QS-9000 3rd Edition Acceptance Sampling - Reduces RI (Return on Investment)

Under the previous QS-9000 revision (2nd) our company applied the Chrysler 'Lot Inspection Sampling Table' to inprocess and final (outgoing) inspection, and since we test all product 100% we felt we adequately met this requirement. We did not, however, apply this table to Receiving Inspection, chossing our own sample size based upon supplier performance and C=0. Now Chrysler's PSO requirement is spelled out in QS-9000 (3rd edition)& PSO clearly states in section 6) Incoming & Outgoing Material Qualification/Certfication Plan that the sampling table applies. This will impose a huge burden on RI at a time when our customers are pushing to REDUCE RI, not increase that activity. Any thoughts?

[Marc]

Have you spoken with your Chrysler rep about it yet?

It is Chrysler SQA who is stating this requirement. In the past, Chrysler accepted the sampling plan as applying to outgoing product inspection only. Now SQA is imposing the PSO manual and its Chysler-established sampling plan for RI in lieu of our previous established quality system procedure. As I see it, the danger of a customer imposed sampling plan is that with multiple customers, a company could conceivably have multiple sampling plan requirements. Many of our parts are interchangeable, regardless of customer. In the past we based our RI sampling on supplier delivered product. Inspecting 125-200 o-rings or stampings because the sampling plan chart requires it is not cost effective.

[Marc]

No one commented on this when posted last November. Anyone have any comments now that some of the 3rd edition issues have been addressed?

[Dawn]

This is kinda the same subject, but different topic.
We use three pieces for setup inspection and in-process checks
The auditor states we shall determine why 3 pcs. is a statistcally sound sampling plan or use 4 or 5 samples instead. The QS SPC book does not state what makes 4 or 5 pieces statistically sound-it just states to use it. How do I determine 3 pieces, 4 or 5 are statistically sound? OC curves and the like are only confusing me.

Can anybody provide some answers on the question of sampling methods. How to establish the minimum amount of sample to be collected and establish a viable sample plan?
Thanks.

[CarolX]

You might want to start with a copy of MIL-STD-105. You can find a copy here:
http://stinet.dtic.mil/

good luck!

CarolX

[lday38]

Chrysler 'Lot Inspection Sampling Table' to inprocess and final (outgoing) inspection, where would I get a copy of that? I need a smaping plan after 100% verification of sorted work. If I use C=o I dont have any AQL level just they donot want any defects.We are trying to get Q-9000 registered.