You have to make sure your routines are as bulletproof as possible. Ideally, the fixturing would be mounted permanently. Fixture would also be designed so that there is only one way to load the part. Have some way to double check that the part is loaded against the fixture stops. Design the Program so that all that the operator needs to do is load the part and tell it to start measuring. Don't be surprised if it takes a year to figure out most of the stupid things that can go wrong.
I worked at a place that implemented having CNC operators use the CMMs to measure their own parts. As Schlag says, Operator mode is the way to go.
I was expecting all sorts of problems, but it actually worked out great. The operators liked it because they could get quicker feedback on their parts and they could work together to prioritize which parts to measure first when there was a backlog. It really helped to ease tensions between the Manufacturing and Quality department.
I liked it because the operators learned for themselves how important it is to thoroughly clean the parts before putting them on the CMM. After a while they were policing each other about cleanliness. No one wanted to be blamed about junk building up on the probes so they would clean their parts and the probe tips more thoroughly than I ever had time for. And, if they ever wanted a part to be remeasured, they could just do it themselves. It was funny to see operators run the same part over and over on different CMMs to try and get a part to squeak into tolerance.
The downside was maintaining the CMMs. Broken probe tips became far more frequent, but that got better with time, training, and program revisions.
Worked with someone once who could misread any kind of possible instructions. It finally got to the point that I would have her review things before I released it out into the wild. If it passed the "Kathy" test then it was perfect. Made me much better at explaining things and making routines idiot resistant.
Depends on the company though. I wanted bosses to buy renishaw equators to have floor inspectors & machine operators to check parts themselves. I got a big fat no-no. So either CMM progr has to check pieces or floor inspectors sit at their stations checking pieces with hard gauges. Otherwise, on average, people do want to lean new stuff (better look on the resume and/or get better pay).
