Companies

I-SEE

A Ukrainian firm building passive, edge-AI computer-vision drone detection that automates net-launchers and is growing into a full anti-drone active-protection system.

I-SEE is a Ukrainian company building drone detection around computer vision rather than radio. Its software pairs optical cameras with edge AI that runs locally on standard GPU hardware, with no internet connectivity, turning an ordinary optical device into a smart sensor. Because it watches rather than emits, the system is described as optically passive and effectively invisible to the electronic-reconnaissance gear that hunts radar and radio sources. It can pick out targets as small as four pixels and, in daylight, reports detection ranges from 250 to 300 metres for FPV drones using a 2x camera, to about a kilometre with a 5x lens, and up to 2.5 kilometres against a Mavic 3T at maximum zoom.

The detection layer is built to feed a kill chain. Running at 30 to 60 frames per second and tracking as many as 100 targets at once depending on the hardware, the software pushes an alert with a photograph, coordinates and bearing to a unit’s existing Telegram or Discord channels within milliseconds — closing the gap between a drone appearing on camera and a human noticing it. By calculating a target’s speed and trajectory, it can also drive net-launching systems automatically, computing the lead and firing a net to neutralise an incoming FPV or “drop” drone without an operator in the loop. The company markets the same core for fixed sites such as warehouses and fortified positions, for mounting on pickups and armoured vehicles, and for integration into interceptor drones; it describes a fuller anti-drone active-protection system in development, a high-speed robotic turret pairing the AI tracker with net-launchers or small arms.

The work sits in the same kinetic counter-FPV space as net-turret builders such as Postup Solutions and shotgun-module maker VARTA , and reflects a broader Ukrainian push — coordinated in part through the Brave1 defence cluster — toward EW-resistant defences against fibre-optic and jamming-immune drones.

counter-fpv computer-vision edge-ai passive-detection net-launcher active-protection

Products

Software

  • I-SEE detection software

    A computer-vision, edge-AI tracker that turns optical devices into passive (non-emitting) drone sensors, processing video locally; reported detection of FPV drones at 250-300 m with 2x cameras, about 1 km with 5x, and up to 2.5 km for a Mavic 3T at maximum zoom. Calculates target speed and trajectory to automate net-launchers.

Hardware

  • Anti-drone active protection system

    An in-development high-speed robotic turret pairing the AI tracker with net-launchers or small arms to detect, aim and engage incoming FPV and "drop" drones automatically.

Sources