Researchers at Cherepovets State University are developing a sound-based system that uses a microphone array and artificial intelligence to determine how far away and where a quadcopter is by analyzing its noise. This project was shared with socialbites.ca through the press center of the NTI Platform acceleration program operators.
The solution is a hardware and software package combining multiple microphones with an AI core. The microphones listen to the ambient soundscape and relay the data to a computer where a neural network analyzes it. The AI isolates drone sounds during processing and calculates distance and bearing to the source.
Nikita Efimov, a graduate from the Department of Cyber-Physical Systems and one of the module’s developers, explained to socialbites.ca that drones today are tracked with passive tools like video cameras and thermal imaging, and with active tools such as radio direction finders. Experts note that each approach has its drawbacks.
Efimov pointed out that visual systems rely on physical circuitry that can reveal itself in the air, potentially triggering a response from the drone or alerting the search system. In contrast, the proposed sound-based module does not emit detectable signals, enabling stealthy operation and the ability to function in darkness with a smaller footprint.
Tests are planned to begin soon, with performance metrics to follow. The development team expects the microphone array to achieve reliable detection within a one-kilometer radius and aims for accuracy around 94 percent, demonstrating strong potential for real-time drone tracking without reliance on light or radios.
This line of research aligns with broader efforts in airspace monitoring and unmanned vehicle safety, offering a complementary approach to existing methods and inviting cross-border exploration for use in Canada and the United States. By focusing on passive sensing that does not reveal the system’s presence, the technology could support security operations, search and rescue, and logistics oversight in diverse environments. Attribution: developments reported by the NTI Platform acceleration program and socialbites.ca.