Intelligent Mission-Planning Technologies: Procedures and Findings from Phase 1 and 2 Experiments

Author(s):

Anthony L. Baker, David Chhan, Bret L. Kellihan, Sarah Thomas, Aaron Necaise, Joshua Foldes, C. Shawn Burke, Katherine R. Cox, and Joe T. Rexwinkle

Abstract:

We are developing intelligent mission-planning technologies to predict performance and workload in mission areas using real-time data. These technologies will enhance mission planning and optimize Soldier capabilities during execution. Two studies, Phases 1 and 2, have been conducted to evaluate this capability. Phase 1 focused on combining participant feedback from after-action reviews with in-mission data to predict workload and performance in subsequent missions, validating the workload-prediction algorithm. Phase 2 expanded this concept to two participants working together and tested a new cooperative planning interface—demonstrating a learning pipeline that adapts in multi-agent scenarios. From these studies, we demonstrated that an evolving algorithm can generalize workload predictions across varying mission contexts. This approach enables the synthesis of multiple diverse data streams to accelerate the mission-planning process and optimize the performance of multi-human multi-agent teams. Ultimately, the technology aims to inform Soldiers of challenging situations in the battlespace, which will enhance mission outcomes in complex, dynamic future conflicts.