Quantifying robustness

Ever asked an integrator to quantify the performance of a machine vision system? If you have you probably heard him growl, “Grr” at you through gritted teeth.

Only he wasn’t growling, he was saying “GR & R.”

For the uninitiated, GR&R, or gage repeatability and reproducibility, is a way to qualify the performance of a measuring system. By quantifying the sources of variation – part, operator and measuring system itself – a GR&R study tells you how well a measuring system will perform, and more importantly, where the variation comes from.

I found some good GR&R primers on the web sites of Quality Digest and Quality Magazine. Start with “Letting You In On a Little Secret” by Steven Ouellette which discusses the basics of “measurement systems analysis” (MSA for short,) then move on to “Quality 101: An Introduction to Gage R&R” by William Kappele and John Raffaldi, before wrapping up with “Understand Gage R&R” by Rick Sloop.

When you’ve finished reading, sit back and think about how you’ll include GR&R criteria in your next vision system RFQ. And if you really want to stretch yourself, consider how GR&R can be applied to attributes rather than dimensions.