A German-based team won the race at the Indianapolis Motor Speedway on Saturday – not by out-accelerating their opponents or driving them off the road, but rather by having better-performing code for their autonomous vehicle.
The Technical University of Munich team won the Indy Autonomous Competition, out of nine teams from a total of 21 universities throughout the world.
Each team raced their cars on the Brickyard track one after another on Saturday afternoon, and the winner was determined by clocking in at the fastest speed in two full-speed laps.
It is unclear how fast the Technical University of Munich’s car was driving, but in timed trials it was clocking in at about 129.2 kilometers per hour, or 80 miles per hour.
Following their victory on Saturday, the Technical University of Munich team was seen kissing the famed Yard of Bricks as they held up their giant check for $1 million.
The team from the Technical University of Munich won $1 million in the Indy Autonomous Competition on Saturday
The team designed a blue autonomous vehicle using a Dallara IL-15 car frame
They were seen kissing the famed Yard of Bricks following their victory
The challenge was designed to prove ‘autonomous technology can work at extreme conditions,’ said Paul Mitchell, CEO of the Energy Systems Network, which co-sponsored the event.
As part of the competition, each team received a Dallara IL-15 car, which looks like a small Formula One car, as well as sensors, cameras, a GPS and radars.
They then had to program about 40,000 lines of code to kickstart the engine and power a computer wedged into a bucket where the driver normally sits, AFP reports.
The entire process took about two years.
According to the TUM team’s page, the team focused on ‘dynamic path planning and control of the vehicle at the handling limits.’
‘In addition, we are experienced in tuning the algorithms to perform better at high velocities and accelerations,’ it wrote, noting: ‘We developed a global optimal path planners vehicle dynamics simulation that can be used for autonomous driving.’
Each team had to program up to 40,000 lines of code to kickstart the engine and power a computer wedged into a bucket where the driver normally sits
They were equipped with sensors, a GPS, cameras and radars to use however they saw fit
A group of students are seen fixing the code for their car on the side of the track
The entire process took two years to develop the cars
The Dallara IL-15 car frame looks like a small Formula One car
The MIT-PITT-RW team, comprising students from the Massachusetts Institute of Tecnology, University of Pittsburgh, Rochester Institute of Technology and University of Waterloo, was the only team without any adult supervision.
They had only received their Dallara car six weeks prior to Saturday’s race, according to the AFP, and engineering student Nayana Suvarna, 22, was elected to be the team’s manager – despite not having a drivers’ license.
‘I don’t know anything about car racing,’ she said, ‘but I’m becoming a fan.’
Their car hit a whopping 81mph in testing, and the car belonging to the PoliMOVE team, a partnership between the University of Alabama and Politecnico in Milan, drove past the pits at around 155mph on Thursday.
But it skidded at the next turn spinning in a full circle, before coming to a stop on the inside lawn.
‘It was a miracle we didn’t crash,’ said Sergio Matteo Savaresi, a professor at Politecnico. ‘We actually reached the very limit of the car.’
He added: ‘A professional driver at that speed with tires like these would have done exactly the same.’
TUM’s winning car was seen crossing the finish line at the race
Spot the Dog, by Boston Dynamics, waved the starting flag
Nine teams from a total of 21 universities throughout the world competed on Saturday
But speed is not the goal of the competition, according to AFP.
Instead, it is meant to get more people to trust autonomous cars, as a Morning Consult survey in September found that 47 percent of Americans considered autonomous vehicles less safe than those driven by humans.
‘If people get used to seeing cars like these going 300 kilometers per hour [about 184mph] and they don’t crash,’ Savaresi said, they may eventually think the cars are safe ‘at 50 kilometers per hour.’
The competition is also designed to enable tech sharing among the engineering community, with several teams planning to make their code publicly available and open source after the competition.
‘So you’re going to take some of the most advanced AI algorithms ever developed for autonomous vehicles and put it out there for industry, for startups, for other universities to build on,’ said Mitchell.
And the project aims to ‘develop a generation of talent,’ Savaresi said.
‘The people who are competing in this challenge are going to go and start companies, they’re going to go work for companies,’ he said. ‘And so think the innovations from this competition will live on for many years.’