Ukrainian startup integrates AI into drones for autonomous final approach

Ukrainian defense firms are integrating artificial intelligence into first-person-view drones to automate the final stage of attacks. The system allows the drone to fly itself to a target during the last 500 meters, or 0.3 miles, of its flight path. This technology is being tested by dozens of Ukrainian military units as the war with Russia enters its fifth year.

Yaroslav Azhnyuk, a 37-year-old computer scientist, runs The Fourth Law and Odd Systems. He previously worked in Silicon Valley and founded Petcube, a company that sells AI-powered pet cameras. Azhnyuk pivoted to defense after Russia's full-scale invasion in 2022. He stated that the new systems are cameras that fling explosives at occupiers, replacing the treat-dispensing cameras of his past.

The Lupynis-10-TFL-1 drone combines a standard FPV frame with an autonomous terminal guidance unit. This unit contains a microcomputer running a deep-learning neural network. The software is trained on hours of combat footage to recognize specific targets, including tanks, light-armored vehicles, and flying objects. During a test in a field outside Kyiv, a pilot named Professor selected a dilapidated car as a target. After the drone locked on, indicated by a red flash on the monitor, the pilot released the control stick. The drone completed the flight and approached the car without further input.

Stab, the head of the test and integration team at The Fourth Law, noted that the drone performs all actions itself once the target is selected. He explained that the AI makes these systems difficult to jam compared to traditional radio-controlled drones. The company names its weapons after Ukrainian intellectuals persecuted by Russia. The Lupynis drone is named for Anatoliy Lupynis, a poet who suffered decades of Soviet imprisonment.

President Volodymyr Zelenskyy has warned that AI could lead battlefield decisions as early as 2027. Azhnyuk acknowledges this development but compares the technology to a calculator rather than a sentient robot. He argued that if adversaries use AI for offensive decision-making, Ukraine must use it defensively to keep pace. He noted that AI systems can make decisions in microseconds, while human control often takes minutes or hours.

Azhnyuk emphasized that humans still select the initial target, serving as a safeguard against rogue behavior. He stated that if Ukraine loses the war, there will be no time to debate the ethics of AI. The immediate priority, he said, is survival.