Task Assignment Optimization in Operational Mission Planning for Scalable Cross-Echelon Command and Control of the Future

Author(s):

Angelique Scharine, Mark Dranias, Gregory M. Gremillion, Sarah Al-Hussaini, Ahmed Khalil, Yoonjae Lee, and Efstathios Bakolas

Abstract:

The military decision-making process has evolved to ensure that all factors of planning are accounted for. Its current form is complex, labor intensive, and slow. To streamline this process, ARL researchers have developed an integrated technology system consisting of a large language model artificial intelligence tool to generate courses of action (COAs), and a constructive simulation tool to allow commanders to forecast outcomes of potential COAs. To augment the capability of that system, this effort examines the development of a greedy task assignment model to optimize the allocation of military units to mission tasks in a simplified but applied domain of brigadelevel mission planning. This application yielded mixed results, as mission planning depends on more than combat power; commanders consider tactical roles, senior commander guidance, and terrain features, which are challenging to account for in a simplified utility model. Here, two implementations are described—a simultaneous assignment and a phased greedy assignment—and the strengths and drawbacks of each method are discussed.