Statistical Process Control

Variation is inherent in any process

No two products will be identical
Variation is due to equipment, materials, operator & environment

Variation assignable causes, chance causes

Assignable causes
larger in magnitude, easily traced & detected

Chance Causes
Inevitable in any process, difficult to trace & detect

S. Ajit

Statistical Process Control

When assignable

causes (Special causes) are present

- Process is out of control
-diff among machine, materials, processes, power

shutdown
When chance

causes (natural causes) are present

- process is in statistical control
- human variability, minor variation in raw material,

machine vibration etc

S. Ajit

Assignable Causes
A process is said to be in statistical control when the location,
spread, or shape of its distribution does not change over time.
After the process is in statistical control, managers use SPC
procedures to detect the onset of assignable causes so that they
can be eliminated.

Location

Spread

Shape
S. Ajit

Statistical Process Control

Statistics: a science that deals with the collection, tabulation,
analysis, interpretation & presentation of quantitative data
SPC: application of statistical methods to the measurement and
analysis of variation in any process

S. Ajit

Two categories of measurement

Variables
Continuous (length, weight, voltage, temperature)
Measurement concerned with degree of conformance to
specification
Expressed in averages & std. deviation

Attributes
Performance characteristic either present / absent
Counted
No. of defectives, no. of red cars produced in a month

S. Ajit

Statistical Process Control

Population (Universe)
Statistics used to refer to any collection of individuals or
of their attributes etc
An unknown pattern of variation from which known
sample can be drawn

Sample
A part or small section selected from the population

S. Ajit

Measures of central tendency- Numerical value that

describes the central position of the data

Mean the arithmetic average. This is used for continuous

data.

Median a value that splits the data into two halves, that is,
one half of the data is smaller than that number, the other half
larger.

Mode this is the category that has the most data (frequently
occurring value)

S. Ajit

Mean

To find the mean, add all of the values, then divide by the
number of values.

Population

Sample

N
n

S. Ajit

Mean example
listing
1
2
3
4
5
6
7
8
9
10
11
12
13
14
n = 15
total

X
14
17
31
28
42
43
51
51
66
70
67
70
78
62
47
737

737/15 =

S. Ajit

x-bar
49.13333

Median

The median is a number chosen so that half of the values in

the data set are smaller than that number, and the other half
are larger.
To find the median
List the numbers in ascending order
If there is a number in the middle (odd number of values)
that is the median
If there is not a middle number (even number of values)
take the two in the middle, their average is the median

S. Ajit

Median Example
listing
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15

X
14
17
28
31
42
43
47
51
51
62
66
67
70
70
78

listing
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16

X
14
17
28
31
42
43
47
51
53
57
62
66
67
70
70
78

51+53 = 52
2

S. Ajit

Mode

The mode is simply the category or value which occurs the

most in a data set.
If a category has radically more than the others, it is a mode.
Generally speaking we do not consider more than two modes
in a data set.

S. Ajit

Mode Example Uni-modal

- The mode is the class, not the frequency.
176
174
172
170
168

A
B
C
D
E

166
164
162
160
158

E
S. Ajit

Mode Example Bi-modal

- The mode is the class, not the frequency.
176
174
172

A
B
C
D
E
F

170
168
166
164
162
160
158

F
S. Ajit

Measures of Dispersion

How the individual observations are spread on either side of

the centre

Measures of dispersion are descriptive statistics that describe

how similar a set of scores are to each other
The more similar the scores are to each other, the lower
the measure of dispersion will be
The less similar the scores are to each other, the higher the
measure of dispersion will be
In general, the more spread out a distribution is, the larger
the measure of dispersion will be

S. Ajit

Measures of Dispersion
Which of the
distributions of scores
has the larger dispersion?
The upper distribution has
more dispersion because the
scores are more spread out
That is, they are less
similar to each other

125
100
75
50
25
0

1 2 3 4 5 6 7 8 9 10

125
100
75
50
25
0
1 2 3 4 5 6 7 8 9 10
16
S. Ajit

Measures of Dispersion

Range
Mean deviation
Standard deviation
variance

S. Ajit

The Range
The range is defined as the difference between the largest
score in the set of data and the smallest score in the set of
data, XL XS
What is the range of the following data:
4 8 1 6 6 2 9 3 6 9
The largest score (XL) is 9; the smallest score (XS) is 1;
the range is XL - XS = 9 - 1 = 8

S. Ajit

When To Use the Range

The range is rarely used in scientific work as it is fairly

insensitive
It depends on only two scores in the set of data, XL and XS
Two very different sets of data can have the same range:
1 1 1 1 9 vs 1 3 5 7 9

S. Ajit

Variance
Variance is defined as the average of the square
deviations
Variance is the mean of the squared deviation scores
The larger the variance is, the more the scores deviate
from the mean
The smaller the variance is, the less the scores deviate
from the mean

N
S. Ajit

Standard Deviation
Since squared units of measure are often awkward to deal
with, the square root of variance is often used instead
The standard deviation is the square root of variance
Standard deviation = variance
Variance = standard deviation2

S. Ajit

Computational Formula Example

X
9
8
6
5
8
6

X2
81
64
36
25
64
36

X-
2
1
-1
-2
1
-1

(X-)
4
1
1
4
1
1

= 42

= 306

=0

= 12

S. Ajit

Computational Formula Example

12

6
2

S. Ajit

Normal Curve
The normal curve is often called the Gaussian distribution,
after Carl Friedrich Gauss, who discovered many of its
properties. Gauss, commonly viewed as one of the greatest
mathematicians of all time.

S. Ajit

Properties of Normal Curve

Also called Bell Curve or Gaussian Curve
Perfectly symmetrical normal distribution
The mean of a distribution is the midpoint of the curve
The tails of the curve are infinite
Mean of the curve = median = mode
The area under the curve is measured in standard deviations
from the mean

S. Ajit

Normal Curve
Has a mean as mid point and standard deviation = 1.
General relationships:
1 s = about 68.26%
2 s = about 95.44%
3 s = about 99.72%

68.26%

95.44%
99.72%

-5

-4

-3

-2

-1

3
S. Ajit

Normal Curve (Z-Score)

Measured in standard deviations relative to the mean of the

curve.
The Z-score can be used to determine an area under the curve
known as a probability.

Xi X
z

S. Ajit

Control chart

Walter Shewhart

Developer of Control Charts in the late 1920s

S. Ajit

Control Chart

A Control Chart is a graph that displays data taken over time

and the variations of this data

Control charts are particularly useful for monitoring quality

and giving early warnings that a process may be going Out of
Control and on its way to producing defective parts.

Objective is to restrict the variation of the process to variation

due to chance causes, by detecting and eliminating the
assignable causes
S. Ajit

Control Chart

The basis is that the laboratory runs control samples together

with the routine samples.

The control values are plotted in a control chart.

S. Ajit

Control Chart
Upper Control Limit
(UCL)
Central Line (CL)

Lower Control Limit

(LCL)

S. Ajit

Control Chart Examples

Control Chart - Types

1. Variables X bar chart, R chart and chart
2. Attributes p chart, np chart, c chart and u chart

S. Ajit

Control Chart - Variables

X bar and R charts
For small samples usually less than 25
Overall mean = averages of sample means

Overall range = average of ranges

For X bar chart

For R chart

UCL = X + (A2 * R)

UCL = D4 * R

LCL = X - (A2 * R)

LCL = D3 * R

A2, D4, D3 constant for a

subgroup size

S. Ajit

Control Chart - Variables

A Pharma company
manufactures a certain
brand of capsules. The
company randomly
picks up samples of 5
capsules from
production at regular
intervals and the
weights in gms of 10
such samples are
given below. Construct
the X bar and R chart
and comment on the
process

Sample
No

Sari
don

colp
ol

met
acin

cold
act

cetri
zin

42

60

65

75

70

39

30

36

45

72

55

66

45

72

78

65

44

50

33

29

75

19

24

80

76

63

48

54

72

36

37

32

48

39

57

75

65

40

45

70

33

40

25

20

50

10

60

63

60

81

55

A2 = 0.31, D4 = 1.78, D3 = 0.22

S. Ajit

Control Chart - Variables

Sample
No

X1

X2

X3

X4

X5

42

60

65

75

70

39

30

36

45

72

55

66

45

72

78

65

44

50

33

29

75

19

24

80

76

63

48

54

72

36

37

32

48

39

57

75

65

40

45

70

33

40

25

20

50

10

60

63

60

81

55

Mean

S. Ajit

Range

Control Chart - Variables

Sample
No

X1

X2

X3

X4

X5

Mean

42

60

65

75

70

62.4

39

30

36

45

72

55

66

45

72

78

65

44

50

33

29

75

19

24

80

76

63

48

54

72

36

37

32

48

39

57

75

65

40

45

70

33

40

25

20

50

10

60

63

60

81

55

S. Ajit

Range

Control Chart - Variables

Sample
No

X1

X2

X3

X4

X5

Mean

42

60

65

75

70

62.4

39

30

36

45

72

44.4

55

66

45

72

78

65

44

50

33

29

75

19

24

80

76

63

48

54

72

36

37

32

48

39

57

75

65

40

45

70

33

40

25

20

50

10

60

63

60

81

55

S. Ajit

Range

Control Chart - Variables

Sample
No

X1

X2

X3

X4

X5

Mean

42

60

65

75

70

62.4

39

30

36

45

72

44.4

55

66

45

72

78

63.2

65

44

50

33

29

75

19

24

80

76

63

48

54

72

36

37

32

48

39

57

75

65

40

45

70

33

40

25

20

50

10

60

63

60

81

55

S. Ajit

Range

Control Chart - Variables

Sample
No

X1

X2

X3

X4

X5

Mean

42

60

65

75

70

62.4

39

30

36

45

72

44.4

55

66

45

72

78

63.2

65

44

50

33

29

44.2

75

19

24

80

76

63

48

54

72

36

37

32

48

39

57

75

65

40

45

70

33

40

25

20

50

10

60

63

60

81

55

S. Ajit

Range

Control Chart - Variables

Sample
No

X1

X2

X3

X4

X5

Mean

42

60

65

75

70

62.4

39

30

36

45

72

44.4

55

66

45

72

78

63.2

65

44

50

33

29

44.2

75

19

24

80

76

54.8

63

48

54

72

36

54.6

37

32

48

39

57

42.6

75

65

40

45

70

59

33

40

25

20

50

33.6

10

60

63

60

81

55

63.8

S. Ajit

Range

Control Chart - Variables

Sample
No

X1

X2

X3

X4

X5

Mean

Range

42

60

65

75

70

62.4

33

39

30

36

45

72

44.4

55

66

45

72

78

63.2

65

44

50

33

29

44.2

75

19

24

80

76

54.8

63

48

54

72

36

54.6

37

32

48

39

57

42.6

75

65

40

45

70

59

33

40

25

20

50

33.6

10

60

63

60

81

55

63.8

S. Ajit

Control Chart - Variables

Sample
No

X1

X2

X3

X4

X5

Mean

Range

42

60

65

75

70

62.4

33

39

30

36

45

72

44.4

42

55

66

45

72

78

63.2

65

44

50

33

29

44.2

75

19

24

80

76

54.8

63

48

54

72

36

54.6

37

32

48

39

57

42.6

75

65

40

45

70

59

33

40

25

20

50

33.6

10

60

63

60

81

55

63.8

S. Ajit

Control Chart - Variables

Sample
No

X1

X2

X3

X4

X5

Mean

Range

42

60

65

75

70

62.4

33

39

30

36

45

72

44.4

42

55

66

45

72

78

63.2

33

65

44

50

33

29

44.2

75

19

24

80

76

54.8

63

48

54

72

36

54.6

37

32

48

39

57

42.6

75

65

40

45

70

59

33

40

25

20

50

33.6

10

60

63

60

81

55

63.8

S. Ajit

Control Chart - Variables

Sample
No

X1

X2

X3

X4

X5

Mean

Range

42

60

65

75

70

62.4

33

39

30

36

45

72

44.4

42

55

66

45

72

78

63.2

33

65

44

50

33

29

44.2

36

75

19

24

80

76

54.8

63

48

54

72

36

54.6

37

32

48

39

57

42.6

75

65

40

45

70

59

33

40

25

20

50

33.6

10

60

63

60

81

55

63.8

S. Ajit

Control Chart - Variables

Sample
No

X1

X2

X3

X4

X5

Mean

Range

42

60

65

75

70

62.4

33

39

30

36

45

72

44.4

42

55

66

45

72

78

63.2

33

65

44

50

33

29

44.2

36

75

19

24

80

76

54.8

61

63

48

54

72

36

54.6

36

37

32

48

39

57

42.6

25

75

65

40

45

70

59

35

33

40

25

20

50

33.6

30

10

60

63

60

81

55

63.8

26

S. Ajit

Control Chart - Variables

Sample
No

X1

X2

X3

X4

X5

Mean

Range

42

60

65

75

70

62.4

33

39

30

36

45

72

44.4

42

55

66

45

72

78

63.2

33

65

44

50

33

29

44.2

36

75

19

24

80

76

54.8

61

63

48

54

72

36

54.6

36

37

32

48

39

57

42.6

25

75

65

40

45

70

59

35

33

40

25

20

50

33.6

30

10

60

63

60

81

55

63.8

26

522.6

357

Total

S. Ajit

Control Chart - Variables

Sample
No

X1

X2

X3

X4

X5

42

60

65

75

70

62.4

33

39

30

36

45

72

44.4

42

55

66

45

72

78

63.2

33

65

44

50

33

29

44.2

36

75

19

24

80

76

54.8

61

R = 357 / 10

63

48

54

72

36

54.6

36

= 35.7

37

32

48

39

57

42.6

25

75

65

40

45

70

59

35

33

40

25

20

50

33.6

30

10

60

63

60

81

55

63.8

26

522.6

357

Total

Mean Range

S. Ajit

X = 522.6 / 10
= 52.26

Control Chart - Variables

For X bar chart
UCL

X = 52.26

= X + (A2 * R)
= 52.26 + (0.31 * 35.7) = 63.327

R = 35.7
LCL
A2 = 0.31,
D4 = 1.78,
D3 = 0.22

= X - (A2 * R)

= 52.26 - (0.31 * 35.7) = 41.193

For R chart

UCL = D4 * R = 1.78 * 35.7 = 63.54

LCL = D3 * R = 0.22 * 35.7 = 7.854

S. Ajit

Control Chart
UCL = 63.32

CL = 52.26

LCL = 41.19

S. Ajit

Control Chart
UCL = 63.32

CL = 52.26

LCL = 41.19

S. Ajit

Control Chart
UCL = 63.32

CL = 52.26

LCL = 41.19

S. Ajit

Control Chart
UCL = 63.32

CL = 52.26

LCL = 41.19

S. Ajit

Control Chart
UCL = 63.32

CL = 52.26

LCL = 41.19

S. Ajit

Control Chart
UCL = 63.32

CL = 52.26

LCL = 41.19

S. Ajit

Control Chart
UCL = 63.32

CL = 52.26

LCL = 41.19

S. Ajit

Control Chart
UCL = 63.32

CL = 52.26

LCL = 41.19

S. Ajit

Control Chart for Attributes

Attributes cannot be measured but can be observed and counted.
Attributes data are usually easy to collect, often by visual
inspection incorrect color, cracks, blow holes etc
Defect: A single non-conforming quality characteristic of an item
Defective: Item having one or more defects

S. Ajit

Control Chart for Attributes

p Chart
Monitors the proportion of defective items, variable subgroup size
25 to 30 samples , each of size around 100 are taken over a time
period to analyse for non - conformities.
p fraction defectives ratio of the no. of defective articles found in
any inspection to the total no. of articles actually inspected.

p
p - central line - average fraction defective p
n

p(1 p)
UCL p 3
n

p(1 p)
LCL p 3
n
S. Ajit

Control Chart for Attributes

p Chart

p
p - central line - average fraction defective p
n

p(1 p)
UCL p 3
n

p(1 p)
LCL p 3
n

S. Ajit

Control Chart for Attributes

np Chart
Monitors the proportion of defective items, Constant subgroup size

n p- central line - average fraction defective

UCL n p 3 n p(1 n p)

LCL n p 3 n p(1 n p)

S. Ajit

Example No. 4
Record of defective observed
during the inspection process of
an automatic machine producing
small capacitors of std size.
(i) Find 99.7% control limits for
the process
(ii) Draw the control chart &
comment on the state of
control of the process

Sample
No

Sample
Size

No. of
defectives

25

50

45

55

35

40

50

65

30

10

25

11

55

12

40

13

50

14

25

15

40S. Ajit

Control Chart for Attributes

Soln:
99.7% means 3
Also size of the sample not uniform
Average sample size (n) = sum of size of each sample / total no. of samples
n = 25+50+45+55+35++40 / 15
= 630 / 15

= 42

S. Ajit

Control Chart for Attributes

Average fraction defective

Central line

p = total no. of defectives / total no. of items inspected

= 3+5+1+2+0++2+2 / 630
= 44 / 630

= 0. 07
Upper Control Limit

p(1 p)
UCL p 3
n

Lower Control Limit

p(1 p)
LCL p 3
n
S. Ajit

Control Chart for Attributes

p(1 p)
UCL p 3
n
0.07 (1 0.07 )
UCL 0.07 3
42
UCL = 0.188

S. Ajit

Control Chart for Attributes

p(1 p)
LCL p 3
n
0.07 (1 0.07 )
LCL 0.07 3
42
LCL = -0.048
Since count cannot be less than 0
LCL = 0
S. Ajit

Fraction defectives
= no. of defectives / Sample size

No. of
defectives

Sample
Size

25

50

45

55

35

40

50

65

30

25

55

40

50

25

40S. Ajit

fraction
defectives

Fraction defectives
= no. of defectives / Sample size

No. of
defectives

Sample
Size

fraction
defectives

25

3/25 = 0.12

50

45

55

35

40

50

65

30

25

55

40

50

25

40 S. Ajit

Fraction defectives
= no. of defectives / Sample size

No. of
defectives

Sample
Size

fraction
defectives

25

3/25 = 0.12

50

5/50 = 0.10

45

55

35

40

50

65

30

25

55

40

50

25

40 S. Ajit

Fraction defectives
= no. of defectives / Sample size

No. of
defectives

Sample
Size

fraction
defectives

25

3/25 = 0.12

50

5/50 = 0.10

45

1/45 = 0.022

55

35

40

50

65

30

25

55

40

50

25

40 S. Ajit

Fraction defectives
= no. of defectives / Sample size

No. of
defectives

Sample
Size

fraction
defectives

25

3/25 = 0.12

50

5/50 = 0.10

45

1/45 = 0.022

55

0.036

35

40

0.025

50

0.018

65

0.03

30

0.010

25

0.08

55

0.09

40

0.10

50

0.06

25

0.08

40 S. Ajit

0.05

Control Chart
UCL = 0.19

CL = 0.07

LCL = 0

S. Ajit

Control Chart
UCL = 0.19

CL = 0.07

LCL = 0

S. Ajit

Control Chart
UCL = 0.19

CL = 0.07

LCL = 0

S. Ajit

Control Chart
UCL = 0.19

CL = 0.07

LCL = 0

S. Ajit

Control Chart
UCL = 0.19

CL = 0.07

LCL = 0

S. Ajit

Control Chart
UCL = 0.19

CL = 0.07

LCL = 0

S. Ajit

Control Chart
UCL = 0.19

CL = 0.07

LCL = 0

All points lie inside the control limits, process is in control

S. Ajit

Control Chart for Attributes

C Chart
No. of defects in a subgroup of constant sample size
(a defective TV defects in board, tuner, color control, volume control)
c no. of defects; n no. of samples

total no of defects in all sample

c
total no. of sample

UCL c 3 c

c
c
n

LCL c 3 c
S. Ajit

Control Chart for Attributes

u Chart
No. of defects in a subgroup of variable sample size
(textiles, films, paper continuous production)
u no. of defects;

total no of defects in all sample

u
no. of units in a sample

u
UCL u 3
n

u
u
n

u
LCL u 3
n
S. Ajit

Control Charts
1. Control charts are designed to be used by employees in their
work areas rather than by inspectors or quality control
personnel
2. Under SPC burden of quality rests with the employees
3. Control charts can also be applied to services organization
(cheque processing accuracy in banks, on-time delivery of
meals in restaurants, etc)

S. Ajit

Process Capability
Control limits are determined using averages but specifications are
for individual values
Even when the process in control, the individual item may not be
within specification limits
Process capability is the ability of the process to meet the design
specifications for a service or product.
Nominal value is a target for design specifications.
Tolerance is an allowance above or below the nominal value.
Process Capability

=6

= 6
d
2

d2 statistical factor

S. Ajit

Process Capability
Nominal
value
Process distribution
Lower
specification

20

Upper
specification

25
Process is capable

30

Minutes

Process Capability
Nominal
value
Process distribution
Lower
specification

20

Upper
specification

25
Process is not capable

30

Minutes

Effects of Reducing
Variability on Process Capability
Nominal value
Six sigma

Four sigma

Two sigma

Lower
specification

Upper
specification

Mean

Process Capability
Measures of process capability:

1. Process Capability Index Cp:

USL LSL
Cp
6

(i) If Cp > 1

Process variation is less than specification

Process is capable

(ii) If Cp < 1

Process not capable of meeting specification

(iii) If Cp = 1

Process is just meeting specification

Normally accepted value is 1.33, nowadays companies stress for Cp = 2

S. Ajit

Process Capability
Measures of process capability:
2. Process Performance Capability Index Cpk : Cpk measures how
close you are to your target and how consistent you are to around
your average performance
Cpk is an index (a simple number) which measures how close a
process is running to its specification limits, relative to the natural
variability of the process. The larger the index, the less likely it is
that any item will be outside the specs." Neil Polhemus

USL X or X LSL
C pk min

S. Ajit

Process Capability
A Pharma company
manufactures a certain
brand of capsules. The
company randomly
picks up samples of 5
capsules from
production at regular
intervals and the
weights in gms of 10
such samples are
given below.
Determine whether
the process is capable
with a tolerance of 5

Sample
No

X1

X2

X3

X4

X5

42

60

65

75

70

39

30

36

45

72

55

66

45

72

78

65

44

50

33

29

75

19

24

80

76

63

48

54

72

36

37

32

48

39

57

75

65

40

45

70

33

40

25

20

50

10

60

63

60

81

55

A2 = 0.31, D4 = 1.78, D3 = 0.22, d2 = 2.326

S. Ajit

Process Capability
Sample
No

X1

X2

X3

X4

X5

42

60

65

75

70

62.4

33

39

30

36

45

72

44.4

42

55

66

45

72

78

63.2

33

65

44

50

33

29

44.2

36

75

19

24

80

76

54.8

61

R = 357 / 10

63

48

54

72

36

54.6

36

= 35.7

37

32

48

39

57

42.6

25

75

65

40

45

70

59

35

33

40

25

20

50

33.6

30

10

60

63

60

81

55

63.8

26

522.6

357

Total

Mean Range

S. Ajit

X = 522.6 / 10
= 52.26

Process Capability
X = 522.6 / 10
= 52.26
R = 357 / 10

USL = 52.26 + 0.5 = 57.26

LSL = 52.26 0.5 = 47.26

USL LSL
Cp
6

= 35.7

Process Capability

=6

= 6

d
d2 statistical factor
2

S. Ajit

Process Capability
X = 522.6 / 10

USL = 52.26 + 5 = 57.26

= 52.26

LSL = 52.26 5 = 47.26

R = 357 / 10
= 35.7

R

d
2

= 35.7 / 2.326 = 15.35

S. Ajit

Process Capability
USL = 52.26 + 5 = 57.26
LSL = 52.26 5 = 47.26

R

d
2

= 35.7 / 2.326 = 15.35

USL LSL
Cp
6
57 .26 47 .26
Cp
6 *15 .35

= 10 / 92.1 = 0.108

S. Ajit

Process Capability

USL X or X LSL
C pk min

USL X

= 57.26 52.26 = 5

X LSL

= 52.26 47.26 = 5

C pk
3 *15 .35