Optimal Proactive Path Planning for Collaborative Robots in Industrial Contexts

The coexistence of humans and robots in the future production plants is one of the pillars of Industry 4.0. Humans and robots will collaborate to accomplish common tasks in order to mutually compensate their deficiencies. In recent years, many efforts have been spent to develop safe motion planning strategies, designed to prevent robots from injuring humans. Most of the previous techniques are classifiable as reactive, since the considered motion controllers impose some local corrective actions in order to dodge the space occupied by the human. In this paper, a proactive approach is adopted, optimizing robotic paths according to a prediction of the volume occupied by the human when collaborating with the robot. The validity of the approach is shown in a realistic use-case involving the collaboration of a human operator with a 7 degrees robotic arm, the ABB YuMi.