Application of Reinforcement Learning in Autonomous Mobile Robots

With the rapid rise of the deep reinforcement learning and progressive development of sensors for autonomous mobile robots it is feasible to teach robots to solve tasks of autonomous navigation in unknown environments. Reinforcement learning itself is a technique that solves this problem. In this work we discuss the reinforcement leaning approach, as well as the requirements needed for the mobile robots to use it. It also give example of modern autonomous mobile robot and its specification.