Dedicated Counter UAS Opposing Forces Utilizing AI/ML: CORTEX

Big Picture: In December 2023, dozens of unidentified drones entered Langley Air Force Base airspace, uncontested, for more than two weeks. None were shot down, and the identity of the operators remains unknown. The United States' counter-UAS (c-UAS) capabilities have failed to prevent hundreds of drone incursions at military installations such as Langley, Hill, and Wright-Patterson AF Bases over the past few years, according to NORTHCOM. Alarmingly, these reported incidents are just the tip of the iceberg; General Gregory Guillot warns that "there's significantly more incursions that we don't see."The United States stands on the brink of a disastrous three-front war that will be dominated by drone warfare, leaving US national security alarmingly vulnerable due to the inadequacy of counter-UAS capabilities. If c-UAS systems are not rapidly trained and developed to meet this threat, the warfighter will be overrun by hordes of adversarial drone swarms. Technical Problem: US counter-UAS systems are falling behind the rapid advancements in UAS technology. Training programs are outdated and fail to simulate the intelligent and quick systems encountered in the field. This results in c-UAS capabilities lagging, creating security gaps and unvalidated defenses. Without effective training and evaluation methods, national security is at significant risk. An advanced training solution is crucial to develop robust c-UAS systems. Solution: Our solution is an AI/ML solution that enables coordinated autonomous swarms of heterogeneous UAS platforms through advanced reinforcement learning and multi-agent coordination. It enables a single operator to control 10 or more diverse UAS platforms simultaneously, executing sophisticated tactics, techniques, and procedures based on documented adversary behaviors. By deploying our solution as a Red Team Opposition Force, the system provides standardized, repeatable c-UAS testing capabilities across DoD installations, enabling comprehensive evaluation of defensive systems against realistic coordinated drone attacks representing EW, EMP, ballistic, and cyber vulnerability profiles. The AI/ML foundation ensures continuous adaptation and learning from each engagement, allowing our solution to evolve with emerging threats while simultaneously training both c-UAS systems and operators through authentic adversarial scenarios. This dual-use capability helps the USAF develop more effective c-UAS technologies by providing validated performance metrics and identifying defensive gaps that traditional testing methods cannot reveal, ultimately enhancing force protection and defending US personnel and assets against increasingly sophisticated UAS threats.