On the Use of AIS and Meteorological Data to Identify Adrift Ships and Extrapolate their Drift Characteristics
Please login to view abstract download link
The wealth of information that constitutes AIS database has been used extensively to produce approaches for anomaly detection. A specific type of anomaly that can be of interest is a ship going adrift. Spotting a ship adrift is important if it does not signal itself rapidly. It can then be useful to be able to predict its trajectory, especially in regions with intense maritime traffic. The method commonly used to predict the trajectory of a drifting object is based on the use of object-specific drift coefficients in addition to wind and current field. However, such coefficients are obtained through field study and are not readily available for large size ships. We thus propose a method to identify a ship going adrift based on its AIS data and to then use these AIS data together with meteorological data to determine the ship drift characteristics.
