In a manufacturing context (I guess Six Sigma?) what does “100% visual inspection” mean?
Does this mean 100% of the manufactured products are inspected (as opposed to a random sampling of, say, 10% of all the manufactured products)?
Or does this mean that inspection is only done visually?
Or what?
I am an engineer in a manufacturing plant, and we use this terminology on Certificates of Compliance. When we use it, we mean that every parts have been visually inspected for defects that would be detectable by the human eye. We are really looking for problems associated with finish, such as discoloration or light paint. The inspection may also catch damage or missing parts. This inspection method would not address things like dimensional checks or material properties.
However, we don’t use Six Sigma, so it’s possible that there is some specific meaning in that system.
So the “100%” basically means that every single part that comes off the line is inspected?
Yes, that’s how I would read it. The thing that’s unknown is what exactly they mean by visual inspection. It could be Bob sitting next to the line looking for the obvious problems or it could be a sophisticated computer vision system carefully examining each piece and comparing it to a known good sample.
Cool. Thanks for your help!
BTW, a “six sigma” type program would be working towards eliminating 100% inspection steps, whether visual or otherwise. Quality improvement programs like Motorola’s famouse “six sigma” program are about building quality into the manufacturing process so that inspections become less and less necessary. The whole idea is that you simply can’t screen for quality, and there is something wrong with your manufacturing process if that is how you achieve a given level of quality. Instead, you use sophisticated statistical techniques to set “process control limits” at each step of the manufacturing process. By keeping your machinery operating within those limits you will eliminate, or reduce, the need for quality screens or inspections.
“100% visual inspection” is supposed to mean human inspection, and that’s not a good thing, either. It’s said that human inspection is on average only 50% effective! Also 100% inspection is overkill unless the process is completely screwed up; implementation of such is usually a reaction to a quality concern and only a temporary measure (unless, as I said, there’s something fundamentally wrong in the process).
6-Sigma isn’t about replacing humans with machinery. It’s about using statistical tools to study and eliminate variation, which ultimately improves quality and would as a consequence – but not an aim – reduce or eliminate human inspection. There are a lot of 6-Sigma detractors around, because aside from its benefits, some of its followers seem to have discovered it as their religion. But 6-Sigma is only a method. You still need the tools and automation capable of working within your control limits, and even then you still need some type of extraordinary measures to deal with potential special causes. Maybe you have a perfect 6-Sigma process, but a special cause is causing 100% visual inspection? That’s not so out of this world.
I worked in a production plant that practiced Six Sigma and did 100% visual inspection.
I this plants case the product we were working with were finished microchips of all dimensions. All products were electronically tested for internal and external compliance with standards. Then every chip was meant to be visually inspected by the operator to check for obvious damage caused by the electronic testing eg. damage to the pins. This visual inspection was done with a magnifier lamp while all chips were in their trays.
We found that we were getting a lot of visual fails reported at end of line which carried out a 10% check on finished goods. We then had to put in place OCR software and cameras on the test machine because we just couldn’t trust the operators to actually do this inspection. When I left the place manual inspection was still happening but the machines also ‘looked’ at the chip as well.