Distributional Reinforcement Learning based Integrated Decision Making and Control for Autonomous Surface Vehicles

TL;DR

This paper introduces a novel distributional reinforcement learning-based navigation system for autonomous surface vehicles, enabling safe and efficient obstacle avoidance and COLREGs compliance in complex maritime environments.

Contribution

It presents a new distributional RL approach that integrates onboard sensors for continuous control in maritime navigation, outperforming existing methods in safety and efficiency.

Findings

Superior navigation safety in simulations

Effective obstacle avoidance and COLREGs compliance

Outperforms state-of-the-art RL and classical methods

Abstract

With the growing demands for Autonomous Surface Vehicles (ASVs) in recent years, the number of ASVs being deployed for various maritime missions is expected to increase rapidly in the near future. However, it is still challenging for ASVs to perform sensor-based autonomous navigation in obstacle-filled and congested waterways, where perception errors, closely gathered vehicles and limited maneuvering space near buoys may cause difficulties in following the Convention on the International Regulations for Preventing Collisions at Sea (COLREGs). To address these issues, we propose a novel Distributional Reinforcement Learning based navigation system that can work with onboard LiDAR and odometry sensors to generate arbitrary thrust commands in continuous action space. Comprehensive evaluations of the proposed system in high-fidelity Gazebo simulations show its ability to decide whether to follow COLREGs or take other beneficial actions based on the scenarios encountered, offering superior performance in navigation safety and efficiency compared to systems using state-of-the-art Distributional RL, non-Distributional RL and classical methods.