Towards continuous control of flippers for a multi-terrain robot using deep reinforcement learning

TL;DR

This paper develops a deep reinforcement learning-based control algorithm for multi-terrain robots with flippers, enabling continuous and adaptable control in complex, partially observable environments.

Contribution

It introduces an end-to-end deep RL control method for multi-terrain robot flippers, addressing challenges of partial observability and environment complexity.

Findings

Successful application of DDPG in complex simulation environments

Smooth control of multi-terrain robots demonstrated

Potential benefits for rescue and assistive robotics

Abstract

In this paper we focus on developing a control algorithm for multi-terrain tracked robots with flippers using a reinforcement learning (RL) approach. The work is based on the deep deterministic policy gradient (DDPG) algorithm, proven to be very successful in simple simulation environments. The algorithm works in an end-to-end fashion in order to control the continuous position of the flippers. This end-to-end approach makes it easy to apply the controller to a wide array of circumstances, but the huge flexibility comes to the cost of an increased difficulty of solution. The complexity of the task is enlarged even more by the fact that real multi-terrain robots move in partially observable environments. Notwithstanding these complications, being able to smoothly control a multi-terrain robot can produce huge benefits in impaired people daily lives or in search and rescue situations.