Identifying drone threats in noisy soundscapes.
We develop microphone housings for your specific threats and soundscape. When the threat or environment changes, we change the housing.
Why listen
Radar finds drones well, but it transmits, and a transmitter can be located and targeted. Russia used anti-radiation missiles against Ukrainian air-defence radars.
Source: RUSI, November 2022
Why listen
A microphone emits nothing, so there is nothing to locate. It is also cheap. Ukraine's Sky Fortress network reportedly runs about 14,000 acoustic sensors at $400 to $1,000 each.
Source: United24 Media
Real event · December 2018 · Gatwick, UK
In December 2018, drone sightings closed Gatwick Airport for about 33 hours. About 1,000 flights and 140,000 passengers were affected. No one was charged.
Sources: BBC, 2019; The Guardian, 2020
The hard part is noise
Every place a drone could strike sounds different, and each faces different drones. Consider an airport, a power station and a substation.
The hard part is noise
An airport is loud: jet engines, service trucks, generators and traffic on the access road. Any drone is a threat there, even a small quadcopter. Microphones along the perimeter face outward, toward where a drone would come from, so most of the airport's own noise arrives from behind them.
The hard part is noise
A power station roars and hums: turbines, generators, transformers and falling water in the cooling towers. The drones sent against power stations are larger fixed-wing attack drones that fly hundreds of kilometres to their target.
Source: Euronews, 2026
The hard part is noise
A substation hums at 120 Hz and its harmonics, from the vibration of its transformer cores. The threat there is a drone that can carry a length of wire into the equipment, as one did in Pennsylvania in 2020.
Sources: Siemens Simcenter; DroneDJ, 2021, reporting the FBI, DHS and NCTC bulletin
The hard part is noise
Picking a drone out of this is hard. A small drone is quiet next to a jet or a transformer. Noise comes from every side and changes through the day. A passing car's pitch drops as it goes by, and echoes off concrete and paved ground bring noise in from directions where there is no source.
What we do
We build a 3D acoustic model of the target including its ground, buildings and surfaces, and the traffic and machinery that emit noise. Then we simulate attacks from different directions to optimize a microphone housing for your specific target against its specific threats.
What we do
A housing's shape decides which directions and which pitches the microphone hears best, so we shape each one for its mission. The housings can be any shape we can 3D print. A housing for a substation will differ greatly from one for an airport.
Reported · 2025 to 2026 · Ukraine
Threats change. Russia's Shahed-type drones used a piston engine, and Ukraine's acoustic sensors were tuned to its sound. Russia now also flies jet-powered versions. In June 2026, Ukraine's commander-in-chief said Russia plans to make half of its strike drones jet-powered. A jet sounds different from a piston engine.
Sources: Meduza, 2026; Euronews, 2026; IEEE Spectrum, 2026
What we do
When a threat changes, we change the housing. We add the new drone's sound to the model, shape a new housing and print it. The microphone and electronics stay, only the housing, which is cheap and quick to print, is swapped. In Ukraine, drone designs now change in weeks, so a sensor has to be able to change with them.
Source: Hudson Institute, 2026. The last sentence is our assessment.
How we design one
Each candidate housing is tested in our acoustic model of its place before anything is printed. The model shows what the microphone inside would hear from each direction: the drone, and everything else.
How we design one
Each candidate shape is scored on how well it separates the drone from the rest of the soundscape. We change the shape, test it again and keep the best.
How we design one
The chosen shape is 3D printed and measured from every direction in the real world. The measurements go back into the models, so the next simulation starts closer to reality.
Where we are
We are at the prototype stage working on housings to detect drones at greater distances than current passive systems in noisy places.
Targets
Current targets: small quadcopters at airports and substations. Please reach out if you would like more information.