Human-Guided Robotic Comanipulation: Two Illustrative Scenarios

摘要

Emerging applications of robot systems that involve physical interaction with humans have opened up new challenges in robot control. While various control techniques have been developed for human-robot interaction, existing methods do not take advantages of both human knowledge and the robot’s ability. In this paper, a human-guided manipulation problem is formulated and solved. The workspace is divided into a human region, where the human plays a more active role in the manipulation task, and a robot region, where the robot is more dominant in the manipulation. The proposed formulation allows the human to take control actions to deal with unforeseen changes or uncertainty in the environment, and also allows the robot to take the lead where the environment is exactly known. We mainly consider two basic scenarios, i.e., robot control first and human control later (R-H) or human control first and robot control later (H-R), to illustrate the concept of human-guided comanipulation. Based on a smooth transition between the human region and the robot region, an adaptive tracking controller is developed to take advantages of both human knowledge and the robot’s ability. The experimental results are presented to illustrate the performance of the proposed control method.