Accident Prevention Based on Maritime Collision Cases in a Port: AI-based Measures to Prevent Ship Collision Accidents

Abstract

This study investigates a collision incident that occurred in Nagoya Port on 12 July, 2022, highlighting the weaknesses of passing agreements that depend on VHF international radiotelephony and the role of human error in frequent port collisions. In this incident, diminished course-keeping ability at low speeds and minor course adjustments were incorrectly interpreted as the other vessel initiating a manoeuver. This misunderstanding resulted in a delay or failure to execute the agreed starboard-to-starboard passing, communicated via VHF, ultimately contributing to the collision. To address this, the authors propose a ‘digital passing agreement’ system integrated with an Automatic Identification System (AIS). This system converts verbal passing agreements into machine-readable data, displaying the agreed status, counterpart’s intent and navigational assumptions on the Electronic Chart Display and Information System/radar/AIS. In addition, it detects early deviations from agreements based on tracks, bearings and speed, facilitating alerts, cancellations or renegotiations to visualise agreements and ensure compliance. The design also enables fast incorporation of third-party advice and alerts through integration with Vessel Traffic Services. Analysis from the 4M perspective indicates that this proposal effectively targets man/management factors, capable of reducing discrepancies in situational awareness under media conditions within the port. This proposal effectively shifts decision-making from a voice-dependent model to a data-driven approach, thereby significantly reducing the impact of human error on vessel operations within the port.