AS9100D Sampling Plan


We were recently issued a minor non-conformance for failure to use a sampling plan per AS9100D, 8.5.1 c., 2.:
ensuring that when sampling is used as a means of product acceptance, the sampling plan is justified on
the basis of recognized statistical principles and appropriate for use (i.e., matching the sampling plan to
the criticality of the product and to the process capability).

The majority of the time our orders consist of prototype lots which are 2-4 pieces with a large number of dimensional features. (Occasionally, we have orders for production lots and we use C=0 sampling for these orders.) Some of our customers define their requirements for inspection and we follow their direction. When requirements are not stated and the quantity is small, we need to implement a “plan that is justified on the basis of recognized statistical principles”. We have a First Article Inspection requirement for all jobs, so we have confidence that the part is running as required. We do in-process inspection, but not all features are checked 100% on the second, third, etc. as required by a sampling plan. Running these parts thru the inspection department for a 100% inspection would be quite a challenge. (Just a note: we have an extremely high quality rating with our customers). Our process seems to be working. Any ideas about other options that would conform to the AS9100D requirement would be appreciated.

Mike S.

Happy to be Alive
Trusted Information Resource
If inspecting more than 1 piece in a 2-4 piece lot is not practical, your "sampling" plan is essentially 1 piece per lot. I would document the "quality parameters" or measures of customer protection of such a plan (ERP or AOQL) and ensure that “individuals who understand the consequences of nonconformance” approve the plan. SAE standard ARP 9013 may provide some assistance.

Bev D

Heretical Statistician
Super Moderator
The key word here is capability. there are some processes that don't change appreciably after set-up. In statistical terms the output variation is not large compared to the specifications. If the process is highly capable, statistical sample plans are not helpful or required. statisics doesnt' apply in a vacuum from physics.

If the characteristics is very critical, then only 100% inspection is useful but only for that characteristic. not all of them.

I suspect that you are doing the right thing but either: can't articulate it in appropriate terms to the auditor or the auditor is checkign the 'statistics' box.
Top Bottom