MNO 3701 Production & Operations
Management
Chapter 10
Statistical Process Control (SPC), Six Sigma & Acceptance Sampling
SPC
Concerned with checking a product or service during its creation.
Control charts
The significant value of SPC, however, is not just to make checks of a single sample but to monitor the
results of many samples over a period of time.
Does this by using control charts, to see whether the process looks as though it is performing as it
should, or alternatively whether it is going out of control. If the process does seem to be going out of
control, steps can be taken before there is a problem.
Most operations chart their quality performance in some way
The control chart is plotting an attribute measure of quality (satisfied or not).
Alternatively, the chart could just as easily represent the average impact resistance of samples.
If the impact resistance is moving above the 'necessary' level, it could indicate that too much material is
being used in the process.
If the reasons for the upward trend are unknown, the management of the operation should want to
investigate the causes.
Looking for trends is an important use of control charts. If the trend suggests the process is getting steadily
worse, it will be worth investigating the process.
If the trend is steadily improving, it may still be worthy of investigation to try to identify what is happening
that is making the process better.
Variation in process quality
Common causes
All processes vary to some extent.
All materials vary a little
Even the environment in which the processing takes place will vary. Given this, it is not surprising that the
measure of quality (whether attribute or variable) will also vary.
1
, Variations which derive from these common causes can never be entirely eliminated
Usually this type of variation can be described by a normal distribution with 99.7 per cent of the variation
lying within ± 3 standard deviations.
The obvious question for any operations manager would be: 'Is this variation in the process performance
acceptable?'
The answer will depend on the acceptable range of weights which can be tolerated by the operation.
This range is called the specification range. If the weight of rice in the box is too small then the organisation
might infringe labelling regulations; if it is too large, the organisation is 'giving away' too much of its product
for free.
Process capability
Process capability is a measure of the acceptability of the variation of the process. The simplest measure
of capability (Cp) is given by the ratio of the specification range to the 'natural' variation of the process
(i.e. ± 3 standard deviations):
Cp= UTL x LTL
6s
Where:
UTL = the upper tolerance limit
LTL = the lower tolerance limit
s = the standard deviation of the process variability.
Generally, if the Cp of a process is greater than 1, it is taken to indicate that the process is 'capable', and
a Cp of less than 1 indicates that the process is not 'capable', assuming that the distribution is normal
The simple Cp measure assumes that the average of the process variation is at the mid-point of the
specification range. Often the process average is offset from the specification range, however
In such cases, one-sided capability indices are required to understand the capability of the process:
Assignable causes of variation
Not all variation in processes is the result of common causes.
There may be something wrong with the process which is assignable to a particular and preventable
cause. Machinery may have worn or been set up badly.
An untrained member of staff may not be following the prescribed procedure for the process.
The causes of such variation are called assignable causes.
2
Management
Chapter 10
Statistical Process Control (SPC), Six Sigma & Acceptance Sampling
SPC
Concerned with checking a product or service during its creation.
Control charts
The significant value of SPC, however, is not just to make checks of a single sample but to monitor the
results of many samples over a period of time.
Does this by using control charts, to see whether the process looks as though it is performing as it
should, or alternatively whether it is going out of control. If the process does seem to be going out of
control, steps can be taken before there is a problem.
Most operations chart their quality performance in some way
The control chart is plotting an attribute measure of quality (satisfied or not).
Alternatively, the chart could just as easily represent the average impact resistance of samples.
If the impact resistance is moving above the 'necessary' level, it could indicate that too much material is
being used in the process.
If the reasons for the upward trend are unknown, the management of the operation should want to
investigate the causes.
Looking for trends is an important use of control charts. If the trend suggests the process is getting steadily
worse, it will be worth investigating the process.
If the trend is steadily improving, it may still be worthy of investigation to try to identify what is happening
that is making the process better.
Variation in process quality
Common causes
All processes vary to some extent.
All materials vary a little
Even the environment in which the processing takes place will vary. Given this, it is not surprising that the
measure of quality (whether attribute or variable) will also vary.
1
, Variations which derive from these common causes can never be entirely eliminated
Usually this type of variation can be described by a normal distribution with 99.7 per cent of the variation
lying within ± 3 standard deviations.
The obvious question for any operations manager would be: 'Is this variation in the process performance
acceptable?'
The answer will depend on the acceptable range of weights which can be tolerated by the operation.
This range is called the specification range. If the weight of rice in the box is too small then the organisation
might infringe labelling regulations; if it is too large, the organisation is 'giving away' too much of its product
for free.
Process capability
Process capability is a measure of the acceptability of the variation of the process. The simplest measure
of capability (Cp) is given by the ratio of the specification range to the 'natural' variation of the process
(i.e. ± 3 standard deviations):
Cp= UTL x LTL
6s
Where:
UTL = the upper tolerance limit
LTL = the lower tolerance limit
s = the standard deviation of the process variability.
Generally, if the Cp of a process is greater than 1, it is taken to indicate that the process is 'capable', and
a Cp of less than 1 indicates that the process is not 'capable', assuming that the distribution is normal
The simple Cp measure assumes that the average of the process variation is at the mid-point of the
specification range. Often the process average is offset from the specification range, however
In such cases, one-sided capability indices are required to understand the capability of the process:
Assignable causes of variation
Not all variation in processes is the result of common causes.
There may be something wrong with the process which is assignable to a particular and preventable
cause. Machinery may have worn or been set up badly.
An untrained member of staff may not be following the prescribed procedure for the process.
The causes of such variation are called assignable causes.
2
