A small team of military researchers recently pushed a new AI-powered counter-unmanned aerial system (C-UAS) into maritime testing at NATO’s Bold Machina exercise.
Built by officer-scholars from the US Naval Postgraduate School, the system was deployed aboard a Dutch Navy combatant craft operating in European waters.
Its mission: detect and identify small, hard-to-track Class 1 drones without emitting signals that could compromise the vessel’s position.
To do that, the system uses artificial intelligence (AI) to combine data from multiple independent sensors, allowing operators to spot incoming drones at range while remaining completely passive.
Its architecture integrates four core subsystems, beginning with customizable sensor platforms, and the Tactical Hybrid Operational Router (THOR), which provides power and networking.
The system also features an AI-driven sensor-fusion engine, and a navigation display that overlays drone detections in real time for operators.
“NATO required us to create a system that was passive so that operators who were on a small boat wouldn’t give off any sort of detectable signatures or emissions,” said Max Leutermann, a US Navy lieutenant commander.
“We spent the beginning of the year figuring out solutions. Now, we started figuring out how to build it and who to build it with.”
The Detection Engine
During the exercise, the C-UAS system brought together a layered mix of sensing technologies to spot small drones across multiple domains.
Short-range acoustic and electro-optical/infrared (EO/IR) sensors from US-based MARA Technologies handled close-in detection, while radio-frequency sensors from DroneShield and Silvus Technologies helped identify and locate drone signals.
Trakka Systems supplied long-range EO/IR sensors, complemented by DSPNor’s low-probability-of-intercept radar.
An Nvidia Jetson developer kit runs the AI,improving detection in real time and updating threat libraries as drone designs evolve.
