Statistical

Process Control
 Statistical Process Control (SPC)

 Uses statistics & control charts to tell

when to adjust process
 Developed by Shewhart in 1920’s
 Involves
 Creating standards (upper & lower limits)
 Measuring sample output (e.g. mean wgt.)

 Taking corrective action (if necessary)

 Done while product is being produced

Statistical Process Control Steps
Statistical Process Control Steps

Start
Statistical Process Control Steps

Produce Good
Start
Provide Service
Statistical Process Control Steps

Produce Good
Start
Provide Service

Take Sample
Statistical Process Control Steps

Produce Good
Start
Provide Service

Take Sample

Inspect Sample
Statistical Process Control Steps

Produce Good
Start
Provide Service

Take Sample

Inspect Sample

Create
Control Chart
Statistical Process Control Steps

Produce Good
Start
Provide Service

Assign.
Take Sample Causes?

Inspect Sample

Create
Control Chart
Statistical Process Control Steps

Produce Good
Start
Provide Service
No
Assign.
Take Sample Causes?

Inspect Sample

Create
Control Chart
Statistical Process Control Steps

Produce Good
Start
Provide Service
No
Assign.
Take Sample Causes?
Yes
Inspect Sample Stop Process

Create
Control Chart
Statistical Process Control Steps

Produce Good
Start
Provide Service
No
Assign.
Take Sample Causes?
Yes
Inspect Sample Stop Process

Create
Find Out Why
Control Chart
 Types of data

 Variables data - quality characteristics

that are measurable values.
 Measurable and normally continuous; may
take on any value.
 Attribute data - quality characteristics
that are observed to be either present or
absent, conforming or nonconforming.
 Countable and normally discrete; integer
 Variability

 Random  Non-Random
 common causes  special causes
 inherent in a  due to identifiable
process factors
 can be eliminated  can be modified
only through through operator or
improvements in the management action
system
 Descriptive statistics

 Measures of Central Tendency

 Describes the center position of the data
 Mean Median Mode

 Measures of Dispersion
 Describes the spread of the data
 Range Variance Standard deviation
Measures of central tendency: Mean
N
1
N
 xi
Arithmetic mean x = i1

h
X  
i1
fi  xi
n

So, for example, if the data are : 0,2,5,9,12 the mean is

(0+2+5+9+12)/5 = 28/5 = 5.6
S where xi is one observation,  means “add up what
follows” and N is the number of observations
 Measures of central tendency:
Median - mode
 Median = the observation in the ‘middle’
of sorted data
 Mode = the most frequently occurring
value
 Median and mode

100 91 85 84 75 72 72 69 65

Mode

Median

Mean = 79.22
 Measures of dispersion:
range
 The range is calculated by taking the
maximum value and subtracting the
minimum value.

2 4 6 8 10 12 14

Range = 14 - 2 = 12
 Measures of dispersion:
variance
 Calculate the deviation from the mean for
every observation.
 Square each deviation
 Add them up and divide by the number of
observations
n

 ( xi  X  
 
 i 1
n
 Measures of dispersion:
standard deviation
 The standard deviation is the square root of
the variance. The variance is in “square
units” so the standard deviation is in the same
units as x.

(x  X i

For ungrouped Data
 i 1

n
h h
n( fi.xi )  ( fi.xi
2
For grouped Data
 i 1 i 1

n(n 1)
 Control charts

 Control charts are powerful aids to

understanding the performance of a process
over time.

PROCESS
Input Output
What’s causing variability?
 Process Control Chart

 Graph of sample data plotted over time

 Process Control Chart

 Graph of sample data plotted over time

X
60
40
20
0
1 3 5 7 9 11
Time
 Process Control Chart

 Graph of sample data plotted over time

X
60
40 Process
20 Average

0
1 3 5 7 9 11
Time
 Process Control Chart

 Graph of sample data plotted over time

X
60
UCL
40 Process
20 Average
± 3s
0 LCL

1 3 5 7 9 11
Time
 Process Control Chart

 Graph of sample data plotted over time

X
60
UCL
40 Process
20 Average
± 3s
0 LCL

1 3 5 7 9 11
Time
 Process Control Chart

 Graph of sample data plotted over time

X
60
UCL
40 Process
20 Average
± 3s
0 LCL

1 3 5 7 9 11
Time
 Process Control Chart

 Graph of sample data plotted over time

X
60
UCL
40 Process
20 Average
± 3s
Chance 0 LCL
Variation
1 3 5 7 9 11
Time
 Process Control Chart

 Graph of sample data plotted over time

X
60
UCL
40 Process
20 Average
± 3s
Chance 0 LCL
Variation
1 3 5 7 9 11
Time
 Process Control Chart

 Graph of sample data plotted over time

X
60
Assignable UCL
Cause 40 Process
Variation Average
20
± 3s
Chance 0 LCL
Variation
1 3 5 7 9 11
Time
 Process Control Chart

 Graph of sample data plotted over time

X
60
Assignable UCL
Cause 40 Process
Variation Average
20
± 3s
Chance 0 LCL
Variation
1 3 5 7 9 11
Time
 Process Control Chart

 Graph of sample data plotted over time

X
60
Assignable UCL
Cause 40 Process
Variation Average
20
± 3s
Chance 0 LCL
Variation
1 3 5 7 9 11
Time
 Control Chart Purposes

 Show changes in data pattern

 Example: Trends
 Make corrections before process is out of control
 Show causes of changes in data
 Special or assignable causes
 Data outside control limits or trend in data
 Chance or common causes
 Random variations around average
 Control charts identify variation

 Chance causes - “common cause”

 inherent to the process or random and not
controllable
 if only common cause present, the process

is considered stable or “in control”

 Assignable causes - “special cause”
 variation due to outside influences
 if present, the process is “out of control”
 A Process Is in Control If …

1. … no sample points outside limits

2. … most points near process average
3. … about equal number of points above and
below centerline
4. … points appear randomly distributed
 Control Chart Patterns

UCL

UCL

LCL

Sample observations
consistently below the LCL
center line
Sample observations
consistently above the
center line
 Control Chart Patterns (cont.)

UCL

UCL

LCL

Sample observations
consistently increasing LCL

Sample observations
consistently decreasing
 Where to Use Control Charts

 Process has a tendency to go out of control

 Process is particularly harmful and costly if it goes out of
control
 Examples
 at the beginning of a process because it is a waste of
time and money to begin production process with bad
supplies
 before a costly or irreversible point, after which product
is difficult to rework or correct
 before and after assembly or painting operations that
might cover defects
 before the outgoing final product or service is delivered
 Control chart components

 Centerline
 shows where the process average is
centered or the central tendency of the data
 Upper control limit (UCL) and Lower
control limit (LCL)
 describes the process spread
 Control Chart Types

Continuous Categorical or
Numerical Data Control Discrete Numerical
Charts Data

Variables Attributes
Charts Charts

R X P np
Chart Chart Chart Chart
 Control Charts
 A graph that establishes
control limits of a
process  Types of charts
 Control limits  Attributes
 upper and lower bands of
a control chart
 p-chart
 c-chart
 Variables
 range (R-chart)
 mean (x bar –
chart)
Control Charts for Variables

 Mean chart ( x -Chart )

 uses average of a sample
 Range chart ( R-Chart )
 uses amount of dispersion in a
sample
 X-bar and R charts

 The X-bar chart - used to detect changes

in the mean between subgroups
 tests central tendency or location effects
 The R chart - used to detect changes in
variation within subgroups
 tests dispersion effects
`X Chart
 Control Chart Types

Continuous Categorical or
Numerical Data Control Discrete Numerical
Charts Data

Variables Attributes
Charts Charts

R X P np
Chart Chart Chart Chart
 `X Chart

 Type of variables control chart

 Interval or ratio scaled numerical data
 Shows sample means over time
 Monitors process average
 Example: Weight samples of coffee &
compute means of samples; Plot
 Determine centerline

 The centerline should be the population

mean, X
 Since it is unknown, we use X double
bar, or the grand average of the
subgroup averages.
m

 X i

X  i1
k
 UCL, LCL calculation

UCL  X  3
LCL  X  3
  standard deviation
 `X Chart
Control Limits
 `X Chart
Control Limits

UCL X = X + A2 × R
 `X Chart
Control Limits

UCL X = X + A2 × R
Sample
mean at
time i

å Xi
i =1
i =1
X =
k
 `X Chart
Control Limits
From
Table
UCL X = X + A2 × R
Sample
mean at
time i

å Xi
i =1
i =1
X =
k
 `X Chart
Control Limits
From
Table
UCL X = X + A2 × R
Sample
mean at
time i

k k Sample

å Xi å Ri range at
time i
i =1
i =1 i =1
i =1
X = R =
k k No. of
Samples
 `X Chart
Control Limits
From
Table
UCLX  X  A2  R
Sample
mean at
LCLX  X  A2  R time i

k k Sample
 Xi  Ri range at
time i
i 1 i 1
X  R 
k k No. of
Samples
 UCL and LCL Calculation

A2 is a Factor depends on subgroup size, refer from Table

= åx Grand Average
x=
k

=
UCL = x + A2 R
=
LCL = x – A2R
 Appendix:
Determining Control Limits for x-bar and R-Charts

SAMPLE SIZE FACTOR FOR x-CHART FACTORS FOR R-CHART

n A2 D3 D4

Fact 2
3
1.88
1.02
0.00
0.00
3.27
2.57
4 0.73 0.00 2.28
ors 5
6
0.58
0.48
0.00
0.00
2.11
2.00
7 0.42 0.08 1.92
8 0.37 0.14 1.86
9 0.44 0.18 1.82
10 0.11 0.22 1.78
11 0.99 0.26 1.74
12 0.77 0.28 1.72
13 0.55 0.31 1.69
14 0.44 0.33 1.67
15 0.22 0.35 1.65
16 0.11 0.36 1.64
17 0.00 0.38 1.62
18 0.99 0.39 1.61
19 0.99 0.40 1.61
20 0.88 0.41 1.59
 `X Chart
Thinking Challenge
You’re manager of a 500-
room hotel. You want to
analyze the time it takes
to deliver luggage to the
room. For 7 days, you
collect data on 5
deliveries per day. Is the
process in control?

Alone Group Class

 R &`X Chart
Hotel Data
Sample
Day Delivery Time MeanRange
1 7.30 4.20 6.10 3.45 5.55 5.32 3.85
2 4.60 8.70 7.60 4.43 7.62 6.59 4.27
3 5.98 2.92 6.20 4.20 5.10 4.88 3.28
4 7.20 5.10 5.19 6.80 4.21 5.70 2.99
5 4.00 4.50 5.50 1.89 4.46 4.07 3.61
6 10.108.10 6.50 5.06 6.94 7.34 5.04
7 6.77 5.08 5.90 6.90 9.30 6.79 4.22
 `X Chart
Control Limits Solution*
 `X Chart
Control Limits Solution*
kk
å X ii 5 .32 + 6.59 + L + 6.79
i =1
i =1
X = = = 5 .813
k 7
 `X Chart
Control Limits Solution*
kk
å X ii 5 .32 + 6.59 + L + 6.79
i =1
i =1
X = = = 5 .813
k 7
k
å R ii 3.85 + 4 .27 + L + 4 .22
i =1
i =1
R = = = 3.894
k 7
 `X Chart
Control Limits Solution*
kk
å X ii 5 .32 + 6.59 + L + 6.79
i =1
i =1
X = = = 5 .813
k 7
k From Table
å R ii
3.85 + 4 .27 + L + 4 .22
(n = 5)
i =1
i =1
R = = = 3.894
k 7
UCL X = X + A2 × R = 5 .813 + 0.577 XX 3.894 = 8.060
 Partial Table for
Control Chart Limits

n A2 D4 D3
2 1.880 3.268 0
3 1.023 2.574 0
4 0.729 2.282 0
5 0.577 2.114 0
6 0.483 2.004 0
7 0.419 1.924 0.076
 `X Chart
Control Limits Solution*
k

X i
5.32 6.59    6.79
X i 1
  5.813
k 7
k From Table
R3.85 4.27    4.22
i
(n = 5)
R 
i 1
 3.894
k 7
UCLX  X  A2  R  5.813  0.577 3.894  8.060
LCLX  X  A2  R  5.813  0.577 3.894  3.566
 `X Chart
Control Chart Solution*

X, Minutes UCL

8
6
4
2 LCL
0
1 2 3 4 5 6 7
Day
R Chart
 Control Chart Types

Continuous Categorical or
Numerical Data Control Discrete Numerical
Charts Data

Variables Attributes
Charts Charts

R X P np
Chart Chart Chart Chart
 R Chart

 Type of variables control chart

 Interval or ratio scaled numerical data
 Shows sample ranges over time
 Difference between smallest & largest
values in inspection sample
 Monitors variability in process
 Example: Weight samples of coffee &
compute ranges of samples; Plot
 R Chart
Control Limits
 R Chart
Control Limits

UCL R = D 4 ×R
 R Chart
Control Limits

From Table
UCL R = D 4 ×R
 R Chart
Control Limits

From Table
UCL R = D 4 ×R

å Ri
Sample range at
time i
i =1
i =1
R =
k No. of
Samples
 R Chart
Control Limits

UCLR  D4  R From Table

LCLR  D3  R

k
 Ri Sample range
at time i
i 1
R 
k No. of
Samples
 R Chart
Example
You’re manager of a 500-
room hotel. You want to
analyze the time it takes
to deliver luggage to the
room. For 7 days, you
collect data on 5
deliveries per day. Is the
process in control?
R &`X Chart
Hotel Data
 R &`X Chart
Hotel Data
Sample
Day Delivery Time MeanRange
 R &`X Chart
Hotel Data
Sample
Day Delivery Time MeanRange
1 7.30 4.20 6.10 3.45 5.55
 R &`X Chart
Hotel Data
Sample
Day Delivery Time MeanRange
1 7.30 4.20 6.10 3.45 5.55 5.32

7.30 + 4.20 + 6.10 + 3.45 + 5.55

Sample Mean =
5
 R &`X Chart
Hotel Data
Sample
Day Delivery Time MeanRange
1 7.30 4.20 6.10 3.45 5.55 5.32 3.85
Largest Smallest

Sample Range = 7.30 - 3.45

 R &`X Chart
Hotel Data
Sample
Day Delivery Time MeanRange
1 7.30 4.20 6.10 3.45 5.55 5.32 3.85
2 4.60 8.70 7.60 4.43 7.62 6.59 4.27
3 5.98 2.92 6.20 4.20 5.10 4.88 3.28
4 7.20 5.10 5.19 6.80 4.21 5.70 2.99
5 4.00 4.50 5.50 1.89 4.46 4.07 3.61
6 10.108.10 6.50 5.06 6.94 7.34 5.04
7 6.77 5.08 5.90 6.90 9.30 6.79 4.22
 R Chart
Control Limits Solution
 R Chart
Control Limits Solution
k

å Ri 3.85 + 4 .27 + L + 4 .22

i =1
R = = = 3.894
k 7
 R Chart
Control Limits Solution
k

å Ri 3.85 + 4 .27 + L + 4 .22

i =1
R = = = 3.894
k 7

UCL R = D 4 × R = 2.114 x 3.894 = 8.232

From Table
(n = 5)
 Partial Table for
Control Chart Limits

n A2 D4 D3
2 1.880 3.268 0
3 1.023 2.574 0
4 0.729 2.282 0
5 0.577 2.114 0
6 0.483 2.004 0
7 0.419 1.924 0.076
 R Chart
Control Limits Solution
k

 Ri
3 . 85  4 . 27    4 . 22
R  i 1
  3 . 894
k 7

UCL R  D 4  R   2 . 114  3 . 894   8 . 232

From Table
LCL R  D 3  R   0  3 . 894  0 (n = 5)
 R Chart
Control Chart Solution

R, Minutes UCL
8
6
4
2
0
1 2 3 4 5 6 7
Day
 Using x- bar and R-Charts Together

 Process average and process variability

must be in control.
 It is possible for samples to have very
narrow ranges, but their averages is
beyond control limits.
 It is possible for sample averages to be in
control, but ranges might be very large.