Cost-effective metrological evaluation and statistical modeling of thermal and payload effects on industrial robot repeatability

This study presents a comprehensive metrological investigation of the repeatability performance of two industrial robots, a FANUC LR Mate 200iD and a KUKA KR 6 R700 Sixx, using a cost-effective measurement system designed in compliance with ISO 9283 standards. The setup enables precise tracking of the end-effector position while maintaining low instrumentation costs, ensuring both accuracy and industrial feasibility. Repeatability was evaluated under varying warm-up durations, payloads, and pose configurations, with 300 consecutive measurements recorded for each condition. Results show that thermal stabilization plays a dominant role, reducing repeatability dispersion from 0.27 mm to below 0.005 mm for the FANUC robot and from 0.15 mm to below 0.004 mm for the KUKA robot after four hours of warm-up. Payload variations produced substantial deviations in the FANUC robot but minimal effects in the KUKA one, revealing distinct robot-specific sensitivities to loading conditions. ANOVA and correlation analyses indicated that warm-up duration contributed more than 50% of repeatability variance for the FANUC robot and up to 98% for the KUKA one. Quadratic regression models further improved predictive accuracy, reaching R² values of 0.90 for the first robot and 0.86 for the second. This work establishes a measurement-science foundation for robot performance characterization that supports uncertainty-informed decision making in precision manufacturing applications, while highlighting critical pathways for future metrological research in industrial robotics.