Motion planning in human robot cooperation via deep reinforcement learning

Abstract

An approach to motion planning for human robot cooperation based on Deep Reinforcement Learning in simulated environments is proposed. This approach aims at solving some of the typical problems of motion planning in human robot cooperation such as the need of inferring the human movements or the need of continuous re-planning trajectories to avoid collisions. The approach tested shows that is able to solve a simple scenario with success rates around 90% and collision rates below 10%.