Exploring interactive simulator in collaborative multi-site teleoperation

Abstract

We have developed collision-free coordination technologies for multiple robots in a common environment remotely controlled from multiple operators over the network with large physical separation. Sensor guided approaches would make remote robot systems rather complicated and hard to control accordingly. The paper reviews human supervisory control approaches based on a delay-tolerant predictive graphics simulator, where different coordination strategies are incorporated to cope with the operator's delayed visual perception over the network. We have built an experimental test bed, where operators control their master to get two slave robots to cooperate in a common task. They detect and/or feel a priori the possibility of collision through the local graphics simulator that runs in near real-time. We have demonstrated various tasks by two slave robots and two operators over a LAN subject to the simulated delays and evaluated the simulator guided approach in multi-site tele-collaboration.