Shore-Side Collision Risk Surveillance for Autonomous Surface Vehicles Using CPA/TCPA Encounter Assessment

Abstract

Autonomous surface vehicles (ASVs) require reliable encounter assessment to support remote supervision in restricted and coastal waters. This paper proposes a shore-side collision-risk surveillance approach based on Closest Point of Approach (CPA) and Time to Closest Point of Approach (TCPA). The proposed framework combines own-ship telemetry, target-track data, relative-motion estimates, and communication status to produce a four-tier alert status for the monitoring operator. Evaluation used records from waypoint navigation, dynamic obstacle trials, web-based telemetry, and fail-safe testing. Simulation and field navigation produced mean waypoint errors of 0.49 and 1.15 m and mean heading errors of 4.21° and 10.8°, respectively. Ten dynamic obstacle trials achieved 100.0% avoidance success and mission recovery, with a mean activation delay of 0.004 s, observation-to-maneuver time of 0.881 s, and avoidance duration of 54.184 s. The communication infrastructure reported a GPS error of 3.66 m, data-fetch time of 199 ms, transmission time of 165 ms, and end-to-end monitoring latency below 250 ms. The combined records demonstrate adequate data readiness for encounter surveillance, while full CPA/TCPA validation still requires continuous target trajectories. The resulting approach supports remote situational awareness without changing the existing vehicle-control architecture.