Inspection – what's a healthy level?
We had a very interesting seminar the other day on the relative merits of mass inspection, directly relating to one of Deming's 14 points that was highly critical of it. I found this all particularly interesting actually, because of the number of different levels that are possible. Firstly there's zero inspection. This just won't work, there's no net to catch the incredibly obvious flaws, and people that are paid by the unit will be more likely to create a less than excellent product if there is nobody to see their shoddy work. Then there's 100 percent inspection. This obviously doesn't work because the inspector will not catch all flaws, and may decide that some units meet the standards depending on their mood. It also catches all mistakes too late, and doesn't look to arrest flaws in the process. Finally it's clearly a de-motivator. What also interested me was the idea of 200% inspection, and it took me a little while to get to grips with the idea that this could be counter-productive. But this was explained, and I need to remember to focus on how each person in the process will be acting; the workers will think that all mistakes will get caught, so take less time to find them themselves, the first inspector will believe that the man after him will catch the mistake, and the second inspector won't expect anything to have gotten past the first inspector. So how can we analyse these numerous problems. 0 doesn't work, 100 doesn't work, 200 doesn't work. It seems to me that somewhere between 0 and 100 is the ideal situation, but I'm not completely sure about this. Does anyone have any ideas what would be the perfect amount of inspection? Of course there won't be a correct answer, otherwise they'd all be using a standard percentage.
If only there were standard answers to these sorts of questions