Reducing Performance Variability and Overcoming Limited Spatial Ability: Targeted Training for Remote Robot Teleoperation

In this paper, we present a targeted training approach for remote teleoperation aimed at achieving consistent proficiency levels across users with varying capabilities. Our approach begins by assessing users' abilities to perform robot motion control, workspace adaptation, and gripper control. It then provides tailored training based on identified skill gaps to enhance the learning effectiveness and user experience. To demonstrate our approach, we conducted a user study, with one group undergoing conventional, free-form training and the other engaging in targeted training in accordance with their skill gaps; after the training phase, participants teleoperated a robotic arm in a simulated medication preparation task for performance evaluation. Our results show that the targeted training approach effectively reduces performance variability and mitigates the influence of spatial ability on both training and task completion time. We discuss the implications of our results for practical teleoperation training and future research.