Cross-Domain Strategic Planning with Adaptation for Risk Uncertainty and Opportunistic Targeting

Elder Research, Inc. (ERI) and the Virginia Tech Hume Center bring together a data science team with operational context in advanced AI/ML workflow development and implementation, and successful research in semantic reasoning, planning under uncertainty, and hierarchical multi-agent tactical planning. Through applying autonomy to tactical planning and execution, we have come to understand the importance of developing models and plans that are inherently flexible, and allow adaptation to the changing battlefield environment. Our proposed approach for operational and strategic planning is based on an agnostic concurrent, hierarchical graph search planning algorithm that accounts for the curse of dimensionality, environmental and adversarial uncertainty, cross-domain optimization, domain-specific optimization algorithms, and the risk inherent in tactical execution. The key innovation of this algorithm is its ability to generate multiple courses of action across domains and hierarchies that inherently account for risk in the battlespace environment, reducing the planning cycle and enabling opportunistic targeting based on C4ISR.