It’s not which measurement has the procedural error that matters. It’s what you do when you discover it.
Alternatively, what should the robot do if it gets one measurement that’s far different than all the others, and it doesn’t know why?
It’s not which measurement has the procedural error that matters. It’s what you do when you discover it.
Alternatively, what should the robot do if it gets one measurement that’s far different than all the others, and it doesn’t know why?