Brief design overview of a new situational awareness system for autonomous ships

Abstract

This paper proposed a new design and its core algorithms of a situation awareness system, named Korea au tonomous surface ships (KASS) intelligent situational awareness system iSAS. iSAS consists of deep-learning algorithms for detecting marine objects by using camera, radar, and light detection and ranging (LiDAR), a probability-based data association algorithm and Bayesian filtering-based multi- target tracking algorithms, and a semantic information-based collision risk evaluation algorithm, which are beyond the functions of acquiring environmental information performed by seafarers of a manned ship. We verified the feasibility of our design with preliminary results and continue developing and verifying it to achieve a higher level of completion at sea.