The use of reinforcement learning to plan the path of a construction robot

Autonomous mobile construction robots are a promising way to increase construction work efficiency, quality, and safety. However, to fulfill their functions, such robots must build their path from the initial point to the final one in the highly complex conditions of a cluttered construction site. This paper proposes a system for planning the optimal path of a construction mobile robot in an unknown environment with obstacles based on reinforcement learning methods. The system uses data obtained from sensors that determine the pose of the robot, its motion parameters, and distances to obstacles to plan the path.