A shared autonomy of multiple mobile robots in teleoperation

摘要

We propose a shared autonomy between a human at an operation site and multiple mobile robots at a remote site. For partial intelligence, we use a distributed sensor-based path-planning algorithm of multiple mobile robots. Each robot smartly avoids its neighbors by vision-based feedback. It reaches its destination by position-based feedback under GPS (Global Positioning System). A human operator picks a robot to do a special task. In the teleoperation, transmission delay must be considered. Since an operator cannot observe the remote site, we prepare an animated predictive display simulating behaviors of all the robots. Therefore, an operator comfortably controls an operated robot in a real-time manner. Moreover in our teleoperation system, a human selects an adequate operation with the help of a repulsive force from the nearest surrounding robot. This force is felt by a joystick with a force feedback. These are considered as a partial intelligence at an operation site. Finally, an usefulness of our shared autonomy is ascertained in a few simulation results.