Optimized joint motion planning for redundant industrial robots

The paper presents a model and solution method for optimized robot joint motion planning of redundant industrial robots that execute a set of tasks in a complex work environment, in face of various technological and geometric constraints. The approach aims at directly exploiting redundancy to optimize a given performance measure, e.g., cycle time. Alternative configurations along the path are captured in a graph model, whereas bi-directional transition between task and configuration spaces facilitates generating relevant, collision-free configurations only. Re-parametrization of the trajectory warrants compliance with the robot's kinematic constraints. Successful application of the method is demonstrated in remote laser welding.