FA865822CB006

Small Business Innovation Research Program (SBIR) Phase II

WEB WEASELS: Autonomous Swarming AI Munitions

Web Weasels (WW) is a multi-pronged RDT&E program designed to meet one of the stated goals from the USAF's 2030 Science & Technology Strategy document: Overwhelm adversaries with complexity, unpredictability, and numbers through a collaborative and autonomous network of systems and effects The near-peer conflict environment is far deadlier to 4th Gen air systems and even 5th Gen can be challenged through cyber and EW methods. The next DoD Offset' calls for smart, autonomous, and connected munitions to prepare, seize and maintain air superiority over the Battlespace. WW builds on a unique swarm planning and deployment capability (Air Commons - Swarm or AC-S) by applying cutting edge AI/ML training techniques and edge computing. WW redefines a smart bomb' to be a highly trained, thinking, reactive, and team-playing asset in the fight. In addition, WW has been built to address the AFRL 2022 SBIR requirements: The Armament Directorate is pursuing ideas that permit Blue Forces to command various collaborative weapons systems and coordinated tactics to ensure success. A dynamic battlespace requires automated, adaptive weapons systems and cooperative tactics Artificial intelligence algorithms with dialable human influence Squadrons of autonomous collaborative munitions operating at the far edge of the C2 infrastructure, need the training, tactics, techniques, and procedures (TTPs) to handle the speed-of-datalink environment that occurs in modern combat. Teamwork, communication, shared mental models, and a robust set of tried and tested strategies are required to survive and to dominate. The parent program, AC-S, allowed commanders to plan, task and manage multiple swarming assets through a 'Swarm ATO' and a Swarm Engine. WW is designed to overlay AI/ML training algorithms onto AC-S to provide pre-launch munitions with a series of TTPs in a Playbook' for a given mission set (i.e., SEAD). The munitions will launch with AI-on-the-Edge hardware, firmware and comms architecture (OXYGEN and DROIDISH ) which allows them to collaboratively assess the combat environment, vote on the most suitable Play, call an Audible' change, and execute the improved plan. WW uses three teams, one building and training AI/ML Playbook algorithms; one focused on AI-on-the-Edge, voting and Audibles; and the third one testing Human-Swarm Interfaces (HSwI) configurations. WW will use AirSim to run trial iterations and Mission Rehearsals, and then fly suitably equipped small UAS (emulating autonomous stand-off munitions) against Liteye System's SHIELD Counter-UAS systems (emulating an Enemy Force integrated radar threat) in the time-honored traditional mission sets of the original USAF Wild Weasels: First In, Last Out .

The goal of phase II is to continue the R&D efforts initiated in Phase I. Funding is based on the results achieved in Phase I and the scientific and technical merit and commercial potential of the project proposed in Phase II.

AF22Z-PDCSO2

F2D-4696

22.4

Keven Gambold