A Risk Assessment and Alerting System for Maritime Attacks

The impact of piracy on the maritime economy has been increasing dramatically over the last decade, which encouraged actors of maritime security to support research works on detection and protection against this threat. However, providing automated support for attacks in the maritime domain is highly dependent on the context and on the information available to the system. In this paper, we present an algorithmic module and its underlying rule-based risk model, designed to detect abnormal behaviours such as maritime attacks from heterogeneous sources of information. The module uses a combination of kinematic and non-kinematic analysis coupled with coherence checks to reach acceptable operational performance, and has been tested on simulations, field tests and real traffic recordings.