Dec 17 2020
“Recently, Michel Baudin […] invited practitioners to further contribute to the knowledge on QRQC, among which myself. As I feel a brief answer on LinkedIn would not do justice to the richness of QRQC, I decided to dedicate a post to the topic. Without ambition, however, to try and be complete in this post, which I feel is not possible with a vast topic like QRQC. But let’s dive in and share some of my experiences with and views on QRQC, the way I experienced and lived it at Valeo at the time.”
Michel Baudin‘s comments: Thanks to Rob van Stekelenborg for stepping up and sharing all of these details.
Of particular interest is his description of Kazuo Kawashima’s rejection of the Pareto principle:
“I also still remember one of Kawashima-san’s strong coaching sessions in 2002 when he proverbially threw the well-known principle of Pareto or 80-20 analysis into the garbage bin; still a quality 101 tool you would say. He explained in his own personal way (he was quite a strong personality, I can tell from experience…) how stupid we were to first collect enough data for a long enough period of time to be able to make such a chart, and to only then start problem-solving.”
Michel Baudin‘s comments: According to Franck Vermet, Pareto was a bone of contention between Kawashima and Ken Sato, who brought QRQC into Faurecia. Sato used Pareto analysis to select problems to solve.
In Vade retro, Pareto!, Cécile Roche weighed in on Kawashima’s side. In her experience, Pareto diagrams were not nearly as useful as expected in setting priorities. I also put out a paper in IE Magazine on Revisiting Pareto, back in 2011.
If defects are sufficiently rare, then running to ground every defect makes sense. On the other hand, if you are overrun with defects, you need triage. This is where Pareto analysis is supposed to help.
Relying exclusively on occurrence counts, however, fails to take into account the challenge of problem-solving. A less frequent defect that is easier to eliminate may be a lower-hanging fruit than the most frequent one.