The University of Toronto’s self-driving car team, aUToronto, beat out seven other universities in the second annual AutoDrive Challenge in Ann Arbor, Michigan, which took place earlier this month.
“We are very proud of the work we’ve done so far and have our eyes on the future.”
The team of engineering and computer science students used the university’s self-driving car, Zeus, to place first in all but one category of the competition. The University of Toronto placed first in every category besides concept design, where it placed second.
“I am extremely proud of our team,” said Tim Barfoot, a University of Toronto professor and faculty advisor. “It was fantastic to see all their hard work pay off. The car performed pretty much exactly as planned. The team really came together and did a brilliant job of supporting each other.”
Zeus uses a laser-based detector to create a three-dimension map of its environment, using the many cameras placed along the vehicle to identify and read traffic signs, lights, and signals. The self-driving car’s software combines various AI and machine learning techniques to identify signs, locate pedestrians and determine fastest routes. The vehicle itself comes with a manual control mode as well as a fully autonomous mode, acting as the physical interface for the system’s components.
aUToronto competed against seven other universities, including the University of Waterloo, and Virginia Tech, Michigan-based Kettering University, Michigan State University, Michigan Tech University, North Carolina A & T State University, and Texas A&M University.
This year’s competition focused on urban environment driving scenarios. The teams were tasked with completing four driving challenges that demonstrated the vehicles’ autonomous capabilities. These challenges included obeying traffic signs, following traffic lights at intersections, and navigating pedestrian crosswalks.
The University of Toronto placed in the following categories:
- Concept Design Presentation: 1st place
- Social Responsibility Report: 1st place
- Social Responsibility Presentation: 1st place
- Mapping Challenge: Tied for 1st place
- Traffic Control Sign Challenge: 1st place
- Pedestrian Challenge: 1st place
- Intersection Challenge: 1st place
- MCity Challenge: 1st place
- Concept Design: 2nd place
“We are very proud of the work we’ve done so far and have our eyes on the future,” said Keenan Burnett, technical team lead of aUToronto.
He attributed the team’s success to the enhancements made to Zeus over the last year. The enhancements included the use of light detection and ranging localization to precisely determine the car’s positioning, and employing Deep Neural Networks to accurately detect pedestrians and traffic lights, which required the collection of more than 70,000 images.
Burnett said to prepare for the final competition next year, the team will improve Zeus’ object-detection capabilities, focus on developing a fully-reliable and safe operating system, and test the vehicle on public roads within the next year.
Image courtesy University of Toronto via AutoDrive
StartUp HERE Toronto is a publishing partner of Betakit and this article was originally published on their site.