Cyber-resilient multi-modal sensor fusion for autonomous navigation
The lecture will present recent advances on multi-modal sensor fusion for cyber-resilient
navigation of autonomous ships in the presence of physical malfunctioning and cyber
threats. Leveraging state-of-the-art methods in machine learning and sensor fusion, virtual
redundancies are developed for the different sensing modalities allowing for the design of
monitoring algorithms to ensure the information integrity of key sensors for navigation and
target tracking. The theoretical framework will be introduced and specific examples for
autonomous navigation of commercial vessels will be expanded, showing the validity of the
research framework on full scale data collected during experimental campaigns of the first
Danish autonomous ferry. The lecture will conclude presenting current and future directions
of research.