When it comes to transportation data, it may be possible to have too much of a good thing. With the rise of sensors and the Internet of Things (IoT), urban planners have more traffic information available than ever before. For some, it’s more than they can use.
“Traditionally, cities have been starved for data. If we wanted to refine an intersection we would count six hours of data every five years. With IoT you can get a continuous stream of data in real time, all the time, from all your intersections,” said Kurtis McBride, CEO and co-founder of transportation analytics firm Miovision
But that massive influx of information isn’t necessarily a clear win. “Now they have so much data, but many of them don’t have the tools to ask the right questions,” he said. Miovision is looking to help remedy that with the launch of Miovision Labs, a new division within the company focused on leveraging transportation data and smart technologies.
The lab team will include technologists and product strategists with expertise in communications, computer vision, deep learning, artificial intelligence (AI), big data analytics and embedded device design. The company plans to work directly with cities to explore possible uses for connected traffic signals and related technologies.
The lab already has a pilot project in the works with the University of Toronto looking at data as a remedy to bicycle accidents.
“There have been a lot of vehicle-bicycle collisions,” McBride said. “Traditionally with these you would look at all the accidents that have occurred, you would look at the areas with the highest incident rate. But you get a very small sample set that way, and you need collisions to occur before you can study it.”
The data team will instead focus on “conflict zones,” places where bikes and vehicles may likely cross paths. “By studying those zones and looking even at near-collisions," he said, "you can find ways to design bike lanes more effectively."
University officials are eager to see how the data can be put to work. The partnership “will give insights that can lead to better decisions about infrastructure. The important piece with this project is that it's using real-world data, not a simulation,” civil engineering professor Matthew Roorda said in a press release.
Traffic is at the heart of many smart city initiatives, in part because of the potentially large impact such projects can spark. With smart traffic management and other measures, cities could save some 4.2 billion man-hours annually by 2021, Jupiter Research
reports. That’s equivalent to saving each city driver roughly a full working day each year.
A number of cities already are working along these lines. San Francisco
received grants for $11 million and $10.9 million respectively last year from the U.S. Department of transportation. The former is looking at new ways to time traffic signals, among other efforts, while the latter wants to craft “smart spines” that would use sensor data to better manage traffic flow.
Cities have used data to manage traffic in the past, but it’s been a clunky endeavor, with surveys typically updated only every five years. “So you get the data, you program the lights and the next day a new factory opens, a school closes, the economy changes. Now the traffic pattern is all different, but we won’t reflect those changes for another five years,” McBride said.
Urban planners have sought to collect real-time data, but it’s been cost-prohibitive in the past. “When you think of how many lights are in the city, how many intersections, and how many movements are in those intersections," McBride said, "you need thousand of sensors and complex, expensive analytics.”
IoT delivers on that need at an increasingly affordable price point. But civic planners must now find a way to put that data to use, which may mean rethinking the uses of data and the analytic framework around it. “All the ways they ask questions are limited and simplistic, because that’s how the data was," McBride said. "So while IoT solves one problem — now they have enough data — it creates a new problem. We need new tools to ask questions of that data."
Miovision Labs is looking to evolve those new tools, for instance in a project with Ontario, Canada's Region of Waterloo district, which is responsible for the operation of about 500 traffic lights. The research team is collecting data in order to develop better algorithms to determine the timing of lights. That would be a big step forward from past analytic efforts.
“Instead of having a human look at the data, this offers the possibility of adjusting the lights automatically based on what is actually happening in the network,” McBride said.
The team also is working with management consulting firm CPCS to look at the ways in which freight moves through metropolitan areas. With the steady rise in Internet-based shopping, cities are seeing a surge in freight traffic, and they’re looking for ways to incorporate this effectively into the local traffic flow.
Data can help. “What are the total miles being driven by delivery trucks? How much delay does that add to the network? Is there a way to move that distribution center to reduce overall traffic time?” McBride said. “The pitch to the freight companies is that if you move your distribution center five miles up the highway you can save a lot of money in terms of travel times or the number of trips. The municipality wins because there is less traffic on the road.”
Maybe the environment benefits too. McBride points to studies showing that optimizing traffic lights can drive environmental impacts equal to taking 10 percent of cars off the road. That could go a long way toward meeting international greenhouse gas reduction goals.
To reap these benefits, city planners will need new toolsets, new analytic capabilities, that help them take full advantage of the data available in the IoT ecosystem. “We view the lab as a way to ensure that we continue innovating, even as we provide customers with the services that they need,” McBride said. “It’s a way to be a little more forward-facing.”