Online Planning of Path-Consistent Stop Trajectories for Collaborative Robotics

This paper presents a method for the online planning of path-consistent stop trajectories for collaborative robotics. The proposed approach considers the manipulator to be enclosed in bounding volumes, whose sizes account for a protecting separation distance and are minimized online considering the robot dynamics and its torque constraints. In case a collision danger between bounding volumes enclosing human and robot is detected, the robot is guided to stop along a safe trajectory that preserves the geometrical path. Simulation results on a validated model of a 7-degree-of-freedom manipulator show the feasibility of the method and provide a more effective human-robot collaboration with respect to a similar approach that does not preserve path consistency during safety stops.