Reinforcement learning in single robot hose transport task: A physical proof of concept

Abstract

In this paper we address the physical realization of proof of concept experiments demonstrating the suitability of the controllers learned by means of Reinforcement Learning (RL) techniques to accomplish tasks involving Linked Multi- Component Robotic System (LMCRS). In this paper, we deal with the task of transporting a hose by a single robot as a prototypical example of LMCRS, which can be extended to much more complex tasks. We describe how the complete system has been designed and built, explaining its different main components: the RL controller, the communications, and finally, the monitoring system. A previously learned RL controller has been tested solving a concrete problem with a determined state space modeling and discretization step. This physical realization validates our previous published works carried out through computer simulations, giving a strong argument in favor of the suitability of RL techniques to deal with real LMCRS systems.