Sharing ML Models at the Edge with the Frontline Perception System (FPS)

The scope of this SBIR effort will focus on proximal model & alert data transfer in denied environments to enable autonomous vehicles to automatically transfer ML capabilities (such as models or alerts of objects/signals of interest) amongst themselves. Conceptually, this works like Appleĺs Airdrop technology - key data is able to be transferred between assets with proximal connectivity in austere locations or communications denied environments without the reliance of traditional Line of Sight (LOS) technology. This new feature, which will be developed under the scope of this Phase II. provides the Air Force the capability to utilize expendable, highly attritable assets to quickly notify and pass critical data to other aircraft or sensors in the stack for enhanced decision-making and action. Outcomes include improved battlespace awareness and comprehensive collaboration to increase the lethality and performance of combat aircraft and ISR platforms.