Revolutionary AI Controls High-Energy Lasers to Boost Drone Defense by NPS

With the escalating threat posed by inexpensive uncrewed autonomous systems (UAS), researchers at the Naval Postgraduate School (NPS) and their partners are leveraging artificial intelligence to revolutionize laser weapon systems (LWS) for enhanced UAS tracking. By automating processes like target classification and maintenance, the LWS can more effectively neutralize hostile drones.

The sophisticated tracking sequence in an LWS can be cumbersome when executed manually, especially under the pressure of multiple drone threats and fast-approaching missiles. This situation is being tackled by an innovative collaboration involving the Naval Surface Warfare Center, Lockheed Martin, Boeing, and the Air Force Research Laboratory (AFRL). They have developed a system that places operators on-the-loop rather than in-the-loop, optimizing control through AI.

Professor Brij Agrawal of NPS emphasizes the cost-efficiency of the LWS. While expensive to build, these systems cost only a few dollars per shot, making them a viable alternative to costly interceptor missiles. The key to this efficiency is advanced AI training, which has been put into practice with thousands of drone images, leading to a robust AI model that has been validated for real-world applications.

Supported by the Joint Directed Energy Transition Office and the Office of Naval Research, the initiative aligns with the CNO’s strategic NAVPLAN, focusing heavily on AI and directed energy applications. The system detects drones via radar, with infrared sensors taking over for precise targeting, enabling the LWS to maintain a lock on targets through dynamic conditions.

Beyond simply identifying targets, the system distinguishes among drone types and poses, pinpointing vulnerable aimpoints. However, atmospheric challenges can complicate these efforts, emphasizing the importance of AI in automating and maintaining precision targeting.

A significant component of this development is the High Energy Laser Beam Control Research Testbed (HBCRT) at NPS. This testbed simulates shipboard laser systems and has undergone enhancements to study the effects of atmospheric conditions on laser precision. The research has also been enriched by contributions from NPS students, with investigations into AI algorithms and optical advancements.

The improved AI model is built on comprehensive datasets containing both real-world and synthetic drone images. The training process utilized advanced deep learning techniques, achieving greater fidelity through diverse image conditions. The culmination of this effort is a highly reliable model now being tested at the Naval Surface Warfare Center.

This research represents a strategic leap in military technology, offering potential applications across various platforms. As military defenses face growing threats from drone swarms, the advancement reflected in this AI-enabled LWS points towards a future of improved speed and effectiveness in threat assessment and engagement.