Integrating rule awareness and semantic reasoning in collision-free vessel path planning

This paper presents the design of an intelligent guidance framework for collision-free navigation of Autonomous Surface Vessels (ASVs), integrating traffic rule awareness and reasoning characteristics. The proposed framework leverages the available qualitative information related to traffic rules and the operational environment(s), in the form of semantic information, as well as sensor information to make online path planning decisions. A modular finite-state machine assigns traffic roles, while the path planner computes a collision-free envelope and reasons over a path, considering both vessels and infrastructure. A Line-of-Sight algorithm and controller enforce the selected path. The method’s effectiveness is demonstrated in a multi-environment case study involving head-on and crossing encounter scenarios, showcasing its adaptability and efficiency across short-sea and inland waterway operations.