A Rule-based Heuristic Method for COLREGS-compliant Collision Avoidance for an Unmanned Surface Vehicle

With an increasing demand throughout industry for autonomous vehicles with more intelligent decision-making features, the issue of Unmanned Surface Vehicle (USV) navigation is addressed with respect to automatic obstacle avoidance. This Obstacle Detection and Avoidance (ODA) system design approach addresses the primary concern, the safety of a USV in operation. Due to the desire for improved autonomy and thus reduced human interaction, it is essential to incorporate vessel avoidance protocols into the ODA architecture for operation amongst maritime traffic, implementing the marine Rules of the Road. A heuristic Rule-based Repairing A* (R-RA*) algorithm is proposed, which is integrated within a decision-making framework. The integrated system has proven effective in simulations for path planning in real-time, in accordance with the International Regulations for Preventing Collisions at Sea.