Knowledge-Based Vessel Position Prediction using Historical AIS Data

The improvement in Maritime Situational Awareness (MSA), or the capability of understanding events, circumstances and activities within and impacting the maritime environment, is nowadays of paramount importance for safety and security. Enhancing coverage of existing technologies such as Automatic Identification System (AIS) provides the possibility to integrate and enrich services and information already available in the maritime domain. In this scenario, the prediction of vessels position is essential to increase the MSA and build the Maritime Situational Picture (MSP), namely the map of the ships located in a certain Area Of Interest (AOI) at a desired time. The integration of de-facto maritime traffic routes information in the vessel prediction process has the appealing potential to provide a more accurate picture of what is happening at sea by exploiting the knowledge of historical vessel positioning data. In this paper, we propose a Bayesian vessel prediction algorithm based on a Particle Filter (PF). The system, aided by the knowledge of traffic routes, aims to enhance the quality of the vessel position prediction. Experimental results are presented, evaluating the algorithm in the specific area between the Gibraltar passage and the Dover Strait using real AIS data.