The University of Toronto and the University of Waterloo are two of eight North American universities chosen to participate in the AutoDrive Challenge, a three-year autonomous vehicle competition sponsored by SAE International and General Motors. The two Ontario schools are the only Canadian universities invited to compete in the inaugural competition, joining Kettering University, Michigan State University, Michigan Technological University, North Carolina A&T University, Texas A&M University and Virginia Tech.
“It's pretty rare for students to have access to self-driving cars for experiments, so this is a really unique opportunity,” explains Keenan Burnett, a Master of Science student and Technical Team Lead for the University of Toronto's AutoDrive Challenge team. Burnett is one of more than 50 students that have joined the team referred to as “aUToronto.”
3 stages to reach level 3 automation
The first of the competition's three stages will culminate at GM's Desert Proving Ground in Yuma, Arizona from April 30, 2018 to May 5, 2018. Teams will be judged on their familiarity with sensing and computation software, with simple missions including straight roadway driving and object detection/avoidance. By the end of year two, the vehicles will need to be able to avoid static obstacles while changing lanes and observing the rules of the road. Finally, by the end of year three, teams will demonstrate an autonomous passenger vehicle with level 3 automation.
Level 1 and 2 automation already in market
“Level one and two automation is already on the market and includes capabilities such as ‘active cruise control' and ‘lane-keeping,'” says second-year engineering undergraduate, Robert Adragna, who is also the team's youngest member. “‘Active cruise control' allows your car to adjust its speed when it senses cars ahead of it that are moving more slowly, while ‘lane-keeping' will similarly sense when your car is drifting out of its lane, responding by giving the car a nudge.” Safety regulators around the world are beginning to include these level one and two autonomous features as prerequisites for receiving high safety ratings. In fact, any 2018 vehicle in Europe must be equipped with automatic emergency braking, road edge detection and lane keeping as a part of its base price in order to achieve a four or five-star safety rating.
The competition comes on the heels of a June 2017 announcement by General Motors Canada that it will hire about 700 engineers in the next few years, many of whom will be based out of its Markham, Ontario software development centre.
GM and Uber play a major role in sparking interest in self-driving cars
Members of the team we spoke with say they weren't initially interested in self-driving cars in particular, but with recent investments by General Motors, Uber and others looking to get a foothold in Ontario's AI innovation ecosystem, that has quickly changed.
“So much has happened in the last few years,” says Mona Gridseth, University of Toronto Computer Science PhD candidate and aUToronto team member. “The confluence of IT and automotive sectors has made this a great place for cutting-edge automotive companies to perform their R&D. You just have to look at General Motors, Uber – and of course the launch of the Vector Institute for artificial intelligence. Vector is driving even more interest from companies in a variety of sectors, especially automotive companies and auto parts manufacturers.”
The opportunity for the two Ontario universities to participate in the challenge will also help market their renowned engineering and computer science programs, as evidenced by one aUToronto team member that signed up for a one year exchange from Technical University of Munich, a master's student named Andreas Schimpe. “When I was looking into opportunities to take part in projects I found this one the most appealing,” he says. “It's definitely the reason I decided to come to Toronto.”