The Abu Dhabi Autonomous Racing League (A2RL) showcased just how far the championship has progressed with its 2025 Abu Dhabi event, where teams were competing for a $2.25million prize pool.
A2RL held its inaugural event at the Yas Marina Circuit in April last year, having adapted a Dallara Super Formula SF-23 chassis into a fully autonomous machine in just seven months.
Now with over 18 intense months of further development, A2RL's EAV-25 - an upgraded version of the predecessor used last year in Abu Dhabi - compared well with ex-Formula 1 driver Daniil Kvyat in the 'human vs AI' showcase and allowed for a six-car autonomous race that produced a pass for the lead plus an incident that then knocked the new race leader out of contention.
Here's everything you need to know about A2RL's Abu Dhabi return...
In pure performance terms, the difference between season one and two has been astronomical.
The technological advancement for season one was already impressive when you consider the cars went from lapping three and a half minutes slower than the human reference in testing to within 10 seconds of Kvyat at the 2024 event.
But season two has brought performance gains that make the AI-driven machines much harder to distinguish from their human-driven counterparts.
Kvyat called it a "day and night difference" versus last year, fresh from chasing the AI down in a special showcase event on Saturday in Abu Dhabi and finding it only around a second slower than him.
"Now I can push properly, I can chase [and it will lap] within a second and it's a great achievement and that's impressive. I congratulated everyone at A2RL, that's great progress," Kvyat told The Race.
"Who knows where we'll be next year. I'm super excited about this technology, it's a fantastic platform for the people to develop this sort of technology right here.
"It's braking very late, quite spot on. Some areas like tyre sliding, that's where maybe the AI is lacking, but they're starting to go more into quantum mechanics.
"Every temperature degree starts to change the behaviour of the car, aerodynamics and so on.
"It's very impressive we're already talking about this now, what was once thought of as science-fiction has now become reality."
Before the six-car autonomous showdown, season one champion TUM faced off against Kvyat in the human vs AI contest.
The TUM car was given a 10-second headstart with Kvyat's goal to chase it down within 10 laps.
It was more of a showcase of the advancement in technology and the improved speed of the AI vs the human driver, so the goal wasn't for Kvyat to overtake and 'win' the race.
The AI therefore took the chequered flag first with Kvyat following behind, but the AI had lapped impressively with a fastest laptime of 59.154s, only 1.585s than Kvyat’s best - a huge improvement on 2024's 10s deficit.
Plus, TUM would later complete a 58.183s during the fully autonomous race, only half a second slower than Kvyat's fastest lap from earlier that day.
"It's a much smaller [gap] than last year," Kvyat told The Race. "That's the most important thing to see the progress A2RL has made in these past couple of years.
"I've always been a big fan of this kind of tech, it's where the future is. The more you go into detail, the harder it gets, and to close these last few tenths of a second is always harder.
"It requires a lot more technology work for the engineers to discover even more details."
Eleven teams from 10 different countries were competing for their share of the $2.25million prize pool in A2RL this year.
The series provides the EAV-25 and the autonomous stack, which quite literally occupies the cockpit and is essentially the brain of the car, but it's down to the teams to code their AI machines to be the fastest.
One of the biggest areas to improve from season one was how the machines coped in a racing environment, with yellow flags and safety cars scenarios proving particularly tricky last year.
This was a key area of focus for the championship and its teams in between seasons, and it all built up to a six-car, 20-lap race on Saturday evening.
After a rolling start the polesitting defending champion TUM, a German-based team that's led by Dr Ing Markus Lienkamp with 14 PhD researchers, led the Italian team Unimore as the two clear benchmark teams.
Unimore was able to dart down the inside of TUM and capture the lead on the second lap, a daring overtake that's arguably the best wheel-to-wheel moment of A2RL's existence so far.
But TUM was able to follow Unimore closely for the following 10 laps, with both cars setting an impressive pace, not too dissimilar to Kvyat’s pace earlier that day - until racing disaster struck for Unimore.
Unimore found a lapped Team Constructor car in its path at Turn 1, and just as the Constructor car went to the outside of the corner, so too did the Unimore entry, wrecking the front right of its car and spearing into the barriers.
That brought out the red flags, which halted the race on lap 12 of 20.
The race restarted with TUM now in a commanding lead that it was able to convert into victory in the main event of A2RL 2025 - repeating its 2024 success and even setting its best laptime on the very final tour.
The team behind TUM remained calm when Unimore overtook it on lap two, with Dr Lienkamp telling The Race: "I expected they would overtake us because they are very fast at the beginning.
"We were calm because we knew our car took five or six laps to really pick up speed."
As for Unimore, it was, of course, disappointed to lose the victory, having also lost out to TUM last year, but it did set the fastest laptime of the whole 2025 event.
"The performance we showed, we are amazingly stable and reaching the professional level. Kvyat did a high 57-second [laptime], we did a 58s laptime, we are there, at a professional level. It's amazing at this point," Unimore's team principal Marko Bertonga told The Race.
"The overtake we showed and set up in place was amazing.
"The accident, I guess, even a professional driver couldn’t have avoided a very slow car crossing into a new direction while you’re being overtaken by someone else.
"I'm incredibly happy about the technology result, less about the final result.
"But I think what we showed is a pretty good success."
You can watch all of the action from the 2025 A2RL event on The Race's YouTube channel.
Four of the six cars were classified, with TUM leading TII Racing home by 37s. Polimove completed the podium.
Kinetiz ended up fourth, recovering from a spin on cold tyres following the red flag restart.
Unimore and Constructor retired with damage from their lap-12 incident.
The improvements have been across the board, but there are some areas where it's been particularly stark.
"For many teams, the biggest [gain] - and it may be shocking to some - was the ability of going full throttle on the straightline," A2RL technical director Nicola Palarchi said when asked by The Race where the biggest chunks of laptime have been gained.
"It's something that looks very simple, but the control you need to apply to the steering actuator at speeds above 200km/h is incredible.
"So many teams last year were struggling to go full throttle and use the full power of the engine.
"So a big [gain] for let's say the teams in the midfield and at the back, was being able to refine their control and go fast.
"For the others was the optimisation of braking and cornering. Acceleration is still tricky; it's still the part where maybe the human has the edge.
"The first year for the vast majority of the team was the learning process.
"[This year] they stepped up their understanding of vehicle dynamics and how to use the car."
There have been over 3000 laps of testing conducted at the Yas Marina Circuit since the start of October.
But that doesn't mean the machines couldn't race elsewhere - the cars are not driving to a set GPS route, but are reacting to the environment around them.
So with the right testing, they could soon race on circuits other than Abu Dhabi and Suzuka at some stage in the future of A2RL.
It could be the gateway to a multi-round, fully autonomous racing series around the world.
And while watching the action in Abu Dhabi, The Race couldn't help but wonder whether AI and humans could even be working in the same race teams in the future as team-mates.
"I was thinking about it, and yes as a lap driving itself," Kvyat replied when The Race put it that idea to him.
"But there are so many procedures, yellow flags, wheel-to-wheel, it might become a bit messy at this point, but maybe in a couple of years' time.
"I think sometimes people misinterpret it. I saw before [in reaction to] Human vs AI as the start of The Terminator days, it's not like that.
"It's more like this is a fantastic platform, I'm the benchmark, the reference, engineers take my data, try to come as close as possible to it.
"So in the end that’s the goal and to bring it to current technology we have for space, for automotive and for transportation, that’s how I see this technology."
In terms of raising the bar for A2RL season three at the end of next year - Kvyat's reference time this year is the target champion TUM are chasing.
"Next year the bar I would say was set by Kvyat, he showed us today that a 57s [laptime] was possible. So I expect more teams to push for 57s laptime next year," Dr Lienkamp added.
"The biggest improvement next year will hopefully be that the interaction of the car next year is safe."
Given the impressive trajectory A2RL's technology is on, it's intriguing to wonder what kind of leap it will be able to take ahead of its third event in Abu Dhabi next year.
Images from A2RL/Spacesuit Media