AI-Powered Self-Driving Cars Race in Abu Dhabi's Autonomous Racing League

Using AI, the race cars learned how to maneuver around the track

Graham Hope

April 30, 2024

3 Min Read
Racing cars of varying colors on a race track

Motorsport embraced a new era over the weekend as the new Abu Dhabi Autonomous Racing League (A2RL) roared into action.

A team from Germany’s Technical University of Munich (TUM) claimed victory at the Yas Marina circuit in an event that was described as “bringing a science experiment to the racetrack” by pushing the boundaries of autonomous technology.

And while the racing clearly has some way to go before it can compete with the spectacle offered by Formula One or NASCAR, organizers had no hesitation in hailing A2RL as a success.

View post on X

Eight teams representing groups from countries as diverse as the United States, Germany, Switzerland, China, the United Arab Emirates, Hungary, Singapore and Italy took part vying for a $2.25 million prize fund.

All had access to the same car, the 2023 Dallara Super Formula, which in its standard form is recognized as the fastest open-wheel race car in the world behind an F1 machine, delivering in the region of 550 bhp and capable of speeds of up to 186 mph.


In A2RL, it is equipped with seven Sony cameras, four ZF radar sensors and three Seyond lidar units as part of an autonomous driving stack that also features a Nvidia GPU. The variable between the teams lies in how they use coding skills, AI algorithms and machine learning to teach the cars how to drive.

Related:Self-Driving AI Traffic Tech Secures Florida Approval, Earns U.S. Patent

Among the elements the AI has to master are understanding grip levels, managing tire temperatures, predicting the movements of rival cars and overtaking.

While we have seen autonomous racing elsewhere previously, A2RL was the first time that four AI-driven race cars have taken to the track simultaneously, having qualified for an eight-lap shootout via a time trial, which was punctuated by some erratic behavior from the entrants that included spins and random stops.

The race itself was also something of an odd spectacle, taking nearly an hour to complete after one car – from Italy’s Polimove – spun out, prompting ‘yellow flag’ conditions that saw speeds curtailed to 25 mph, plus a host of technical issues.

The TUM racer ultimately claimed victory after it passed the entry from Italy’s Unimore, which slowed down then came to a halt while in the lead.

The event was organized by ASPIRE, a subsidiary of Abu Dhabi’s Advanced Technology Research Council (ATRC), and despite the obvious teething problems, CEO Stephane Timpano was pleased with how it turned out.

“The launch of A2RL has reshaped the landscape of sports and technology, leaving a lasting impact for years to come,” Timpano said. “Moving forward, we’re broadening our focus to include different vehicle types, while actively attracting top talent worldwide to showcase their mettle.”

Related:Developer of AI Tech for Self-Driving Vehicles Valued at $6B

Around 10,000 spectators were in attendance at Yas Marina. The main supporting race, meanwhile, saw former Formula One racer, Russia’s Daniil Kvyat, perform a demo with an autonomous car in an “AI vs Human” challenge. Although Kyvat was clearly comfortably faster, he praised the tech afterward.

This article first appeared in AI Business' sister publication IoT World Today.

About the Author(s)

Graham Hope

Graham Hope has worked in automotive journalism in the U.K. for 26 years, including spells as editor of leading consumer news website and weekly Auto Express and respected buying guide CarBuyer.

Keep up with the ever-evolving AI landscape
Unlock exclusive AI content by subscribing to our newsletter!!

You May Also Like