A Lithuanian startup launched an open-source drone detection system on June 20, 2026, that works by listening to drone sounds. The system aims to recruit 10,000 people across the Baltic region to help spot Shahed-type drones — unmanned aircraft used in attacks.

This is the largest citizen-powered listening network of its kind in the area. The timing matters. Lithuanian officials had conducted practice drills in May to test the detection system, and earlier, real drone incidents had made the public aware of the problem. In March 2026, a suspected drone crashed into a lake near the Belarus border. Months before that, in July 2025, Lithuanian leaders were evacuated to a shelter after an alert about a combat drone crossing from Belarus — though it turned out to be a false alarm. Those real events showed people why detection matters more than any government announcement could.

How It Works

The system listens for the distinctive sound of drone rotors and motors. Shahed drones, for example, have a characteristic pusher-propeller sound that acoustic sensors can pick up. Since the sensors are passive — they only listen, they do not broadcast anything — they cannot give away their location to adversaries.

The main challenge is sorting out real drone sounds from all the other noise in cities and towns. With 10,000 separate listening devices across the region, there will be a lot of false alarms to filter through. The startup has not publicly explained how it will handle this problem — specifically, how much processing happens at each sensor versus how much happens at a central location, or what noise level has to be met before an alert gets sent out.

An open-source design means that ordinary people can install a sensor at home, on their building, or in their neighborhood. This approach is much cheaper than hiring contractors to install a grid of professional equipment everywhere. The tradeoff is that the sensors will vary in quality, they could be damaged or interfered with, and someone who understands how the detection works could design a false signal to confuse it. The startup has not publicly detailed how it plans to defend against these risks.

Money and a Broader Shift

The drone-listening network launch comes as money is flowing into Lithuanian defence technology more broadly. A Lithuanian defence engineering company called PDKINEMATICS closed a €2 million funding round in mid-June 2026. A Lithuanian drone startup also secured €2 million to expand across NATO. It is unclear whether these are the same company seen from different angles or two separate funding deals, but the pattern is clear: Baltic defence technology is attracting serious investment money, not just government grants.

For comparison, a commercial option exists. Fortem Technologies makes a system called DroneHunter F700 that shoots down drones like the Shahed — more expensive and complicated than a listening network, but part of the layered approach NATO countries are building: detect, track, classify, shoot down. The Lithuanian citizen network handles the first two steps at a cost that lets it cover a lot of ground.

There is a practical question worth asking before this network grows larger. Continuous audio monitoring across 10,000 locations will pick up a lot of environmental sound — people talking, traffic, everything. Open-source code means anyone can read it and audit what the system records, keeps, and sends to a central server. But open-source code alone does not guarantee privacy. The startup should be clear about what it deletes immediately and what it stores. That is something the public and regulators should ask about now, before the system goes to scale.

The broader direction is clear. The Baltic states have moved from treating drone breaches as rare surprises to treating detection systems as essential infrastructure that ordinary citizens help maintain. This follows a pattern we have seen before — volunteer weather stations, crowd-sourced earthquake reporting, distributed computing projects — where many people contribute something small to solve a problem no single organization can afford to solve alone. Whether the startup's technology can turn 10,000 sensors into reliable alerts is a question that real operational results will answer, not funding rounds or press releases.