Detection of anomalous behaviour in ship reporting data for improved maritime security

Nowadays, most of the world’s medium and large ships use equipment to self-report their positions. The data from these systems add up to large amounts, and provide detailed insight in ship traffic and ship behaviour. It is sought to exploit these data to find security risks in the maritime domain, by finding anomalies, i.e., behaviours that are different from what is normal. A few examples of different methods and their results are discussed, using data acquired over the Mediterranean Sea and the Western Indian Ocean. It is concluded that automatic algorithms for anomaly detection already provide some functionality, but that their design needs further improvement; and that for the final interpretation a human analyst is still needed.

VAN WIMERSMA GREIDANUS Herman;  VESPE Michele;  ALVAREZ ALVAREZ Marlene; 

2016-09-23

Fraunhofer Verlag

JRC100956

978-3-8396-1011-4,   

2364-3986,   

http://www.future-security2016.de/,    https://publications.jrc.ec.europa.eu/repository/handle/JRC100956,