Anomaly Detection and Intelligent Sensor Resource Management

Recent DoD programs have furthered intelligence, surveillance and reconnaissance (ISR) objectives by advancing the state-of-the-art in sensor resource management and target tracking. The majority of this work, however, has focused upon the problem of tracking a handful of high value targets for lengthy periods. There is a need to develop systems that detect anomalous activities that have the potential to be threatening or to be militarily significant. Such detections may result in high value target tracks, but the objective of this effort is to manage a constrained set of ISR assets to detect anomalous events. This requires the an automated resource manager to actively command sensors and modes, and to do so in a manner that de-conflicts parallel, competitive tasks. The sensor suite must simultaneously build normalcy models, detect anomalies, command sensor timeline to evaluate anomalies, and manage concurrent, high value tracks. Black River Systems proposes to develop a system architecture that manages sensor resources to execute four concurrent objectives: normalcy monitoring, anomaly detection, anomaly interrogation, and high value target track. Leveraging our tracking, sensor resource management, and simulation and modeling experience, we will demonstrate and analyze system performance through trade studies and simulations.