Fast and curious, Virginia pioneers autonomous racing

The University is home to some of the best coaches in the world. Lars Tiffany, Andres Pedroso and Todd DeSorbo are regarded as arguably the very best in their respective sports. The Cavaliers also boast a handful of other prestigious coaches, though, and one of them teaches mechanical engineering in Olsson Hall. 

When one thinks of collegiate athletics, racing self-driving cars is certainly not the first event that comes to mind. Yet in the School of Engineering and Applied Science, Associate Professor Madhur Behl is rising as one of the University’s best coaches, leading a cohort of all-star mathletes. Currently composed of 21 undergraduates and a handful of graduate veterans, the Virginia autonomous car racing team is a pioneer on the frontier of artificial intelligence, with an autonomous racecar that has reached a top speed of nearly 150 miles per hour. 

Autonomous racing is certainly unique. It is a sport, art and science combined into one easily defined task — getting an autonomous AI car to cross the finish line first. Racing is ingrained in American media, demonstrated by famous quotes like, “If you ain’t first, you’re last.” Despite this common line from a Will Ferrell movie, sometimes progress is worthy of praise as well. Rising from a mere concept to near-champions, the Virginia autonomous racing team has become a global leader in just four years. 

After completing his doctorate from the University of Pennsylvania back in 2015, Behl wanted to research the possibility of commercial driverless cars. Behl’s accolades are extensive, as he is a respected leader in mechanical engineering. When he arrived on Grounds, he sought to experiment with autonomous vehicles — setting out to autotomize remote control miniature race cars. By using relatively inexpensive cars, Behl created a new discourse of accessibility in AI by using AI toy racecars as entertainment. Then, he upped the ante.

Working initially with graduate students, Behl formed a team of engineers and began working on a life-sized, fully autonomous AI racecar. While Behl himself is incredibly humble, it must be said that he is clearly one of the founding fathers of autonomous racing. The esteemed professor discussed the incredible story of his program with The Cavalier Daily.

The racing team first began officially competing in 2020, when the Indy Autonomous Challenge invited Behl’s team to a competition against other collegiate programs — marking the very first fully autonomous racing competition. The Indy Autonomous Challenge features two cars at once — an attacker and a defender. The competition begins with the defending car coasting at up to 80 miles per hour, and the attacker must pass them within two laps. If the attacker clears, then the roles switch. If both attackers succeed, then the speed is increased until one car cannot keep pace or commits a penalty. 

The AI can commit infractions such as accelerating past the coasting speed as the defender or operating in an unsafe manner with braking or steering. Again, the car is fully autonomous — the engineers cannot communicate with the car and are essentially high-profile witnesses for the race. 

At its inception, Virginia was an underdog program, having never finished higher than fifth place at any competition from 2021-2023. However, much like the classic comeback of other Cavalier athletic programs, the team began to rise exponentially. At the famous Las Vegas Motor Speedway Challenge in January, Virginia made history.

The Cavaliers entered the competition unseeded — meaning they had to earn a spot through a trial race. In that trial, they finished first and were awarded the top seed. Virginia kept advancing, eventually facing off against a team of students from Massachusetts Institute of Technology, Pittsburgh, Rochester Institute of Technology and University of Waterloo — dubbed MIT-PITT-RW — in the semifinal. The Cavaliers and MIT-PITT-RW fought in a tight battle, but Virginia ultimately triumphed due to a game-ending technical penalty on MIT-PITT-RW. The final then positioned the Cavaliers against the world-leading University of Munich team. 

“This is the race we have been preparing for,” Behl said. “We were happy to be fighting for the top position … this race was going to go the distance, no one was backing out.”

In a heated offensive round, the Virginia car was running out of time to pass Munich. With mere seconds remaining, the car moved onto the rigid bank side of the track and passed Munich at an incredible 120 miles per hour in the middle of a difficult turn. While Behl’s squad wanted to push even further, Virginia unfortunately fell victim to game-ending defensive penalties — a similar circumstance to how it advanced to the final but with the roles reversed. Despite the loss, Behl’s team proved they belong among the world’s best.

“I’m very happy the team was able to demonstrate their capability,” Behl said. 

The Munich team was 40 people strong and featured significantly more graduate-level experience than a Cavalier squad mostly composed of undergraduate students. However, despite a series of untimely penalties in the championship round, Virginia had reached its greatest peak yet.

“What they have achieved is nothing short of extraordinary,” Behl said.

The close loss aside, the Cavaliers have made their mark on the international stage. While Behl continues to work towards his goal of safe autonomous vehicles for commercial use, Virginia aims for a shot at glory on one of the most prestigious racing tracks in the world. In June, the Cavaliers will compete at the historic Monza Circuit in Italy — home of the Italian Grand Prix and several Formula 1 campaigns. Virginia is redefining the boundaries of sport, art and science as the team continues to soar. 

“I see faster cars, more cars, and a brighter future in auto racing … Other leagues are popping up. This is not a one-off,” Behl said.

Behl then posed questions of potential races against human-controlled remote racing cars or even actual NASCAR drivers. Future races could also take place outside of a standard oval track or even in harsh weather conditions, according to Behl. As autonomous racing grows, Behl noted that the team is currently accepting applications. 

What was intended to be a contained science experiment has morphed into an entirely new sport — and the future of AI. As the entire world continues to forge new horizons regarding AI, a crucial global leader in the field can be found here on Grounds — and he happens to be one of the greatest coaches at the University.