Formal Verification of Path Planning Safety and Reachability in Unmanned Surface Vehicles

Yu Lu, Pan Sun, Shijie Shi, Meng Li, Huilin Ge

Abstract: In the realm of maritime technology, Unmanned Surface Vehicles (USVs) are increasingly crucial, particularly in their use for complex tasks requiring efficient and reliable path planning. This study addresses the critical intersection of path planning and energy efficiency in USVs, focusing on the challenge of navigating dynamic oceanic environments while minimizing energy consumption. Building upon our previous work, which developed an energy-efficient path planning algorithm incorporating the Voronoi diagram, Visibility graph, Dijkstra's search, and an energy consumption function, we extend our research to the formal verification of these path planning algorithms. We employ model checking as a robust formal verification method. This paper presents a novel framework for verifying key correctness properties of USV path planning, including deadlock-freeness, safety, reachability, and energy efficiency, using linearly priced time automata (LPTA) and the UPPAAL model checking tool. Our approach not only demonstrates the potential of model checking in enhancing the reliability and operational viability of USVs but also pioneers its application in validating the real-world efficacy of USV path planning algorithms. The findings offer significant insights for advancing safer, more efficient, and economically viable USV operations, setting a new benchmark in the field of maritime autonomous systems.