From: Mark Maral
Date: 15 Apr 2000
Time: 10:05:57
Hello,
I work for a service organization that receives repaired product, in this case disk drives from the OEM. The problem I face is that the organization I work for does a 100% functional inspection of all product received from the repair source. The measure of the quality output of the repair process is calculated as from the number of defective drives from the total number received for the month. There is no regard given to the various part numbers received and no regard given to when units are received during the month. Bascically, the OEM's repair quality score is based on the number of defects received during the month over the total number of units recieved. My problem is that I see a lot of statistical methods for sampling in the mfg process where there appears to be controlled lots, but how do I implement a statiscally valid sampling plan in the case where there does not appear to be a standard lot? The problem is I want to be able to establish a valid sampling plan for repaired product and I am not quite sure how to go about it based on the manner in which we receive product from the repair supplier.
Thanks, Mark