Human augmentation in teleoperation of arm manipulators in an environment with obstacles

Discusses an approach to combining human and machine intelligence in teleoperation of 6 degree of freedom (DOF) arm manipulators. Two algorithms are presented that can operate in near real time and can deal with an unknown environment. We capitalize on the fact that the operator needs help primarily in the areas where the arm is near the obstacles; this lowers the computational costs. The main, major linkage algorithm operates on the principle of a greedy search. Perception of the hints provided by the operator is used in a few basic rules that limit the directions of the search. Overall, the intention is to allow the operator to concentrate on “global” motion planning tasks and leave collision detection and local path finding to the machine intelligence. An experimental study in progress is presented that is expected to demonstrate the algorithms´ performance and to suggest the types of situations where it will be the most effective