HERA (Hypothesis Evaluation and Reasoning Assistant)

HERA (Hypothesis Evaluation and Reasoning Assistant) is a innovative decision support system leveraging Multi-Entity Bayesian Networks (MEBN) and the Probabilistic Web Ontology Language (PR-OWL) to autonomously manage multiple hypotheses in complex, uncertain decision spaces typical of modern defense environments. HERA addresses key challenges in hypothesis management, including explainable hypothesis generation for adversary intentions and friendly courses of action (CoAs), low-latency processing of exponentially growing adversary/friendly decision spaces, and maximizing user trust through advanced AI explanation capabilities. HERA builds upon a TRL7 MEBN/PR-OWL framework to provide a robust foundation for probabilistic reasoning under uncertainty. The proposed system features a modular, plug-and-play architecture supporting deployment across multiple domains. In Phase I, the Interactive Aptitude LLC and George Mason University team will develop proof-of-concept MEBN models, prototype core decision support components, and demonstrate the technical feasibility of HERA for a space/ground design reference mission, setting the stage for further maturation and transition in subsequent phases.