Complex event processing applied to early maritime threat detection

In recent decades the globalization has caused a huge increase of ship movements carrying goods and passengers between countries. It makes quite difficult to detect manually critical or relevant situations that may occur in marine traffic. In this paper, we propose an event-driven service-oriented-architecture that combines the complex event processing and data distribution system, building a high performance and available system for analyzing and correlating data provided by ship's automatic identification systems in real time. This architecture will be able to detect automatically, and as soon as possible, abnormal situations occurred in seas all over the world. For this, a set of complex event patterns for detecting AIS hijacking or failure, as well as ship engine malfunction or ship collision is proposed and defined in this work.