Today's Edition
MAYS IBRAHIM (Abu Dhabi)
As the Abu Dhabi Autonomous Racing League (A2RL) returns for its second season this Saturday, teams from across the globe are putting the final touches on their AI-powered race cars, each aiming to shave fractions of a second off lap times and inch closer to human-level performance.
The competition, held at Yas Marina Circuit, continues to push the boundaries of autonomous racing, testing not only speed but strategic decision-making under high-pressure conditions.
For the UAE-based TII Racing team participating in the Grand Final, perception has been the key upgrade this season.
Giovanni Pau, the team principal and Technical Director at the Technology Innovation Institute’s Autonomous Robotics Research Centre, explained that their focus has been on “making the car see as best as we could”, enabling it to detect other cars ahead or behind and accurately estimate their speed.
“Basically, what you do is watch how the gap evolves over time and estimate whether it’s closing or widening,” Pau told Aletihad.
He stressed how this improved situational awareness allows the car to time overtakes and respond to unpredictable obstacles.
While human drivers rely on instinct and feel, AI can process thousands of decisions per second, making real-time adjustments based on sensor data from GPS, LiDAR, radar and cameras, Pau added.
Although the AI has not yet matched a professional driver’s ability to strategise dynamically, he believes the technology offers a significant advantage in analysing data and making split-second calculations.
Representing the US in the Silver Race, Code19 Racing team has concentrated on pure speed. Lawrence Walter, co-founder and CEO, said their team has focused on tyre management, braking algorithms, and precise track localisation to optimise lap times in the silver category’s time trials.
“We’ve built an AI interface that lets us communicate with the car during testing through a natural language interface,” Walter told Aletihad. “This allows us to quickly analyse data and fine-tune performance.” Walter stressed that AI is still learning the subtle nuances of “driver feel” that professional pilots develop over years, but it compensates with sheer computational speed.
“Human and AI performance are very close right now,” he said.
Walter believes AI may soon outpace humans in racing, but rather than replacing drivers, it will enhance their performance and make the sport more thrilling for fans.
The competition will be split into two categories: a Silver Race for emerging teams, and a Grand Finale for the top performers, competing for a $2.2 million prize pool.
Following a rigorous October qualification, TUM, Unimore, Kinetiz, TII Racing, PoliMOVE, and Constructor will battle in a 20-lap race for the Season 2 championship.
The remaining five teams – RAPSON, Code 19, Fly Eagle, FR4IAV, and TGM Grand Prix – will compete in the Silver Race.
Last season’s reigning champion TUM is set to race against former F1 driver Daniil Kvyat in A2RL’s second Human vs AI showcase.
