Providing Haptic ''Hints'' to Automatic Motion Planners

摘要

In this paper, we investigate various paradigms for enabling a human operator and an automatic motion planner to cooperatively solve a motion planning query. Our work is motivated by our experience that automatic motion planners sometimes fail due to the difficulty of discovering ‘critical’ configurations of the robot that are often naturally apparent to a human observer. In particular, we study how haptic and visual interfaces can be used to enable a user and an automatic motion planner to cooperatively solve a motion planning query. While much progress has been made in the areas of automatic motion planning and haptic interfaces, little attention has been directed towards combining the two. Our goal is to develop cooperation paradigms that are natural for the human operator and beneficial to the automatic planner. Some of the issued addressed in our study include determining ‘natural’ ways for the user to understand the progress made by the motion planner (e.g., visualization techniques) and developing techniques by which the automatic planner can utilize (easily generated) user-input. In this paper, we propose several possible approaches to these problems, compare them according to ease of use and effectiveness for solving motion planning queries, and provide recommendations for further development.