LiDAR Co-Simulation for Maritime Assistance System V&V

Kurzfassung

Maritime Autonomous Surface Ships (MASS) increasingly rely on LiDAR-based perception systems for maneuver automation, collision avoidance, and berthing assistance. However, systematic V&V of these safety-critical, sensor-driven systems remain constrained by the lack of scalable, automated, and transparent simulation environments. While commercial maritime simulators offer high-fidelity traffic environments, they typically rely on proprietary, non- interchangeable sensor models. Research-oriented platforms, on the other hand, often operate offline and lack real-time integration with bridge systems, limiting their applicability for end-to-end testing. To address this gap, this paper presents a sensor-centric co- simulation framework that integrates a modular LiDAR simulation environment with S-100 maritime traffic data or real onboard bridge systems. The proposed approach supports data- driven scenario generation based on maritime standards, high- fidelity raycasting-based LiDAR simulation, and real-time coupling with both SiL and HiL testing workflows. An experiment using the research vessel Sally demonstrates the framework’s ability to inject virtual LiDAR data into live operational scenarios, enabling systematic evaluation of assistance functions under realistic conditions. The results highlight the potential of sensor-centric co-simulation as a critical enabler for scalable, reproducible, and evidence-based V&V in the maritime domain.