Statistical

SUPPLEMENT
Process
Control

PowerPoint presentation to accompany

Heizer and Render
Operations Management, Eleventh Edition
Principles of Operations Management, Ninth Edition

PowerPoint slides by Jeff Heyl

2014 Pearson Education, Inc.

Learning Objectives
1. Apply quality management tools for
problem solving
2. Identify the importance of data in
quality management

6S2
6S2
 Introduction
Statistical Process Control (SPC)

Statistical
Quality Control
Acceptance Sampling (AS)

Statistical process control is a statistical technique that is

widely used to ensure that the process meets standards.
Acceptance sampling is used to determine acceptance or
rejection of material evaluated by a sample.

6S3
6S3
 Introduction
Pottery Making Process

Preparing
the clay for Wedging Throwing Pinching Painting Firing
throwing pots

6S4
6S4
Introduction

6S5
6S5
Statistical Process Control
Chart (SPC)
Variability is inherent in every process.
Natural Variation
Variation
Assignable Variation

Natural variation can not be eliminated

Assignable variation -- Deviation that can be
traced to a specific reason: machine vibration, tool
wear, new worker.

6S6
6S6
 Statistical Process Control
Chart (SPC)
The essence of SPC is the application of statistical
techniques to prevent, detect, and eliminate
defective products or services by identifying
assignable variation.

6S7
6S7
 Statistical Process Control
Chart (SPC)
A control chart is a time-ordered plot obtained from an ongoing process

Abnormal variation Out of

due to assignable sources control
UCL

Mean
Natural variation
due to chance
LCL
Abnormal variation
due to assignable sources

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Sample number

6S8
6S8
 Statistical Process Control
Chart (SPC)
x -charts (for controlling central tendency)
Control Charts
for Variable Data
R-charts (for controlling variation)
Control
Charts
p-charts (for controlling percent defective)
Control Charts for
Attribute Data
c-charts (for controlling number of defects)

Variable Data (continuous): quantifiable conditions along a

scale, such as speed, length, density, etc.
Attribute Data (discrete): qualitative characteristic or condition,
such as pass/fail, good/bad, go/no go.

6S9
6S9
Statistical Process Control
Chart (SPC)
1. Take random samples
2. Calculate the upper control limit (UCL) and the lower
control limit (LCL)
3. Plot UCL, LCL and the measured values
4. If all the measured values fall within the LCL and the UCL,
then the process is assumed to be in control and no actions
should be taken except continuing to monitor.
5. If one or more data points fall outside the control limits, then
the process is assumed to be out of control and corrective
actions need to be taken.

6S-10
x-Charts

Upper control limit (UCL) = x + A2R

Lower control limit (LCL) = x - A2R

where R = average range of the samples

A2 = control chart factor
x = average of the sample means

6S11
6S11
x-Charts
Hour 1 Hour 2
Box Weight of Box Weight of
Number Oat Flakes Number Oat Flakes
1 17 1 14
2 13 2 16
3 16 3 15
4 18 4 14
5 17 5 17
6 16 6 15
7 15 7 15
8 17 8 14
9 16 9 17
Range=18-13=5 Range=17-14=3
R = (5+3)/2 = 4
6S12
6S12
x-Charts

Upper control limit (UCL) = x + A2R

Lower control limit (LCL) = x - A2R

where R = average range of the samples

A2 = control chart factor from Table
S6.1 (page241)
x = average of the sample means

6S13
6S13
 x-Charts
Hour 1 Hour 2
Box Weight of Box Weight of
Number Oat Flakes Number Oat Flakes
1 17 1 14
2 13 2 16
3 16 3 15
4 18 4 14
5 17 5 17
6 16 6 15
7 15 7 15
8 17 8 14
9 16 9 17
Average=(17+13++16)/9=16.11 Average=(14+16++17)/9=15.22

x = (16.11+15.22)/2 = 15.665
6S14
6S14
x-Charts

Upper control limit (UCL) = x + A2R

Lower control limit (LCL) = x - A2R

where R = average range of the samples

A2 = control chart factor
x = average of the sample means

6S15
6S15
x-Charts
Sample Size Mean Factor Upper Range Lower
Range
n A2 D4 D3
2 1.88 3.27 0
3 1.02 2.58 0
4 .73 2.28 0
5 .58 2.12 0
6 .48 2.00 0
7 .42 1.92 0.08
8 .37 1.86 0.14
9 .34 1.82 0.18
10 .31 1.78 0.22
11 .29 1.74 0.26

6S16
6S16
x-Charts

Upper control limit (UCL) = x + A2R

Lower control limit (LCL) = x - A2R

where R = average range of the samples

A2 = control chart factor from
x = average of the sample means

6S17
6S17
 x-Charts
Example 1: Eight samples of seven tubes were taken at random
intervals. Construct the x-chart with 3- control limit. Is the
current process under statistical control? Why or why not? Should
any actions be taken?

Sample size = n = 7

A2 = ?

6S18
6S18
x-Charts
Sample Size Mean Factor Upper Range Lower
Range
n A2 D4 D3
2 1.88 3.27 0
3 1.02 2.58 0
4 .73 2.28 0
5 .58 2.12 0
6 .48 2.00 0
7 .42 1.92 0.08
8 .37 1.86 0.14
9 .34 1.82 0.18
10 .31 1.78 0.22
11 .29 1.74 0.26

6S19
6S19
 x-Charts
Example: Eight samples of seven tubes were taken at random
intervals. Construct the x-chart with 3- control limit. Is the
current process under statistical control? Why or why not? Should
any actions be taken?

A2 = 0.42

6.36 6.38 ... 6.34

x 6.36 oz
8
0.16 0.18 ... 0.18
R 0.17 oz
8
UCL x A2 R 6.36 0.42(0.17) 6.43 oz
LCL x A2 R 6.36 0.42(0.17) 6.29 oz 6S20
6S20
 x-Charts
Control Chart
for sample of
7 tubes

6.43 = UCL

6.36 = Mean

6.29 = LCL

| | | | | | | | | | | |
1 2 3 4 5 6 7 8 9 10 11 12

Sample number
It is assumed that the central tendency of process is in control with
99.73% confidence. No actions need to be taken except to continuously
monitor this process.
6S21
6S21
Steps in Creating Charts
1. Take samples from the population and
compute the appropriate sample statistic
2. Use the sample statistic to calculate control
limits
3. Plot control limits and measured values
4. Determine the state of the process (in or out of
control)
5. Investigate possible assignable causes and
take actions

6S22
6S22
R-Charts

Upper control limit (UCL) = D4R

Lower control limit (LCL) = D3R

where
R = average range of the
samples
D3 and D4 = control chart factors

6S23
6S23
R-Charts
Sample Size Mean Factor Upper Range Lower
Range
n A2 D4 D3
2 1.88 3.27 0
3 1.02 2.58 0
4 .73 2.28 0
5 .58 2.12 0
6 .48 2.00 0
7 .42 1.92 0.08
8 .37 1.86 0.14
9 .34 1.82 0.18
10 .31 1.78 0.22
11 .29 1.74 0.26

6S24
6S24
 R-Charts
Example 2
Average range R = 5.3 pounds
Sample size n = 5
From Table 1 D4 = ? D3 = ?

6S25
6S25
R-Charts
Sample Size Mean Factor Upper Range Lower
Range
n A2 D4 D3
2 1.88 3.27 0
3 1.02 2.58 0
4 .73 2.28 0
5 .58 2.12 0
6 .48 2.00 0
7 .42 1.92 0.08
8 .37 1.86 0.14
9 .34 1.82 0.18
10 .31 1.78 0.22
11 .29 1.74 0.26

6S26
6S26
 R-Charts
Example 2
Average range R = 5.3 pounds
Sample size n = 5
From Table S6.1 D4 = 2.12, D3 = 0

UCLR = D4R UCL = 11.2

= (2.12)(5.3)
= 11.2 pounds Mean = 5.3

LCLR = D3 R
LCL = 0
= (0)(5.3)
= 0 pounds

6S27
6S27
 R-Charts
Example 3: Refer to Example 1. Eight samples of seven tubes
were taken at random intervals. Construct the R-chart with 3-
control limits. Is the current process under statistical control? Why
or why not? Should any actions be taken?

0.16 0.18 ... 0.18

R 0.17 oz
8
n=7
D3 =? D4 = ?

6S28
6S28
R-Charts
Sample Size Mean Factor Upper Range Lower
Range
n A2 D4 D3
2 1.88 3.27 0
3 1.02 2.58 0
4 .73 2.28 0
5 .58 2.12 0
6 .48 2.00 0
7 .42 1.92 0.08
8 .37 1.86 0.14
9 .34 1.82 0.18
10 .31 1.78 0.22
11 .29 1.74 0.26

6S29
6S29
 R-Charts
Example S6.3: Refer to Example S6.1. Eight samples of seven
tubes were taken at random intervals. Construct the R-chart with
3- control limits. Is the current process under statistical control?
Why or why not? Should any actions be taken?

0.16 0.18 ... 0.18

R 0.17 oz
8

D3 =0.08, D4 = 1.92

UCL D4 R 1.92(0.17) 0.33 oz

LCL D3 R 0.08(0.17) 0.01 oz
6S30
6S30
 R-Charts
Control Chart
for sample of
7 tubes

0.33 = UCL

0.17 = R

0.01 = LCL
| | | | | | | | | | | |
1 2 3 4 5 6 7 8 9 10 11 12

Sample number
The variation of process is in control with 99.73% confidence.

6S31
6S31
 Mean and Range Charts
(a) The central tendency of process is in control, but its
variation is not in control.
UCL
(R-chart detects
R-chart increase in
dispersion)
LCL

UCL
(x-chart does not
x-chart detect dispersion)

LCL

6S32
6S32
Mean and Range Charts
(b) The variation of process is in control, but its
central tendency is not in control.
UCL
(R-chart does not
R-chart detect changes in
mean)
LCL

UCL
(x-chart detects
x-chart shift in central
tendency)
LCL

6S33
6S33
 R-Chart and X-Chart
Example S6.4: Seven random samples of four resistors each are taken to
establish the quality standards. Develop the R-chart and the x-chart both
with 3- control limits for the production process. Is the entire process
under statistical control? Why or why not?

n=4

D3 = 0, and D4 = 2.28

R = (3 + 2 + + 4)/7 = 3.0

UCL D4 R 2.28(3.0) 6.84

LCL D3 R 0(3.0) 0

6S34
6S34
 R-Chart and X-Chart
Control Chart
for sample of
4 resistors
6.84 = UCL

3.0 = R

0 = LCL
| | | | | | | | | | | |
1 2 3 4 5 6 7 8 9 10 11 12

Sample number
The variation of process is in control with 99.73% confidence.

6S35
6S35
R-Chart and X-Chart

n = 4, A2 = 0.73

R = (3 + 2 + + 4)/7 = 3.0

X= (100.5 + 101.5 + + 101.0)/7 99.79

UCL x A2 R 99.79 0.73(3.0) 101.98

LCL x A2 R 99.79 0.73(3.0) 97.6 6S36
6S36
 R-Chart and X-Charts
Control Chart

101.98 = UCL

99.79 = Mean

97.6 = LCL

| | | | | | | | | | | |
1 2 3 4 5 6 7 8 9 10 11 12

Sample number
The central tendency of process is not in control with 99.73%
confidence.
In conclusion, with 99.7% confidence, the entire resistor
production process is not in control since its central tendency is
out of control although its variation is under control.
6S37
6S37
Problem
A part that connects two levels should have a distance between the two
holes of 4. It has been determined that x-bar chart and R-chart should
be set up to determine if the process is in statistical control. The
following ten samples of size four were collected. Calculate the control
limits, plot the control charts, and determine if the process is in control
No. of Sample Mean Range
1 4.01 0.04
2 3.98 0.06
3 4.00 0.02
4 3.99 0.05
5 4.00 0.06
6 3.97 0.02
7 4.02 0.02
8 3.99 0.04
9 3.98 0.05
10 4.01 0.06
6S38
6S38
 R-Chart and X-Chart
Example S6.5: Resistors for electronic circuits are manufactured at Omega
Corporation in Denton, TX. The head of the firms Continuous Improvement
Division is concerned about the product quality and sets up production line
checks. She takes seven random samples of four resistors each to establish
the quality standards. Develop the R-chart and the chart both with 3-
control limits for the production process. Is the entire process under statistical
control? Why or why not?
# of sample Readings of Resistance (ohms)
1 99 100 102 101
2 101 103 101 101
3 98 102 101 99
4 99 100 99 100
5 99 99 98 100
6 95 100 97 96
7 101 99 101 103
6S39
6S39
R-Chart and X-Chart
# of Sample 1 2 3 4 5 6 7
Sample range 3 2 4 1 2 5 4
Sample mean 100.5 101.5 100.0 99.5 99.0 97.0 101.0

3 2 ... 4 n=4 D3 =0 D4 = 2.28

R 3.0
7
UCL D4 R 2.28 3.0 6.84

6.84 = UCL LCL D3 R 0

variation of process
3.0 = R is in control with
99.73% confidence.
0 = LCL
| | | | | | | | | | | |
1 2 3 4 5 6 7 8 9 10 11 12

Sample number 6S40

6S40
 R-Chart and X-Chart
# of Sample 1 2 3 4 5 6 7
Sample range 3 2 4 1 2 5 4
Sample mean 100.5 101.5 100.0 99.5 99.0 97.0 101.0
3 2 ... 4 X= (100.5 + + 101.0)/7 99.8
R 3.0
7
A2 =0.73 UCL x A2 R 99.8 0.73(3.0) 102.0
n=4
LCL x A2 R 99.8 0.73(3.0) 97.6
102.0 = UCL central tendency of
process is not in control
99.8 = X with 99.73% confidence.
97.6 = LCL Thus, entire process is
| | | | | | | | | | | | not in control.
1 2 3 4 5 6 7 8 9 10 11 12

Sample number 6S41

6S41
EX 2 in class
A quality analyst wants to construct a sample mean chart for
controlling a packaging process. Each day last week, he randomly
selected four packages and weighed each. The data from that
activity appears below. Set up control charts to determine if the
process is in statistical control

Day Package 1 Package 2 Package 3 Package 4

Monday 23 22 23 24
Tuesday 23 21 19 21
Wednesday 20 19 20 21
Thursday 18 19 20 19
Friday 18 20 22 20

6S42
6S42
 Statistical Process Control
Chart (SPC)
x -charts (for controlling central tendency)
Control Charts
for Variable Data
R-charts (for controlling variation)
Control
Charts
p-charts (for controlling percent defective)
Control Charts for
Attribute Data
c-charts (for controlling number of defects)

Variable Data (continuous): quantifiable conditions along a

scale, such as speed, length, density, etc.
Attribute Data (discrete): qualitative characteristic or condition,
such as pass/fail, good/bad, go/no go.

6S43
6S43
Control Charts for Attribute
Data
Categorical variables
Good/bad, yes/no, acceptable/unacceptable
Measurement is typically counting defectives
Charts may measure
Percentage of defects (p-chart)
Number of defects (c-chart)

6S44
6S44
P-Charts
p (1 p )
UCL p z p p z
n

p (1 p )
LCL p z p p z
n
where p = mean percent defective overall
the samples
z = number of standard deviations =
3
n = sample size

6S45
6S45
 P-Charts
Example S6.6: Data-entry clerks at ARCO key in thousands of insurance
records each day. One hundred records entered by each clerk were
carefully examined and the number of errors counted. Develop a p-chart
with 3- control limits and determine if the process is in control.
Sample Number Percent Sample Number Percent
Number of Errors Defective Number of Errors Defective
1 6 .06 11 6 .06
2 5 .05 12 1 .01
3 0 .00 13 8 .08
4 1 .01 14 7 .07
5 4 .04 15 5 .05
6 2 .02 16 4 .04
7 5 .05 17 11 .11
8 3 .03 18 3 .03
9 3 .03 19 0 .00
10 2 .02 20 4 .04
Total = 80 6S46
6S46
P-Charts
n = 100
Total number of errors 80
p 0.040
Total number of records examined (100)(20)

Total fraction defective 0.06 0.05 ... 0.04

or , p 0.040
Number of samples 20

p (1 p ) 0.04(1 0.04)
UCL p z 0.04 3 0.099
n 100

p (1 p ) 0.04(1 0.04)
LCL p z 0.04 3 0.019 0
n 100

Because we cannot have a negative percent defective

6S47
6S47
 P-Charts Possible
assignable
causes present
.11
.10 UCL= 0.10
.09
Percent defective

.08
.07
.06
.05
.04 p = 0.04
.03
.02
.01
| | | | | | | | | | LCL= 0.00
.00
2 4 6 8 10 12 14 16 18 20
Sample number
Possible good assignable
causes present
The process is not in control with 99.73% confidence.
6S48
6S48
C-Charts
A c-chart is used when the quality cannot be
measured as a percentage.
Number of car accidents per month at a particular
intersection
Number of complaints the service center of a hotel
receives per week
Number of scratches on a nameplate
Number of dimples found on a metal sheet

6S49
6S49
C-Charts

UCL = c + 3 c LCL = c - 3 c

where c = mean number defective overall th

6S50
6S50
 C-Charts
Example S6.7: Over 9 weeks, Red Top Cab company received the following
numbers of calls from irate passengers: 3, 0, 8, 9, 6, 7, 4, 9, 8, for a total
of 54 complaints. Determine the 3- control limits of a c-Chart.
UCL = 13.35
c = 54 / 9 = 6 complaints /week 14

Number of defect
12
10
UCL = c + 3 c 8
=6+3 6 6 c= 6
= 13.35 4
2 LCL = 0
LCL = c - 3 c 0 | | | | | | | | |
=6-3 6 1 2 3 4 5 6 7
Week Number
8 9

= - 1.35 => 0
The process is in control
Because we cannot have the negative with 99.73% confidence.
number of defective records 6S51
6S51
Managing Quality Summary
1. Effective quality management is data
driven
2. There are multiple tools to identify and
prioritize process problems
3. There are multiple tools to identify the
relationships between variables

6S52
6S52