When to use it

Calibration corrects Cartesian movements, so it pays off wherever target positions are calculated rather than taught directly. The four typical use cases are:

Palletizing & Grids. Drop positions across a pallet or grid are computed from a few taught corner points, so every placement depends on accurate Cartesian positioning. Without calibration, the error grows across rows and layers.

Vision. Camera coordinates must be turned into accurate Robot moves at the picked location. Calibration is what makes this hand-off between camera and Robot consistent.

Insertion. Mating parts under tight tolerances, where even a small XYZ error causes jams.

Conversely, applications that only replay taught joint positions do not benefit from calibration. If precision is required at specific points, you can use a workaround by saving the location twice: first as a joint movement for speed, and then as a Cartesian movement to leverage the calibration.