Skip to main content

Traffic moves along Dundas Street West in downtown Toronto on March 3 2014.FRED LUM/The Globe and Mail

Toronto mayoral candidate Karen Stintz wants the city to test pilot a locally developed "smart traffic light" project, which she claims would cut 25 per cent off of most residents' commute time.

The MARLIN system, developed by a group of University of Toronto researchers, uses artificial intelligence to enable traffic lights to "learn" traffic patterns and adjust to conditions in order to ultimately reduce gridlock.

Traffic congestion has emerged as a key issue in the Toronto mayoral race, and Ms. Stintz's Friday announcement was just the latest in a slew of ideas from the leading candidates on how to cut gridlock.

"This innovation is happening right in our backyard, at the University of Toronto," she told reporters at a press conference Friday morning. "This is the kind of modern, new and innovative thinking that we need in a mayor and that I will deliver."

Other cities across North America are already using similar "smart traffic" technology to measure traffic flow and curb congestion. But Baher Abdulhai, the lead U of T professor on the project, says that MARLIN – a chip the size of a smartphone and plugged into a controller at each intersection – is more advanced because unlike other products, it can coordinate traffic decisions "in two dimensions" – meaning it can account for traffic in more than one direction.

He also said that because MARLIN continues to "learn" and update every day, "it is always current."

U of T has already pledged to fund one-third of the cost to install MARLIN at 10 of the city's intersections for a pilot project, while Ms. Stintz said she's secured an additional one-third from Metrolinx. The decision on whether to put in the remaining $100,000 for a total of $300,000 will be put in front of city council next month, she said.

Ms. Stintz specifically identified Lakeshore Boulevard, just west of the downtown core, as a problematic area she'd like to see MARLIN tested in.

But Steve Buckley, Toronto's general manager of transportation, said the city has already been in talks with the MARLIN developers for close to a year, and remain skeptical of some of their promises.

Ms. Stintz and Mr. Abdulhai say the MARLIN technology can reduce intersection delays by 40 per cent, but Mr. Buckley said compared with similar systems in place in other cities, that figure "doesn't appear to be too realistic."

Mr. Buckley said the city already has in place adaptive responsive technology at 360 of the roughly 2,500 intersections that have traffic lights, and has already embarked on the process of looking at other products on the market to upgrade the existing system.

And while MARLIN is one of the products the city has looked at, Mr. Buckley said that city staff have concerns about the fact that it is as yet untested in any other cities. "Basically, they have a software system, but it hasn't been tested in the field," he said.

Councillor Denzil Minnan-Wong, chair of the public works and infrastructure committee, said he'd also had meetings with the U of T group over the past year as part of the city's overall plan to decrease traffic.

"The city is investing $60-million over five years implementing a suite of solutions across the city," he said. "Councillor Stintz's solution is not new to the city, but I appreciate her enthusiasm."

Report an error

Editorial code of conduct