Ukrainian Firm Launches Offline AI Drone Detection System for Frontline Use

A Ukrainian software company has introduced I-SEE, an AI-powered drone detection system designed specifically for frontline military operations. The software runs entirely offline on standard consumer hardware equipped with GPUs, addressing the critical gap in detecting fiber-optic FPV drones that evade traditional radio frequency sensors. By utilizing computer vision, I-SEE identifies targets as small as four pixels at distances up to 2.5 kilometers, depending on camera optics. It processes video feeds at 30 to 60 frames per second, tracking up to 100 simultaneous targets without requiring internet connectivity, thus ensuring resilience in electronically contested environments. Upon detection, the system instantly alerts military units via Telegram or Discord, providing photographs, coordinates, and trajectory data to facilitate rapid counter-measures. This innovation aims to reduce reaction times from seconds to milliseconds, significantly enhancing soldier survival rates against mass drone attacks. The solution integrates with existing camera infrastructure via standard protocols, offering a cost-effective alternative to complex radar systems. Field tests indicate varying detection ranges based on zoom levels, with thermal cameras achieving maximum range. This development highlights the increasing role of accessible AI technology in modern asymmetric warfare.