Designing Path of Collision Avoidance for Mobile Manipulator in Worker Safety Monitoring System Using Reinforcement Learning

Mobile manipulator is a robot for the next generation of processes. The mobile manipulator can increase process efficiency by being combined with the collaborative robot system. However, due to worker safety issues, simultaneous control of the mobile manipulator is still difficult. In this paper, we present the designing path of collision avoidance of a mobile manipulator for the safety of workers. By monitoring the position and velocity of a worker modeled as a cylinder, the mobile manipulator designs a path to avoid the worker using reinforcement learning. We learned the robot using simulation, and we checked the mobile and manipulator move to the destination by avoiding the worker.