If you had a tool to predict traffic, would it change your driving habits?
York University PhD student Tim Leonard recently spent an entire weekend with virtual strangers developing a web-based platform to do just that when he participated in TrafficJam, a 48-hour hackathon designed to help solve Toronto’s traffic problems.
The idea conceived by Leonard and his four teammates – who call themselves and their project “RueView” – earned a second-place finish at the event that saw 150 participants operating in more than 20 different teams.
Co-sponsored by Evergreen CityWorks and the City of Toronto, TrafficJam brought together a cohort of “engaged citizens, digital creatives and data detectives” to build tools that will kick-start the development of real traffic solutions in Toronto. The city is said to have the longest commute times in North America, with an annual cost of $6 billion.
The event began Oct. 2 at 8pm and continued right through to Oct. 4 at 2pm. Coders, developers, data scientists and designers were given free and open access to Toronto’s traffic data, and an opportunity to collaborate with local traffic analysts, government officials and data collectors.
“We had to ask ourselves, how can we distill large volumes of really complex traffic data into a very simple form that people can use?” said Leonard. “We thought, if you can engage people with traffic data, they will change their behaviour and we can make an impact.”
The idea, he said, was to present the information in the same way that weather data is used.
Leonard, who met his teammates just prior to the event in an online forum for TrafficJam participants, contributed to the project as a data scientist. RueView’s other members are: Ashkan Parcham Kashani, developer; Sharmarke Noor, designer; Joanne Dang, designer/idea generator; and Zifan Zhang, idea generator.
With the clock ticking, RueView wasted no time getting to work when the hackathon kicked off, and Leonard pushed through the first 24 hours on very little sleep while his role as data scientist was at the forefront of the project’s development.
“Once we had the basic idea in place, we sat down and formulated a plan on how we could make it work, and started prioritizing and putting code together,” he said. “It took pretty much the full 48 hours to get a final product out the door.”
And the final product earned the team a $2,000 prize and the second place title.
The prototype, explained Leonard, uses the Gardiner Expressway westbound and predicts the traffic based on traffic and weather history. The team is now working on including events, road closures and holidays as part of the data that predicts how vehicles will move along the busy Toronto highway.
“We’re really excited,” he said. “The proof of concept worked and now we are working to modify it, scale it up and polish the code. There’s so much room for improvement, it is always going to be evolving.”
His first hackathon was an intense but immensely positive experience, said Leonard, noting that organizers provided a very comfortable and accessible environment, and the participants, although operating in different teams, worked to share ideas and information across the competition.
Leonard, who earned his undergrad and master’s degrees in psychology at York, is working toward a PhD in psychology (functional area: brain, behaviour and cognitive science) and said he hopes to pursue a career in data science.
A field that is still “struggling to define itself,” he said, data science is a hybrid career that combines elements of programming, researching and problem solving, and can be applied to any discipline.
“Though I had experience programming as a hobby, graduate school has been a wonderful opportunity for me to hone my data science skill set,” he said. “Learning to program and analyze big data sets is an absolute must in our lab – we deal with a massive amount of data.”
In particular, he said the statistic classes offered through the quantitative methods specialization in York’s Graduate Program in Psychology have been an “invaluable resource” for learning how to interpret and explore data sets.
He will also work to apply his background in psychology to find ways to encourage people to use RueView.
“If we can change traffic behaviour in five per cent of people, that’s a major shift on the roads, and we are confident we can make a change.”
Leonard suggested his exposure to so many different areas of study at York has helped him define his education and career goals.
“It’s something I was open to and seeking out, and it’s definitely been an influence,” he said.
To use RueView and follow the team’s progress, visit rueview.io or follow Leonard on Twitter at twitter.com/letimle.
For more on TrafficJam, visit trafficjam.to.
By Ashley Goodfellow Craig, YFile deputy editor