M
mungnr
My company is a small job shop, and currently a typical run for us is 1-8 pcs. on average. However, we also serve as a distributor for several "off the shelf" items. On these "catalog" type pcs. we will typically get in shipments of 1000-3000 pcs which we stock and pull as we need them. A typical pull will be 128 pcs. - 512 pcs. Currently the method is to pull 1 pc. per 50 to inspect when an order is processed (i.e. for a 128 pc order, (3) pcs are pulled at random and inspected). No one here can tell me why this number was chosen, other than "it seemed about right" and the president wanted a standard rule of thumb to use. I want to use some statistics for generating a sample size rather than pulling a number out of thin air (MIL-STD-105).
My question is, if I use a lot sampling plan, do I need to sample each order as it is pulled, or can I use a standard inspection on the 1000 pc lot that comes in so that as long as I am pulling parts from that lot I don't need to pull parts from the smaller order size when it ships? To put it in simpler terms, do I sample the larger "lot" when it is received, or the smaller "lot" when it ships?
My question is, if I use a lot sampling plan, do I need to sample each order as it is pulled, or can I use a standard inspection on the 1000 pc lot that comes in so that as long as I am pulling parts from that lot I don't need to pull parts from the smaller order size when it ships? To put it in simpler terms, do I sample the larger "lot" when it is received, or the smaller "lot" when it ships?
