A machine learning method for the evaluation of ship grounding risk in real operational conditions

• A Machine learning method utilising big data analytics is used to evaluate ship grounding risk. The analysis based on an Avoidance Behaviour-based Grounding Detection Model (ABGD). Results are in good agreement with real accident records available for routes between the ports of Tallin (Estonia) and Helsinki (Finland). proposed may assist identification operational vulnerability a fleet at knowledge developed can support development future decision systems. Ship groundings often lead damages resulting oil spills or flooding subsequent capsizing. Risks be estimated qualitatively through experts’ judgment quantitatively maritime traffic data. Yet, studies using remain limited. In this paper, we present evaluation risk environmental conditions. makes use streams from Automatic Identification System (AIS), nowcast data, seafloor depth General Bathymetric Chart Oceans (GEBCO). evasive action Ro-Pax passenger ships operating shallow waters idealized under various patterns that link side - forward scenarios. Consequently, (ABGD-M) introduced identify potential scenarios, probabilistic quantified observation points along voyages. applied over 2.5 years ice-free period Gulf Finland. indicate estimation extremely diverse depends voyage routes, points, It concluded (1) better critical scenarios underestimated existing databases; (2) improved understanding avoidance behaviours conditions; (3) profile life cycle operations (4) waterway complexity indices vulnerability.