Young engineers at RTU MIREA are developing an acoustic drone detection system. It allows tracking drones that remain invisible to classic radars due to their small size, low altitude, or the use of fiber optics instead of a radio channel, the university told "Perviy Tekhnicheskiy".
The acoustic system uses microphones distributed across the protected area, which continuously listen to the airspace. Special software extracts characteristic sounds of drones from the noise and compares them with a database of acoustic signatures. Upon detection, the system uses triangulation to determine the coordinates, speed, and direction of flight, issuing an alarm signal.
We are not trying to compete with radars — we are closing their blind spots. If a drone is flying low, if it is small, if it does not emit a radio signal — the radar will not see it. But it still makes noise. Propellers, engine, body — each device has its own acoustic signature. We are teaching the machine to hear these sounds, distinguish a drone from the wind or a passing bird, and accurately determine where it is flying from. It's like putting an invisible hearing network around an object.
The advantage of the approach is complete passivity: the system emits nothing, it cannot be detected, it is not afraid of interference. Analysis of the sound spectrum allows not only to detect an object, but also to identify its type, and sometimes even the specific model — each drone has its own "voice".
Currently, the team is developing detection and identification algorithms, collecting a database of acoustic signatures of various UAV models, and preparing to create the first prototype of a distributed sensor network. The project received a grant from the RTU MIREA acceleration program, which will allow completing the development of a minimum viable product and applying for a patent.