A probabilistic analytics method to identify striking ship of ship-buoy contact at coastal waters

Abstract

The identification of the ship that contact with the buoy can provide evidence for accident accountability. To this aim, the paper develops a probabilistic analytics method to evaluate the ship-buoy contact risk for the striking ship identification at the coastal areas by combining buoy domain and bounding box models. The method makes use of Automatic Identification System (AIS) data and navigational buoy data. Firstly, an AIS-based probabilistic buoy domain model is adopted for the determination of the safety boundary of the buoy to detect potential striking ships with a higher contact probability. Then, the bounding boxes of the navigational buoy and the detected potential striking ships are developed to detect the real striking ship by analyzing the interaction between the ship bounding box and the buoy bounding box. Finally, the probabilistic analytics method is demonstrated in the South China Sea and validated using historical ship-buoy contact records. Results indicated that, from a probabilistic perspective, the safety buoy domain (critical boundary) existed with diverse distances dynamically. The proposed method could assist the identification of striking ships while aiding the definition of the safety buoy domain for preventing ship-buoy contacts. As a result, it has the potential to support the development of ship-buoy contact risk management and assist surveillance operators and master on board by improving their cognitive abilities in dangerous traffic scenarios.