Florida International University researchers have developed a new way to detect incoming drones using the unique magnetic signatures generated by their motors — offering a detection method that is stealthier, safer and drastically cheaper than traditional radar.
As small drones become increasingly common on the battlefield, the need for affordable and effective counter-drone technologies is growing. The Department of War has requested $74 billion for drone and counter-drone technologies in its fiscal 2027 budget.
The research findings, published in IEEE Sensors Journal, show that drones can be identified by an unavoidable byproduct of flight: the extremely low frequency magnetic fields generated by their spinning electric motors. Because different drone models rely on distinct motor configurations, each model leaves its own low-frequency magnetic signature.
By recording these signatures with a specialized antenna and processing the data with a machine learning model, the FIU team identified outdoor drones with up to 98.52% accuracy at ranges up to 15 meters — detecting not only a drone's presence, but its specific make and model.
Unlike standard systems, FIU's technique requires no active signal transmission. It is a potential leap forward in stealthy drone detection, said Arjuna Madanayake, a professor at FIU's College of Engineering & Computing.
"Every detection method currently in use gives something away. Radar transmits a signal. Acoustic systems get overwhelmed by background noise. We went a different direction entirely," Madanayake said. "While the range of our method is limited for now, it costs only a few hundred dollars to develop and could run continuously without any specialized support, unlike radar systems."
The study focused specifically on micro-Unmanned Aerial Systems (micro-UAS) — small, mass-producible drones that can carry explosives or spy on adversaries while remaining nearly invisible on conventional radar. Because FIU's device passively receives low-frequency emissions without emitting a return signal, it leaves no trace for adversaries to track.
The research group is currently working with U.S. industry to further develop the technology.
"Knowing the type of drone that's coming matters," Madanayake said. "Different drones can carry different payloads and have different threat profiles. The more you can tell a defender about what's coming, the better they can respond."
The full team, led by Madanayake of FIU's Department of Electrical & Computer Engineering, includes Sri Lanka-based researcher Chatura Seneviratne and Brookhaven National Laboratory researcher Soumyajit Mandal.
