Traffic sucks. Can technology from Miovision get us unstuck?

Globe and Mail

a day ago

In 2007, Toronto—forever a city where the car is king—set out to do the impossible: ban private automobiles from a major downtown street. King Street West runs from the Financial District to the Fashion District and beyond, and is lined with some of the city's glitziest venues—Roy Thomson Hall, the Princess of Wales Theatre, the TIFF Lightbox. It's also home to the Toronto Transit Commission's busiest streetcar route, and that year, in a bid to relieve overcrowding and improve travel times on the line, Toronto launched a pilot project that restricted motorists from travelling along King, eliminated on-street parking and improved pedestrian infrastructure.
It was, not surprisingly, controversial. Drivers complained that the new regulations were confusing and unfair, pedestrians grumbled that motorists flouted the rules, local restaurants insisted the project was killing their businesses. The whole thing threatened to blow up in the city's face.
But Toronto had a secret weapon: technology made by a company in Kitchener, Ont., called Miovision. Founded in 2005, Miovision pioneered the use of computer vision in traffic management, allowing municipalities to understand their traffic flow in granular detail and adjust infrastructure, mainly traffic signals, to make it move more efficiently. Co-founder and CEO Kurtis McBride (no relation) likes to call his company 'the smartphone of the intersection.'
Toronto installed dozens of Miovision's proprietary camera systems along King and the surrounding streets, and these gathered, 24-7, high-resolution video of each passing vehicle and pedestrian. Those same systems then used algorithms to translate that footage into data about the volume of traffic, but also how long cars were delayed at specific lights, how many near-miss accidents there were, and whether drivers complied with the new rules in general.
'There were a lot of questions, and we needed to have KPIs to measure the effectiveness of what we were doing,' says Roger Browne, the city's director of traffic management. 'At the time, there was no other technology as advanced as Miovision's. No one was operating at that same degree of accuracy.'
A year and a half later, the city declared that daily weekday ridership on the King streetcar had jumped 16%, with travel times improved by four to five minutes. Based on that data, Toronto decided to make King a permanent transit priority corridor.
This wasn't Miovision's first success story, and it wouldn't be its last. If anything, its tech has only become more urgently needed (which has helped drive three-year revenue growth of nearly 350%). While the King streetcars still run with relative ease, traffic in Toronto has, notoriously, accelerated from headache to crisis. According to the Toronto Region Board of Trade, the city's streets are now congested for 11 hours a day, up from nine hours in 2019.
Torontonians have the longest average travel time in North America. Earlier this year, the Dutch geolocation company TomTom, which releases an annual Traffic Index, ranked Toronto the eighth-worst city in the world for congestion (Vancouver was No. 5), with Torontonians losing 77 hours a year stuck in rush-hour traffic.
The economic, environmental and social effects are profound. Productivity and competitiveness are crippled, investment evaporates, carbon emissions increase, and overall quality of life is severely diminished—nobody who's sat for hours on the 401 needs to be told that gridlock's been linked to stress and depression. A report released last year by the Canadian Centre for Economic Analysis estimated that all told, congestion costs Ontario an eye-watering $56.4 billion a year—and growing.
The reasons for this are legion and sadly familiar: a rapidly growing population and aging infrastructure; construction on virtually every street; underfunded public transit; an endless parade of delivery vans and e-bikes; entitled BMW owners who treat red lights like suggestions. But a lesser-known cause is the fact that Toronto, like most cities around the world, still relies on an old-school system of traffic signals set to fixed schedules, unable to quickly adjust to increasing or declining numbers of road users. It's a system that hasn't kept pace with increasing demand for complete, multimodal streets or the explosion in last-mile delivery services. It's a system that Miovision was expressly created to change.
If you think sitting in a traffic jam is tedious, try spending your weekends counting the cars in one.
In 2004, when he was an engineering student at the University of Waterloo, McBride got a job doing exactly that. As part of a summer co-op program, he did some software development for a transportation engineering company in Toronto, and to make a little beer money, he joined a bunch of other students on a weekend traffic count. It wasn't exactly a fun task—standing by the side of the road in downtown Toronto in the middle of July, manually counting cars with an electronic clicker board—but it was a life-changing one.
'You can imagine the accuracy of the data that was coming out of that,' he tells me. 'And then you'd see the important decisions that were being made with that data on the other side.' Those decisions, chiefly around how traffic signals were timed in an intersection, were also, McBride realized, being made too infrequently—every five years or so—and not keeping pace at all with how quickly traffic was evolving in the city.
There had to be a better way, McBride thought, and in 2005, while doing a master's at Waterloo, he created it: AI-powered software that could take video footage of traffic and extract from it a more precise vehicle count. It could also tell whether those vehicles were cars, buses, trucks or bicycles; how fast they were going; and how they interacted with each other and with intersections. Such data points, arranged by the system into accessible charts and graphs, were extremely valuable to traffic agencies; it could guide them when deciding to widen a road, change the timing of a traffic signal or reduce speed limits. It could even provide 'near-miss' analytics, allowing a traffic engineer to identify potential safety risks, anticipate future collisions, and adjust roadways or intersections accordingly.
That same year, to commercialize this technology, McBride, who's now 45, co-founded Miovision with two other Waterloo students. They moved into hardware, as well, building a self-contained, battery-powered unit that now comes equipped with a 360-degree camera that can capture, process and upload video on the spot and in real time. They called it the Miovision Scout. 'We changed the model,' McBride says. 'Now we had one person in a truck with 20 cameras who could do the same amount of counts it used to take an army to do.'
Scout was a relatively niche product, designed largely for traffic engineers. But it was a hugely successful one. Ten years later, it was being used in 17,000 municipalities in 50 countries, and that year, Miovision raised $30 million in a venture funding round led by Montreal's MacKinnon, Bennett & Co., then the company's largest shareholder. As its manufacturing needs grew, it moved from downtown Kitchener into a former tire factory that McBride, along with a couple of other investors, transformed into a massive tech hub they called Catalyst 137. In addition to Miovision's own offices and 20,000-square-foot factory, the 465,000-square-foot building now houses about 25 startups, a large co-working space and a cluster of financial services.
A decade in, McBride began to recognize that traffic infrastructure writ large was also ripe for disruption. Increasing numbers of cities were becoming 'smart'—that is, leveraging digital technology and data collection to improve the lives of residents. But intersections were still, in a word, dumb. Traffic signals were (and largely still are) low-tech, analog systems—basically, a bank of light switches on a timer, turning on and off. Signals could be adjusted to behave differently during specific times of the day, but they couldn't reflect real-life, real-time traffic conditions. And if you wanted to add functionality to an intersection—say, installing an induction loop in the pavement so a car could trigger a light to change—each instance of that meant a new capital purchase, a new truck roll-out, a new maintenance issue.
By that point, though, the iPhone had also been around for almost a decade, transforming a simple cellphone into a nearly infinite array of different devices: a camera, a flashlight, a computer, a jukebox. What if Miovision did something similar, McBride thought, consolidating all the functionality of an intersection into a single piece of hardware that could meet the most common intersection needs at once?
'If you look at a problem from first principles,' McBride says, 'there's often a better way to solve it than the way it's being solved in the market.' That insight led to a whole new generation of integrated Miovision products that included more sophisticated cameras and myriad software applications, all of which could be plugged into an intersection and bring it into the 21st century. The company built a lot of this functionality itself; other capabilities came through acquisitions—between 2021 and 2024, Miovision acquired six different companies.
Some of the capabilities were relatively simple, like video detection. Instead of digging up the pavement for the aforementioned induction loop, a Miovision camera could do the work of activating a green light on a little-used side street. Or a Miovision system could be used to communicate to a traffic team that a signal was malfunctioning. Detroit's traffic signals were constantly being knocked out by rolling brownouts, and crews spent an inordinate amount of time just driving around looking for those. It was a drain on the traffic department's budget and slowed down emergency vehicles. Miovision systems plugged into the city's intersections sent text messages indicating when the signals went down, and crews could immediately respond to those specific problems.
Other applications were more sophisticated. The sequencing of traffic lights is traditionally co-ordinated by an electro-mechanical traffic signal controller whose timing is programmed by a traffic agency based on historical traffic patterns that can be minutes, days or even years old. In contrast, Miovision's so-called adaptive signal technology enables a traffic light to respond as needed to unexpected congestion or reduced capacity. Miovision deployed the systems in Boston, Pittsburgh and Portland, reducing average travel times by 25% and emissions by 20%.
As part of a pilot in Peterborough, Ont., adaptive signals deployed on the city's busiest street resulted in almost $1 million in travel-time savings, 106,700 litres less fuel used and a 273-ton reduction in carbon emissions. In 2022, Peterborough city staff recommended expanding the tech to 20 other intersections throughout the city.
Two other Miovision technologies—emergency vehicle pre-emption and transit signal priority—essentially allow first responders and public buses or streetcars to directly communicate with traffic signals to request green lights or extend green-light cycles under certain conditions. This has been successfully implemented in Waterloo, as well as in smaller cities in California and Texas.
Such technology might also soon be available in your own car. Last year, Miovision acquired the U.S. company Traffic Technology Services, which pioneered so called vehicle-to-everything (V2X) technology. (It also has insight into 80,000 intersections.) Its systems allow private vehicles to 'talk' directly with traffic signals through the dashboard, telling drivers the optimal speed to use to avoid red lights. Audi is currently Miovision's largest customer for this technology, but McBride sees the biggest market in delivery fleets, for which such systems would reduce delivery times, fuel usage and carbon emissions.
To get to the point where that's possible, however, Miovision needs to be deployed in far more intersections. Out of about 400,000 existing intersections in North America with stoplights, McBride says that his company can currently be found in around 66,000 of them. 'We really need to get to scale with the number of intersections that we have on our platform,' McBride says. 'That's a big focus for us now, to just build the density of our network. We've done pilots with almost all the automotive manufacturers, but we need penetration for them to want to roll out.'
Given Toronto's ongoing traffic woes, it might surprise you to learn that the city currently has 213 people working in its traffic management department, the one Roger Browne runs. This April, however, faced with a public increasingly frustrated and infuriated by gridlock, Mayor Olivia Chow created a new position that would effectively rule this department. The new traffic czar would be tasked with streamlining all its moving parts, and accountable for its successes and failures. A candidate is set to be announced this summer.
Who this czar will be is still anybody's guess, but if there's one thing they might consider when they take the job, how about earmarking some more money for Miovision? At the moment, the company's systems are deployed in roughly 100 intersections across Toronto (out of more than 2,500). But given the sheer enormity of the traffic problem, and Browne's enthusiasm for Miovision, a question inevitably arises: Why aren't they in every intersection?
According to McBride, Miovision began selling its systems to Toronto around the same time it began selling them to Chicago. Today, its coverage of Chicago is 10 times that of Toronto—about 40% of intersections in the Windy City. Browne's explanation for this is unsurprising: 'It's always a benefit-cost type thing when it comes to these deployments.' A Miovision system can cost upwards of $50,000 per intersection, and the city has limited resources. Sometimes, Browne says, you also don't need all the functionality that comes with that price tag: 'If all I want to do is provide an advance left turn for turning vehicles, I'm going to cut an $800 loop in the pavement.'
But Miovision can also help direct those dollars. In 2016, the city created a new traffic agent program, deploying these agents to guide vehicular and pedestrian traffic through particularly troublesome intersections. A data collection survey determined that such agents reduced blocked intersections by 90%. By the end of the year, the city will expand that program from 20 agents to 100. While not specifically committing to Miovision, Browne also adds that a 'key pillar' of Toronto's congestion management strategy is to increase the use of detection technology in general.
Traffic hasn't just increased, though; it's also become more complex. With the rapid emergence and convergence of AI and robotics, McBride is certain that complexity is only going to grow. He expects that, within a few years, the old-fashioned traffic signal will be a thing of the past; intersections will be completely digitized and automated. To keep up, Miovision is doubling down on its own complexity. V2X technology was one major step in that development, but McBride's larger ambition for the company is to become, in a decade's time, infrastructure-to-everything, with infrastructure including whatever kind of autonomous vehicle might be delivering your afternoon cortado.
'The amount of mode shifting that's going to happen in the next 10 years is going to be dramatic,' McBride says. 'We almost won't recognize what's driving down the road.' With any luck, or a few more Miovision systems installed, you won't be stuck behind it.
Your time is valuable. Have the Top Business Headlines newsletter conveniently delivered to your inbox in the morning or evening. Sign up today.