KINGFISHER: Total Maritime Awareness System (Demonstration)

Continuous monitoring of naval activities of vast areas, hundreds of nautical miles away from shoreline, is one of todays most challenging problems. To enable total maritime awareness in these areas, the use of satellite-based sensors is an efficient and cost-effective way to perform this task. The KINGFISHER (Figure 1) is a maritime multi-layer intelligence system created for an in-depth analysis of large maritime areas of interest in order to provide maritime awareness and detect illegal activities such as illegal fishing and immigration. The major novelty of KINGFISHER is the detection of abnormal vessel behavior by analyzing and correlating various data sources such as Electro-Optics (EO) imagery, Synthetic Aperture Radar (SAR) imagery, Automatic Identification System (AIS) and Open Source Intelligence (Osint), to track moving vessels with satellite sensors in order to detect uncooperative ones, and select the most suitable satellite and its best time frame for vessel detection. Satellite imagery covers a relatively small area, and can be acquired only at predefined acquisition opportunities. Thus, effective usage of satellite sensor will lead to a more economic solution and is mostly required in this domain. The Research and Innovation team at ImageSat International (ISI) developed algorithms for KINGFISHER based on Artificial Intelligence (AI) techniques such as Multi Agent System and Deep learning. The objective of the AI techniques is to enable decision support for maritime operators and provide an optimal, autonomous and economic solution for maritime system. Detection of a moving vessels with satellite sensors, is a challenging problem which requires a large amount of satellite imageries to enable vessel detection. We propose to minimize this coverage area by using the prediction process of the moving target. Using multi-agent based modelling and simulation, we developed a prediction algorithm for the vessel behavior and selection algorithm which recommends on the best satellite and its observation window for this mis-