AIS-based near-collision database generation and analysis of real collision avoidance manoeuvres

Abstract Economic and technological development has increased the amount, density complexity of maritime traffic, which resulted in new challenges. One challenge is conforming to distinct evasion manoeuvres required by vessels entering into near-collision situations (NCSs). Existing rules are vague do not precisely dictate which, when how collision avoidance (CAMs) should be executed. The automatic identification system (AIS) widely used for vessel monitoring traffic control. This paper presents an efficient, scalable method processing large-scale raw AIS data using closest point approach (CPA) framework. NCSs identified create a database historical data. Important features describing CAMs defined, estimated analysed. Applications on high-quality real-world set show promising results subset situations. Future applications may play significant role regulatory framework, navigation protocol compliance evaluation, risk assessment, avoidance, algorithm design testing autonomous vessels.