MASS, COLREG, and technologies

Ship collision avoidance is crucial in maritime navigation and a significant challenge is translating the Collision Regulations (COLREG) into collision avoidance algorithms suitable for fully autonomous ships. To investigate, a mixed-method experimental research of criteria-based subjective performance was undertaken. This mixed approach involved gathering quantitative data through bridge simulator experiments and qualitative data through subjective assessments conducted by maritime simulation lecturers. Bridge simulators were effectively adapted to serve as a testing bed for autonomous navigation. During the algorithm validation stage, it became obvious the main challenge was to overcome the variety and subjectivity of COLREGs and the diverse ways seafarers take to avoid collisions in various situations. ‘Seamanship ordinary practice of seamen’ or by ‘the special circumstances of the case’ was a huge challenge for developing machine interpretation algorithms. The results achieved in the simulation trials uniquely identified the benefits of using iterative testing and development in collaboration with experienced seafarers and software developers.