During the winter in Boston, school buses inevitably end up getting stuck in the snow. When that happens, the City wants to tell parents not only that students will be delayed by inclement weather but also how late their children will be. In 2011, this was a problem because there were less than half a dozen City staff available to make calls to parents in the school district. But why were parents being contacted by phone, and how did the City know how long it would take for each child to arrive? It turned out that the Boston school buses were being tracked by GPS—a detail that was instantly promising to the volunteers from Code for America who had been presented with the problem. They came back the next day with a draft app that parents could use to locate school buses on their smart phones.
This anecdote, provided by Abhi Nemani of Code for America at the most recent meeting of the Urban Systems Collaborative, provides a window into the emerging revolution of “big data”, the trail of data points generated by individuals, infrastructure, and the natural world that promises to fundamentally change our cities. With a surplus of available information, often what is needed most is for someone to connect the dots.
The Urban Systems Collaborative (USC) occupies a unique niche in the ecosystem of urban data. Coordinated by a group from academia and industry, including IBM Smarter Cities researchers Colin Harrison and Jurij Paraszczak, USC identifies itself as being “engaged in study, evaluation and modification of real-world information to reveal emerging patterns of urban behavior that are changing the ways that people live in cities and how these changes affect the planning, design, development, governance, and operation of cities.” USC met at UC Berkeley in September in their first West Coast meeting, having assembled previously in New York and Chicago. The meeting’s stated themes, which functioned more as starting points than strictly followed directives, were to map information flows for emergency management and citizen engagement, and to consider the roles of centralized and distributed approaches to operations and open data.
The keynote speech was delivered by Steven Koonin, a theoretical physicist who was recently named the Director of the Center for Urban Science and Progress (CUSP) at New York University (NYU). In his address, Koonin spoke about the ability of new information sources to help us understand the relationship between psychology, human behavior and the physical sciences—or in his words, “from sensors to sociologists.” But as Koonin noted, while big data allows for better identification of outlier values and a finer stratification of data, the flood of information raises issues about the quality of the data and who has access to it.
Visualization of Taxi Pick-ups (Orange) and Drop-offs (Blue) in New York City (NYU Center for Urban Science and Progress)
A series of short presentations and panel discussions followed Koonin’s keynote address. Speakers came from IBM, Esri, UC Berkeley, Stamen Design, the City of Minneapolis, Carnegie Mellon, Code for America, Fehr & Peers, Ushahidi, the City of Oakland, and Skidmore, Owings, & Merrill. To start, several perspectives were put forward on how best to understand big data. Colin Harrison suggested that systems in a city are like systems in a human body, where some processes are inherently centralized, while others are inherently diffuse. Mark Baker of Esri approached the topic of urban information through the lens of geography, saying, “big data tied to geography becomes important information.” Jerry Walters of Fehr & Peers, noted that big data gives cities greater options for faster, more efficient decision making. Jennifer Pahlka of Code for America, spoke about data in terms of public engagement, saying that “open data is about a connection between citizens and government.” Presentations were distinct, with each speaker engaging their area of interest, but in contrast to earlier USC meetings, a degree of mutual awareness and understanding between disciplines had emerged.
In addition to theoretical constructs, symposium participants also presented a variety of technical solutions already being employed to address urban problems:
Multimodal Network Simulation in Silicon Valley (Fehr & Peers)
For some of these solutions, new data and analysis provided insight into previously unknown trends. But more frequently, technology was used to improve communication with the public. In fact, reference to the human element became one of the underlying themes of the symposium, with speakers considering the topic as it pertained to systemic change, citizen engagement, and the teaching of city science.
In his keynote address, Koonin said that innovation on its own would not be enough to affect change. The true societal levers were things like behavior, perception, and regulation. These areas, he proposed, were equally if not more important than technological advancement for altering the status quo.
Oakland Crimespotting (Stamen Design)
Nicholas de Monchaux of UC Berkeley led a workshop tasked with proposing content for a chapter on citizen engagement in a hypothetical textbook on urban systems. Several examples were discussed, including High Speed Rail in California, the open-source Bogota taxicab database, Nextbus copyright issues, and Solar Boston environmental calculators. Among the group, there was some skepticism about whether a single chapter would be sufficient to cover the topic of citizen engagement and whether it belonged in a volume for students alongside topics like city operations. Ultimately, participants were more comfortable providing caveats than fully-formed recommendations or outlines.
The amount of available urban data will soon overwhelm the number of people trained to work with it. The McKinsey Global Institute has projected that “The United States alone faces a shortage of 140,000 to 190,000 people with analytical expertise and 1.5 million managers and analysts with the skills to understand and make decisions based on the analysis of big data.” Several academic institutions are working to meet this growing demand. NYU is bullish enough on big data that they will be supporting 50 full-time senior researchers in this area through their Center for Urban Science and Progress over the next five years, and in 2013 they will begin offering a graduate degree in Urban Science and Informatics geared toward producing graduates for public agencies, private data providers, startups, and non-governmental organizations. What is the best training for these students? If the Urban Systems Collaborative were to suggest a curriculum, would it be mathematical, spatial, design-savvy, computational, or engineering-based? Would the foundational skills taught in urban planning programs be equally prized? As Otto Doll, Chief Information Officer for the City of Minneapolis, noted at the symposium in Berkeley, big data in the abstract is not an end in itself. “The key,” he observed, “is turning information into knowledge.”