Perfection Through Mistake-Proofing | IndustryWeek

The article makes the point that mistake-proofing must be “effortless.” The way I usually say it is that a mistake-proofing/poka-yoke device must not add labor, a point that is frequently missed in discussions of this topic in the US.

Why is it essential? Because any device that adds labor is guaranteed to be by-passed under pressure. If preventing a mistake requires one more gesture, on any day where “we have to ship all this by 6:00PM,” the organization will find a way around it.

Mistake-proofing makes a difference in any process where human error is a major cause of failure. Many processes qualify, but not all. If the main cause of defects is the machine’s  inability to hold tolerances consistently, mistake-proofing will not do much good.

Yes, a device that is fallible cannot be considered mistake-proofing. Usability engineering, for example, provides user interfaces  that make mistakes unlikely, but not impossible. Sometimes it is sufficient, but it is not mistake-proofing.

The one criterion I have an issue with is simplicity. A mistake-proofing device must be simple to use, I agree, and its design should not be anymore complex than necessary. However, where the stakes in human error are high, as in airliner cockpits or semiconductor process equipment, preventing mistakes may require elaborate technology. If a device for this purpose  works every time and adds no labor, I see no reason to deny it the “mistake-proofing” label.