Assistive Positioning via Integrating User Input and Dynamic Movement Primitives in Shared Teleoperation

This paper proposes an imitation learning-based methodology for assistive teleoperation to achieve intuitive and precise operation for object grasping. The proposed method integrates user input and the output of Dynamic Movement Primitives according to the distance to the target object so that the robot trajectory will be attracted to the desired position and orientation for grasping. Experimental results suggest the effectiveness of the proposed method for assistive teleoperation on a simulated teleoperation system.