Pose planning of autonomous surface vehicle for passing through or around objects

Abstract

Autonomous Surface Vessels, also referred to as Unmanned Surface Vessels (USVs), are revolutionizing maritime operations. The self-navigating vessels make use of GPS, LiDAR, and cameras for navigation and decision-making independent of human input. These vessels have found increased applications in areas of use as varied as military surveillance to oceanographic research. The Maritime RobotX Challenge is a prestigious biennial global competition for university students that advances innovation in autonomous maritime robotics and builds technological advances with strengthened collaboration between academia and industry through its teams from around the world. A key aspect of this challenge is the structuring of powerful USVs that can most easily perform a set of tasks. The particular research is focused on the design and implementation of an appropriate pose-planning system for RobotX challenges: avoiding obstacles, following paths, and circling around the targets. The hybrid pose planning system was realized: An improved A* algorithm made for global pose planning and a modified artificial potential field approach for local pose planning. The system has been implemented on an ROS basis, and its key components-task global path planner, local path planner, and planning bus have been validated through simulations enabled within this environment. It was further deployed on the WAM-V16 USV and tested on-site during the competition. It provides valuable lessons in the design and application of USV pose planning systems.