AI-based navigation systems meet collision avoidance rules, project concludes
MAXCMAS stands for machine executable collision regulations for marine autonomous systems and other participants in the study were Lloyd’s Register, Warsash Maritime Academy (WMA), Queen’s University Belfast and Atlas Elektronik.
The partners demonstrated that the use of newly developed algorithms allowed existing collision avoidance regulations to remain relevant in a crewless environment, enabling Artificial Intelligence-based navigation systems to comply with the rules to avoid collisions effectively, even when manned vessels in the vicinity were interpreting the rules differently.
The research was undertaken using WMA’s networked bridge simulators which analysed reactions from crews faced with real-world situations to design and refine the MAXCMAS algorithms. The project partners adapted a bridge simulator of commercial specification for this purpose.
“Through MAXCMAS, we have demonstrated autonomous collision avoidance that is indistinguishable from good seafarer behaviour,” said Rolls-Royce Future Technologies Group’s Eshan Rajabally, who led the project. “We’ve confirmed this by having WMA instructors assess MAXCMAS exactly as they would assess the human.”