Corrective Actions - Getting to the True Root Cause

[Jim Wynne]

Re: Corrective Actions getting to the root cause?

It is not a “standard”, it is a maximum allowable value.

I used "standard" in its sense that's synonymous with "criterion."

The system owners want to take the easy way out and blame the operator when, in fact, the system failed the operator. The system did not provide for the proper tooling (do you have a left-handed operator operating a right-handed machine?), proper flow (if non-conforming material does not flow 180 degrees away from the conforming material flow path, it WILL find its way back into standard flow), adequate space, good lighting, etc. Basically, the operator cannot cause a failure. They can only point out weaknesses of the system. The 3% opportunity is to allow the system owners and management to lay some blame on the operator and, thereby, feel better.

The subject of operator error has been extensively discussed here. Have a look at Operator Error, System Error, or Both? (for example), so there's no point in doing it again here.

If the operator truly caused a failure, they had to do so purposefully. Follow Deming’s writings and fire them if necessary.

That's not right. It's not even wrong.

P.S.: By following the above concept to identify product and tooling changes, a team reduced scrap rates of a product line that used 288,000 parts and assemblies per day from 40% to 0.05% in 3 months and achieve almost zero process failures (0.03 DPMO) within 12 months and making 55% EBIT.

That's very impressive, but I must say that a product line that's scrapping >115,000 units a day seems like a pretty easy target. Nonetheless, if you were able to reduce operator error to next to nothing, that's great. You shouldn't project your own relatively narrow experiences (especially ones that are surely atypical) onto the world at large, though.

 

[adamsjm]

Re: Corrective Actions getting to the root cause?

..., but I must say that a product line that's scrapping >115,000 units a day seems like a pretty easy target.

I agree. But why did they allow a five-step deterioration to occur before trying to fix it? I do not know. My first assignment with that company was to fix the non-conformance problem.

You shouldn't project your own relatively narrow experiences (especially ones that are surely atypical) onto the world at large, though.

Fortunately, they are not atypical.

The plant where that line is located went on to win a Shingo Prize for Excellence in Manufacturing and those lean manufacturing processes were recognized by TPS guru, James P. Womak, author of The Machine That Changed the World – The Toyota Production System, as “The best production system I have seen in the world.” (~140 manufacturing cells, producing 950 p/n assemblies, ~58,000,000 assemblies shipped/yr.)

I worked in a another plant in a different industry and that plant improved to the point that they produced 2.3 PPM failures (true six sigma) while producing 220,000,000 parts and assemblies of 1800 different part numbers per year (~185 different part combinations per day). I never tried to calculate the average number of operations to calculate DPMO. It also was a profitable plant while giving their customers 2-5% cost reductions year-over-year.

Recently I have been working with a large (>$20B) corporation and we reduced the time to develop their APQP documents by 90% while significantly improving their development collaboration.

The one failure, which I have seen, occurred in a highly industrial city where the workers and supervisors had been led through the process poorly in the past. They “knew that it was not possible to reproduce that kind of results. We’d tried it before”. Indeed, they proved it to be true.

Moral: If you don’t think you can, you’re right!

 

[Jim Wynne]

Re: Corrective Actions getting to the root cause?

Moral: If you don’t think you can, you’re right!

If you think you can when you should realize that you really can't, you're a victim of human error.