MSIM Team Helps Seattle Wrangle Its Parking Data

Students walking on the University of Washington campus

As anyone who has ever tried to find on-street parking in downtown Seattle knows, it can be complicated. And while it can be difficult for drivers, it’s even more challenging for public officials who are trying to manage the supply of on-street parking without all the information they need.

Four iSchool students are using their information management skills to help the Seattle Department of Transportation answer two crucial questions. How many parking spots are open? And what factors affect parking availability?

The Master of Science in Information Management students – Nathan Cunningham, Allison Chapman, Sahil Aggarwal and Shreya Sabharwal (pictured at top, left to right) – are tackling the problem for Capstone, the culminating project in which iSchool students apply their knowledge to solve real-world problems. The team members wanted a project that would allow them to use their skills to think creatively.

“All of us wanted to work on a project with a social goal,” Aggarwal said. “We were looking for people who had a drive to do something for the social good.”

Around the time the students were looking for a Capstone project, the city released a huge data set of parking information as part of Seattle’s Open Data Program, which encourages the development of innovative technology systems that improve people’s lives. The team saw a chance to help the city, as well as people who park downtown.

The city had been working to better understand on-street parking for a while. But the complexity of information and the size of the data set made it a challenge. The team took on the project to help the city’s transportation department better understand — and therefore improve — its parking situation.

“We wanted to see if we could use our data-science and machine-learning skills to solve this problem creatively,” Aggarwal said.

The team began defining the project and figuring out how to wrangle millions of records. They interviewed Seattle Department of Transportation staff and looked at work the city had already done. While there is plentiful data from parking meters, SDOT can’t accurately predict parking-spot occupancy on its own — some drivers don’t pay, some leave before their time runs out, etc. The team then looked at other factors that could affect parking, such as the weather, the time of year and whether a major event was scheduled nearby.

“This is a really innovative project that is pushing the boundaries of what can be done."

“Getting the data together as one unified data source was challenging,” said Sabharwal. “The data we were interested in was enormous — every parking transaction across all the streets of Seattle for the last two years. We had to be smart about how we extracted data.”

The team limited their project to Belltown North, which gave them a more reasonable amount of data to work with. Sabharwal and Aggarwal spent a lot of time troubleshooting and finding ways to access the information effectively. Once they had the data they needed, they were able to start using their machine-learning and data skills to make connections.

Cunningham says the iSchool gave the team members the knowledge they need to successfully complete a project that required a solid foundation in data science, critical analysis and communications skills.

“The iSchool prioritizes interdisciplinary thinking in its curriculum, so we’ve learned to use it in our project, too,” he said.

The team is wrapping up the work now. When they are done, they will give the city a model for Belltown North with analysis about factors that affect parking occupancy. They’ll also share information to help the city re-create or expand their method to other areas, along with recommendations for future work.

This project is important to the city, said Mary Catherine Snyder, the parking strategist for SDOT. The city needs information to set appropriate rates for parking. Correct pricing will help create appropriate demand for parking — so there are just enough spots open to allow drivers to park quickly, without endless circling that causes congestion and releases greenhouse gases, among other problems.

Snyder knows this project won’t answer all the city’s questions, but it’s an important step and sets the stage for more progress.

Paul Alley, the open data manager for Seattle’s IT department, and Snyder have been impressed by the students’ skills, persistence and creativity. They appreciate the team’s ability and willingness to keep going and not get stuck, even when faced with challenges.

“This is a really innovative project that is pushing the boundaries of what can be done,” Alley said. “It’s a very exciting use of the Open Data project that can have a really big effect.”

For the students, the work has been a satisfying way to use their skills beyond the classroom as they get ready to graduate.

“Applying the skills I’ve learned in the MSIM program to a complex, real-world problem has been the most exciting aspect of the project for me,” Chapman said. “It’s been fun to move beyond using a pre-selected data set to answer a clearly defined set of questions like we often do in our classes when we’re learning new skills.”