These U of T students are 'pioneering' the autonomous vehicle of tomorrow

A self-driving car, built by a team of University of Toronto engineering and computer science students, is "pioneering" the future for how autonomous vehicles are designed.

The team's self-driving car, Zeus, won the 1st phase of an international driverless vehicle competition

These University of Toronto students are leading how driverless vehicles will be designed. (Submitted by University of Toronto)

A self-driving car, built by a group of University of Toronto engineering and computer science students, is "pioneering" the future for how autonomous vehicles will be designed. 

Their autonomous vehicle, named Zeus, won the first phase of the AutoDrive Challenge, an international driverless vehicle competition.

"We're elated, a little tired, but very happy that we made it through the competition," said Zachary Kroeze, the team's leader. 

The three-year long competition is designed to lead innovation in the burgeoning field of autonomous vehicles. 

The University of Toronto team's car outperformed seven others from Canada and the U.S., including an entry from the University of Waterloo, earlier this month.

The week-long competition, that wrapped last Saturday in Yuma, Ariz., required teams to test how their self-driving car could navigate an urban environment. 

Zeus, a self-driving car designed by a group of University of Toronto students, outperformed autonomous vehicles built by other teams. (Submitted by the University of Toronto)

Zeus was converted from a Chevrolet Bolt EV. The vehicle was provided by its manufacturer, General Motors.

The team equipped Zeus with around $200,000 worth of modifications that took them almost a year to perfect. The modifications included a 3D LiDar laser system that detects obstacles on the roadway and a camera that recognizes road signs and ground markings. 

"We can't just go with what the industry is doing," Liam Horrigan, the team's mechanical engineer, said. 

"There's issues, there's accidents and what not. Even the companies that are doing best are very, very secretive about what's going on, so it means that we're really pioneering this in a sense." 

This team of University of Toronto engineering and computer science students won the first phase of the AutoDrive Challenge. (Submitted by the University of Toronto)

The AutoDrive Challenge required each self-driving car to complete nine tasks, including geolocating 50 locations in the U.S., driving along a straight line, detecting and avoiding objects, and obeying road signs.   

Horrigan, 22, considers his teammates to be leaders in the field of self-driving cars, highlighting that big companies are far from finding solutions that allow these vehicles to perform well in cities.   

"We are really at the forefront, and we try to solve the problem ourselves," he said. 

The autonomous vehicle competition will continue for another two years. The goal is to develop cars that are capable of driving routes in urban areas by 2020. 

It's something that can take companies decades to build properly, Kroeze explained, noting how fast self-driving technology is currently advancing. 

The AutoDrive Challenge tested each team's self-driving vehicle on its ability to compete nine tasks. (Submitted by the University of Toronto)

Angela Schoellig, an aerospace professor at the University of Toronto, has coached the team from the beginning. She anticipates the next phase of the AutoDrive Challenge will test the students' design capabilities. 

"It's extremely challenging because the environment changes so much. We have snow, we have rain, there may be fog or other people in the way," she said.  

The University of Toronto team plans to update the car's software using their $30,000 prize, so it can handle more complex tasks, such as following a path between two points and bypassing moving objects. 

Kroeze wants Zeus to be able to handle pedestrians crossing the road by the end of the competition. 

With files from Philippe de Montigny