Statistical Evaluation of Robot Trajectories in Automated Dimensional Measurements

The influence of a robot’s manipulation can be observed in a robotic measurement system. Different robot end-effector trajectories yield different robot end-effector accuracy and repeatability errors. Trajectory parameters, robot motion type, velocity, and length of motion were identified as influential sources. A robot arm was used to insert measuring objects into the measurement device for dimensional measurements. In the first part, the measurement datasets for linear and joint robot motions were compared for three different velocities and four motion lengths. The influence of the number of active joints in the robot’s motion was compared for two velocities and four magnitudes of joint rotation. Dimensional measurement variability was analysed using measurement system analysis (MSA), and the statistical influence of trajectory parameters was further addressed by analysis of variance (ANOVA). All identified trajectory parameters have a statistically significant impact on measurement variability, reflecting the robot end-effector’s accuracy and repeatability errors. Linear motion provides higher measurement variability up to 20%, a velocity increase that is typically up to 25–35% and motion length that is typically up to 15–35%.