Current Monitoring-based Diagnosis of Drive Belt Degradation in Wafer-handling Robots

Drive belt degradation affects the positional accuracy and repeatability of wafer-handling robots, leading to positional errors. This study developed a method for noninvasive diagnosis of drive belt degradation in wafer-handling robots using features extracted from motor current data. The motor current generating the torque is monitored for diagnosis, eliminating the need for additional sensors. Features are extracted by monitoring instantaneous changes in the motor current based on the operating characteristics of the wafer-handling robot. The diagnosis relies on a feature called the instantaneous current variation rate and a modified sequential probability ratio test (SPRT). An experiment with a five-axis wafer-handling robot confirmed that the distribution of the instantaneous current variation rate varies depending on the level of drive belt degradation and the modified SPRT successfully diagnoses drive belt degradation without generating false alarms.