Setting AQL Levels with Multiple Suppliers

[sunbike]

When working with multiple suppliers, each of whom is providing a subassembly for the finished product, what is the general approach to setting AQLs? I'm assuming you start with the finished product AQL and try to work backwards.

Let's say you have a 0.20% AQL max for Critical defects for the finished product. Supplier A provides subassembly A and Supplier B provides subassembly B where A + B = final product. Can the AQL be split between suppliers (e.g. Supplier A has 0.15% and Supplier B has 0.05% or can each supplier use 0.20%?

 

[Jim Wynne]

Please describe what you mean by AQL, and what you do with it.

 

[Bev D]

We also need to understand how defects affect production - there is a difference between final product inspection/test and in process inspection/test.

 

[sunbike]

Please describe what you mean by AQL, and what you do with it.

Thanks for the reply. In this context, I'm referring to the highest defect rate that's acceptable in a sampling plan for final release. Assuming I have a maximum allowable rate (%) in mind for my final finished product, what max defect rate would subassembly suppliers need to set for themselves in order to achieve the defect rate for the finished product?

For the example in the original post, ignoring contribution from the final assembly process, in my mind each supplier would need to meet 0.20% or better. However, I've also heard of "splitting" the defect rate between suppliers to meet the desired maximum allowable defect rate for the final product.

We also need to understand how defects affect production - there is a difference between final product inspection/test and in process inspection/test.

Mainly looking at final assembled product release and subassembly product release from suppliers. I didn't consider subassembly release as in process but maybe this is an incorrect view? In my mind, each supplier would have their own separate in process sampling plan that they would develop to meet the max allowable defect rate for their subassembly, which in turn would be based on the max allowable defect rate for the final assembled product.

Apologies if I'm mixing up terms/concepts. Still learning and happy to try to clarify some more if needed.

Thanks!