Setting up MSA system in a machine shop environment.

Guys, I have been charged with setting up an msa system in a machine shop environment on an ongoing basis. We have 15 operators all using digital calipers, we currently have minitab access as well. Any advice on the best way to approach this, please?

MSA is the combination of Gage R&R and calibration. Gage R&R is used to determine if the gage is appropriate for the dimension being measured. Calibration assured the gage is still good, by reading accurately. Only calibration has a time function.

Gage R&R Tips
Picking operators:
You should not need to "test" every operator as a part of gage R&R (although some folks choose to.) The most robust use of gage R&R is to find 3 operators that represent the scope of skill levels on the shop floor. I recommend a new operator (or temporary, if you use them), a longer term employee, and then one considered to be the most "expert" in the gage's use. If you do a good job of picking people with this range of skill levels, the gage R&R should be meaningful.

Picking Parts:
You may not have the luxury of performing gage R&R on every dimension you are measuring. If that is the case, pick those dimensions that are the most critical, the dimensions used to determine adjustments and tool change, the tightest tolerances and the dimensions that are the most difficult to measure due to their location.

Thanks Bob. I am currently running a Gage R&R on each operator to build up some baseline data. I may then go to sampling different op's on different months.

nealet01 said:

Thanks Bob. I am currently running a Gage R&R on each operator to build up some baseline data. I may then go to sampling different op's on different months.

Click to expand...

That is certainly acceptable. Gage R&R is a toolbox, not a tool. Depending on how you set it up, it can be used to answer several different questions. The key is to set it up correctly for the question you are trying to answer!

nealet01 said:

Thanks Bob. I am currently running a Gage R&R on each operator to build up some baseline data. I may then go to sampling different op's on different months.

Click to expand...

In addition to Bob's good advice, you should be aware that there is no need or reason to keep doing GR&R, especially when you've done them with all of the operators. Once it's established that the device/operator combination is appropriate to the task, that shouldn't change.

Thanks guys!