[OpenArm Troubleshooting] End-Effector Z Check Fails After Table Move (Intermediate)
OpenArm end-effector Z-height failure after table move: fixture plane checks, geometry-vs-calibration diagnosis, and production release thresholds.
A practical OpenArm workcell problem appears after moving the table, fixture, or platform under the robot. The arm may still home correctly, but a simple Z-height check suddenly fails because the surrounding geometry changed while your assumptions did not.
How are you diagnosing OpenArm cases where the end-effector Z height check fails after a table move?
Please share how you distinguish a shifted environment from a robot-side calibration problem, which fixture or reference checks you trust most, and when you choose a fast workcell update versus a full recalibration.
If you reply, include one exact Z-check symptom and one exact workspace or fixture check that exposed the shift.
Intermediate context: Our symptom was consistent: home passed, but the Z touch-off on the same fixture came in low by 1.4 mm after the table move. XY still looked reasonable, which made it easy to miss at first.
Intermediate context: What exposed it fast was a dial-indicator sweep on the fixture plate corners. We measured a clear tilt introduced during remount, so the issue was workspace geometry, not encoder drift.
Intermediate context: We allow partial recovery only if a 4-point plane fit leaves under 0.25 mm residual and the Z check passes three fixtures. Otherwise we redo full calibration and collision clearances before releasing the cell.
Intermediate context follow-up: Follow-up question: after table remount, what was your measured plane tilt and your final accepted Z residual threshold before releasing production?