Thanks for the tip, but I'm asking more about practicality as you come up against the limits of the arm's accuracy during use.
Right now, I'm using it to measure, in simplest terms, a 9'-0" cube volumetrically speaking that requires (4) moves. I've had trouble with repeatability in the past, but believe the problem has been solved through more careful setup and measurement procedure. I'm wondering if my movement philosophy might need changed for better accuracy from movement-to-movement.
It's just difficult for me to get my head to go from +/- 0.004" point-to-point accuracy to how accurate the arm can locate, say, an 11/16" dia. hole on the last move of the program. I'm not sure anyone can tell me definitively... That's a pretty complicated problem to solve geometrically and statistically.
--------------------------------
Fitter, happier, more productive