AIAG/VDA’s FMEA Manual Is a Major Advance (my take on this subject)

Bill Levinson

Industrial Statistician and Trainer
The article uses a Shigeo Shingo example in which the failure mode is a flange not contacting three reference planes in a jig. The Occurrence rating is determined not by frequency of occurrence (e.g. percent, or per million opportunities) but rather by the nature of the prevention control, which relies on worker vigilance--never very comforting in a Shigeo Shingo case study. The same goes for the Detection rating, because the detection control relies on the worker to notice the problem.

A jidoka detection control reduces the Detection rating to 2, again because of the nature of the control rather than any estimate or calculation of the likelihood that the problem will escape detection. Installation of plungers that hold the part squarely in the jig (if achievable) is similarly a prevention control that ought to reduce the occurrence rating to 2 or maybe 3 (previously determined quantitatively, as in occurrences per thousand or millions of opportunities, which is unlikely to be available in this application).
Top Bottom