Acceptance Sampling
Acceptance Sampling
Acceptance Sampling
This product may be used as a component orraw material used in the company's production process.
A sample is taken from the lot, and some quality characteristic of the units in the sample is inspected.
On the basis ofthe information given by the sample, a decision regarding disposition of the lot is made.
Accepted lots are put into production whereas the rejected lots may be returned back to the supplier or may be
subjected to some other lot disposition action.
Page 3 of 26
1) Acceptance sampling is used for lot sentencing only and not for the estimation of lot quality.
2) Acceptance sampling plans do not provide any direct form of quality control. Acceptance samnpling
simply accepts or rejects lots. Process control method controls and systematically improves the lot
3) Acceptance sampling is not effective to "inspect quality into the production" but is used as an audit tool
It is useful when the supplier's process is so good that the defective items/units are almost never found or
when to find defective units is not economically justified. For instance, if the capability ratio of the supplier
is 3 or 4 then acceptance sampling is unlikely to discover any defective units.
2) 100% inspection:
It is generally used when the product is extremely critical and passing any defectives would result in an
unacceptably high failure cost at subsequent stages OR when the supplier's capability ratio is too low to trust
the product.
Page 5 of 26
3) Acceptance Sampling:
It is generally used in the following situations:
a) When the testing is destructive. For ex. in case of crackers, bulbs, batteries etc.
b) When the cost of 100 % inspection is extremely high.
c) When 100% inspection is not technologically feasible OR would require so much time that the
production proces gets delayed.
d) When the inspection error rate is high as there are very large number of items to inspect.
e) When the supplier has an excellent quality history and hence 100% inspection is not necessary. But
the supplier's process capability 1s not that satsfactory as to allow no inspection.
Page 6 of 26
6) Rejection of the entire lot provides a motivation to the supplier for quality improvement.
(psychological pressure)
Page 7 of 26
1) As the decision regarding lot sentencing is taken based only on a sample, there are risks of
2) As only a sample is inspected, less information is generated about the product or the product
manufacturing process.
3) Acceptance sampling requires planning and documentation of the procedure whereas 100%
Lot Formation:
Lot formation influences the effectiveness of the acceptance sampling plan.
1)
1) The units in the lot should be produced by the same machines, the same operators, and from
2) Larger lots are preferred over smaller lots. It is economically efficient to inspect larger lots.
3) Lots should be packaged so as to minimize shipping and handling risks of the units.
Page 10 of 26
Outgoing lots
Incominglots Inspection
Fraction defective activity Fraction defective
Po
PiPo
Fraction
Accepted defective
Acceptance sampling program usually requires corrective action when the lots are rejected. This generally
takes the form of 100% inspection or screening of rejected lots, with all discovered defective items either
removed for subsequent rework or returned to the vendor or replaced from a stock of good items. Such
sampling programs are known as rectifying inspection plans. Because the inspection activity affects the
final quality of the outgoing product. This is illustrated in the above figure.
Page 11 of 26
Suppose that incoming lots that are submitted for acceptance inspection have fraction defective po. Some of
the lots will be accepted and some will be rejected. The rejected lots will be sereened and their final fraction
defective will be zero. However, accepted lots have fiaction defective po. Consequently, the outgoing lots
from the inspection activity are a mixture of lots with proportion defective po and proportion defective zero.
So the average fraction defective in the outgoing lots is pi Po- This rectifying inspection program serves to
results from the application of a rectfyng inspection plan. It is the average value of the lot
quality that will be obtained over a long period of time from the sequence of lot from a process,
Thus, lots in the outgoing stage of inspection have an expected number of defective units equal to Pap(N-n).
This can be expressed as an average fraction defective, called the average outgoing quality (AOQ).
A0Q= P(N-n)
P, is probability if acceptingthe lot
N
If the lot size N is very large relative to n, then formula for AOQ reduces to:
Step 2) Inspect all the items included in the sample. Let 'd' denote the number of defectives in that sample.
Step 3) If dsc, then accept the lot and replace all defective items found in the sample by non-defective items.
OR
Ifd>c, then reject the lot. In this case the entire lot is inspected and all the defective items in the lot are replaced by non-defective items.
Page 16 of 26
In such a plan the chance of 100% inspection increases as the percentage of defective items in the
lot increases. Thus, the amount of inspection automatically increases as the lot quality
deteriorates.
Page 17 of 26
Probability of acceptance:
Let the probability that a particular lot submitted for inspection is accepted is denoted by P. If p' be
the fraction defective ofthe lot then for largen and small p',
d~Poisson (m = np')
Then,
e-1P' (np'
a2d=o P(d) = Za=0 (np)t
d!
Page 18 of 26
OC Curve of SSIP:
The important measure of performance of an acceptance sampling plan is the OC curve. The curve plots the
probability of accepting the lot with proportion defective p against the lot fraction defective p. The OC curve
deseribes the diseriminative power of the sampling plan. It shows the probability that a lot submitted with certain
proportion defcctive will be accepted or rejected. Suppose that N is the lot size then distribution of d i.c. no. of
defectives in a sample of size n, is a random varialble following Poisson distribution (as discussed above) with
parameter (m = np) i.e. d- Poisson (m = np), where p is the proportion defective.
Let P,(p) or L{p) denote the probability of accepting the lot of quality p. Then
Then.
If the lot is accepted, only the sample of size n will be inspected i.c. only n units will be inspected.
If the lot is rejected, then the entire lot will be inspected for defectives i.e. N units will be inspected.
nPa + N -
NPa
N+ (n-N)Pa
Page 20 of 26
The value of A0Q depends upon the incoming lot quality p. If the incoming lot quality is very good i.e. p
is small, then most of the lots will be accepted and AOQ is small.
When the incoming lot quality is very bad i.e. p is larger, then most of the lots will be rejected and which
leads to very good level of quality in outgoing lots i.e. A0Q is small.
Page 22 of 26
Thus, AOQ will vary depending upon the value of p. In between these two extremes (i.e. very
small and very large value of p) the A0Q curve rises and passes through a maximum and the
decreases. The maximum ordinate of AOQ curve represents the worst possible average quality
that would result from a rectifying inspection plan and this point is known as AOQL (Average
outgoing Quality Limit). Thus, A0QL is maximum value or peak of AOQ curve.
Sampling plans are designcd such that A0QL does not exceed a certain specified value.
Page 23 of 26
Thus, AOQ will vary depending upon the value of p. In between these two extremes (i.e. very
small and very large value of p) the AOQ curve rises and passes through a maximum and the
decreases. The maximum ordinate of AOQ curve represents the worst possible average quality
that would result from a reetilying inspeetion plan and this point is known as AOQL (Average
outgoing Quality Limit). Thus, AOQL is maximum value or peak of AOQ curve.
Sampling plans are designed such that AOQL does not exceed a certain specified value.
Page 24 of 26
sampling inspection plan. If the lot contams no defective, then no lot will be rejected and annount of inspection for
If the items are all defective, then every lot vill be rejected then amount of inspection will be N.
Ifthe lot quality is 0 <p<1, the average amount of inspection per lot will vary between the sample size and the
lot size N.
ATI is the average number of units inspected per lot while applying the acceptance sampling plan.
Page 25 of 26
Consumer's Risk:
Any sampling scheme would imvolve risk on the part ol the consumer in the sense that he has to
Producer's Risk:
The producer has to face the situation that some good quality lots will be rejected. Producer's risk
is denoted by P, or a. If p is the submitted lot quality and the lot quality is considered as a