S
Snarf
First piece inspections are intended to give us some assurance that our process is set up correctly. Typically one part is measured from each tool cavity, if each part is within spec there is a quality sign off and the machine operator is off and running. I've often wondered whether there is a better way to assess initial set up without having to put every process on an SPC chart. What if your first piece measurement is within 1 SD of the the target mean, but the underlying process mean is significantly different from the target mean?
Ex. A machine is used to cut aluminum rails to length. A capability study is performed during PPAP on the length cutting process with the following result (see Cpk.png). Cp is 1.43 Cpk is 0.44. Process is obviously not centered. Looking at the data on an IMR (attached as SPC.jpg) chart there is a significant difference between the mean of the values taken during regular in-process checks and the mean of the values used for the special PPAP capability study.
The standard suggestion to the supplier in this case would be "improve your set up" but many of the points measured during the capability study are within 1 standard deviation of the target mean. If those particular points were the ones that were measured during first piece inspection there would be no signal there telling the quality engineer or the operator that the process is not capable.
My thoughts are that a first piece inspection or even a regular in-process check is not helpful unless the data point is placed into context using SPC. That said SPC is often not something a company is willing to put into place on every process so what are the alternatives?
Ex. A machine is used to cut aluminum rails to length. A capability study is performed during PPAP on the length cutting process with the following result (see Cpk.png). Cp is 1.43 Cpk is 0.44. Process is obviously not centered. Looking at the data on an IMR (attached as SPC.jpg) chart there is a significant difference between the mean of the values taken during regular in-process checks and the mean of the values used for the special PPAP capability study.
The standard suggestion to the supplier in this case would be "improve your set up" but many of the points measured during the capability study are within 1 standard deviation of the target mean. If those particular points were the ones that were measured during first piece inspection there would be no signal there telling the quality engineer or the operator that the process is not capable.
My thoughts are that a first piece inspection or even a regular in-process check is not helpful unless the data point is placed into context using SPC. That said SPC is often not something a company is willing to put into place on every process so what are the alternatives?
