Normality Modeling and Change Detection for Space Situational Awareness (SSA)

ABSTRACT:Applied Defense Solutions (ADS) has embarked upon a new approach to data correlation and aggregation. The ADS Hierarchical Reasoning Tool (HRT) provides a set of unique signatures for automatically recognizing and classifying a resident space object (RSO). Here, we seek to leverage hierarchical reasoning to provide innovative and automated analysis capabilities that capture and learn the normal status and behavior of satellites, detect changes, and assess the implications all within the context of events in the space domain. To assess orbital events and provide for timely decision analysis and courses of action, ADS proposes the development of a scalable automated workflow to support an Orbital Event Characterization Tool (OECT) within a distributed service oriented based architecture. Our primary research goal will be to connect the RSO feature hypothesis generation capabilities of HRT into the OECT capability to provide high level hypothesis management of events, characterize anomalous events, and detect changes in both object appearance and behavior all within context provided by a multi-INT event timeline. Our proposed approach will model and/or learn the normal behavior of space-based objects via Bayesian update process and leverage ADS operational expertise of satellite operations to generate a hierarchy reflecting mission level object life cycles.BENEFIT:Hierarchical reasoning capabilities provide a structured and mathematically rigorous methodology to correlate and aggregate sparse data from disparate sensors. The completed software tools could be used by both government and commercial entities that wish to not only provide indications of and attribution for anomalous events but also predict the likelihood of future intention and warn of possible threats.