In late 2024 a Ukrainian start-up began flying drones that, once given a target area, talked to each other. They decided who struck first. They adapted when one of them failed.
Swarmer, founded in May of 2023, fielded autonomous swarms in active combat in late 2024 — drones that, once given a target area, communicate among themselves to decide which strikes first and which adapts if one fails. By September of 2025 the company's software had been used in over eighty-two thousand combat missions. In the same month it raised a fifteen-million-dollar Series A — backed by Eric Schmidt's D3 Ventures, Broadband Capital, and Network VC — the largest publicly disclosed Ukrainian defence-tech round of the war. A three-person crew can manage twenty-five drones; the platform is scaling toward a hundred.
Russia's Shahed and Geran are also acquiring autonomy. Ukrainian investigators have found NVIDIA Jetson modules and thermal-vision cameras in 2025 wreckage, allowing some onboard target recognition and terminal guidance at night. Russia is now equipping Shaheds with cluster munitions and mines.
Former Ukrainian commander-in-chief Valery Zaluzhnyi has warned that whichever side first cracks reliable autonomous swarming — he suggests 2027 as a horizon — will gain a transformative battlefield advantage. The race is on. The current state of the art is impressive but partial; the headlines outrun the systems. A drone that picks its own target is not the same thing as a drone that picks its own target reliably under jamming, weather, and counter-fire.
This sheet is the only one in the atlas drawn with a warm-coloured border. It is the only sheet whose future tense outweighs its past.