Sunday, July 5, 2009

Cpk vs Ppk - Part 1

I have a new product launch to production few month a go. For any new product ramp up, there are a series of data collection to monitor the process stability and capability. The whole purpose of the monitoring is to ensure consistency of machines and peoples variations. I used to call the exercise as 10 Lots monitoring.

Here the example of a data set collected from the 10 Lot monitoring. And I will report the process performance to the management in this form.



If you are familiar with Cpk and Ppk stuff, you will not agree with me.


Actually, I also dislike the format because we only measure Cpk value of each lot and conclude the process performance based on the Cpk value estimated. No choice, it is the yard stick I asked to follow.


The Cpk value of each monitoring lot cannot be use to present the overall process performance because the standard deviation use for Cpk calculation is the standard deviation within the sub group. Variations between machine, peoples, material batch and others were not taking into consideration in Cpk calculation.


Therefore, Ppk is a better yard stick to measure the overall process performance for new product ramp up.

Last update : 02/06/08

0 comments: