In this situation a motorcycle running 3 feet away from the barrier would have a capability of 2.0 while the compact car would have a capability of 1.0.

Please explain this with calculation.

]]>All capability indices are just probabilities of meeting specifications. This is never said. IMO, it’s a significant omission.

]]>All the vehicles are running straight down through the lane your process is centered on nominal (the target value). This is equivalent to what is typically called Cp your process potential calculated as the width of your specification (the vehicle) divided by your variability (the barrier).

Should that not be “the width of your specification (the barrier) divided by your variability (the vehicle)”?

]]>However, I have a real problem with two statements.

First, don’t get these guys judging Cp without looking at Cpk also. Take the example of a motorcycle driving in the same lane going in the opposite direction. The motorcycle I have described still has the same Cp as your motorcycle from Monday but is entirely out of specification (when taking into the driving rules of the road). Cp gives the process owner an idea of potential but doesn’t imply anything about what’s actually IN THE PROCESS which is why we need to look at Cpk also and graphical representations.

Also, the statement of why to use Cpk is because we can only get damaged on the closest side implies you can’t get defects on the other side. I’m sorry but you have to look at the other side. You CAN be making no defects but often you will be depending on how much variation is in the process. I always tell my students to look at a histogram with the specs shown on it if they don’t have good statistical software like Minitab which gives them the Cp and Cpk.

]]>