
0
0.111
1
0.355
2
0.623
3
0.819
4
0.928
5
0.975
6
0.993
7
0.998
8
1.000
9
2.4 0.091 0.308 0.570 0.779 0.904 0.964 0.988 0.997 0.999 1.000
2.6 0.074 0.267 0.518 0.736 0.877 0.951 0.983 0.995 0.999 1.000
2.8 0.061 0.231 0.469 0.692 0.848 0.935 0.976 0.992 0.998 0.999
3.0 0.050 0.199 0.423 0.647 0.815 0.916 0.966 0.988 0.996 0.999
3.2 0.041 0.171 0.380 0.603 0.781 0.895 0.955 0.983 0.994 0.998
3.4 0.033 0.147 0.340 0.558 0.744 0.871 0.942 0.977 0.992 0.997
3.6 0.027 0.126 0.303 0.515 0.706 0.844 0.927 0.969 0.988 0.996
3.8 0.022 0.107 0.269 0.473 0.668 0.816 0.909 0.960 0.984 0.994
4.0 0.018 0.092 0.238 0.433 0.629 0.785 0.889 0.949 0.979 0.992
4.2 0.015 0.078 0.210 0.395 0.590 0.753 0.867 0.936 0.972 0.989
4.4 0.012 0.066 0.185 0.359 0.551 0.720 0.844 0.921 0.964 0.985
4.6 0.010 0.056 0.163 0.326 0.513 0.686 0.818 0.905 0.955 0.980
4.8 0.008 0.048 0.143 0.294 0.476 0.651 0.791 0.887 0.944 0.975
5.0 0.007 0.040 0.125 0.265 0.440 0.616 0.762 0.867 0.932 0.968
5.2 0.006 0.034 0.109 0.238 0.406 0.581 0.732 0.845 0.918 0.960
5.4 0.005 0.029 0.095 0.213 0.373 0.546 0.702 0.822 0.903 0.951
5.6 0.004 0.024 0.082 0.191 0.342 0.512 0.670 0.797 0.886 0.941
5.8 0.003 0.021 0.072 0.170 0.313 0.478 0.638 0.771 0.867 0.929
6.0 0.002 0.017 0.062 0.151 0.285 0.446 0.606 0.744 0.847 0.916
10 11 12 13 14 15 16
2.8 1.000
3.0 1.000
3.2 1.000
3.4 0.999 1.000
3.6 0.999 1.000
3.8 0.998 0.999 1.000
4.0 0.997 0.999 1.000
4.2 0.996 0.999 1.000

4.4 0.994 0.998 0.999 1.000
4.6 0.992 0.997 0.999 1.000
4.8 0.990 0.996 0.999 1.000
5.0 0.986 0.995 0.998 0.999 1.000
5.2 0.982 0.993 0.997 0.999 1.000

5.4 0.977 0.990 0.996 0.999 1.000
5.6 0.972 0.988 0.995 0.998 0.999 1.000
5.8 0.965 0.984 0.993 0.997 0.999 1.000
6.0 0.957 0.980 0.991 0.996 0.999 0.999 1.000

Table III
Poisson Distribution (Continued)
~ 6.2
0
0.002
1
0.015
2
0.054
3
0.134
4
0.259
5
0.414
6
0.574
7
0.7'16
8
0.826
9
0.902
6.4 0.002 0.012 0.046 0.119 0.235 0.384 0.542 0.687 0.803 0.886
6.6 0.001 0.010 0.040 0.105 0.213 0.355 0.511 0.658 0.780 0.869
6.8 0.001 0.009 0.034 0.093 0.192 0.327 0.480 0.6~~8 0.755 0.850
7.0 0.001 0.007 0.030 0.082 0.173 0.301 0.450 0.5~19 0.729 0.830
7.2 0.001 0.006 0.025 0.072 0.156 0.276 0.420 0.5ti9 0.703 0.810
7.4 0.001 0.005 0.022 0.063 0.140 0.253 0.392 0.5~19 0.676 0.788
7.6 0.001 0.004 0.019 0.055 0.125 0.231 0.365 0.5110 0.648 0.765
7.8 0.000 0.004 0.016 0.048 0.112 0.210 0.338 0.4~I1 0.620 0.741
8.0 0.000 0.003 0.014 0.042 0.100 0.191 0.313 0.453 0.593 0.717
8.5 0.000 0.002 0.009 0.030 0.074 0.150 0.256 0.366 0.523 0.653
9.0 0.000 0.001 0.006 0.021 0.055 0.116 0.207 0.324 0.456 0.587
9.5 0.000 0.001 0.004 0.015 0.040 0.089 0.165 0.2fi9 0.393 0.522
10.0 0.000 0.000 0.003 0.010 0.029 0.067 0.130 0.2~!0 0.333 0.458
10 11 12 13 14 15 16 17 18 19
6.2 0.949 0.975 0.989 0.995 0.998 0.999 1.000
6.4 0.939 0.969 0.986 0.994 0.997 0.999 1.000
6.6 0.927 0.963 0.982 0.992 0.997 0.999 0.999 1.000
6.8 0.915 0.955 0.978 0.990 0.996 0.998 0.999 1.000
7.0 0.901 0.947 0.973 0.987 0.994 0.998 0.999 1.000
7.2 0.887 0.937 0.967 0.984 0.993 0.997 0.999 0.9919 1.000
7.4 0.871 0.926 0.961 0.980 0.991 0.996 0.998 0.9919 1.000
7.6 0.854 0.915 0.954 0.976 0.989 0.995 0.998 0.9919 1.000
7.8 0.835 0.902 0.945 0.971 0.986 0.993 0.997 0.9919 1.000
8.0 0.816 0.888 0.936 0.966 0.983 0.992 0.996 0.9918 0.999 1.000
8.5 0.763 0.849 0.909 0.949 0.973 0.986 0.993 0.997 0.999 0.999
9.0 0.706 0.803 0.876 0.926 0.959 0.978 0.989 0.995 0.998 0.999
9.5 0.645 0.752 0.836 0.898 0.940 0.967 0.982 0.991 0.996 0.998
10.0 0.583 0.697 0.792 0.864 0.917 0.951 0.973 0.986 0.993 0.997
20 21 22
8.5 1.000
9.0 1.000
9.5 0.999 1.000
10.0 0.998 0.999 1.000

Table IV Binomial Distribution

Probability of r or fewer occurrences of an event in n trials
p (the probability of occurrence on each trial)
n r 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50
2 0 0.9025 0.8100 0.7225 0.6400 0.5625 0.4900 0.4225 0.3600 0.3025 0.2500
1 0.9975 0.9900 0.9775 0.9600 0.9375 0.9100 0.8775 0.8400 0.7975 0.7500
3 0 0.8574 0.7290 0.6141 0.5120 0.4219 0.3430 0.2746 0.2160 0.1664 0.1250
1 0.9928 0.9720 0.9392 0.8960 0.8438 0.7840 0.7182 0.6480 0.5748 0.5000
2 0.9999 0.9990 0.9966 0.9920 0.9844 0.9730 0.9571 0.9360 0.9089 0.8750
4 0 0.8145 0.6561 0.5220 0.4096 0.3164 0.2401 0.1785 0.1296 0.0915 0.0625
1 0.9860 0.9477 0.8905 0.8192 0.7383 0.6517 0.5630 0.4752 0.3910 0.3125
2 0.9995 0.9963 0.9880 0.9728 0.9492 0.9163 0.8735 0.8208 0.7585 0.6875
3 1.0000 0.9999 0.9995 0.9984 0.9961 0.9919 0.9850 0.9744 0.9590 0.9375
5 0 0.7738 0.5905 0.4437 0.3277 0.2373 0.1681 0.1160 0.0778 0.0503 0.0312
1 0.9774 0.9185 0.8352 0.7373 0.6328 0.5282 0.4284 0.3370 0.2562 0.1875
2 0.9988 0.9914 0.9734 0.9421 0.8965 0.8369 0.7648 0.6826 0.5931 0.5000
3 1.0000 0.9995 0.9978 0.9933 0.9844 0.9692 0.9460 0.9130 0.8688 0.8125
4 1.0000 1.0000 0.9999 0.9997 0.9990 0.9976 0.9947 0.9898 0.9815 0.9688
6 0 0.7351 0.5314 0.3771 0.2621 0.1780 0.1176 0.0754 0.0467 0.0277 0.0156
1 0.9672 0.8857 0.7765 0.6554 0.5339 0.4202 0.3191 0.2333 0.1636 0.1094
2 0.9978 0.9842 0.9527 0.9011 0.8306 0.7443 0.6471 0.5443 0.4415 0.3438
3 0.9999 0.9987 0.9941 0.9830 0.9624 0.9295 0.8826 0.8208 0.7447 0.6562
4 1.0000 0.9999 0.9996 0.9984 0.9954 0.9891 0.9777 0.9590 0.9308 0.8906
5 1.0000 1.0000 1.0000 0.9999 0.9998 0.9993 0.9982 0.9959 0.9917 0.9844
7 0 0.6983 0.4783 0.3206 0.2097 0.1335 0.0824 0.0490 0.0280 0.0152 0.0078
1 0.9556 0.8503 0.7166 0.5767 0.4449 0.3294 0.2338 0.1586 0.1024 0.0625
2 0.9962 0.9743 0.9262 0.8520 0.7564 0.6471 0.5323 0.4199 0.3164 0.2266
3 0.9998 0.9973 0.9879 0.9667 0.9294 0.8740 0.8002 0.7102 0.6083 0.5000
4 1.0000 0.9998 0.9988 0.9953 0.9871 0.9712 0.9444 0.9037 0.8471 0.7734
5 1.0000 1.0000 0.9999 0.9996 0.9987 0.9962 0.9910 0.9812 0.9643 0.9375
6 1.0000 1.0000 1.0000 1.0000 0.9999 0.9998 0.9994 0.9984 0.9963 0.9922
8 0 0.6634 0.4305 0.2725 0.1678 0.1001 0.0576 0.0319 0.0168 0.0084 0.0039
1 0.9428 0.8131 0.6572 0.5033 0.3671 0.2553 0.1691 0.1064 0.0632 0.0352
2 0.9942 0.9619 0.8948 0.7969 0.6785 0.5518 0.4278 0.3154 0.2201 0.1445
3 0.9996 0.9950 0.9786 0.9437 0.8862 0.8059 0.7064 0.5941 0.4770 0.3633
4 1.0000 0.9996 0.9971 0.9896 0.9727 0.9420 0.8939 0.8263 0.7396 0.6367
5 1.0000 1.0000 0.9998 0.9988 0.9958 0.9887 0.9747 0.9502 0.9115 0.8555
6 1.0000 1.0000 1.0000 0.9999 0.9996 0.9987 0.9964 0.9915 0.9819 0.9648
7 1.0000 1.0000 1.0000 1.0000 1.0000 0.9999 0.9998 0.9993 0.9983 0.9961
9 0 0.6302 0.3874 0.2316 0.1342 0.0751 0.0404 0.0207 0.0101 0.0046 0.0020
1 0.9288 0.7748 0.5995 0.4362 0.3003 0.1960 0.1211 0.0705 0.0385 0.0195
2 0.9916 0.9470 0.8591 0.7382 0.6007 0.4628 0.3373 0.2318 0.1495 0.0898
3 0.9994 0.9917 0.9661 0.9144 0.8343 0.7297 0.6089 0.4826 0.3614 0.2539
4 1.0000 0.9991 0.9944 0.9804 0.9511 0.9012 0.8283 0.7334 0.6214 0.5000
5 1.0000 0.9999 0.9994 0.9969 0.9900 0.9747 0.9464 0.9006 0.8342 0.7461
6 1.0000 1.0000 1.0000 0.9997 0.9987 0.9957 0.9888 0.9750 0.9502 0.9102
7 1.0000 1.0000 1.0000 1.0000 0.9999 0.9996 0.9986 0.9962 0.9909 0.9805
8 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 0.9999 0.9997 0.9992 0.9980
10 0 0.5987 0.3487 0.1969 0.1074 0.0563 0.0282 0.0135 0.0060 0.0025 0.0010
1 0.9139 0.7361 0.5443 0.3758 0.2440 0.1493 0.0860 0.0464 0.0232 0.0107
2 0.9885 0.9298 0.8202 0.6778 0.5256 0.3828 0.2616 0.1673 0.0996 0.0547
3 0.9990 0.9872 0.9500 0.8791 0.7759 0.6496 0.5138 0.3823 0.2660 0.1719
4 0.9999 0.9984 0.9901 0.9672 0.9219 0.8497 0.7515 0.6331 0.5044 0.3770
5 1.0000 0.9999 0.9986 0.9936 0.9803 0.9527 0.9051 0.8338 0.7384 0.6230
6 1.0000 1.0000 0.9999 0.9991 0.9965 0.9894 0.9740 0.9452 0.8980 0.8281
7 1.0000 1.0000 1.0000 0.9999 0.9996 0.9984 0.9952 0.9877 0.9726 0.9453
8 1.0000 1.0000 1.0000 1.0000 1.0000 0.9999 0.9995 0.9983 0.9955 0.9893
9 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 0.9999 0.9997 0.9990
VILLANOVA UNIVERSITY XII-7 LEAN SIX SIGMA HANDBOOK

Table V t Distribution
d.f. t o.100 t o.050* t o.025** t o.010 t o.005 d.f.

1 3.078 6.314 12.706 31.821 63.1)57 1
2 1.886 2.920 4.303 6.965 9.925 2
3 1.638 2.353 3.182 4.541 5.841 3
4 1.533 2.132 2.776 3.747 4.604 4
5 1.476 2.015 2.571 3.365 4.032 5
6 1.440 1.943 2.447 3.143 3.707 6
7 1.415 1.895 2.365 2.998 3.499 7
8 1.397 1.860 2.306 2.896 3.355 8
9 1.383 1.833 2.262 2.821 3.250 9
10 1.372 1.812 2.228 2.764 3.169 10
11 1.363 1.796 2.201 2.718 3.106 11
12 1.356 1.782 2.179 2.681 3.055 12
13 1.350 1.771 2.160 2.650 3.012 13
14 1.345 1.761 2.145 2.624 2.977 14
15 1.341 1.753 2.131 2.602 2.947 15
16 1.337 1.746 2.120 2.583 2.921 16
17 1.333 1.740 2.110 2.567 2.898 17
18 1.330 1.734 2.101 2.552 2.878 18
19 1.328 1.729 2.093 2.539 2.861 19
20 1.325 1.725 2.086 2.528 2.845 20
21 1.323 1.721 2.080 2.518 2.831 21
22 1.321 1.717 2.074 2.508 2.819 22
23 1.319 1.714 2.069 2.500 2.807 23
24 1.318 1.711 2.064 2.492 2.7~97 24
25 1.316 1.708 2.060 2.485 2.7:87 25
26 1.315 1.706 2.056 2.479 2.7'79 26
27 1.314 1.703 2.052 2.473 2.7'71 27
28 1.313 1.701 2.048 2.467 2.763 28
29 1.311 1.699 2.045 2.462 2.7S6 29
info 1.282 1.645 1.960 2.326 2.5'76 info
There is only a 5% probability that a sample with 10 degrees of freedom will have
a t value greater than 1.812.
* one tail 5% a risk ** two tail 5% a risk
VILLANOVA UNIVERSITY XII·8 LEJ!\N SIX SIGMA HANDBOOK

Table VI
Critical Values of the Chi-Square (X2) Distribution
I OF I 2
X O•99
I 2
X 0.9S
I 2
X 0.90
I 2
X 0.10
I X20.os
I 2
X O•01
I
1 0.00016 0.0039 0.0158 2.71 3.84 6.63
2 0.0201 0.1026 0.2107 4.61 5.99 9.21
3 0.115 0.352 0.584 6.25 7.81 11.34
4 0.297 0.711 1.064 7.78 9.49 13.28
5 0.554 1.15 1.61 9.24 11.07 15.09
6 0.872 1.64 2.20 10.64 12.59 16.81
7 1.24 2.17 2.83 12.02 14.07 18.48
8 1.65 2.73 3.49 13.36 15.51 20.09
9 2.09 3.33 4.17 14.68 16.92 21.67
10 2.56 3.94 4.87 15.99 18.31 23.21
11 3.05 4.57 5.58 17.28 19.68 24.73
12 3.57 5.23 6.30 18.55 21.03 26.22
13 4.11 5.89 7.04 19.81 22.36 27.69
14 4.66 6.57 7.79 21.06 23.68 29.14
15 5.23 7.26 8.55 22.31 25.00 30.58
16 5.81 7.96 9.31 23.54 26.30 32.00
18 7.01 9.39 10.86 25.99 28.87 34.81
20 8.26 10.85 12.44 28.41 31.41 37.57
24 10.86 13.85 15.66 33.20 36.42 42.98
30 14.95 18.49 20.60 40.26 43.77 50.89
40 22.16 26.51 29.05 51.81 55.76 63.69
60 37.48 43.19 46.46 74.40 79.08 88.38
120 86.92 95.70 100.62 140.23 146.57 158.95
The above table addresses both the left and right tails of X2. 950/0 of the area under
the X 2 distribution lies to the right of X 20.9S ' Example: for a right tail evaluation with
D.F. = 5, only 5 % of the values will exceed the critical value of 11.07.
VILLANOVA UNIVERSITY XII·9 LEAN SIX SIGMA HANDBOOK

f(F)
Table VII
Distribution of F a
Fa
F Table a = 0.05
v1(DF)
v2(DF) 1 2 3 4 5 6 7 8 9 10 12 15
1 161.4 199.5 215.7 224.6 230.2 234.0 236.8 238.9 240 . 5 241.9 243.9 245.9
2 18.51 19.00 19.16 19.25 19.30 19.33 19.35 19.37 19.:S8 19.40 19.41 19.43
3 10.13 9.55 9.28 9.12 9.01 8.94 8.89 8.85 8.S11 8.79 8.74 8.70
4 7.71 6.94 6.59 6.39 6.26 6.16 6.09 6.04 6.00 5.96 5.91 5.86
5 6.61 5.79 5.41 5.19 5.05 4.95 4.88 4.82 4.7'1 4.74 4.68 4.62
6 5.99 5.14 4.76 4.53 4.39 4.28 4.21 4.15 4.10 4.06 4.00 3.94
7 5.59 4.74 4.35 4.12 3.97 3.87 3.79 3.73 3.6i8 3.64 3.57 3.51
8 5.32 4.46 4.07 3.84 3.69 3.58 3.50 3.44 3.3,9 3.35 3.28 3.22
9 5.12 4.26 3.86 3.63 3.48 3.37 3.29 3.23 3.18 3.14 3.07 3.01
10 4.96 4.10 3.71 3.48 3.33 3.22 3.14 3.07 3.02 2.98 2.91 2.85
11 4.84 3.98 3.59 3.36 3.20 3.09 3.01 2.95 2.90 2.85 2.79 2.72
12 4.75 3.89 3.49 3.26 3.11 3.00 2.91 2.85 2.80 2.75 2.69 2.62
13 4.67 3.81 3.41 3.18 3.03 2.92 2.83 2.77 2.71 2.67 2.60 2.53
14 4.60 3.74 3.34 3.11 2.96 2.85 2.76 2.70 2.65 2.60 2.53 2.46
15 4.54 3.68 3.29 3.06 2.90 2.79 2.71 2.64 2.59 2.54 2.48 2.40
v 1(DF)
v2 (DF) 20 30 40 50 60 00
20 2.12 2.04 1.99 1.96 1.95 1.84

30 1.93 1.84 1.79 1.76 1.74 1.62
40 1.84 1.74 1.69 1.66 1.64 1.51
50 1.78 1.69 1.63 1.60 1.58 1.44
60 1.75 1.65 1.59 1.56 1.53 1.39
00 1.57 1.46 1.39 1.35 1.32 1.00
VILLANOVA UNIVERSITY XII - 10 LEJI.N SIX SIGMA HANDBOOK

f(F)
Table VIII
Distribution of F a
Fa
F Table a =0.025
v1(DF)
v2(DF) 1 2 3 4 5 6 7 8 9 10 12 15
1 647.8 799.5 864.2 899.6 921.8 937.1 948.2 956.7 963.3 968.6 976.7 984.9
2 38.51 39.00 39.17 39.25 39.30 39.33 39.36 39.37 39.39 39.40 39.41 39.43
3 17.44 16.04 15.44 15.10 14.88 14.73 14.62 14.54 14.47 14.42 14.34 14.25
4 12.22 10.65 9.98 9.60 9.36 9.20 9.07 8.98 8.90 8.84 8.75 8.66
5 10.01 8.43 7.76 7.39 7.15 6.98 6.85 6.76 6.68 6.62 6.52 6.43
6 8.81 7.26 6.60 6.23 5.99 5.82 5.70 5.60 5.52 5.46 5.37 5.27
7 8.07 6.54 5.89 5.52 5.29 5.12 4.99 4.90 4.82 4.76 4.67 4.57
8 7.57 6.06 5.42 5.05 4.82 4.65 4.53 4.43 4.36 4.30 4.20 4.10
9 7.21 5.71 5.08 4.72 4.48 4.32 4.20 4.10 4.03 3.96 3.87 3.77
10 6.94 5.46 4.83 4.47 4.24 4.07 3.95 3.85 3.78 3.72 3.62 3.52
11 6.72 5.26 4.63 4.28 4.04 3.88 3.76 3.66 3.59 3.53 3.43 3.33
12 6.55 5.10 4.47 4.12 3.89 3.73 3.61 3.51 3.44 3.37 3.28 3.18
13 6.41 4.97 4.35 4.00 3.77 3.60 3.48 3.39 3.31 3.25 3.15 3.05
14 6.30 4.86 4.24 3.89 3.66 3.50 3.38 3.29 3.21 3.15 3.05 2.95
15 6.20 4.77 4.15 3.80 3.58 3.41 3.29 3.20 3.12 3.06 2.96 2.86
v1(DF)
v2 (DF) 20 30 40 50 60 00
20 2.46 2.35 2.29 2.25 2.22 2.09

30 2.20 2.07 2.01 1.97 1.94 1.79
40 2.07 1.94 1.88 1.83 1.80 1.64
50 1.99 1.87 1.80 1.76 1.72 1.55
60 1.94 1.82 1.74 1.70 1.67 1.48
00 1.71 1.57 1.48 1.43 1.39 1.00

Table IX
Control Chart Factors
CHART FOR CHART FOR STANDARD

CHART FOR RANGES
AVERAGES DEVIATIONS
Sample Center Center

Control limit Control Limit Control Limit
Observations Line Line
Factors Factors Factors
Factors Factors
n A2 A3 C4 B3 B4 d2 03 04
2 1.880 2.659 0.7979 0 3.267 1.128 0 3.267

3 1.023 1.954 0.8862 0 2.568 1.693 0 2.574
4 0.729 1.628 0.9213 0 2.266 2.059 0 2.282
5 0.577 1.427 0.9400 0 2.089 2.326 0 2.114
6 0.483 1.287 0.9515 0.030 1.970 2.534 0 2.004
7 0.419 1.182 0.9594 0.118 1.882 2.704 0.076 1.924
8 0.373 1.099 0.9650 0.185 1.815 2.847 0.136 1.864
9 0.337 1.032 0.9693 0.239 1.761 2.970 0.184 1.816
10 0.308 0.975 0.9727 0.284 1.716 3.078 0.223 1.777
15 0.223 0.789 0.9823 0.428 1.572 3.472 0.347 1.653
20 0.180 0.680 0.9869 0.510 1.490 3.735 0.415 1.585
25 0.153 0.606 0.9896 0.565 1.435 3.931 0.459 1.541
The above table is extracted from Table 27 in ASTM Manual on P"esentation of Data
and Control Chart Analysis (1976). ASTM Publication STP15D. American Society for
Testing and Materials, Philadelphia, pp. 134-135. Used with permission.
-
X - R Charts X - S Charts
CLx == X :I: ~R Cli =X ± AaS
UCl R = 1~4S
lCl R =
-
1~3S
Approximate capability Approximate capability

XII. APPENDIX· AUTHOR/NAME INDEX
LSS Author/Name Index

Adams, S. 1-2 Garro, E. Intro-4, XI-6
Adriano, S. Intro-3 Garvin, D. V-51
Akao, Y. 111-24 Geddes, L.IV46
Albrecht, K. III-53, IV46 Gee, G.lntr04, 1149, IV46, V-51, VI-59
Altshuller, G. X-30, X-37 George, M.L. 1-2, 11-12, 11-21, 1149, III-53, VI-59, VIII-70
Alukal, G. VIII-70 Gleason, G. III-53, IV-47
Anderson, P.F. V-51 Goetsch, D.L. III-53, VIII-70, IX-61
Ashley, S. X-37 Goldratt, E. 1-2, VIII-29, VIII-70
Auvine, B. IV-46 Gordon, J.R. IV46
Ballis, J.I-2 Gosset, W.S. VII-38
Barney, M. 1148 Grant, E.L. VI-59, IX-61
Bateman, T.S. IV46 Griffin, A. III-53, IV47
Beitz, W. X-38 Gryna, F.M. II-50, III-54, VI-59, X-38
Belbin, M. IV-22 Hahn, G.J. 1149, VIII-70
Besterfield, D.H. 1148, III-53, IV46, V-51, IX-61 Hamel, G. X-38
Bogan, C. III-53 Harold, D. X-38
Bongiorno, J. X-37 Harry, M.J. 1-3, 11-7, 11-22, 11-38, 1148, 1149, III-54, IV-47, VI-59
Boothroyd, G. X-37 Hashimoto, T. V-14, V-51
Bralla, J. X-37 Hauser, J.R. III-54, IV-47, X-38
Brenton, T. Intro-3 Hayslip, W.R. III-54, IV-47
Breyfogle, F.W., III 1-2, 11-22, 11-38, 11-48, II-52, III-53, IV46, VI- Heskett, J.L. IV-47
59,IX-6i Hill, T. V-51
Brown, M.G. 11-48 HiII,W.II49
Burman, J.P. VIII-71 Hirano, H. 1-3
Bylinsky, G. 1148 Hockman, K. X-38
Carlson, D.R. IV-48, VI-60, VIII-72, IX-62 Hoerl, R. 11-49
Cavanagh, R.R. 1-3, II-51, III-54, IV48, V-52 HOisington, S. 1-3
Chadwick, G. 11-48 Huang, G. X-38
Clausing, D. 111-24, III-54, IV47, X-38 Huck, S. VII-60
Conner, G. V-51, VI-59, VIII-70 Hunter, M.R. III-54, IV-47
Coombs, C.F. X-38 Imai, M. 1-3, V-i, V-51, VII-60, VIII-70, IX-61
Cooper, R. X-37 Ireson, W.G. X-38
Cormier, W. VII-60 Ishikawa, K. 11-22, 11-30, 11-31, 1149, IV-47, V-14, V-51
Cox, J. VIII-70 Izadi, M. V-51, VI-59
Crabtree, R. Intr04, 11-3, 11-4, 1148 Janis, I. IV-47
Crawford, C. X-37 Jiro, K. V-11
Crosby, P.B. 11-22-11-24, 1148 Joiner, B.L. IV-48
Cross, N. X-37 Jones, D.T. 1-3, 11-12, 11-20, 1149, II-52, V-52, VIII-72
Cutler, A.N. 1148 Jones, T.O. IV47
Davidson, P. VIII-70 Juran, J.M. 1-3, 11-22, 11-32, 11-33, II-50, 111-1, III-54, IV47, V-iS,
Davis, S.B. III-53, VIII-70, IX-61 V-51, VI-60, VII-60
DeCarlo, N. 11-48 Kanigel, R. II-50
Delahaye, B. IX-62 Kano, N. V-23
Delavigne, K.T. 1149 Kaplan, R.S. 1-3, II-50
Delbecq, A.IV40 Kastle, B. III-53
Deming, W.E. 1-2, 11-22, 11-25-11-27, 1149, III-53, IV46, IX-1 Kerzner, H. III-54
Dennis, P. 1-3 Kirkpatrick, D. IX-61
Densmore, B. IV46 Kleiner, A. II-50
Dewhurst, P. X-37 Knight, W. X-37
Dovich, R.A. X-37 Knutzen, B. X-39
Eckes, G. III-53, V-51 Kobayashi, I. 1-3
Edenborough, N.B. V-51, VI-59 Kolsar, P.J. II-50
Eitington, J. IV-46 Kudlacik, E. X-40
Emiliani, M. 11-49 Kuyatt, C.E. VIII-72
Endres, A. X-38 Langley, G. X-39
English, M. III-53 Langley, R. VII-60
Extrom, M. 1V-46 Lao-Tzu 1-1
Feigenbaum, A.V. 11-22, 11-28, 11-29, 1149, III-53, IV-46 Launsby, R.G. VI-60
Ford, H.J. 11-12, 11-14, II-iS, 1149, X-1 Leavenworth, R.S. VI-59, IX-61
Frank, B. III-53 Lennon, J. Intro-3
Franklin, L.A. VI-60, VII-60 Levinson, W.A. 1-3
Furlong, C.B. 1V-46 Liker, J. II-50, V-51, VIII-71 , IX-61
Futrell, D. III-53, IV46 Little, J.D.C. VI-9
Gaebler, T. IV48 Loog, K.IV-47
Galvin, R.W. 11-7, 11-37, 11-38 Lorber IV-47
Gantt, H. 11144 Lorenz, M.O. V-iS
ViLLANOVA UNIVERSITY XII-13 LEAN SIX SIGMA HANDBOOK

XII. APPENDIX - AUTHOR/NAME INDEX
Loveman, G.W. IV-47 Schenk, V.L. IX-S1

Lowenstein, M.W. IV-47 Schermerhorn, J.R., Jr. IV-48, IX-62
MaGuire, M. II-50 Schlesinger, L. IV-47
Mak, K. X-38 Schmidt, S.R. VI-SO
Manos A. VIII-70 Scholtes, P.R. IV-48
Marriott, P. III-53 Schroeder, R. 1-3, 11-38, 11-49, 1II-~i4, IV-47, VI-59
Martin, P. III-54 Seneca 11-1
Maxey, J. VI-59 Senge, P.M. II-51
May, M. III-54, V-51 Shainin, D. IX-38
McCartney, P. Intro-3 Shanklin, M. IV-4S
McGrath, P. V-51, VI-59 Shapiro, S.S. VIII-70
Meier, D. V-51 Sharma, A. 1-3, 11-12, 11-20, 11-21, VI-SO, VII-SO, VIII-71 , IX-S2
Metcalf, M.D. Intro-5, VII-SO Shaw, V. Intro-3
Miller, W.B. IX-61 Shevenaugh, D. III-53, IV-47
Mizuno, S. II-50 Shewhart, W.A. 11-22, 11-34, 11-35, II-51, IX-11, IX-62
Moen, R. V-52 Shina, S. X-39
Moody, P.E. 1-3, VI-SO, VII-SO, VIII-71 , IX-S2 Shingijutsu Co., Ltd. 11-21
Mora, O. Intro-5 Shlngo, S. 1-3, 11-12, 11-19, II-51, VIII-71 , VIII-72, IX-S2
Moss, R. X-38 Shook, J. 1-3, II-51, V-52, VI-SO
Nakajima, S. IX-61, IX-S2 Simmerman, S.J. IV-48
Naumann, E. 1-3 Simon, K. X-39
Neuman, P.R. 1-3, II-51, 111-54, IV-48, V-52 Skaggs, T. VIII-72
Nierenberg, G. IV-47 Smith, B. 11-7,11-22, 11-37, IX-62
Noble, T. 11-5 Smith, G.M. VI-SO
Nolan, T. V-52 Snee, R.D. II-52, III-54, IV-48
Norton, D.P. II-50 Sobek, D.K., II. V-52, XI-32
Ohno, T. 11-2, 11-12, 11-17, 11-18, II-50, VIII-1S Stamatls, D. X-39
Omdahl, T.II-51, V-52 Stein, P.G. VI-60
Osada, T. VIII-71 Stengel, C. IV-1
Osborn, A.F. X-33, X-38 Stoecklein, M. X-39
Osborne, D. IV-48 Streibel, B.J. IV-48
Pahl, G. X-38 Suri, R. VIII-27, VIII-72
Pande, P.S. 1-3, II-51, III-54, IV-48, V-52 Suzaki, K. VIII-72, IX-62
Paradise, T.R. Intro-5 Swanson, H.S. VIII-1
Pareto, V. V-15 Taguchi, G. 11-22, 11-35-11-37, II-52, X-14-X-1S, X-39
Pearson, T.A. II-51, IV-48 Tague, N.R. 1-3
Phadke, M.S. X-38 Tate, K. III-54
Phadnis, S. X-38 Taylor, B.N. VIII-72
Pitcher, M. VIII-71 Taylor, F.W. 11-12, 11-13, II-52
Plackett, R.L. VIII-71 Thomas, K. IV-48
Poole, S. IV-4S Tjosvold, D. IV-48
Pope, A. XII-1 Toyoda, E. 11-12, 11-17, II-52
Porter, M. V-52 Toyoda, K. 11-12, 11-1S
Potts, M. II-51 Toyoda, S. 11-12, 11-15, 11-1S
Prahalad, C. X-38 Treffs, D. X-39
Preiss, R. III-53, IV-47 Triola, M.F. VI-SO, VII-SO
Price, M. VI-59 Tuckman, B.W.IV-48
Provost, L. V-52, X-39 Van deVen, A. IV-40
Rabeneck, L. X-39 Van Landingham, R.D. III-54, IV-47
Rabeneck, S. X-39 Vandervort, C. X-40
Radcliffe, D. X-40 Wald, A. VI-SO
Ramberg, J. II-51 Warzusen, C. III-53
Reid, R. X-39 Watson, B. X-40
Reiling, J. X-39 Watt, W.W. VII-1
Rerick, R.A. 1-3 Wheeler, D.J. II-52
Richardson, W.R. Intro-S, 11-49, IV-4S, IV-48, V-51, VIII-72, IX- Whiteley, R.C. IV-48
S2 Wile, D. IX-S2
Robertson, J.D. 11-49 Wilkins, Jr., J. X-40
Robinson, A. V-52, VIII-71 Womack, J.P. 1-3, 11-12, 11-20, II-52, V-52, VIII-72
Robinson, J. IX-62 Woolsey, R.E.D. VIII-1
Roos, D. 1-3, II-52, V-52 Wortman, B.L. Intro-3, Intro-6, 1-2D, 11-49, II-52, IV-4S, IV-48,
Rosenau, M. X-39 V-51, VI-60, VII-60, VIII-72, 1)(-62, X-37
Rother, M. 1-3, II-51, V-52, VI-SO Wu, Y.II-52
Rowlands, D. III-53, VI-59 Yang, K. VIII-72
Rybarczyk, P. IX-62 Zeithaml, C.P.IV-4S
Sabatini, J. II-51 Zemke, R. IV-48
Sasser, Jr., W.E.S. IV-47 Zinkgraf, S. 11-49
Schaaf, D. IV-48
VILLANOVA UNIVERSITY XII -14 LEtlN SIX SIGMA HANDBOOK

XII. APPENDIX· SUBJECT INDEX
LSS Subject Index

1.5 sigma shift 1-19,11-7,11-8, VI-14, XII-3 Are the residuals well behaved? VIII-40
14 step approach to quality improvement 11-24 Arrow diagram 111-46
1FAT (one factor at a time) VIII-34 Assembly X-28
2 mean equal variance t test VII-51 Attribute
2 mean unequal variance t test VII-51 charts IX-23, IX-24
5 Whys 1-13, VII-59 charts, formulas IX-24
5 Ws and H (or 2 H) VII-59 data VI-15-VI-17
5-M and E example V-13 data, conversion to variables VI-17
55 data, process capability VI-58
defined 1-13 Attributions IV-31
deployment VIII-13 Autonomation 1-13, IX-49
follow-up VIII-13 Autonomous TPM small group activities IX-46
housekeeping and workplace organization VIII-2-VIII-6 Availability IX-43
implementation road map VIII-7 Average IX-13
launch VIII-9 and range method VI-32-VI-35
measuring the impact VIII-7 Avoiding IV-38, IV-39
operational level VIII-8-VIII-13
phlosophy level VIII-8 B
pre-launch VIII-8, VIII-9 Balanced design VIII-41
process level VIII-14, VIII-15 Balanced scorecard 11-44, 11-45
shine VIII-15 Bar charts 111-44, 111-45
shine (sweep) VIII-11 , VIII-12 example 111-45
sort VIII-14 Basic 111-22
standardize VIII-12 Basic concepts of hypothesis testing VII-24-VII-27
straighten VIII-11, VIII-14 Belbin team roles IV-22, IV-23
sustain VIII-12, VIII-13 Benchmarking 111-27-111-29
what 5S is not VIII-6 performance 111-27
80 problem solving 111-12 process 111-27
project 111-27
A strategic 111-27
A3 medical center improvement XI-27-XI-32 Benefit effort matrix 111-4
A3 report V-7-V-10 Beta error VII-25
Accommodating IV-38, IV-39 Bias VI-29
Accuracy VI-29 Binomial distribution IX-24, XII-7
Active learning IX-59 Black belts IV-15
Activity network diagrams (AND) 111-46, 111-47 Block VIII-41
Ad hoc teams IV-6 Blocking VIII-41
Adding new team members 1V-13 Box-Behnken VIII-41
Additional classes of wastes VII-5 Brainstorming IV-42, VIII-58
Additional team problem areas 1V-30 Buffer VIII-31, VIII-32
Adjourning IV-25 stock V-40
Adjust 111-11 Business
Aesthetics X-29 case V-4
Affinity diagrams V-11, V-12 level 111-18
example V-12 level metrics 11-46
Agenda 1V-12
Alpha error VII-25
Analysis IX-56
c
c chart IX-23, IX-24, IX-27,IX-28
Analysis of variance (AN OVA) VII-52-VII-57 example IX-27, IX-28
comparison of three or more means VII-52 formulas IX-24
method VI-32, VI-36-VI-38 crr (cycle time) V-38
one-way VII-53-VII-55 Calculating an estimate of al. VII-18
two-way VII-55-VII-57 Capability
Analysis techniques VII-2 Index VI-53
hypothesis testing VII-24 index failure rates VI-56
references VII-60 Index with process centering VI-53
root cause analysis VII-58 post-improvement analysis VIII-59
seven classical wastes VII-2 ratio VI-53
variable relationships VII-6 short-term versus long-term VI-57
Analyze 11-9,111-10, X-8 Case study
AND (activity network diagrams) 111-46, 111-47 insurance claim flow improvement XI-20-XI-26
Andon board 1-13 lean six sigma implementation at Ludovico XI-11-XI-19
ANOVA (analysis of variance) 1-13, VII-52-VII-57 medical center improvement using A3 problem solving
comparison of three or more means VII-52 XI-27-XI-32
method VI-36-VI-38 multi-vari VII-10, VII-11
one-way VII-53-VII-55 QCI spaghetti studies XI-2-XI-5
two-way VII-55-VII-57 reference XI-32
Appearance X-29 setup reduction VIII-24
Appropriate vehicles for different situations IV-34 software license Improvement at Exactus XI-6-XI-10
VILLANOVA UNIVERSITY XII ·15 LEAN SIX SIGMA HANDBOOK

XII. APPENDIX - SUBJECT INDEX
Catastrophic X-23 references IX-61, IX-62

Cause·and-effect diagrams V-13, V-14 standard work IX-50
example V-14 total productive maintenance IX-40
Cellular manufacturing 1-13 training requirements IX-54
CFM (continuous flow manufacturing) VIII-25, VIII-26 visual systems IX-47
defined 1-13 Controlling 111-33
Champions IV-16 Conversion of attribute data to variables measures VI-17
Changeover time (COT) V-38 Coordinator IV-22, IV-23
Charter negotiation V-5 Correlation coefficient (r) VII-19, VIII-42
Check sheets VI-20, VI-21 Correlation example VII-20, VII-21
recording VI-20, VI-21 COT (changeover time) V-38
Checklists VI-21 Covariates VIII-42
Chi-square distribution XII-9 Cp VI-53, VI-54
Chi-square test VII-42-VII-47, VII-51 Cpk VI-53, VI-54
case I VII-44 Cpm VI-54
case II VII-45 CPM (critical path method) 111-40··111-43
Classic team problem solving steps 111-8, 111-9 defined 1-13
Coding, data VI-19 example 111-41-111-43
Coefficient of determination (r2) VII-20 CrVI-53
Collaborating IV-38, IV-39 Crash an activity 111-40-111-43
Collecting customer data 111-18 Creative design ideas X-33, X-34
Collection methods VI-18-VI-21 Critical X-23
Collinear VIII-41 Critical path 111-36, 111-37, 111-39-111·43, 111-46
Common team problems IV-31 Critical path method (CPM) 111-40-111-43
Communication defined 1-13
downward flow IV-32 example 111-41-111-43
formal and informal IV-34 Critical to quality (CTQ) tree 1-14,. V-19, V-20
horizontal IV-33 Crosby, P.B. 11-23, 11-24
other forms IV-34, IV-35 14 step approach to quality improvement 11-24
techniques IV-32-IV-36 four absolutes of quality management 11-24
upward flow IV-33 Cross functional teams IV-7, IV-SI
Company team benefits IV-2, IV-3 CTQ (critical to quality) tree 1-14, V-19, V-20
Companywide quality control (CWQC) 11-30, 11-31 Current state map V-41
Comparison of three or more means VII-52-VII-55 Current state mapping V-38
Competing IV-38, IV-39 Curvature VIII-42
Competitive shopper 111-19 Customers 111-13, 111-17
Completer IV-22, IV-23 collecting data 111-18
Completion rate VI-9, VI-10 data analysis V-21 , V-22
Compromising IV-38, IV-39 expectations 111-22
Concept design X-17 needs 111-23
Confidence interval service measurement 111-21
for proportion VII-34 surveys 111-20
for standard deviation VII-35 CWQC (companywide quality cOlltrol) 11-30, 11-31
for the mean VII-33, VII-34 Cycle time 11-42, V-30-V-35, V-38, V-40, IX-52
for variance VII-35 calculation VI-10
of 1-alpha for beta1 VII-18 defined 1-14, V-30
Conflict resolution IV-37-IV-39 reduction V-30-V-35
Confounded VIII-41 training V-30
Connector V-44 Cyclical variation VII-9
Continuous data sigma known VII-33
Continuous data sigma unknown VII-33, VII-34 D
Continuous flow manufacturing (CFM) VIII-25, VIII-26 Data
defined 1-13 accuracy VI-24, VI-25
Control 11-9, 111-10 analysis, customer V-21 , V-22
chart analysis IX-31 attribute VI-15
chart factors XII-12 attribute, process capability VI-58
chart interpretation IX-34 boxV-40
charts V-21 , IX-11 box, value stream map V-47
charts for attributes IX-12 coding VI-19
charts for variables IX-12 collecting customer 111-18
factors X-14, X-15 collection VI-15
plan input sources IX-7 collection methods VI-18-VI-2'1
plan line items IX-8, IX-9 conversion of attribute to variables VI-17
plan organization IX-6 examples VI-23
plan summary IX-10 flow chart symbol V-44
plan, illustrative IX-10 instruments to gather 111-19
plans IX-5 locational VI-16
plans, types of IX-5 types of VI-15-VI-17
Control concepts variables VI-15, VI-16
control charts IX-11 Decision V-44
control plans IX-5 Defect relationships VI-11, VI-12
quality controls IX-2 Defectives IX-24

Defects IX-24 DFA (design for assembly) X-28

Defects per million opportunities (DPMO) VI-ii, VI-12 DFM (design for manufacturabllity) X-27
Defects per normalized unit VI-11 DFSS (design for six sigma) 1-14, X-2-X-10
Defects per unit (DPU) VI-11-VI-13 DFX (design for X) X-25-X-29
Defects per unit opportunity (DPO) VI-ii, VI-12 characteristics X-27-X-29
Define 11-9, 111-10, V-2, X-8 defined 1-14
Define product family V-37 techniques and tools X-25, X-26
Define, measure, analyze, design, verify (DMADV) X-8 Discounts IV-31
Define, measure, analyze, improve, control (DMAIC) 111-10 Dissatisfaction sources 111-19
Defining opportunities Dissatisfiers V-23
A3 reportV-7 DMADV (define, measure, analyze, deSign, verify) X-8
cycle time reduction V-30 DMAIC (define, measure, analyze, improve, control) 11-9, 111-
flow charts, process mapping V-43 10
lean thinking V-24 defined 1-14, 11-9
metrics selection V-6 Document archiving III-52
problem definition tools V-11 DOE (design of experiments) VIII-34-VIII-57
process mapping versus VSM V-46 applications VIII-35
project charter V-2 defined VIII-42
references V-51, V-52 Graeco-Latin designs VIII-51
spaghetti diagrams V-49 hyper-Graeco-Latin designs VIII-51
value stream mapping V-36 improvement analysis VIII-62-VIII-67
voice ofthe customerV-19 introduction VIII-34, VIII-35
Degrees of freedom VIII-42 iterative approach VIII-39
DelayV-44 Latin square designs VIII·50
Delighters V-23 Plackett-Burman designs VIII-52
Deliverables 111-34 randomized block plans V111-49
Deming, W.E. 11-25-11-27 steps VIII-36
chain reaction 11-27 terms VIII-41-VIII-47
fourteen obligations of top management 11-26 three factor, three level experiment VIII-48
seven deadly diseases that management must cure 11-26 two·level fractional factorial example VIII-53
Deployment VIII-13 typical checklist VIII-36, VIII-37
Description of control plan line items IX-8, IX-9 Dominant participants IV-31
Design 111-11, X-7, X-8 Downtime IX-43
FMEAX-20 Downward flow of communication IV-32
FMEA example X-23 DPMO (defects per million opportunities) 1·14, VI-ii, VI·12
FMEA process steps X-22 DPO (defects per unit opportunity) 1-14, VI-ii, VI·12
for six sigma, see DFSS X-2 DPU (defects per unit) VI-11-VI-13
using quality function deployment X-11-X-13 defined 1-14
Design for assembly (DFA) X-28 DriftVI-30
Design for manufacturability (DFM) X-27 Drum-buffer-rope VIII-31, VIII-32
Design for six sigma (DFSS) X-2-X-10 Dummy node 111-46
Design for X (DFX) X-25-X-29
characteristics X-27-X-29 E
techniques and tools X-25, X-26 Eight disciplines 111·12
Design improvement Eight symptoms of groupthlnk IV-29
creative design ideas X-33 Electronic flow V-40
deSign for six sigma X-2 Eliminating wastes VIII·2
design for X X-25 Enterprise resource planning (ERP) 1-14
failure mode and effects analysis X-18 Environment X·28
quality function deployment X-11 Environmental factors IV-44
references X-37-X-40 Equipment variation VI-28
robust design X-14 Ergonomics X-29
systematic design X-32 ERP (enterprise resource planning) 1·14
TRIZX-30 Error-proofing 1-14
Design of experiments (DOE) VIII-34-VIII-57 Errors VII·25
applications VIII-35 ES (external setup) VIII-22, VIII-23
defined VIII-42 Evaluator IV-22, IV-23
Graeco-Latin designs VIII-51 Evolutionary operations (EVOP) VIII-61
hyper-Graeco-Latin designs VIII-51 EVOP (evolutionary operations) 1-14, VIII-42, VIII-61
improvement analysis VIII-62-VIII-67 Execute 111·11
introduction VIII-34, VIII-35 Executive sponsors IV·16
iterative approach VIII-39 Expected 111-22
Latin square designs VIII-50 Experiment VIII-42
Plackett-Burman designs VIII-52 Experimental
randomized block plans VIII-49 assumptions VIII-39, VIII-40
steps VIII-36 error VIII-42
terms VIII-41-VIii-47 objectives VIii-38
three factor, three level experiment VIII-48 External performer factors IX·55
two-level fractional factorial example VIII-53 External setup (ES) VIII-22, VIII·23
typical checklist VIII-36, VIII-37 Extract V-44
Desired 111-22
Developing sampling plans VI-48 F

F distribution XII-10, XII-11 Goals 11-42

F test VII-48-VII-51 metrics 11-42, 11-43
Face-to-face interviews 111-19 statement 111-34
Facilitation IV-19 Graeco-Latin designs VIII-51
Facilitator IV-10, IV-12 Green belts IV-16
Failure mode and effects analysis (FMEA) X-18-X-23 Group accomplishments IV-12
design X-20 Groupthink IV-29
design example X-23
design process steps X-22 H
follow-up X-21 Hansei 111-11
process X-20 Heijunka 1-15
service X-21 Hints on regression analysis VII-17
steps and benefits X-18, X-19 Histograms VI-39, VI-40
system X-19 bimodal VI-39
types X-19 construction example VI-40
Failure rates XII-3 negatively skewed VI-39
Features X-29 special causes VI-39
Feedback IX-59 truncated VI-39
Feuding IV-31 Horizontal communications IV-3:1
FIFOV-40 Hoshin planning 1-15
Final report III-52 House of quality 111-24-111-26, X-11-X-13
Finished goods movement V-40 Housekeeping and workplace or!;Janization VIII-2
Finisher IV-23 Hyper-Graeco-Latin designs VIII-51
First order linear model VII-21 Hypothesis
First pass yield (FPy) VI-13 basic concepts VII-24-VII-27
First-order VIII-43 chi-square tests VII-42-VII-47
Fishbone diagrams V-13, V-14 F tests VII-48-VII-50
Five common rules IX-34 null VII-24
Flexible manufacturing system (FMS) 1-14 paired-comparison tests VII-4D
Floundering IV-31 summary of inference tests VIII-51
Flow chart V-43-V-45 testing VII-24
example V-45 tests VII-30, VII-31
nonconforming material VI-4 tests for means VII-36-VII-39
Flow time VI-9 tests for proportions VII-41
FMEA (failure mode and effects analysis) VIII-58, X-18-X-23
defined 1-14 I
design X-20 IDEA (investigate, design, execute, adjust) 111-11
design example X-23 Ideal final result (IFR) X-30
design process steps X-22 Identify X-7
follow-up X-21 Identify, design, optimize, and validate (IDOV) X-7, X-8
process X-20 Identifying
service X-21 characteristics VI-47
steps and benefits X-18, X-19 specifications or tolerances V 1-48
system X-19 IDOV (identify, design, optimize, and validate) X-7, X-8
types X-19 IFR (ideal final result) X-30
FMECA (failure mode, effects, and criticality analysis) 1-14, Illustration of batch versus one p,iece flow VI-8
X-18 Illustrative control plan IX-10
FMS (flexible manufacturing system) 1-14 Implementation planning V-39
Focus groups 111-19 Implementation road map, 5S VIII,·7
Follow-up VIII-13 Implementer IV-22, IV-23
Ford 80 problem solving 111-12 Implementing and validating solutions VIII-58
Formal and informal communications IV-34 Importance of top-down support IX-57
Forming IV-24-IV-27 Improve 11-9, 111-10
Four absolutes of quality management 11-24 Improvement teams IV-6, IV-9
Fourteen obligations of top management 11-26 Improvement techniques
FPY (first pass yield) 1-14, VI-13 design of experiments VIII-34
Fractional VIII-43 eliminating wastes VIII-2
Fractional factorial implementing and validating s,olutions VIII-58
design comparison VIII-66, VIII-67 measurement system re-anal~'sis VIII-68
two-level example VIII-53 references VIII-70-VIII-72
French design model X-10 theory of constraints VIII-29
Full factorial VIII-43 Improving the process VIII-2
Future state map V-39, V-42 Inference tests VII-51
Initiating teams IV-2
G Inner array VIII-44
Gantt charts 111-44, 111-45 Innovator IV-22, IV-23
defined 1-14 Input factor VIII-43
example 111-45 Inputs 111-13
Gasoline mileage experiment VIII-53-VIII-57 Instruments to gather data 111-19
Gemba 1-15 Insurance claim flow improvement XI-20-XI-26
Go see V-40 Interaction VIII-44
Goal factors IV-44 Intercept VII-13

Internal handbook contents Intro-7

customer voice 111-19 handbook question contents Intro-8
performer factors IX-55 methodology - define V-2
setup (IS) VIII-22, VIII-23 project management 111-2
Interval scale VI-22, VI-23 structure IV-17
Inventory V-40, VII-3 teams IV-6, IV-9
turns 1-15 value 11-2-11-6
Investigate 111-11 Lean six sigma goals
Investigate, design, execute, adjust (IDEA) 111-11 lean pioneers 11-12
Investigation IX-56 metrics and goals 11-42
Investigator IV-22, IV-23 organizational leadership 11-39
IS (internal setup) VIII-22, VIII-23 quality and six sigma gurus 11-22
Is the measurement system capable? VIII-39 references 11-48-11-52
Is the process stable? VIII-39 six sigma introduction 11-7
Ishikawa diagrams V-14 value of lean six sigma 11-2
Iterative approach to DOE VIII-39 Lean six sigma implementation at Ludovico XI-11-XI-19
Learning principles IX-59, IX-60
J Least squares VII-14-VII-17
Jidoka 1-15, IX-49 example VII-16, VII-17
JIT (just-in-time) 1-15 Lessons learned III-52
Jump to solutions IV-31 Leve1VIII-44
Jumps in process levellX-36 loading 1-16
Juran, J.M. 11-32, 11-33 Line balancing 1-16
Trilogy 11-33 Line graphs V-21
Just-in-time (JIT) 1-15 Linear model VII-12, VII-13
Linear regression, multiple VII-21
K Linearity VI-30
Kaizen V-25, V-30, V-32, V-35, VIII-28 Linking projects to organizational goals 111-2-111-4
blitz VIII-28 Listening strategies IV-36
burstV-40 Little's Law VI-9
defined 1-15 Load leveling V-40
eventV-32 Loading time IX-43
Kanban Locational data VI-16
batches V-40 Long-term versus short-term capability VI-57
defined 1-15 Lower control limit (LCL) IX-11, IX-13-IX-15, IX-17-IX-21, IX-
postV-40 24, IX-25, IX-27, IX-29, IX-30
production V-40 LSS project management
pull VIII-16, VIII-17 linking projects to organizational goals 111-2
signal V-40 project plan elements 111-33
withdrawal V-40 project selection 111-13
Kano analysis V-23 project selection process 111-5
Kano model 1-15, V-23 references III-53, III-54
KJ Method V-11 risk analysis and management 111-30
KPIV (key process input variables) 1-15
KPOV (key process output variables) 1-16 M
Magnetic disk or storage V-44
L Main effect VIII-44
L-type survey matrix 111-20 Maintainability X-28
UT (lead time) V-38 Management
Laboratory to laboratory variation VI-28 presentations IV-45
Lack of variability IX-37 support 11-39, IV-5
Latin square designs VIII-50 Manual information flow V-40
Launch VIII-9 Manual operation V-44
LCL (lower control limit) IX-11, IX-13-IX-15, IX-17-IX-21, IX- Manufacturability X-27
24, IX-25, IX-27, IX-29, IX-30 Marginal X-23
Lead time V-38, V-40 Marketplace response 11-43
Leader role IV-20 Master black belts IV-15
Lean Material variation VI-28
enterprise 11-11 Matrix diagrams V-22
enterprise, defined 1-16, 11-11 Mean
manufacturing 11-11 comparison of three or more VII-52-VII-55
manufacturing, defined 1-16 confidence interval VII-33, VII-34
metrics VI-9, VI-10 hypothesis tests for VII-36-VII-39
pioneers 11-12 point estimate VII-32
production 11-11 Meaningful material IX-60
techniques in service 11-11 Measure 11-9, 111-10, X-8
thinking V-24-V-29 Measurement
Lean six sigma error VI-27, VI-28
body of knowledge 1-4-1-12 project activity 111-48-111-51
glossary 1-13-1-20 scales VI-22, VI-23
goals 11-2 system analysis VI-26, VI-27
handbook bibliography 1-2, 1-3 system re-analysis VIII-68, VIII-69

systems VI-26 Non-value-added 1-17, VII-2

techniques VI-2 Nonconforming material flow chart VI-4
tenns VI-29-VI-31 Normal distribution VI-41-VI-44, XII-2
Measurement techniques Normalized yield VI-13
data collection VI-15 Norming IV-24-IV-27
measurement systems VI-26 np chart IX-23, IX-24, IX-29, IX-30
process analysis VI-2 example IX-29, IX-3~
process capability analysis VI-39 formulas IX-24
references VI-59, VI-60 NPV (net present value) 1-17
Measuring the impact of 5S VIII-7 Null hypothesis VII-24-VII-27, VII·36-VII-41, VII-43-VII-48, VII-
Median control charts IX-18 50, VII-54-VII-57
Medical center improvement using A3 problem solving XI-
27-XI-32
Meeting
o
OEE (overall equipment effectivlmess) IX-43, IX-45
attendance log IV-12 OFAT (one factor at a time) VIII-~:4
sample fonns IV-12 Offset VI-29
structure IV-11 One piece flow 1-17, VI-8
Members One-tail test VII-26
adding new IV-13 One-way ANOVA VII-53-VII-55
removing IV-13 Operating speed rate IX-43, IX-44
role IV-21 Operation time IX-43
selecting IV-13 Operational level, 5S VIII-8-VIII-1:3
Merge V-44 Operations level 111-18
Method of least squares VII-14-VII-17 metrlcs 11-46
Metrics 11-42, 11-43 Operator V-40
business level 11-46 variation VI-28
goals 11-42, 11-43 Opinions as facts IV-31
operations level 11-46 Optimization VIII-45
perfonnance 11-46 Optimize X-7, X-8
process 11-47 Ordinal scale V1-22, VI-23
selection V-6 Organizational
use of 11-43, 11-44 leadership 11-39-11-41
Milestones 111-34 linking projects to goals 111-2-111-4
reporting III-51 Orthogonal VIII-45
MiniTab results VII-23, VIII-64, VIII-65 Other communication forms IV-:l4, IV-35
MinorX-23 Other unusual patterns IX-34
Misleading infonnation IV-33 Outer array V111-45
Mistake proofing 1-17, VIII-18-VIII-20 Outputs 111-13
defined 1-16 Overall equipment effectiveness (OEE) IX-43, IX-45
Mixture experiments VIII-44 Overbearing participants IV-31
Mizusumashi 1-20 Overlearning IX-59
Monitor evaluator IV-23 Overproduction VII-2, VII-3
MotlonVII-4
MRP (materials requirements planning) 1-16 P
Muda 1-16, VII-2 p chart IX-23-IX-26
Multi-collinearity VIII-44 example IX-25, IX-26
Multi-vari formulas IX-24
analysis VII-6-VII-9 p test VII-41, VII-51
case study VII-10, VII-11 p-value VII-22
re-analysis VIII-59 Packaging X-29
sampling plan procedure VII-6 Paired comparison VIII-45
Multiple linear regression VII-21 Paired t test VII-40, VII·51
Multiple-sense learning IX-60 Paired-cell overlapping loops of ,cards with authorization
Multivoting IV-41 (POLCA VIII-27
Mura 1-17 Paired-comparison hypothesis tE!sts VII-40
MX bar-MR charts IX-21, IX-22 Parallel experiments VIII-45
Parameter design X-17
N Pareto
Nagara 1-17 analysis V-22
Natural work team organization IV-8 analysis, weighted V-17, V-18
Negative nellies IV-31 diagrams V-15, V·16
Negotiation techniques IV-36 diagrams, history V·15
Nemawashi 1-17 PDCA (plan, do, check, act) 1·17, 111-6
Nested experiments VIII-45 PDSA (plan, do, study, act) 1-17, 111-7
Net operating rate IX-44 Pearson product-moment coefficient of correlation VII-19
Network planning rules 111-36 Perfection 1-17, V-29
NGT (nominal group technique) IV-40 Performance
Node 111-46 benchmarking 111·27
NoiseVI-30 efficiency VI-9, IX-43, IX-44
Noise factors X-14, X-15 index VI-55
Nominal group technique (NGT) IV-40 index with process centering VI·55
Nominal scale VI-22, VI-23 measurement considerations 11-47

metrics 11-46 level 111-18

ratio VI-55 mapping V-43, V-44
Performing IV-25-IV-27 mapping versus VSM V-46-V-48
PERT (program evaluation and review technique) 111-36-111- metrlcs 11-47
39 owners IV-16, IV-17
chart example 111-38, 111-39 performance exercise VI-55
defined 1-17 performance indices VI-54, VI-55
PFMEA (process FMEA) X-20 performance versus specifications VI-50
Philosophy level, 5S VIII-8 variation out-of-controIIX-32
Piece-to-piece variaton VII-11 Processing VII-4
Plackett-Burman designs VIII-52 Profit 11-42
Plan, do, check, act (PDCA) 111-6 Program evaluation and review technique (PERT) 111-36-111-
Plan, do, study, act (PDSA) 111-7 39
Planning 111-33 chart example 111-38, 111-39
forDFSS X-9 Project
tools 111-36 activity measurement 111-48-111-51
Plant IV-22, IV-23 benchmarking 111-27
Planter IV-22 charter V-2, V-3
Plops IV-31 document archiving III-52
Point estimate for documentation 111-48-111-52
population mean VII-32 final report III-52
population variance VII-32 lessons learned 111-52
Point of use inventory 1-17 linking to organizational goals 111-2-111-4
Points near or outside limits IX-37 management, lean six sigma 111-2
Poisson plan elements 111-33
distribution IX-24, XII-4-XII-6 scope 111-33
equation VI-12 selection 111-5, 111-13
Poka-yoke 1-17, VIII-18-VIII-20 selection process 111-5
everyday examples VIII-19 teams IV-6, IV-9
POLCA (paired-cell overlapping loops of cards with Proportion
authorization VIII-27 confidence interval VII-34
Positional variation VlI-9 hypothesis test for VII-41
Post-improvement capability analysis VIII-59 Pull
PpVI-55 arrowV-40
PpkVI-55 circle V-40
ppm VI-56 system 1-18
PrVI-55 valueV-28
Practical significance V/I-28 Push arrow V-40
Pre-control technique IX-38, IX-39 Put-downs IV-31
Pre-launch VIII-8, VIII-9
Precision VI-30, VIII-45 Q
Primacy and recency IX-59 QCI spaghetti studies XI-2-XI-5
Principle of least squares VII-15 QFD (quality function deployment) 111-24-111-26, X-11-X-13
Probability of a defect VI-14 design using X-11-X-13
Problem QRM (quick response manufacturing) VIII-27
definition tools V-11 Qualitative VIII-45
solving methodology 111-6 QualityX-27
solving steps 111-8, 111-9 circles IV-8
solving, 80 111-12 control methods IX-3, IX-4
statement V-4 controls IX-2, IX-3
Procedural variation VI-28 gurus 11-22
Procedures VI-2, VI-3 teams IV-8, IV-9
Process 111-13 Quality function deployment (QFD) 111-24-111-26, X-11-X-13
analysis VI-2 design using X-11-X-13
average and variation out-of-controllX-33 Quantitative VIII-45
average out-of-control IX-32 Questioning techniques IV-35
benchmarking 111-27 Queue time 1-18
boxV-40 Quick response manufacturing (QRM) VIII-27
capability analysis VI-39
capability exercise VI-54 R
capability for attribute data VI-58 Random sampling VI-25
capability from control charts VI-51, VI-52 Randomized block plans VIII-49
capability indices VI-53, VI-54 Randomized trials VIII-46
capability studies VI-45-VI-49 Range IX-13
cycle efficiency VI-9, VI-10 method VI-32
design example X-35, X-36 Ranking scale VI-22
elements 111-13, 111-14 Ratio scale VI-22, VI-23
factors IV-44 Recognition and reward IV-28
FMEAX-20 Recorder role IV-21
improvement team IV-6 Recurring cycles IX-36
in control IX-33 References
in control with chance variation IX-35 analysis techniques VII-60

control concepts IX-61, IX-62 Sample

defining opportunities V-51, V-52 meeting forms IV-12
design improvement X-37-X-40 required size VII-29
improvement techniques VIII-70-VIII-72 Sampling
introduction 1-20 developing plans VI-48
lean six sigma bibliography 1-2, 1-3 random VI-25
lean six sigma goals 11-48-11-52 sequential VI-25
LSS project management III-53, III-54 stratified VI-25
measurement techniques VI-59, VI-60 Sanitize 1-13
team organization and dynamics IV-46-IV-48 Satisfaction/complaint cards 111-19
Regression analysis VII-17 Satisfiers V-23
Reinforcement IX-59 Schedule box V-40
Relationship factors IV-44 Scheduling 111-33
Reliability X-27 Scorecard 11-44, 11-45
Removing team members IV-13 Screening experiment VIII-47
Repair, rejects VII-3 Second order linear model VII-2'1
Repeatability Second-order VIII-47
and reproducibility VI-32 Seiban 1-18
calculation VI-34 Seiketsu 1-13, VIII-3, VIII-5
defined VI-30 Seiri 1-13, VIII-3
Repeated trials VIII-46 Seiso 1-13, VII\-3-VII\-5
Reproducibility Seiton 1-13, VIII-3, VIII-4
and repeatability VI-32 Select and scale the process variables VIII-37
calculation VI-34 Selecting team members IV-13
defined VI-31 Selection of problem solving methodology 111-6
Required resources 111-34 Selectivity VI-31
Required sample size VII-29 Self directed teams IV-7, IV-9
Residual error VIII-46 Sensei 1-18
Residual types VIII-40 Sensitivity VI-31
Residuals VIII-46 Sequential experiments VIII-47
Resolution VI-31 Sequential sampling VI-25
Resolution I VIII-46 Service FMEA X-21
Resolution II VIII-46 Serviceability X-28
Resolution III VIII-46 Set in order VIII-4
Resolution IV VIII-46 Setup reduction (SUR) VIII-21-VII\-24
Resolution V VIII-46 case study VIII-24
Resource Investigator IV-23 Setup time 1-18
Resources 11-43 Seven classical wastes VII-2-VII-S
for training IX-58 Seven deadly diseases that management must cure 11-26
Response X-14, X-15 Severity index X-24
surfaces VIII-60, VIII-61 Shaper IV-22, IV-23
variable VIII-47 Shewhart, W.A. 11-34, 11-35
Response surface methodology (RSM) VIII-46 Shewhart cycle 11-35
Rewards and reinforcement 11-41 Shine VIII-ii, VIII-12, VIII-15
Rising ridge VIII-60 Shine, scrub, clean or sweep VIII-4
Risk Shitsuke 1-13, VIII-3, VIII-6
analysis 111-31 Shojinka 1-18
analysis and management 111-30-111-32 Short-term versus long-term capability VI-57
assessment 111-31, X-23, X-24 Shy members IV-31
defined 111-30 Sigma relationships VI-14
handling 111-31 Signal factor X-14, X-15
management continuous process 111-32 Simple linear model VII-12, VII-13
planning 111-31 Simplex design VIII-47
Risk priority number (RPN) X-22 Single minute exchange of dies (SMED) VIII-21
Risky-shift IV-30, IV-31 Single piece flow 1-18
ROA (return on assets) 1-18 SIPOC (suppliers, inputs, processes, outputs, customers)
Robust design VIII-47, X-14-X-17 1-18,111-13,111-14
approach X-14, X-15 Six big losses IX-40, IX-41
example X-15-X-17 Six sigma
introduction X-14 defined 1-19, 11-7
ROI (return on investment) 1-18 failure rates XII-3
Role factors IV-44 gurus 11-22
Rolled throughput yield (RTY) VI-13 introduction 11-7-11-9
Root cause analysis VII-58 metrics VI-11-VI-14
Rope VIII-31 , VIII-32 quality level VI-14
RPN (risk priority number) X-22 results 11-10
RSM (response surface methodology) VIII-46 Skills matrix 1-19
RTY (rolled throughput yield) 1-18, VI-13 Slack time 111-37, 111-39, 111-43, 111-411
Slope VII-13
S Small lot principle 1-19
Saddle minimax VIII-60 SMED (single minute exchange of dies) 1-19, VIII-21
Safety VIII-3, VIII-6, X-27 Software license improvement at Exactus XI-6-XI-10
VILLANOVA UNIVERSITY XII - 22 LE.6.N SIX SIGMA HANDBOOK

Software variation VI-28 facilitation IV-19

Sort 1-13, VIII-9, VIII-14 Improvement IV-6, IV-9
Sort or organize VIII-3 initiating IV-2
Source V-40 lean six sigma IV-6, IV-9
Spaced learning IX-59 life cycle characteristics IV-27
Spaghetti diagrams V-49, V-50 meeting structure IV-11
Special roles IV-34 member benefits IV-3
Specialist IV-22, IV-23 member role IV-21
Stability VI-30, VI-31 natural work organization IV-8
Stage gate process clarification X-5 objectives IV-4
Stakeholder analysis 111-14·111-16 organization and dynamics IV-2
Standard performance checklist IV-43
deviation confidence interval VII·35 performance factors 1V-44
inventory IX-53 process improvement IV-6
normal table XII-2 project IV-6, 1V-9
work IX-50-IX-53 quality 1V-8, IV-9
work, defined 1-19 removing members IV-13
Standardize 1-13, VIII-12 resources IV-3
Standardize or purity VIII-5 roles IV-15-IV-17
Stationary ridge VIII-60 selecting members IV·13
Statistical self directed IV-7, IV·9
significance VII-28 size IV-14
tests V-21 stages IV-24-IV-26
Steering committee role IV-18 synopsis of applications IV-9
Steps for constructing X bar - R charts IX-14 tools 1V-40
Steps to implement TPM IX-45, IX-46 types of IV-6-IV-8
Storming IV-24-IV-27 typical operating guidelines IV-10
Straight line VII-13, VII-14 worker IV-22, 1V-23
Straighten 1-13, VIII-11, VIII-14 Team organization and dynamics
Strategic benchmarking 111-27 conflict resolution IV-37
Stratified sampling VI-25 initiating teams IV-2
Student·s t test VII-37 -VII-39 references IV-46-IV-48
Summary of inference tests VII-51 team dynamics IV-28
Supermarket V-40 team performance checklist IV-43
Suppliers 111-13 team roles IV-15
Suppliers, inputs, processes, outputs, customers (SIPOC) team stages IV-24
111-13,111-14 team tools IV-40
SUR (setup reduction) VIII-21-VIII-24 Teamworker IV-23
case study VIJI-24 Temporal variation VII-9
Surveillance IX-56 Terminator V-44
Surveys 111-19 Test coverage VIII-47
customer 111-20 Test statistic VII-24
L type 111-20 Testability X-27
pitfalls 111-21 The French design model X-10
Sustain 1-13, VIII-12, VIII-13 Theory of constraints (TOC) VIII-29-VIII-33
Sustain or self-discipline VIII-6 Theory of inventive problem solving X-30
Symbols, six sigma VI-11 Therblig symbols VIII-15
Synopsis of team applications IV-9 Thomas-Kilmann conflict mode instrument IV-38
System FMEA X-19 Three factor, three level experiment VIII-48
Systematic design X·32 Throughput rate VI-9
Tile tunnel kiln X-16
T Time line V-40
t distribution XII-8 Time to market X-29
t test VII-37 -VII-39, VII-51 Time-to-tlme variaton VII-11
paired VII-40 Timekeeper role IV-21
Takt time V-30-V-35, VI-5-VI-8, VIII-26 TO (total opportunities) VI-11
defined 1-19, VI-5 TOC (theory of constraints) VIII-29-VII/-33
Tangents & digressions IV-31 Tolerance design X-17
Task forces IV-6 Total defects per unit (TDPU) VI-13
TDPU (total defects per unit) VI-13 Total lead time VI-9, VI-10
Team Total opportunities (TO) VI·11
ad hoc IV-6 Total productive maintenance (TPM) IX-40-IX-46
adding new members IV-13 metrics IX-43-IX-45
additional problem areas IV-30 steps to implement IX-45, IX-46
Belbin roles IV-22, IV-23 Total quality control (TQC) 11-29
common problems IV-31 Toyota Production System (TPS) 11-2, 11-17, 11-18, VIII-21 ,
company benefits IV-2, 1V-3 VIII-27
composition 111-34 TPM (total productive maintenance) 1X-40-lX-46
cross functionaIIV-7, IV-9 metrics IX-43-IX-45
diversity IV-14 steps to implement IX-45, IX-46
dynamics IV-28 TPS (Toyota Production System) 11-2, 11-17, 11-18, VIII-21 ,
empowerment IV-5 VIII-27

TQC (total quality control) 11-29 data VI-15-VI-17

Training 11-40, 11-41 Variance
and development IX-57 confidence interval VII-35
needs analysis IX-56 point estimate VII-32
performance gaps IX-55 VAT (value-added time) V-38
requirements IX-54, IX-55 Verify X-8
resources IX-58 Verifying stability and normality VI-49
Transfer of learning IX-60 Visual
Transport VII-4 control 1-20
Treatments VIII-47 systems IX-47-IX-49
Trends IX-35 Vital few V-15
Triage 1-19 voe (voice ofthe customer) 1-20, 111-19, 111-24, V-19, X-11, X-
Trilogy 11-33 13
Trivial many V-15 Voice of the customer (VOC) 111-19, 111-24, V-19, X-11, X-13
TRIZ 1-19, X-30, X-31 VSM (value stream mapping) 11-3, V-36-V-42
Truck shipment V-40 icons V-40
Two-factor, two-way ANOVA VII-56, VII-57 versus process mapping V-4E~-V-48
Two-level fractional factorial example VIII-53
Two-tail test VII-26, VII-27 W
Two-way AN OVA VII-55-VII-57 Waiting VII-5
Type I error VII-25 Wanderlust IV-31
Type II error VII-25 Waste VII-2-VII-5
Types of additional classes VII-5
charts IX-12 defined 1-20
control plans IX-5 Water spider 1-20
data VI-15-VI-17 WBS (work breakdown structure) 1-20, 111-35
errors VII-25 WCM (world class manufacturin!J) 1-20
teams IV-6-IV-S Weighted Pareto analysis V-17, \/-18
Typical What 5S is not VIII-6
DOE checklist VIII-36, VIII-37 Whole or part learning IX-59
Pareto diagram V-16 Widely used symbols VI-11
team operating guidelines IV-10 WIP (work-in-process) 1-20, VIII-31 , VIII-32
inventory VI-9
U Within piece variaton VII-11
u chart IX-23, IX-24 Work
formulas IX-24 cell 1-20
U-shaped flow V-27, V-34, V-35 center 1-20
UCL (upper control limit) IX-11, IX-13-IX-15, IX-17-IX-21, IX- instructions VI-3, IX-2
24, IX-25, IX-27, IX-29, IX-30 sequence IX-52
Unanticipated 111-22 Work breakdown structure (WBS) 111-35
Upper control limit (UCL) IX-11, IX-13-IX-15, IX-17-IX-21, IX- Work-in-process (WIP) VIII-31 , VIII-32
24, IX-25, IX-27, IX-29, IX-30 Written procedures IX-2
Upward flow of communications IV-33
Usage of DFX techniques and tools X-25, X-26
Use of metrics 11-43, 11-44 X bar and R chart
x
User friendliness X-29 control chart IX-16
Validate X-7, X-S

v data IX-15
steps for constructing IX-14
terms IX-13
Value V-24, V-25 X bar and sigma charts IX-17
added, defined 1-19 X-MR charts IX-19, IX-20
chain V-25
defined 1-19 Y
flowV-27 Yield VI-13
of lean six sigma 11-2-11-6 first pass VI-13
pull V-28 formulas VI-13
stream V-25, V-26 normalized VI-13
stream manager V-37 relationships VI-12, VI-13
stream, defined 1-19 rolled throughput VI-13
stream, see Value stream mapping V-36
Value stream mapping (VSM) 11-3, V-36-V-42 Z
icons V-40 Z long-term VI-14
versus process mapping V-46-V-48 Z short-term VI-14
Value-added time (VAT) V-38 Z test VII-36, VII-37, VII-51
Variability VI-30 Z value VI-14, VI-42-VI-44, VI-51, "I-56
Variable relationships VII-6 examples VI-43, VI-44
Variables table XII-2
charts IX-12

XII. APPENDIX - QUESTION ANSWERS
SECTION SECTION SECTION SECTION SECTION

II III IV V VI
1. d 1. d 1. d 1. b 23. a 1. d 25. a
2. c 2. c 2. c 2. c 24. c 2. a 26. c
3. d 3. d 3. d 3. b 25. d 3. c 27. d
4. d 4. a 4. c 4. a 26. a 4. d 28. a
5. c 5. c 5. d 5. d 27. b 5. c 29. d
6. a 6. a 6. a 6. b 28. b 6. b 30. b
7. c 7. d 7. d 7. d 29. a 7. c 31. b
8. a 8. b 8. d 8. d 30. b 8. b 32. a
9. d 9. c 9. b 9. b 31. a 9. a 33. b
10. d 10. d 10. b 10. b 32. a 10. a 34. b
11. d 11. c 11. a 11. c 33. d 11. c 35. c
12. b 12. d 12. b 12. b 34. c 12. c 36. d
13. c 13. b 13. b 13. c 35. c 13. b 37. c
14. b 14. b 14. a 14. d 36. d 14. c 38. d
15. a 15. c 15. d 15. a 37. c 15. d 39. a
16. c 16. c 16. d 16. c 38. d 16. b 40. b
17. b 17. c 17. d 17. b 39. c 17. d 41. c
18. b 18. c 18. c 18. b 40. b 18. c 42. a
19. a 19. b 19. d 19. a 41. c 19. b 43. a
20. d 20. a 20. a 20. d 42. b 20. b 44. c
21. b 21. c 21. d 21. a 43. c 21. a 45. b
22. d 22. b 22. a 22. b 44. a 22. b 46. b
23. c 23. b 23. b 23. d 47. d
24. b 24. d 24. b 24. c 48. c
25. a 25. c 25. d
26. b 26. a 26. b
27. a 27. a 27. c
28. a 28. d 28. c
29. c 29. a 29. b
30. c 30. a 30. a
31. c 31. a 31. a
32. a 32. c 32. a
33. a 33. d 33. b
34. b 34. d 34. a
35. b 35. a 35. b
36. b 36. a 36. a
37. b 37. d 37. a
38. b 38. a 38. a
39. c 39. c 39. c
40. b 40. a 40. a

XII. APPENDIX - QUESTION ANSWERS
SECTION SECTION SECTION SECTION ,.

VII VIII IX X
1. a 25. b 1. b 31. c 1. d 25. c: 1. d

2. d 26. a 2. a 32. a 2. c 26. d 2. a
3. b 27. c 3. d 33. c 3. a 27. d 3. b
4. a 28. a 4. a 34. c 4. c 28. Cl 4. a
5. c 29. a 5. d 35. d 5. d 29. d 5. c
6. b 30. a 6. a 36. c 6. c 30. c: 6. a
7. b 31. a 7. d 37. a 7. d 31. d 7. a
8. b 32. a 8. a 38. d 8. b 32. cl 8. C
9. d 33. c 9. b 39. d 9. d 33. c: 9. d
10. b 34. d 10. a 40. d 10. d 34. d 10. b
11. b 35. d 11. d 41. c 11. c 35. d 11. b
12. b 36. a 12. c 42. d 12. c 36. cl 12. d
13. a 37. a 13. a 43. d 13. c 37. d 13. c
14. d 38. d 14. a 44. a 14. d 38. b 14. b
15. b 39. c 15. b 45. b 15. c 39. c: 15. c
16. b 40. c 16. b 46. b 16. d 40. c: 16. c
17. b 41. d 17. b 47. d 17. b 41. c: 17. c
18. d 42. a 18. a 48. d 18. b 42. b 18. b
19. c 43. b 19. b 49. d 19. b 43. d 19. b
20. d 44. c 20. b 50. d 20. b 44. cl 20. d
21. b 45. d 21. b 51. b 21. b 45. d 21. b
22. c 46. b 22. d 52. b 22. d 46. cl 22. C
23. b 47. c 23. a 53. c 23. c 47. b 23. d
24. c 48. d 24. b 54. c 24. d 48. b 24. a
25. c 55. c 25. b
26. d 56. c 26. b
27. b 57. b 27. a
28. d 58. d 28. c
29. b 59. a 29. a
30. b 60. d 30. a
31. b
32. b

\
"THE HORROR OF THAT MOMENT," THE KING WENT ON,

"I SHALL NEVER, NEVER FORGETI"
"YOU WILL, THOUGH," THE QUEEN SAID, "IF YOU DON'T

MAKE A MEMORANDUM OF IT."
LEWIS CARROLL
(CHARLES LUTWIDGE DODGSON)
"THROUGH THE LOOKING GLASS"
1872
CLASSROOM NOTES

Six Sigma Handbook - 2007 PDF

Uploaded by

Copyright:

Available Formats

Six Sigma Handbook - 2007 PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Six Sigma Handbook - 2007 PDF

Uploaded by

Copyright:

Available Formats

VILLANOVA UNIVERSITY

LEAN SIX SIGMA

@ by Quality Council of Indiana - All rights reserved

Professional Education Online

JOHN LENNON/PAUL McCARTNEY

This Handbook is produced in collaboration between Villanova University and

First Edition - April 2, 2007

VILLANOVA UNIVERSITY INTRO-3 LEAN SIX SIGMA HANDBOOK

In addition to a B.A. in Management and Organizational Development, Mr. Crabtree

VILLANOVA UNIVERSITY INTRO-4 LEAN SIX SIGMA HANDBOOK

Metcalf Training specializes in all facets of lean (TPS) manufacturing, with an

Mr. Paradise's professional career includes 19 years experience in quality

VILLANOVA UNIVERSITY INTRO-5 LEAN SIX SIGMA HANDBOOK

VILLANOVA UNIVERSITY INTRO-6 LEAN SIX SIGMA HANDBOOK

VILLANOVA UNIVERSITY INTRO-7 LEAN SIX SIGMA HANDBOOK

II. Lean Six Sigma Goals 10% 40

III. Lean Six Sigma 10% 40

IV. Team Organization 10% 40

V. Defining Opportunities 11% 44

VI. Measurement Techniques 12% 48

VII. Analysis Techniques 12% 48

VIII. Improvement Techniques 15% 60

IX. Control Concepts 12% 48

XI. Manufacturing and

Totals 100% 400

Quality Council of Indiana

VILLANOVA UNIVERSITY INTRO-8 LEAN SIX SIGMA HANDBOOK

Professional Education Certifications Online

A JOURNEY OF A THOUSAND MILES MUST BEGIN WITH A

VILLANOVA UNIVERSITY 1-1 LEAN SIX SIGMA HANDBOOK

5IX 5IGMA CON5ULTANT

Lean Six Sigma Bibliography

Ballis, J. (2001). Managing Flow. Dallas: Brown Books.

Dennis, P. (2002). Lean Production Simplified. New York: Produc:tivity Press.

George, M. L. (2001). Lean Six Sigma. New York: McGraw-Hili.

VILLANOVA UNIVERSITY 1- 2 LEA,N SIX SIGMA HANDBOOK

Lean Six Sigma Bibliography (Continued)

Kaplan, R. (1996). The Balanced Scorecard. Cambridge, MA: Harvard Business

Kobayashi, I. (1995). 20 Keys to Workplace Improvement. New York: Productivity

Naumann, E. & Hoisington, S. (2001). Customer Centered Six Sigma. Milwaukee:

VILLANOVA UNIVERSITY 1- 3 LEAN SIX SIGMA HANDBOOK

Lean Six Sigma

• ASQ/Navy LSS BOK

II. Lean Six Sigma Goals

A. Value of Lean Six Sigma

B. Origins of Lean Six Sigma

D. Quality and Six Sigma Gurus

VILLANOVA UNIVERSITY 1- 4 LEA.N SIX SIGMA HANDBOOK

Lean Six Sigma Body of Knowledge (Continued)

III. Lean Six Sigma Project Management

A. Linking Projects to Goals

B. Problem Solving Methodologies

VILLANOVA UNIVERSITY 1- 5 LEAN SIX SIGMA HANDBOOK

Lean Six Sigma Body of Knowledge (Continued)

E. Project Management Techniques

2. Work Breakdown Structure

IV. Team Organization and Dynamics

VILLANOVA UNIVERSITY 1- 6 LE)I,N SIX SIGMA HANDBOOK

Lean Six Sigma Body of Knowledge (Continued)