Improvement Insights Blog
Capability Analysis Issues
Struggling to get a good Cp and Cpk? It might be your data. Here’s why:
“I’m Jay Arthur, author of “Lean Six Sigma Demystified” and QI Macros [software].
“We had a guy call in the other day and he was trying to determine the capability of his process. Unfortunately, he’d taken [measurements of] everything from his startup to his shutdown scrap and ran that as a capability study, and he wasn’t capable. I said, “Well, there you go…” But if you run a control chart of that, you can see the wiggly startup and the wiggly shutdown and then there’s this nice stable process in the middle. If you’re going to do capability analysis, you want to analyze the stuff in the middle where you’re running hot, straight and normal. You may have to scrap the stuff at either end that doesn’t fit with reality and your capability will be very good.
“That’s kind of an observation, and this has happened to me a number of times. Somebody was producing brake assemblies and there was wire in there. The wire changed by 2/10,000 of an inch but… you can’t measure [capability] across that change of material. You can measure it before, you can measure it after, but you cannot measure capability as it crosses over because you’re not going to get the result you want. So these are the kinds of things – these little subtle things – that trip people up.
“The great thing about startup and shutdown scrap is you can use it in your Gage R&R because you need part variation to do a Gage R&R; you cannot use the stuff in the middle that’s good because there’s not enough part variation. So over time, you learn some of these things and you find that maybe we should just rethink how we’re doing this, right? This is where we’re going to measure the capability – we’re going to shave off the startup and shutdown scrap and just send out what’s really good.
“That’s my Improvement Insight. Let’s go out and improve something this week – capability might be a start.”