AQL level

I am hoping to reduce our sampling requirement, so any info or logic behind why you use what you use would be helpful.

At a previous company (ISO 13485 and FDA) I was a QE.

I would routinely work with customers to identify where we could lower our sample size for FINAL INSPECTION.

Rough Process:
1) Identify the dimensions that were jacking up up the AQL
2) Perform process capability studies on each identified dimension. CPK had to be <=1.33 over X number of runs (usually three). If everything checked out the AQL could then be lowered for that dimension only. We'd hit all of the tight dimensions on a print & push it until we had parts moving a lot faster through the whip. Many parts that were 100% inspect jobs for many years were turned into three piece operations. Was really nice. Customer was a stickler about paperwork due to the med registrations but it was worth it in the end.
3) Every 5th, 8th, or 10th lot (depending on the customer) we'd have to jack to AQL back to "normal" to demonstrate our continuous capability.

The above has to be supported by a robust in process inspection system as well as a production team that plays along. Everyone has to check their work and follow their process plans but if they did, it worked.

Most companies I have seen try to do stuff like this fail because people get lax over time. If you're the type of company where things taper off after a couple of months of a certain initiative, this isn't for you.

So a little history. The company I work for has changed their sample plan probably 3 times since I have worked here in the past 2 years. Always by people outside of quality. The last time it was changed I was on my honeymoon and they changed it without me knowing while I was gone because I am sure they knew I would have raised a red flag if I was there. (that person is gone)

I am finally having an opportunity to give some feedback on the situation with a recent change of leadership. Even though we are technically using Z1.4 AQL 1.0 C=0 and MIS-50822 we have a grab bag of numbers for a few of our inspection levels. I am merely trying to gather information so I can make a more informed answer. I am not trying to reduce our sample plan to cheat the inspection, I am trying to have a sample plan that is statistically sound and know that we are not over inspecting.

This can of worms was opened because a customer sent us their preferred sample plan and it was about half the size of our sample plan so we are clearly over measuring these parts.

Although it is not my fault it is my responsibility.

I was not trying to imply you were seeking to cheat LaserJay. I was only pointing out that IMNSHFO proving with data that you can reduce your sample plan without risk of releasing non-conforming product is likely to be a more successful campaign than trying to justify the reduction based on what others are doing.

I was not trying to imply you were seeking to cheat LaserJay. I was only pointing out that IMNSHFO proving with data that you can reduce your sample plan without risk of releasing non-conforming product is likely to be a more successful campaign than trying to justify the reduction based on what others are doing.

Whether you intended it or not, you gave some good feedback to his question. So the question actually produced good input.

Just sayin'