A Deep Spatiotemporal Approach in Maritime Accident Prediction: A Case Study of the Territorial Sea of South Korea

Abstract

Predicting the risk of maritime accidents is crucial for improving traffic surveillance and marine safety. The availability of data sources and development of machine learning and deep learning methodologies have the potential to improve operational risk prediction. This study investigated the application of deep learning in both short- and long-term predictions of maritime accidents in different grid sizes by considering multiple influencing factors. Accordingly, we utilized several big data sources containing data collected from the territorial sea of South Korea, including ocean accident, vessel trajectory, ocean depth, and weather data. Seven machine learning and two deep learning algorithms were implemented and compared for three different grid sizes using daily, weekly, and monthly models. The results revealed that the proposed deep spatiotemporal ocean accident prediction (DSTOAP) model outperformed other traditional machine and deep learning algorithms in most scenarios, not all of them. For the practical application, the results of this study can guide ocean accident management and safety planners in choosing appropriate methods for different time schedules and different grid sizes according to the range of coverage of the patrol ship.

Keywords: deep learning, Maritime Risk, Ocean Accident Prediction, Vessel Trajectory