default The Goal Oriented Decision Support System (GO-DSS) for surveillance asset allocation: integration of different information layers

By

The Goal Oriented Decision Support System (GO-DSS) for surveillance asset allocation: integration of different information layers. Fabbri, Tommaso; Vicen Bueno, Raul; Grasso, Raffaele. CMRE-FR-2015-016. December 2015.

Data mining, AI and optimisation techniques provide helpful information to substantially improve the decision making process. The above mentioned techniques can be exploited to increase the performances of planning operations. This research details the developments of the Goal Oriented Decision Support System (GO-DSS) as a planning system of N controllable moving assets to optimise the coverage of high-risk areas during counter-piracy operations. By means of machine learning, data fusion and multi-objective evolutionary optimisation, this research shows how heterogeneous in-formation sources can be optimally exploited to improve the performance of the DSS. The research illustrates how Meteorological and Oceanographic (METOC) forecast, pirate attack reports and AIS vessel traffic data are integrated as information sources and details the principal components of the planning tool. Preliminary evaluation scenarios are also provided demonstrating the effectiveness of the approach. Test cases performed during this research and development proves a reduction of more than 60%of the distance between the reported piracy activities and the planned asset trajectories compared to the case in which AIS traffic data are not used.