Models

Models for
for Quality
Quality Control/
Control/
Process
Process Control
Control and
and
Improvement
Improvement
 Agenda
Agenda
• Quality Tools
• Control Charts (Variables)
 Overview
Overview of
of Operations
Operations
Management
Management
Customer

Customer Requirements

Marketing Product Design

& Sales
Planning
Process Design
Forecasting
Capacity Distribution
Order Entry Production
Schedule Whse & transport,
Quality Control
by channel
Materials Management
Supply Chain Management,
Purchasing, Inventory Control

Support from Accounting, Finance, Human Resources, Information Systems

Dimensions
Dimensions of
of Quality
Quality Exhibit
Exhibit2-
2-
1
1
Quality
Quality Tools
Tools
Tools
Toolsfor
forProcess
ProcessImprovement
Improvement Exhibit
Exhibit4-8
4-8
Flow
Flow Chart
Chart Symbols
Symbols Exhibit
Exhibit4-
4-99
 Process
Process for
for aa PBJ
PBJ Sandwich
Sandwich

Get Make Eat

Ingredients Sandwich Lunch

• At what level of detail should a process diagram be

developed?
• Is this process diagram helpful?
• For what purposes can it be used?
• Can it be used for training? For process
improvement?
 Process
Process for
for aa PBJ
PBJ Sandwich
Sandwich

Get
Open
from
jelly?
Get refridg Get plate
Get PB
bread from and knife
from pantry Get from cupb
shelf from cupb

Put bread Open Spread Spread

on plate PB jar PB to jelly on other to
desired desired on other
thickness on one piece of bread
 Process
Process Diagrams
Diagrams
• What is the purpose?
• Communication?
• Training?
• Process Redesign?
• ISO 9000?
• Forming consensus?
 Pareto
Pareto Diagram
Diagram Example
Example
Defective Items Number of Defects
Bad Rubber 91
Poor adhesion 128
Cracks 9
Voids 36
Impurities 15
Cuts 23
Other 12
 Ways
Ways to
to Use
Use Pareto
Pareto Diagrams
Diagrams

• Focus on the principal aspect of a

problem.
• Decide the objective of your
improvements and your improvement
items.
• Predict the possible effectiveness of the
improvement.
• Evaluate the importance of causes.
 Ways
Ways to
to Use
Use Pareto
Pareto Diagrams
Diagrams

• Understand the effectiveness of the

improvement.
• Make easy improvements right away.
• Evaluate monetary losses rather than units
or cases.
• Improve your explanations and records.

Source: Ozaki & Asaka, Handbook of

Quality Tools, Productivity Press, 1990.
Pareto
Pareto Analysis
Analysis of
of Complaints
Complaints
Pareto
Pareto Chart:
Chart: Errors
Errors Exhibit
Exhibit4-13
4-13
Fishbone
FishboneChart
Chart--Truck
TruckDelivery
DeliveryFailures
Failures Exhibit
Exhibit4-15
4-15
Scatter
ScatterDiagram
Diagram:: Cure
Curetime
time Exhibit
Exhibit4-17
4-17
Run
RunDiagram
Diagram:: Outer
OuterDiameters
Diameters Exhibit
Exhibit4-18
4-18
Run
RunChart
Chart of
of Customers
CustomersWaiting
Waiting Exhibit
Exhibit4S-2
4S-2
Central
CentralLimit
LimitTheorem
TheoremIllustrated
Illustrated TM
TM4-6
4-6
Normal
NormalCurve
Curve TM
TM4-15
4-15
 Normal
Normal Distribution
Distribution
• Z=(Xi - X-bar)/
• Central Limit Theroem
Control
Control Charts
Charts for
for Variables
Variables
SPC:
SPC: Control
Control Limits
Limits TM
TM4-8
4-8
 Control
Control Chart
Chart Uses
Uses
• Decision making tools
– Provide timely information on recently
produced parts.
– Helps determine process capability
• Problem solving tools
– Help locate and investigate causes of
questionable quality
– Aid operator in monitoring process
 Control
Control Chart
Chart Uses
Uses
• Aid in process design and
improvement
– Data from current process can be
evaluated against design targets
– Help study changes made to process (with
skilled interpretation)
 Common
Common Measures
Measures
• Manufacturing - length, height,
viscosity, color, temperature, and
velocity
• Service - # of errors, # of incorrect trx.,
delivery times, checkout times, cycle
time,
• Software - effort, deliverable dates, #
defects by type, etc.
 X-bar
X-bar &
&RR Charts
Charts
• What is the goal of measurement?
– Is it total product performance or just one
particular dimension?
• Select characteristic to be measured to
achieve goal (consider collection cost)
• Choose rational sub-group & size
– Need a homogeneous sub-group produced
under the same conditions (machine,
operator, mold, etc.)
 X-bar
X-bar and
and R
R Charts
Charts
• Sampling is typically more frequent
when control charts are first used.
• Important to know the following:
– Who will be collecting the data?
– What aspect of the process is to be
measured?
– Where in the process will the sample
be taken?
 X-bar
X-bar and
and R
R Charts
Charts
– When (how freq) will the process be
sampled?
– Why is this sample being taken?
– How will data be collected?
– How many samples and what sub-
group size?
 X-bar
X-bar Chart
Chart Guidelines
Guidelines
• Large sub-group size is more sensitive
to variations in the process average
and increases inspection costs.
• Sub-group sizes smaller than 4 do not
create a representative distribution of
sub-group averages.
• When sub-group size exceeds 10, use
the (s) chart rather than the (R) chart.
X and R Charts TM 4-9
Process
ProcessControl
ControlChart
Chart Factors
Factors Exhibit
Exhibit4-22
4-22
X and R Charts TM 4-10
X and R Charts TM 4-11
 Developing
Developing aa Control
Control Chart
Chart
1. Take 20-30 random samples of
size n where n depends on type of
control chart
2. For each sample calculate
sample statistic such as X-bar, R
or p.
3. Plot the sample statistics
sequentially
 Developing
Developing aa Control
Control Chart
Chart
4. Calculate grand means and
control limits

5. Evaluate results and recalculate

control limits if necessary.
 Stratification
Stratification of
of Data
Data
By material Manufacturer, buyer, brand, place of production, purchase
date, lot received, production lot, components, purity, size,
time stored, storage place, etc.
By machine, equipment or Machine type, number, model, performance & age, by
tool factory, line, tool, mold & die
By operator Individual, team, group, age, experience, etc.

By operating procedures Temperature, pressure, speed, rotational frequency, line

& operating conditions speed, location of operation, illumination, air temperature,
humidity, weather, operating procedure
By measurement & Instrument, measurement procedure, place or person.
inspection Inspection tools, procedure, place, inspector, etc.
By time a.m., p.m., shift, day, week, month, season, just before or
after operation.
By environment & weather Air temperature, humidity, weather condition, sound level,
illumination, etc.
Other New vs. previous product, unit vs. continuously produced
product, good vs. bad product, packaging, transportation
 Possible
Possible Causes
Causes for
for Trends
Trends

X-bar Charts
– A gradual deterioration of tooling or
equipment.
– Worker inattention or fatigue
– Gradual tool wear
– Gradual changes in environmental
conditions
– Accumulation of waste products
– Machine adjustments becoming loose
 Are
Are You
You Out-of-Control?
Out-of-Control?

– eight or more consecutive points move

upward in value or if eight or more
consecutive points move downward in
value.
– there are an unusually small number of
runs above or below the centerline
(sawtooth pattern).
– thirteen consecutive points within the C
zone.
 Possible
Possible Causes
Causes for
for Trends
Trends

R Charts
– Gradual change in material composition
– Accumulative effects of worker fatigue
and inattention
– Effects of learning on worker
performance
 Possible
Possible Causes
Causes for
for Jumps
Jumps

X-bar Charts
– Tool and die changes
– Change or recalibration of testing and
measuring equipment
– Mixing of the output from different
machines
– Change in materials or parts that come
from different suppliers
– Different workers
 Possible
Possible Causes
Causes for
for Jumps
Jumps

R Charts
– Material changes
– Operator changes
– Setup changes
– Operating method changes
 Possible
Possible Causes
Causes for
for Cycles
Cycles

X-bar Charts
– Recurring environmental conditions
– Differences in the composition and
physical characteristics of raw materials
– Rotation on a regular basis of operators
or machines
– Rotation of testing and measuring
equipment
 Possible
Possible Causes
Causes for
for Cycles
Cycles

R Charts
– Maintenance
– Periodic replacement of worn tools
– Differences in raw material composition
and physical characteristics
– Rotation of machine operators
 Control
Control Charts
Charts for
for Variables
Variables
• X-bar & R Charts
– Useful with small sample sizes
– Easy to calculate
• X-bar & s Charts
– Preferable when n>10
– Better measure of process variation
• Individuals Charts
– Low production rate
– Sample size = 1
 Control
Control Chart
Chart Formulas
Formulas

• X-bar & R Chart • Individuals Chart

UCLX = X + (A2)(R) UCLX = X + (3/d2)(R)
LCLX = X - (A2)(R) LCLX = X - (3/d2)(R)
UCLR = (D4)(R) UCLR = (D4)(R)
LCLR = (D3)(R) LCLR = (D3)(R)

• X-bar & s Chart • Multivariate Chart

UCLX = X + (A3)(s)
UCL = [(mnp--mp-np+p)/
LCLX = X - (A3)(s) (mn-m-p+1)]Fo,p,mn-m-p+1
UCLR = (B4)(s)
LCLR = (B3)(s)
Questions?