DroneForge AI Makes FPV Drone Operators Obsolete
A new generation of autonomous FPV drones may soon eliminate the need for human pilots altogether. The U.S. startup DroneForge is developing an AI-driven system that turns any consumer drone into a fully autonomous machine capable of making its own tactical decisions.
AI that sees without GPS
Unlike traditional drones that rely on GPS and complex sensor arrays, DroneForge’s platform operates entirely on monocular vision. Its computer-vision pipeline reacts in under 50 milliseconds, allowing precise maneuvering even in GPS-denied or jammed environments.
Recently, the company integrated EdgeTAM — an optimized version of Meta’s SAM 2 model — which enables drones not only to detect but to continuously track moving targets. The AI uses frame-memory mechanisms to predict object trajectories even when they temporarily disappear from view.
Autonomy as a new standard
According to DroneForge, their SDK already supports prompt-based mission planning and integration with advanced AI models for object recognition and dynamic flight paths. The system can upgrade drones costing as little as $100, making autonomous flight accessible to developers, researchers, and defense contractors alike.
In the long term, such drones could handle complex operations — from search-and-rescue to military missions — without continuous human supervision. The company’s $2.5 million seed funding underscores investor confidence in a future where machines no longer need pilots to dominate the skies.
Editorial Team — CoinBotLab