Case-Based SSA Probabilistic Reasoning Using Multi-Domain Information
In order to help space situational awareness (SSA) decision-makers identify and prioritize threats, the SHERLOC (Ssa HEuRistics using fuzzy LOgic and Case-based reasoning) system is designed to use probabilistic reasoning from multi-domain evidence to interpret the current situation in real-time based on knowledge of past cases. A case-based reasoning (CBR) approach captures complex relationships between evidentiary factors and threat conditions and adapts as new threat examples are added to the case base.
Different information sources, from Indications and Warnings (I&W) alerts to observable tracking data to communications traffic and collected intelligence, can most effectively contribute to situational understanding when considered together for the aggregate picture they present. SHERLOC provides a probabilistic framework for interpreting otherwise hidden interrelationships between multi-domain input data. Prior probability distributions for combined evidence factors form the basis for threat assessments in new situations, with the added benefit of explainability in terms of relevant similar prior cases. Bayesian conditional probability tables provide a method for augmenting directly derived case data with estimated values from subject matter experts.
An initial operational prototype computes dynamic threat assessments for three different kinds of space threats, each with input variables for different combinations of multi-domain evidence unique to the threat categories. A console in the prototype user interface allows operators to perform rapid “what if” experiments to see how different values for input variables affect the current threat estimate.