Cerema is in support of the Department of Maritime Affairs and the Department of Energy and Climate to deal with maritime safety issues around future offshore wind farms. In this context, the study of ships abnormal behavior (such as drifting) is very little studied by lack of tools in this field. The purpose of the challenge is to develop an innovative demonstrator in the form of an algorithm. In the aim of improving the detection of abnormal ship behavior with Artificial Intelligence solutions (or another way). For this challenge, we will focus our development on AIS data. We will analyse, as main features, the dynamic informations about ships contained on AIS data like speed, course... But also static informations like type of ship, dimension, activities... In addition we could use the weather forecast for a global view. The idea is that at the end of the challenge, we will be able to identify (using IA or thresholds) the main abnormal ship behaviours according to the traffic around to improve the maritime safety. |