Ukraine operates a nationwide acoustic detection network that identifies incoming drones and missiles by sound. Distributed microphones mounted on rooftops, mobile-phone towers and in open areas continuously capture ambient audio.  Artificial intelligence is running on each of these devices, filters out background noise such as traffic and weather, and calculating the direction, speed and position of airborne threats. This information is transmitted is relayed back in real time to command centres and mobile air-defence units.

The system runs continuously and functions as a low-cost, passive layer of air defence, much cheaper than radar.  Its latest component, AReS V5, extends this capability by generating real-time flight-path projections, allowing defenders to track and engage targets with greater accuracy

Why sound matters

Since Russia’s full-scale invasion in 2022, Ukraine has faced drones and missiles flying very low to avoid radar, combined with radar jamming, it makes it very difficult for Ukraine to track these drones.

But everything that flies makes noise.

  • Shahed drones sound like distant lawnmowers
  • FPV drones sound like small electric motors
  • Cruise missiles also produce sound

Sound travels at about 340 metres per second. When a drone passes several microphones, the sound reaches each one at slightly different times. The system uses these small differences to calculate a very precise direction and uses AI to determine what is producing that sound.

Early systems were simple but proved effective; and have now been developed into two main networks called Sky Fortress and Zvook.

Always on listening network

This Sky Fortress has more than 14,000 sensors around Ukraine, with plans to add 15,000 more. Zvook adds hundreds of extra sensors, often placed on mobile-phone towers. Each unit is small, costs only a few hundred dollars, and runs on solar or mains power.

These sensors listen and communicate back to base.  The AI has learnt what sounds are threat and which are just everyday background noise.

When a threat is confirmed, the system:

  • Calculates direction, height and speed
  • Shares the data with national air defence
  • Alerts nearby teams using tablet maps

This gives defenders a few minutes to act. This system works in darkness, fog, rain and city noise and cant be jammed or easily detected.  In one attack, 80 out of 84 drones were shot down using acoustic data alone.

AReS V5

This is a more advanced version that provides early warning with precision, over large areas.

It has multiple arrangement of microphones (called an array) as to detect the direction the sounds being heard from the sky. A technique called beamforming lets it focus on different directions, like pointing a “listening beam”.

Operators see a webcam view with a sound-based overlay showing where the objects are that it can hear in the sky. Even when a drone is hidden by clouds or darkness, its path is still tracked very clearly.

Performance includes:

  • FPV drones detected at 200–300 metres
  • Shahed-type drones detected up to 5 km
  • Proven tracking of Shahed-136 at 3.7 km and cruise missiles at 900 km/h

AReS V5 can track multiple targets at once and send coordinates directly to weapons or ground teams. A smaller version is being developed for unmanned ground vehicles to protect supply routes.

Layered defence

Ukraine places these sensors across cities, infrastructure and transport routes.  Together with radar and cameras, they form a wide layered defence system, that provides early warning over a very large area of the country.

In the Chernihiv region, acoustic alerts helped teams destroy 87 drones in one night. In Kyiv, AReS systems have tracked missile paths in real time, even fibre-optic drones too.

The Future

Ukraine is expanding the network with support from the Brave1 programme. AReS is moving into full production, and the United States and NATO are studying it as a low-cost way to counter similar drones.  Smaller versions of this technology will also be used to help protect convoys, infrastructure, and robotic patrol systems.

Sources:

Leave a Reply