A collision avoidance algorithm for ship guidance applications

The paper presents a collision avoidance algorithm for ship open sea navigation, based on an ad hoc modified version of the Rapidly-exploring Random Tree (RRT*) algorithm. The proposed approach is designed to act as the high level layer of the navigation control structure for an autonomous ship. Collision and grounding still represent the primary source of sea accidents, thus an automatic system able to detect static and moving obstacles and plan an evasive route could significantly improve safety during navigation, especially in crowded areas. Focusing on the maritime field, a review of the scientific literature dealing with collision avoidance is presented, showing potential benefits and weaknesses of the different approaches. Among the several methods, details about the RRT and RRT* algorithms are given. The ship path planning problem is introduced and discussed, formulating suitable cost functions and taking into account both topological and kinematic constraints. The algorithm is able to manage multiple moving obstacles with variable speed and course. Eventually, a time-domain ship simulator is used to test the effectiveness of the proposed algorithm over a number of realistic operation scenarios. The obtained results are presented and critically discussed.