Where Industry Meets Innovation

  • Contact Us
  • sign in Sign In
  • Sign in with certificate
mit campus

Resources

Search News

  • View All
  • ILP News
  • MIT Research News
  • MIT Sloan Management Review
  • Technology Review
  • Startup Exchange

ILP Institute Insider

June 6, 2016

Putting a Face on Data

Sarah Williams harnesses diverse data sets to build visualizations that help stakeholders better understand urban environments.

Steve Calechman

While reciting facts helps sway opinion, showing them crystallizes an argument. Modern technology has allowed researchers to find data at a pace and with a precision that wasn’t possible in the last decade. This is where Sarah Williams does her work.

Sarah Williams
Assistant Professor of Urban Planning
Director, Civic Data Design Lab
As an assistant professor of urban planning and director of the Civic Data Design Lab, she uses everyday tools, cell phones in particular, to survey a city’s infrastructure. The crowd sourcing produces real-time information on human behavior that can be put on a chart or map. For her students, it builds the necessary workforce skill of communicating with data for policy change. For governments, it provides crucial insights to develop infrastructure and policy initiatives, and, for industry, the analytics reveal new markets to tap and existing ones that can be better served.

Getting from point a to b
One of Williams’ major projects has been studying transportation in Nairobi, Kenya. The capital city’s public transit system involves a collection of varying size buses and vans, called Matatus. While it has designated lines and stops set by a co-operative of hundreds of owners who manage over 130 routes, “the system is chaotic,” Williams says. Often residents know how to navigate their daily routes but they don’t know how to navigate the overall system.

The problem was that with the number of independent owners, there was a lack of shared data necessary to develop transit models and improve travel patterns. Williams’ solution was to leverage the ubiquitous use of cell phones to collect data on a system that serves 3.5 million residents. By transforming the information into a General Transit Feed Specification (GTFS), she says that she was able to work with Google to make Nairobi the first city to have their informal transit system searchable on Google Maps.

What was once somewhat tacitly understood was now clarified. “It makes the invisible visible,” Williams says. Citizens could see where connection points were and could more easily navigate the city. Owners and drivers were able to make better strategies as they could identify underserved regions. Planners could make more informed decisions by better understanding transit flows, and government officials could finally see it as one system that needed to be managed, especially because of the severely congested streets.

Being able to visualize the data allowed for unforeseen possibilities as well. Since its release in 2015, the data has been downloaded by over 200 entities, which have used it for everything from transit efficiency models, trip planning applications, and more recently as the basis for developing plans for a proposed Bus Rapid Transit (BRT) system. Nairobi’s National Transit Safety Association (NTSA) has also employed it for a new monitoring system, which they hope will help ensure greater safety on the Matatus, she says.

The design of a district
Williams has also used crowdsourced information in Manhattan. The city had been looking at the Garment District. Sections were underused and officials wanted to rezone it in order to allow non-fashion businesses into this prime real estate. Designers resisted, claiming the area was still of prime importance to their industry, she says.

Williams used social media to track 100 designers for two weeks in order to see how the district behaved and the truth behind whether the district was as robust as the fashion designers contended. She found active manufacturing, but not in the assumed areas. The city’s data showed apparel manufacturing was concentrated near Broadway, but Williams’ data showed the manufacturers with the most business were actually closer to 8th Avenue. “The results provided evidence for the city to create greater protections to this part of the district,” she says.

The data also showed that 85 percent of apparel-related business transactions happened within the area’s 12 blocks, and that large and mid-level designers used the district more, making prototypes before shipping them off to Asia for full production. Each finding showed the everyday importance, both locally and in the global marketplace, she says. City officials decided to modify their plan, and went back to the drawing board with much more detailed data about the interworking of the district, Williams says.

Participants of the study used FourSquare to track which businesses they checked in with during the two-week period. The visits were geo-registered and in real-time automatically loaded into Williams’ map database. The social media gaming application allowed involvement, and the development of a dataset, through personal cell phones. “It was the pre-loaded business locations available through social media, which allowed us to pick up on the exact business locations that traditional GPS units can’t in dense urban environments, creating a more precise, easily accessible survey tool,” Williams says.

Playing the lottery
One overall goal for Williams is, in an increasingly data-centric society, students must be data literate. It’s a necessity, she says, to be able to read a map or chart and be able to critique the information, not merely accept it as fact.

In a National Science Foundation grant she worked on with Laurie Rubel from Brooklyn College, Williams and colleagues created a curriculum that teaches math using social justice data to help contextualize and visualize issues for high school students. In the first module, studying the New York City lottery, Williams designed a web-based interactive map that allowed students to explore data about every lottery retailer. Students could see ticket sales versus money won, which neighborhoods spent more, and what percentage of income was spent on the lottery. “Ultimately, students learned the odds were against them as it was clear people win and lose in the same proportions across the city,” Williams says.

The tool also allowed the students to generate geographically tagged field interviews through the creation of videos, audio, and pictures, which were automatically uploaded to the online map. Using interviews with lottery players and retailers, coupled with map data, students developed reasoned arguments about whether the lottery was “good” or “bad,” she says.

On a bigger level, crowd sourcing engages the public with essential civic needs. With the bus system in Nairobi, riders provided feedback that directly led to where the people were going, where they wanted to go, and from that, over five different startups used the data to provide increased transportation access. In Manhattan’s Garment District, technology showed that the section acts as a prototyping center for the global apparel industry.

The insights are new and fresh, and they happen at a scale in space and time that was previously unavailable. The result is that businesses and governments can take that information and make decisions and develop services, tools, and products that are more responsive to what users want. “All of us have sensors, our cell phones,” Williams says, “and these, along with the intertwined social media, allow anyone to better understand the pulse of the city in real-time.”