Pose Repeatability

How closely a robot returns to the same point each time it's commanded there.

Pose repeatability measures how consistent a robot's actual stopping point is across repeated executions of the same command to reach the same target pose, usually written as something like “±0.05 mm.” The international standard ISO 9283 defines the measurement method: the robot approaches the same position repeatedly from the same direction, and the spread of the resulting points is measured statistically. It differs from absolute positioning accuracy, which measures how far off the robot lands from the true target when commanded to a point by coordinates — this is usually much worse than repeatability, and is improved through kinematic calibration. Traditional industrial robots are mostly programmed by teaching (moving the arm to a point and recording it) and returning to that same point, so repeatability matters most for them; when picking up an object at coordinates computed from vision, absolute accuracy and hand-eye calibration error matter more.

ExampleAn arm spec'd at ±0.1 mm repeatability will reliably return to within 0.1 mm of a taught point; but when it moves to a new position computed from a camera, the error can be much larger — that's a question of absolute positioning accuracy instead.

Also called

Repeatability, Repeat Positioning Accuracy

Related

Absolute Positioning Accuracy · Industrial Robot · Teach-and-Playback Programming · Kinematic Calibration · Hand-Eye Calibration · Collaborative Robot

Sources

Industrial robot - Wikipedia（Repeatability 与 ISO 9283 小节） (Chinese)