Is AQL necessary for ISO 9001 receiving inspection?

Ajit Basrur

Leader
Admin
Luciano,

I always prefer to apply ANSI Z1.4 for incoming sampling as it is based on statistical approach. It also helps me to avoid lot of customer queries. When I tell them that I apply ANSI Z1.4, they are comfortable (and definitely I am :D )

In my procedure, I have pasted the ANSI tables and an example to show how to select the total num ber of cartons and no. of units to be sampled.
 

Johnson

Involved In Discussions
Thank you very much for your advice
My question is: How can I know the size of my sample to be relevant ?
Which are the rules to establish the size of my sample ?
Is percentage from the quantity recived ?

Best regards
God bless you



Thank you for your ansver

Hello,

To determine the sample size is really a challege job ! That is mostly related what products you make, quality level of the receiving material, importance of the characteristics to be inspected, difficulties and cost of the inspection etc. Here are some hints from my experience:
- If it is a bulk raw material, you only need to take one sample( fixed weight or volume needed for the inspection) from the incoming bach or lot.
- If the quality level is low, you should take more samples which should be enough for conducting statistical analysis, which can the calulation of process capability such as Ppk. If the continuos receiving inspection result shows good/bad capability( less than 1.33 or greater than 1.67 for example), you should reduce/increase the sample size.
-You can use skip lot philosph depending on quality level.
- To make the sampling method simple and reasonable. You should understand if and why the sample size is to high or too low.
- Both 100% inpection and no-inspection ( but verify supplier inspection report ) is applicable.

Hope this help you to set up your own procedure and control leve.

Best Regards
Johnson Shao:bigwave: :bigwave:
 

Coury Ferguson

Moderator here to help
Trusted Information Resource
Luciano,

I always prefer to apply ANSI Z1.4 for incoming sampling as it is based on statistical approach. It also helps me to avoid lot of customer queries. When I tell them that I apply ANSI Z1.4, they are comfortable (and definitely I am :D )

In my procedure, I have pasted the ANSI tables and an example to show how to select the total num ber of cartons and no. of units to be sampled.

luciano said:
Hello friends

In need some help to understand if I need AQL for incoming inspection or not for ISO 9001.
I am trying to implement ISO 9001 in the factory where I work, and I don't know what to do.
If I don't need AQL, how I will establish the appropriate sample for inspection.
If I need AQL, which is more proper.

Thank all of you very much.

ANSI Z1.4 is what most people use, even though it replaced MIL-STD-105 (almost to a tee).

There is no requirement in ISO9001:2000 requiring a Sample Plan. This is usually specified in the contractual document, under most conditions. In my opinion.
 

Bev D

Heretical Statistician
Leader
Super Moderator
I use the Binomial or Poisson directly for sample size seelction.
One issue with using AQL plans is that they determine sample sizes based on the probability of accepting the stated defect (AQL) level. (and yes, I know I could use the LTPD...but I like things as simple and straigthforward as possible.

I always use "accept on 0, reject on 1".

given the defect level, p, that you want a high confidence level in detecting and the confidence level, P(detection), (the probability that you will detect a process that is 'p' defective or worse...)

the Binomial yields: n = natural log(1-P(Detection)) / natural log((1-p)

the Poisson yields: n = -natural log(1-P(detection)) / p

so if you want to catch a lot that is 0.5% defective 95% of the time
p=.005
P(Detection) = .95

n = 598 with the Binomial
n = 600 with the Poisson
 
S

sharon.kuan

Hello,

To determine the sample size is really a challege job ! That is mostly related what products you make, quality level of the receiving material, importance of the characteristics to be inspected, difficulties and cost of the inspection etc. Here are some hints from my experience:
- If it is a bulk raw material, you only need to take one sample( fixed weight or volume needed for the inspection) from the incoming bach or lot.
- If the quality level is low, you should take more samples which should be enough for conducting statistical analysis, which can the calulation of process capability such as Ppk. If the continuos receiving inspection result shows good/bad capability( less than 1.33 or greater than 1.67 for example), you should reduce/increase the sample size.
-You can use skip lot philosph depending on quality level.
- To make the sampling method simple and reasonable. You should understand if and why the sample size is to high or too low.
- Both 100% inpection and no-inspection ( but verify supplier inspection report ) is applicable.

Hope this help you to set up your own procedure and control leve.

Best Regards
Johnson Shao:bigwave: :bigwave:

Hi Johnson,

Thanks for your post. I have a better understanding about sampling inspection now.

Could you please kindly explain more on 'process capability such as Ppk'?

I'm newbie in this QA work and is currently cracking my head :frust: to determine sampling size of final inspection for product release. Our product is test kits for drug detection. Each batch of product is only about 20 test kits. The test kit is not-reusable after testing.

Thanks,
Sharon
 

Johnson

Involved In Discussions
Hi Sharon,

You can find many documents and templates in the "Post Attacments List" of this forum which explain process capability, PPk, Cpk

Here it is how CPK and PPK is calculated :

Cpk = min { (USL - Xbar) /3 sigma, (Xbar - LSL)/3 sigma}

Pp = (USL - LSL) / (6 Sigma)

Ppk = min {(USL - X bar)/3 Sigma, (X bar - LSL)/3 Sigma}

Where:
USL is the upper specification limmit
LSL is the lower specification limmit
X bar is the average value of the group of inspection data.
Sigma is the standard deviation of the group of the data.

There is also other calculation method depending on how is the data collected

Ppk in QS9000 means Preliminary Process Capability Index. It
should be calculated before Mass Production and based on limited product
quantity. Normally, it should be more than 1.67 because it's a short term
process capability which doesn't consider the long term variation. But, in
QS9000 3rd edition, there's no Compulsory Requirement that the Ppk must be more than 1.67. In QS9000 3rd edition, it states like Ppk/Cpk >=1.33.
Another one is in 6-Sigma. Ppk in 6-Sigma means Process Performance Index. It's a long term process capability covered the long term process variation and based on more product quantity. Generally, in 6-Sigma, the Ppk value is less than Cpk value.
Ppk:Overall performance capability of a process, see Cpk.
Cp:A widely used capability index for process capability studies. It may range in value from zero to infinity with a larger value indicating a more capable process. Six Sigma represents Cp of 2.0.

Hope it will not lead you more confused. But you can read some literatures to understand it.
 

Johnson

Involved In Discussions
Luciano,

From statistical theory, the sample size should be chosen according to the table which have been established in the national standard of many countries.

But as you may awared that it is not good for practical use. The concept of AQL does not reflect the philosphy of "zero defect".

You can keep the sampling size " simple and stupid". But the important thing is you should not accept any defect in receiving inspection. That means if you find one single defect during the inspection, you must sort or inspect the whole lot.

I would not suggest use percentage. You may use either "standard " method like regulated in ANSI Z1.4 , or fixed samle size.
 
S

sharon.kuan

Thank you very much for your reply. I just got a copy of the ANSI standard Z1.4 and Z1.9. The Z1.9 is actually an old version (1980), but I supposed it will be sufficient for me to understand some basic.

Anyway, I still not very sure about which standard to follow: Attributes or Variables? I supposed for our product which QC test are measure in mean, standard deviation & CV, the inspection by variables is the appropriate one. Am I right?
 
T

thomastang

1.apperance, Use ANSI/ASQC Z1.4 Single Sampling Plan for General Inspection Level I AQL: 0.25(critical), 1.0(major), 2.5(minor).
2.packing, IQC inspector draws 13 reels (bag, roll, box, etc packing unit) for checking inner conformance , if receipt packing is less than 13 reels(bag, roll, box, etc packing unit), IQC should check all the receipt. Accept=0, Rej=1
3.function,Use ANSI/ASQC Z1.4 Single Sampling Plan for Special Inspection, LEVELS-3, AQL at 1.0.
 
S

sanjayARYAN

Thank you very much .
There is another way to evaluate by sampling without AQL my incomings part ?
Usualy a recive a plastic raw material (PP, PE-LD, PE-HD, ABS....); colour pigments, metal pipes.
How can I establish the corect size of the sample ?
For me it's Ok tu use AQL, but if it's not necessary I don't want to complicate the sistem (we are not a big factory approximately 100 peoples, ) but allso I want to assure the quality of incomings
Many, many thanks.
God bless you

Is it mandatory to have inspection at incoming stage? if we verify only suppliers' inspection report,it will be accepted by ISO auditor?
 
Top Bottom