A ship collision is an accident that can have a significant impact on life and the environment as well as property loss. These ship collision accidents are occurring continuously and collision accidents due to failure of collision risk situation awareness of the navigators are the majority of ship collision accidents. The reduction of navigator's failure of these type of human error is crucial for collision accident prevention.
This study proposed a evaluation and warning system for collision risk situation awareness of ship navigator. And as a fundamental study of this system, the evaluation algorithm was developed which can classify whether the navigator aware collision risk situation or not.
Navigator behavior data were collected under in a simulated environment using multiple accelerometers to collect behavior patterns of before and after the awareness of collision risk situation of the navigators. Then, the analysis of the collected data and derivation of algorithm was performed using machine learning. The validation of the algorithm was carried out through the validation experiment. The proposed algorithm is expected to provide a basis for the evaluation and warning system for collision risk situation awareness of ship navigator to prevent ship collision accidents in practice.