The Abu Dhabi Autonomous Racing League (A2RL) concluded its second season with a heart-stopping event at the Yas Marina Circuit on November 15, demonstrating how quickly extreme autonomous driving technology is evolving. The final, which saw six autonomous race cars compete on track for the first time in the world, crowned the TUM Team as champion, but it was the entire day that marked a turning point.
The Dynamic Impact: Human vs AI and Progress
One of the most anticipated events was the Human Vs AI demonstration, which saw former Formula 1 driver Daniil Kvyat challenge the car of the reigning champion team, TUM. The goal was to measure AI’s progress compared to 2024, when Kvyat had won by a convincing margin. This year, Kvyat gave the autonomous vehicle a . The outcome left everyone stunned: the AI driver crossed the finish line less than a second behind the human. The gap has drastically reduced, a technological leap that Kvyat himself described as “astounding”, adding: “Being on track with an AI driver is unlike anything else and it was fun to bring an exciting battle to the fans tonight.”
The Kvyat vs AI duel is not just an exhibition, but a clear metric of progress:
- Measuring the Technological Gap: in 2024, Kvyat’s margin of victory was “convincing”, suggesting a gap of tens of seconds. The reduction to less than a second (after giving a 10-second advantage over 10 laps) implies that, on a single lap, the AI is now extremely close to the physical limit of the car and human precision.
- Driving Analysis: the AI’s ability to match the times of an F1 driver like Kvyat on a technical circuit like Yas Marina indicates that the code is not just following the ideal line, but is optimally managing the grip limit and tire wear, crucial aspects that were the Achilles’ heel of early autonomous vehicles. As Kvyat noted, “Being on track with an AI driver is unlike anything else.”
Six-car Final with an Unexpected Ending
With a prize pool of $2.25 million at stake, six elite teams faced off in the 20-lap final race: TUM, Unimore, Kinetiz, PoliMOVE Autonomous Racing Team, TII Racing and Constructor University Racing Team. After the Silver Race, a requalification race won by , the final saw TUM start from pole position. The Italian team Unimore showcased an exceptional performance, executing a bold overtake on TUM and taking the lead.
The Italian team Unimore had built a significant lead, with victory seemingly assured, when drama unfolded on the track. The team (a lapped car) stopped on the track mid-corner; in an attempt to avoid it, Unimore’s car hit it, sustaining front-end damage and being forced to retire. After a , the race restarted. With their main rival out of the race, the TUM team maintained a commanding lead, crossing the finish line to win the A2RL 2025 championship.
- The Bold Overtake: Unimore’s overtake on TUM is not just a simple position change, but proof that the Italian AI had developed a more effective and aggressive planning and maneuver algorithm than the fast TUM (which started from pole). This demonstrates the maturity reached by autonomous racing strategies.
- The Human/AI Error of Do: the real challenge for AI on track is not the fast lap, but managing chaos and outliers (unexpected events) caused by other cars. The accident highlighted that current AI is very strong in predicting rational behaviors, but still vulnerable to failures or totally anomalous maneuvers by other competitors. TUM won not only for speed, but for algorithmic robustness in “Full Course Yellow” conditions and restart. The accident between Unimore and the lapped car (Constructor) was the critical point. Unimore’s AI, while attempting the evasive maneuver, failed to fully manage the unexpected situation of the lapped car stopped mid-corner.
A2rl: the Driverless Formula Car Championship
The A2RL (Abu Dhabi Autonomous Racing League) is the world’s most important autonomous racing league, established by ASPIRE (part of the Advanced Technology Research Council of the UAE government). Its distinctive feature is the use of real driverless formula cars. Its mission is clear: to accelerate the development of Artificial Intelligence and technologies for autonomous vehicles quickly and safely, using motorsport as the ultimate testing ground. The cars used in season 2, called , are based on the Japanese Super Formula chassis, equipped with a , capable of exceeding 290 km/h. Compared to the previous year, the cars have received numerous hardware and software updates, reflecting day and night progress in technology.
A2RL is not just motorsport; it’s a high-speed laboratory where university teams (Unimore, TUM, PoliMOVE) are rewriting autonomous driving protocols. The technology developed in A2RL will have direct implications not only for Level 5 road vehicles, but also for logistics and active safety, making A2RL one of the most important engineering competitions on the planet, as emphasized by ASPIRE’s CEO.
The Technical Platform: EAV-25, not just a Chassis
The heart of A2RL lies in the unified hardware platform: the EAV-25 cars are based on Super Formula (known for exceptional aerodynamics and extreme G-forces).
- Power and Aerodynamics: the use of the Super Formula chassis – among the fastest in the world, even with “only” 500 HP from the 2.0-liter turbo – means that teams are not just optimizing driving, but must manage sensitive aerodynamic load and top speeds exceeding 290 km/h.
- The Hardware/Software Challenge: the real leap is in the mentioned hardware and software updates. Teams don’t modify the car’s mechanics, but the perception systems (LiDAR, cameras, radar) and the brain (the on-board computing system) on which the AI runs. Their ability to improve performance so quickly suggests exponential progress in Machine Learning applied to dynamic vehicle control in extreme conditions.
Bertogna of Unimore Racing
Marco Bertogna, Head of Unimore Racing, expressed his frustration with the final result, while praising his team’s performance: “I was very, very happy with the performance we showed, the overtake we made was at a professional level. I’m incredibly happy with the technological result, less happy with the final outcome.” Stephane Timpano, CEO of ASPIRE, concluded that the event is not only incredible but is “the result of very hard work that has been done by all the teams” and emphasized how A2RL is the only environment in the world where you can witness such an extreme test and such rapid technological development.