Ais-based event detection for offshore wind maintenance vessel operations: A validated framework

Enhancing the efficiency of offshore wind farm (OWF) operation and maintenance requires understanding vessel movements, turbine visit sequences, vessel–turbine interaction duration as vessel–turbine interaction durations, and transit times. However, detailed operational records are often unavailable because manual logging adds workload for operators. This study proposes an automatic identification system (AIS)-based framework for reconstructing turbine-level maintenance vessel operations from vessel trajectories and turbine coordinates. The framework combines heuristic vessel identification, geospatial filtering, and rule-based event detection to infer visited turbines, visit sequences, service timing, and vessel–turbine interaction durations from AIS data. It is developed using AIS data from maintenance operations at Horns Rev 3 in Denmark and externally validated using operator logs from a vessel operating at three German offshore wind farms: Amrumbank West, Kaskasi, and Nordsee Ost. For a representative validation day, the framework exactly reconstructed the full turbine visit sequence and achieved minute-level temporal accuracy. Across three months of validation, it achieved 100% visit-sequence accuracy, a mean Interaction-Duration MAE of 1.36 min, and a mean mission-duration error of 1.29%. The results show that AIS data can support reliable reconstruction of turbine-level offshore maintenance operations when proprietary records are unavailable, enabling improved vessel scheduling, routing, and maintenance planning.