CONTROL TOOLS FOR SIX SIGMA

(MEC 424)

Unit – V

Advance Statistical
Process Control Tools
1
CONTROL CHARTS
 Introduced in 1926 by WALTER SHEWART, who
concluded that a distribution can be transformed
into normal shape by estimating mean and
standard deviation.
 Control chart is a device which specifies the state
of statistical control.
 Control chart detects the variation in processing
and warns if there is any deviation from the
specified tolerance limits.
 The purpose of using control chart is to stabilize
process by keeping it under control and carrying
2
out necessary adjustments (on line).
ADVANTAGES
 A control chart indicate whether the process is in
control or out of control.
 It determines the process variability and detects
unusual variations in a process.
 It ensures product quality level.

 It warns in time and if process is rectified at that

time percentage of rejection can be reduced.
 It provides information about selection of process
and setting up of tolerance limits.

3
CLASSIFICATION

Shewart
Control
Charts

Variable Attribute
Control Control
Charts Charts

X bar & R X bar & S P and nP C and U

Charts Charts Charts Charts

4
VARIATIONS
 No two natural items in any category are the
same.
 Variation may be quite large or very small.

 If variation very small, it may appear that items

are identical, but precision instruments will show
differences.

5
CATEGORIES OF VARIATIONS
 Within-piece variation
• One portion of surface is rougher than another
portion.
 A piece-to-piece variation

• Variation among pieces produced at the same

time.
 Time-to-time variation

• Service given early would be different from that

given later in the day.

6
SOURCES OF VARIATIONS
 Equipment
• Tool wear, machine vibration, …
 Material

• Raw material quality

 Environment

• Temperature, pressure, humidity

 Operator

• Operator performance- physical & emotional

7
X BAR AND S CHARTS
The X- bar chart is developed from the average of
each subgroup data.
• used to detect changes in the mean between
subgroups.

The S- chart is developed from the standard

deviation of each subgroup data
• used to detect changes in variation within
subgroups

8
PROCEDURE
 Determine the variable to monitor.
 At predetermined, even intervals, take samples of
size n (usually n=4 or 5).
 Compute X BAR and R (or s) for each sample, and
plot them on their respective control charts. Use the
following relationships:

9
 After collecting a sufficient number of samples, k
(k>20), compute the control limits for the charts. The
following additional calculations will be necessary

 If any points fall outside of the control limits,

conclude that the process is out of control, and begin a
search for an assignable or special cause. When the
special cause is identified, remove that point and
return to step 4 to re-evaluate the remaining points.
 If all the points are within limits, conclude that the
process is in control, and use the calculated limits for 10

future monitoring of the process.

EXAMPLE
 A large hotel in a resort area has a housekeeping
staff that cleans and prepares all of the hotel's
guestrooms daily. In an effort to improve service
through reducing variation in the time required
to clean and prepare a room, a series of
measurements is taken of the times to service
rooms in one section of the hotel. Cleaning times
for five rooms selected each day for 25
consecutive days appear below:

11
12
13
SOLUTION

14
Calculate control limits

X = 15.94 , s =1.14
x Chart Control Limits
UCL = x + A2 S
LCL = x – A2 S
S Chart Control Limits
UCL = B4 S
LCL = B3 S X-BAR CHART
CL =15.94
UCL = 15.94+ 1.427(1.14) = 17.56
LCL = 15.94 – 1.427(1.14) = 14.31
S CHART
CL = 1.14
UCL = 2.089(1.14) = 2.38 15
LCL = 0 (1.14) = 0
16
17
EXAMPLE
In semiconductor processing, the photolithography
process is used to transfer the circuit design to
silicon wafers. In the first step of the process, a
specified amount of a polymer solution, photoresist,
is applied to a wafer as it spins at high speed on a
turntable. The resulting photoresist thickness x is a
key process variable. Thickness data for 25
subgroups are shown in. Each subgroup consists of
three randomly selected wafers. Construct and s
control charts for these test data and critcially
evaluate the results.
18
No.
x Data x s No.
x Data x s
1 209.6 20.76 211.1 209.4 1.8 14 202.9 210.1 208.1 207.1 3.7
2 183.5 193.1 202.4 193.0 9.5 15 198.6 195.2 150.0 181.3 27.1
3 190.1 206.8 201.6 199.5 8.6 16 188.7 200.7 207.6 199.0 9.6
4 206.9 189.3 204.1 200.1 9.4 17 197.1 204.0 182.9 194.6 10.8
5 260.0 209.0 212.2 227.1 28.6 18 194.2 211.2 215.4 206.9 11.0
6 193.9 178.8 214.5 195.7 17.9 19 191.0 206.2 183.9 193.7 11.4
7 206.9 202.8 189.7 199.8 9.0 20 202.5 197.1 211.1 203.6 7.0
8 200.2 192.7 202.1 198.3 5.0 21 185.1 186.3 188.9 186.8 1.9
9 210.6 192.3 205.9 202.9 9.5 22 203.1 193.1 203.9 200.0 6.0
10 186.6 201.5 197.4 195.2 7.7 23 179.7 203.3 209.7 197.6 15.8
11 204.8 196.6 225.0 208.8 14.6 24 205.3 190.0 208.2 201.2 9.8
12 183.7 209.7 208.6 200.6 14.7 25 203.4 202.9 200.4 202.2 1.6
13 185.6 198.9 191.5 192.0 6.7
19
SOLUTION
The following sample statistics can be calculated
from the data in Table : x = 199.8 Å, s = 10.4 Å.
For n = 3 the required constants from Table 21.1
are c4 = 0.8862, B3 = 0, and B4 = 2.568. Then the
and s control limits can be calculated from Eqs.

20
 Estimation of s Chart
n c4 B3 B4
2 0.7979 0 3.267
3 0.8862 0 2.568
4 0.9213 0 2.266
5 0.9400 0 2.089
6 0.9515 0.030 1.970
7 0.9594 0.118 1.882
8 0.9650 0.185 1.815
9 0.9693 0.239 1.761
10 0.9727 0.284 1.716
15 0.9823 0.428 1.572
20 0.9869 0.510 1.490 21

25 0.9896 0.565 1.435

The traditional value of c = 3 is selected for Eqs. The resulting
control limits are labeled as the “original limits” in Fig.
Figure indicates that sample #5 lies beyond both the UCL for
both the x and s control charts, while sample #15 is very close
to a control limit on each chart. Thus, the question arises
whether these two samples are “outliers” that should be
omitted from the analysis. Table indicates that sample #5
includes a very large value (260.0), while sample #15 includes
a very small value (150.0). However, unusually large or small
numerical values by themselves do not justify discarding
samples; further investigation is required.
Suppose that a more detailed evaluation has discovered a
specific reason as to why measurements #5 and #15 should be
discarded (e.g., faulty sensor, data misreported, etc.). In this
situation, these two samples should be removed and control
limits should be recalculated based on the remaining 23
samples. 22
 Original Limits Modified Limits
(omit samples #5
and #15)
x Chart Control
Limits
UCL 220.1 216.7
LCL 179.6 182.2

s Chart Control
Limits
UCL 26.6 22.7
LCL 0 0

23
24
INDIVIDUAL MEASUREMENTS
 Sometimes repeated measures in a subsample
don't make sense:
 inventory level
 accounts payable – price of an item

 Other reasons for using individual measurements

 Variation in sample only reflects measurement error,
e.g., batch production of chemicals
 Automated inspection – every unit is analyzed
 Production rate very slow – inconvenient to wait for
large enough sample

25
CONTROL CHARTS FOR INDIVIDUAL
MEASUREMENT

 Notes about Individuals Charts:

 Sample points must be relatively frequent
 There is more sampling error (false alarm &
insensitivity)
 Sample points tend to be non-normal
 points are not averages and central limit theorem does not
apply

 Control Chart Types for Individuals:

 MR - Moving Range chart
 MA – Moving Average chart
 EWMA – Exponentially Weighted Moving Average
 CUSUM – Cumulative Sum
26
MOVING RANGE CONTROL CHART
 MR2i = |xi – xi-1| or MR3i = |xi – xi-2|
 Computation of the Moving Range:
Obs i xi MR2 MR3
1 33.75 - -
2 33.05 0.70 -
3 34.00 0.95 0.25
4 33.81 0.19 0.76
5 33.46 0.35 0.54
… … … …
n
27
 General model for moving range chart:
UCL  m MR  Ls MR
CL  m MR
LCL  m MR  Ls MR
 Plot MR2i = |xi – xi-1| or MR3i = |xi – xi-2| on control chart

 Substituting estimates for mR and sR and using “3-sigma”

limits:
CL  m MR  MR
 Where MR is:
n n
MR 2  1
n1  MR2i
i 2
or MR3  1
n2  MR3
i 3
i
28
 For MR2 use d2 for n = 2
For MR3 use d2 for n = 3
UCL  m MR  3s MR  MR  3  d 3s 
MR
 MR  3d 3  D4 MR
d2
LCL  D3 MR  0

 Very similar to Range chart except we’re using

moving range instead of average range

29
X CONTROL CHART
 Plot sample statistic: x

 General model for x chart

UCL  m x  Ls x
CL  m x
LCL  m x  Ls x
 Substituting estimates for mx and sx and using 3-sigma
limits
CL  m x  x
where:
n
mx  x  1
n x
i 1
i 30
 x chart upper and lower limits:

UCL  m x  3s x  x  3s
MR
 x 3
d2

MR
LCL  x  3
d2

31
CAUTIONS
 Always check x’s for normality
 If x’s not normal, control limits are inappropriate
 Very BAD at detecting small shifts, i.e., shifts <
2s
x chart x chart
(n = 5) (n = 1)

size of shift b ARL1 b ARL1

1s 0.78 5.25 0.98 43.96

2s 0.07 1.08 0.84 6.30

32
3s 0.00 1.00 0.50 2.00
 Plot MR2i = |xi – xi-1| on control chart
 From initial data compute x and MR:
mˆ  x  33.52
MR 0.48
MR  0.48 sˆ  
d2 1.128
(since d2 = 1.128 for n =2)
 Control Limits for Moving Range chart (use D4 &
D3 for n =2)

UCL  D4 MR  3.267  0.48  1.57

CL  MR  0.48
LCL  D3 MR  0 33
 UCL  x  3sˆ  34.80
Compute control Limits for x
CL  x  33.52
Chart
LCL  x  3sˆ  32.24
X Cha rt for Viscosity
35
34.5
34
X

33.5
33
32.5
32
0 3 6 9 12 15
Observa tion

MR(2) Chart for Viscosity

1.6

1.2
MR(2)

0.8

0.4
34
0
0 3 6 9 12 15
Observation
MOVING AVERAGE CONTROL CHART
 In situations where data are collected slowly over a
period of time, or where data are expensive to collect,
moving average charts are beneficial.
 Moving Average Charts are a set of control charts for
variables data (data that is both quantitative and
continuous in measurement, such as a measured
dimension or time). The Moving Average chart
monitors the process location over time, based on
the average of the current subgroup and one or more
prior subgroups. The Moving Range chart monitors
the variation between the subgroups over time.
35
ALTERNATIVES FOR SHEWHART CONTROL
CHARTS
 All control charts so far have been Shewhart Control
Charts
 uses information about the process contained in the last
plotted point
 ignores information given by the entire sequence of points,
unless sensitizing rules are used

 Shewhart charts are relatively insensitive to small shifts,

ex. shifts < 1.5s

 Three Alternative charts:

 MA – Moving Average control chart
 EWMA – Exponentially Weighted Moving Average control
chart
 CUSUM – Cumulative-sum control chart 36
MOVING AVERAGE CONTROL CHART
 Plot sample statistic: average of last w data points (Mi )

 Computing point to plot ( Mi ) for the chart:

xi  xi 1  ...  xi  w1
Mi 
w
 Estimate for μ (to find center line):
1 n
μ0   xi
n i 1
 Estimate for s (to find control limits, changes with each
point):
σx
σ 37
w
 General model for MA control chart
σx
UCL  μ0  Lσ  μ0  L
w
1 n OR
CL  μ0   xi
•

n i 1
σx
UCL  μ0  Lσ  μ0  L
w
 Notes:
 Picking w larger makes chart faster to detect to smaller shifts
 Picking w smaller makes chart more sensitive to larger shifts
 MA is better at detecting smaller shifts than a Shewhart
chart, but not as effective as a EWMA or CUSUM chart

38
EXAMPLE
Observations:
1) 100 10) 96.5 19) 100.9
2) 101.7 11) 105.2 20) 98.6
3) 104.5 12) 95.1 21) 105.9
4) 105.2 13) 93.2
5) 99.6 14) 93.6
6) 101.4 15) 103.3
7) 94.5 16) 100.1
8) 1010.6 17) 98.3
9) 99.1 18) 98.5
Take n=2 39
•n= number of measurements
in moving average
•MR= l current measurement
– previous measurement I
•R = total of MRs/ total
numbers of MRs
•X = total of measurements/
total numbers of
40
measurements
 110
M
o
UCL
v 105
i CL
n
g 100

A
95
v
e
LCL
r 90
a
g
e 85
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
41
Sample No.
EXPONENTIAL MOVING AVERAGE (EMA)
 An exponential moving average (EMA) is a type
of moving average (MA) that places a greater
weight and significance on the most recent data
points. An exponentially moving average reacts
more significantly to recent price changes than a
simple moving average (SMA), which applies an
equal weight to all observations in the period.

42
FEATURES
 The EMA is a moving average that places a
greater weight and significance on the most
recent data points.
 Like all moving averages, this technical indicator
is used to produce buy and sell signals based on
crossovers and divergences from the historical
average.
 Traders often use several different EMA days, for
instance, 20-day, 30-day, 90-day, and 200-day
moving averages.

43
EXPONENTIAL MOVING AVERAGE

Where
EMAt = EMA Today
Vt = Value Today
EMAy = EMA Yesterday
s = Smoothening
d = No. of days
44
 To calculate an EMA, you must first compute the
simple moving average (SMA) over a particular
time period. The calculation for the SMA is
straightforward: it is simply the sum of the
stock's closing prices for the number of time
periods in question, divided by that same number
of periods. So, for example, a 20-day SMA is just
the sum of the closing prices for the past 20
trading days, divided by 20.
 Next, you must calculate the multiplier for
smoothing (weighting) the EMA, which typically
follows the formula:
[2 ÷ (selected time period + 1)].
So, for a 20-day moving average, the multiplier
would be [2/(20+1)]= 0.0952.
 Finally, to calculate the current EMA,
45
EXAMPLE
Observations:
1) 100 10) 96.5 19) 100.9
2) 101.7 11) 105.2 20) 98.6
3) 104.5 12) 95.1 21) 105.9
4) 105.2 13) 93.2
5) 99.6 14) 93.6
6) 101.4 15) 103.3
7) 94.5 16) 100.1
8) 1010.6 17) 98.3
9) 99.1 18) 98.5
Take n=2 46
 Observations EMA
Assumed 98.00
 S = 2 / (21+1) = 100 98.46
0.091 101.7 99.19
104.5 100.40
105.2 101.49
99.6 101.06
101.4 101.14
94.5 99.63
101.6 100.08
99.1 99.85
96.5 99.09
105.2 100.48
95.1 99.26
93.2 97.88
93.6 96.91
103.3 98.36
100.1 98.76
98.3 98.65
98.5 98.62
47
100.9 99.14
98.6 99.01
105.9 100.58
 110.00
108.00
UCL
106.00
104.00
102.00
E 100.00 CL
M
98.00
A
96.00
94.00 LCL
92.00
90.00
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23

Sample No.
48
EWMA - EXPONENTIALLY WEIGHTED
MOVING AVERAGE CONTROL CHART
 The Exponentially Weighted Moving Average
(EWMA) is a statistics for monitoring the process
that averages the data in a way that gives less
and less weight to data as they are further
removed in time.
 In statistical quality control, the EWMA chart (or
exponentially-weighted moving average chart) is
a type of control chart used to monitor either
variables or attributes-type data using the
monitored business or industrial process's entire
history of output. While other control charts treat
rational subgroups of samples individually, the
EWMA chart tracks the exponentially-weighted 49
moving average of all prior sample means.
The statistic that is calculated is:

EWMAt = λ Yt + (1 - λ) EWMAt-1
for t = 1, 2, ..., n.

Where:
 EWMA0 is the mean of historical data (target)

 Yt is the observation at time t

 n is the number of observations to be monitored

including EWMA0
 0 < λ ≤ 1 is a constant that determines the depth
of memory of the EWMA.
The equation is due to Roberts (1959). 50
 The center line for the control chart is the target
value or EWMA0. The control limits are:

UCL = EWMA0 + KS λ / (2-λ)

LCL = EWMA0 –KS λ / (2-λ)

where the factor K is either set equal 3 or chosen

using the Lucas and Saccucci (1990) tables. The
data are assumed to be independent and these
tables also assume a normal population.
Factor S is the smoothening factor

51
EXAMPLE
 To illustrate the construction of an EWMA control chart,
consider a process with the following parameters calculated
from historical data:
EWMA0 = 50 s = 2.0539
with λ chosen to be 0.3 so that λ / (2-λ) = .3 / 1.7 = 0.1765
and the square root = 0.4201. The control limits are given by
UCL = 50 + 3 (0.4201)(2.0539) = 52.5884
LCL = 50 - 3 (0.4201) (2.0539) = 47.4115
Consider the following data consisting of 20 points
52.0 47.0 49.6 51.2
47.0 51.0 47.6 52.6
53.0 50.1 49.9 52.4
49.3 51.2 51.3 53.6
52
50.1 50.5 47.8 52.1
53
54
 The red dots are the raw data; the jagged line is
the EWMA statistic over time. The chart tells us
that the process is in control because all EWMAt
lie between the control limits. However, there
seems to be a trend upwards for the last 5
periods.

55
CAPABILITY ANALYSIS
 Capability analysis is a set of calculations used to
assess whether a system is statistically able to
meet a set of specifications or requirements. To
complete the calculations, a set of data is
required, usually generated by a control chart;
however, data can be collected specifically for this
purpose.

56
WHEN TO USE?
 Is it necessary to understand how the
system performs in comparison to
specification limits?
 Does the specification consist of an upper
and lower requirement?
 Are no special causes of variation present?

 Is the data in variables form?

 Do the individual values form a normal

distribution?
 Has the data been collected over a period of
time? 57
PROCEDURE
 Sketch the distribution
 Calculate the estimated standard deviation

 Determine the location of the tails for the

distribution
 Draw the specification limits on the distribution

 Calculate how much data is outside the

specifications
 Calculate and interpret the capability indices

 Analyze the results

58
PROCESS CAPABILITY
 From a conceptual view it is a measure of the
relationship between the voice of the process and
the voice of customer (VOC). It is essentially a
ratio of the customer requirement (specification)
and the expected process variation.
Process capability = Voice of the customer /
Voice of the process

59
60
61
62
TOTAL PRODUCTIVE MAINTENANCE
(TPM)
 Total Productive Maintenance (TPM) is both
 a philosophy to permeate throughout an
operating company touching people of all
levels
 a collection of techniques and practices
aimed at maximizing the effectiveness (best
possible return) of business facilities and
processes

63
TPM

64
TPM & TRADITIONAL MAINTENANCE
 Reactive maintenance inherently
wasteful and ineffective with following
disadvantages:
 No warning of failure
 Possible safety risk
 Unscheduled downtime of machinery
 Production loss or delay
 Possible secondary damage
 Need for:
 Stand-by machinery
65
 A stand-by maintenance team
 A stock of spare parts
 Costs include:
 Post production
 Disrupted schedule
 Repair cost
 Stand-by machinery
 Spare parts
 Real cost of reactive maintenance is
more than the cost of maintenance
resources and spare parts
 Pro-active maintenance (planned,
preventive and predictive) more
desirable than reactive maintenance
66
 TPM enables or provides:
 The traditional maintenance practices
to change from reactive to pro-active
 A number of mechanisms whereby
 Breakdowns are analyzed
 Causes investigated
 Actions taken to prevent
further breakdowns
 Preventive maintenance schedule to be
made more meaningful

67
 To ‘free up’ maintenance professionals
to:
 Carry out scheduled and
preventive maintenance
 Gather relevant information
as important input to the
maintenance system
 Keep the system up to date
 To review cost effectiveness
 To develop and operate a very effective
maintenance system an integral part of
manufacturing 68
IMPORTANCE
1. It guarantees dramatic results (Significant tangible
results)
 Reduce equipment breakdowns
 Minimize idle time and minor stops
 Less quality defects and claims
 Increase productivity
 Reduce manpower and cost
 Lower inventory
 Reduce accidents
2. Visibly transform the workplace (plant environment)
 Through TPM, a filthy, rusty plant covered in oil and grease,
leaking fluids and spilt powders can be reborn as a pleasant and
safe working environment
 Customers and other visitors are impressed by the change 69
 Confidence on plant’s product increases
3. Raises the level of workers knowledge and skills

As TPM activities begin to yield above

concrete results, it helps:
 The workers to become motivated
 Involvement increases
 Improvement suggestions proliferate
 People begin to think of TPM as part of the job

70
TPM POLICY & OBJECTIVES
• To maximize overall equipment
effectiveness (Zero breakdowns and
failures, Zero accident, and Zero defects
etc) through total employee involvement

• To improve equipment reliability

and maintainability as contributors to
quality and to raise productivity

71
• To aim for maximum economy in equipment for its
entire life
• To cultivate equipment-related expertise and skills
among operators
• To create a vigorous and enthusiastic work
environment
TPM Corporate policy for the following purposes:
• To aim for world-class maintenance, manufacturing
performance and quality
• To plan for corporate growth through business
leadership
• To promote greater efficiency through greater
flexibility
• Revitalize the workshop and make the most of
employee talents
72
TPM ACTIVITIES
1. Autonomous maintenance
2. Equipment improvement
3. Quality maintenance
4. MP(Maintenance Prevention) system-building
5. Education and training

73