5 Questions: Richard Plunz on Crowdsourcing Urban Design with Twitter

March 02, 2018

During a storm, several thousand swales (low-lying channels that are shallower than a ditch) across New York City collect rainwater runoff and keep it from flooding the sewers. While working on an app to help monitor this system, Richard Plunz and his colleagues at the Earth Institute’s Urban Design Lab found that community involvement would be just as critical to its success as technology.

“The only practical way to monitor these sites is via social media, allowing people in the neighborhood to report on the status of green infrastructure,” said Plunz, director of the Urban Design Lab and a professor at the Graduate School of Architecture, Planning and Preservation.

Now, he and his colleagues are studying how this crowdsourcing approach might apply to the entire field of urban design. Their project combines mapping techniques with Twitter-usage data to gain a real-time understanding of how people occupy public space.

“Social media provides the first tool for gauging how areas with high-density tweets, such as parks, tourist attractions and transit hubs are actually used,” said Plunz, whose team collaborates on a broad range of multidisciplinary research projects, including investigations of new ecological realities facing cities, next-generation economic development and the intensifying effects of climate change. “Over time, we anticipate that social media will help us to produce a new generation of urban design and planning tools that address how to make these spaces more environmentally and socially resilient.”

Plunz, who is known for a wide range of innovative urban research, development and design projects, has a particular expertise in infrastructure. His landmark 1990 book, A History of Housing in New York City, was republished in a revised edition in 2016. In 2017 he published, City Riffs: Urbanism, Ecology, Place, which traces the changing perspectives of urban design by moving between 16 cities, including New York, Rome, New Delhi, Caracas and Seoul.

Q. What have you learned from analyzing Twitter use in public parks?

A. We have been able to confirm the efficacy of real-time geospatial tracking of information, which is already an important breakthrough in terms of urban design. What we did not anticipate was the potential for understanding sentiment regarding urban settings and in response to real-time tracking, and especially in affirming the accuracy of Twitter data relative to specific events and conditions. We have also considered the potential for predictions related to people and events. For example, in our work on the High Line, we can distinguish between New York residents and tourists. We can anticipate hours before an event that it could be oversubscribed and therefore require special precautions relative to security and logistics.

Q. What made you choose Bryant Park, Washington Square Park and the High Line?

A. These particular parks are representative of three differing ways that public space is used. Bryant Park has the most potential for Twitter-based analytics, given its heavy usage, especially on weekday lunchtimes. Washington Square Park has a much greater full-time occupancy, which gives a more expansive view of its use. The High Line’s long trajectory and large percentage of tourists provides yet another reading and perspective on the relationship between a park and its people. We are researching the effectiveness of Twitter in helping to understand these different characteristics in real time and the possibilities for correlating them with many variables, including weather patterns and social events.

Q. What can Twitter tell us that surveys can’t?

A. Twitter data is 24/7 and in continuum, a conscious stream, a collective picture of social responses to particular situations and contexts. It provides a tool for future planning as opposed to a system focused on specific issues at specific times. It is fundamentally cognitive in nature and, therefore, represents a huge advance in our comprehension of how we interact with our environment and vice versa.

Q. What are the practical applications?

A. There is real-time monitoring of public space density-of-use patterns and even usage prediction related to specific public events. It’s also possible to use Twitter density and sentiment as a tool for long-term design considerations. This can include, at different scales, the redesign of public parks—adding more grass and trees, for example, or new activity areas, or figuring out if nearby buildings or construction are having a negative impact on vegetation. Is a park being overused? Or are users dissatisfied and, if so, why? How can a park be more effective and have more social benefits? The next 10 years will be huge in reaching a new understanding of every aspect of our built environment via social media data.

Q. What are the limitations of this approach? Where would you like to see it go?

A. Perhaps the biggest limitation now is the uneven distribution of Twitter usage throughout the city. In our study we ended up focusing on lower Manhattan, in part because there is sufficient Twitter data for these areas. Central Park had to be eliminated because the Twitter-density patterns are so dispersed. We anticipate that this problem will disappear as the use of Twitter-like social media grows. That will be the moment when techniques like ours become everyday practice, opening a new window on to urban cognition. The cognitive mapping of urban environments entails a mental decoding of information. It has been a preoccupation for centuries, but more as a metaphysical construct. There have been iconic studies, from Italian architect Giambattista Nolli’s 1748 plan of Rome to Baudelaire's 1847 Flâneur—who walks the city in order to experience it—to urban planner Kevin Lynch's Image of the City analytics of the 1950s. We are moving from urban cognition as a primarily literary and conceptual tool into the realm of its becoming an operational tool for spatial design. This is a new moment for an ages-old preoccupation.

—Eve Glasberg