Remote supervision and task-level control of industrial robot systems using augmented vision

摘要

Remote control of robot systems are necessary in case of uncertainties or in hazardous/hostile environments. Mostly, the approaches are based on stereo-vision or synthetic 3D environments. The operator controls the motion of the robots remotely on the motion-level. For a range of industrial applications, these approaches are not suitable. This paper presents an approach of task-level remote control based on augmented vision. The video image is super-composed with synthetic 3D elements in order to provide rich information about the task and the current situation in the remote world. Complementary, a synthetic view of the remote world allows to obtain better spatial orientation. The operator interacts with the 3D elements and controls the robot on the task-level. An underlying model base and the local vision-based robot control make it possible that the robot performs accurate and fast motion and object handling.