Minchan Kim beats AI in Abu Dhabi drone racing final

Minchan Kim beat TII Racing’s AI drone in the Human vs AI finale at the A2RL Drone Championship in Abu Dhabi after the best-of-nine showdown reached 4-4 and came down to the last run. The margin was razor-thin, but the outcome showed the current ceiling of autonomous drone racing: the machines are already fast enough to pressure a world FPV champion, yet Kim still won on line discipline and recovery when the AI struck a gate and could not continue cleanly.

The championship took place on 21-22 January 2026 during UMEX in Abu Dhabi and carried a total prize pool of USD 600,000. A2RL said the event brought together 10-plus global teams, six elite human FPV pilots and a 14-gate course built for AI versus human battles and multi-drone racing. The format moved through testing, qualification, semi-finals, a human-vs-AI benchmark and finals, turning Kim’s win into one part of a broader stress test for autonomy under race pressure.

TII Racing still set the pace in the AI Speed Challenge with a fastest lap of 12.032 seconds, a time that underlined how far the vision-led stacks have come. The autonomous drones in Season 2 used onboard AI-driven vision and real-time computation, with trajectories calculated at more than 130 km/h. Giovanni Pau, TII’s technical director, tied that pace to disciplined software development, while ASPIRE chief executive Stephane Timpano said the teams were reaching higher speeds with better stability and consistency than in Season 1.

That progress has been visible since the first A2RL x DCL Autonomous Drone Championship in April 2025 at ADNEC Marina Hall, where 14 international teams raced on a 170-meter course with 22 gates. The standardized drones in that debut event used a forward-facing camera, an inertial measurement unit and NVIDIA Jetson Orin NX hardware, with MavLab from TU Delft winning the AI vs Human Challenge and TII Racing taking the multi-autonomous drone race. The January 2026 championship pushed the same idea further: fewer sensors, more speed and a tighter contest between machine precision and human racecraft.

A2RL and the Technology Innovation Institute have framed the series as a real-world testbed for high-performance autonomy, and Faisal Al Bannai has said it is meant to prove innovation in practice, not just in theory. The league has also tried to widen the pipeline around the sport, with its Drone STEM Program, run with UNICEF, training more than 100 Emirati students across four locations. In Abu Dhabi, the headline was Kim’s narrow win, but the deeper story was that autonomy is now close enough to force a champion to earn every inch of the track.