Operational Planning Decision Support using Multi-Dimensional Data Farming

Multi-Dimensional Data Farming (MDDF) uses Machine Learning (ML) to automate data farming to allow improved and faster decisions in highly complex multi-scale, multi-domain, and multi-level hybrid war campaigns. This has significant utility when used in support of Operational planning, allowing multiple Courses of Action (CoA) to be rapidly developed and evaluated prior to the execution of any operation. Using MDDF enables decision-makers to explore the problem space and identify multiple optimal solutions significantly faster than current techniques.
MSG-186 has applied MDDF in a sand-box environment to an illustrative combined strategic campaign and tactical hybrid warfare operation resource allocation problem considering the balance between local and global optimal solutions. We have tested the technical feasibility of implementing MDDF within the Federated Mission Network operational environment at Coalition Warrior Interoperability Exercise (CWIX).
Through MDDF, we aim to show it is possible to combine ML techniques exploring operations at multiple scales (Multi-Domain Operations and targeted-fidelity modelling) and optimize the strategic/operational level goal, by selecting the correct resource allocation scheme at the tactical level. This paper describes an ML-based assistant able to conduct MDDF experiments and optimization tasks on an automated basis, which was examined in detail during CWIX in 2024.