High-Speed Data Return for Tactical Environments

The modern battlespace includes incredible complexities and near-peer adversaries that make missions more dangerous than ever to our pilots and co-pilots.Ā Such warfighters rely heavily upon advanced sensor systems to deliver them real-time situational awareness (SA) that allows them to effectively achieve their mission objectives.Ā While the U.S. Navy has continued to increase the number of sensors and storage for data/intelligence onboard their aircraft, they currently lack a time-efficient process to effectively utilize all the data they receive. ĀIn order to utilize the data effectively, two major challenges need to be addressed: Automation of mission data exploitation to extract actionable intelligence automatically Automation of mission re-planning to rapidly inject new intelligence into the current mission. While significant ongoing investments are focused on the automated sensor exploitation need, we assert that more research funding is necessary to solve the mission re-planning challenge.Ā Toyon proposes research and development of machine learning techniques to automate mission planning in ISR platforms. Our proposed approach is the application of Imitation Learning (IL) to develop a mission-planning AI which mimics the behaviors of mission planning subject matter experts. We assert that the IL approach is both novel and promising to solve the incredibly difficult problem.