Robot Calibration
Robot calibration identifies the differences between a robot's nominal kinematic model and the real machine, such as link lengths, joint offsets and mounting position, and corrects for them to improve absolute positioning accuracy.
A measurement device such as a laser tracker, camera system or calibrated probe records actual tool positions at many robot poses. Software compares them with the positions predicted by the nominal model and identifies corrected parameters, sometimes including joint compliance and gear effects, which the controller then uses. Related tasks include axis zero-position calibration, called mastering by some manufacturers, tool centre point calibration and calibration of fixture and camera frames.
Calibration matters when programs come from offline programming, CAD or vision, when a robot is replaced and existing programs must run on the new unit, and in precision tasks such as drilling, measurement and machining. It also makes it easier to share programs between identical cells.
Results depend on measurement accuracy, temperature and the poses used, and the benefit applies mainly within the calibrated region. Zero-position data must be restored after motor or encoder replacement following OEM procedures. Periodic verification checks whether accuracy has drifted after collisions or maintenance.
Key points
- Corrects the kinematic model to match the physical robot
- Uses laser trackers, cameras or probes to measure actual positions
- Improves absolute accuracy for offline and vision-guided programs
- Axis zero calibration follows OEM procedures after part replacement
Where AiVibe comes in
AiVibe designs and manufactures the AiAmbA AI Factory, whose edge devices and AI agents let people talk to robot controllers in plain language. Robotics perception is an AiAmbA use case, and robot safety functions follow ISO 10218 and never depend on the AI layer.