Anthony M. Townsend - Smart Cities - Big Data, Civic Hackers, and the Quest for a New Utopia

Geraci and I had stayed in touch, and throughout the autumn and early winter of 2008, we would meet for long walks around the East Village, looping out from my apartment at Ninth Street and Third Avenue, on a gallery walk of grassroots smart-city projects. Past the free hot spot a crew of NYCwireless volunteers had installed in early October 2001 to provide relief Internet access after the September 11 terrorist attacks. Past the block where Geraci and fellow ITP student Mohit SantRam had launched the first Neighbornode cluster from SantRams apartment. Past the cafe where Crowley and Selvadurai were hard at work coding the first version of Foursquare. As John shared his ideas for building DIYcity, wed use the city as a brainstorming tool, rehashing the lessons of those earlier efforts.

DIYcity mushroomed overnight. Geraci had built the site using Drupal, an open-source system that allowed anyone to easily form a new group devoted to a specific city or a particular problem. In less than a month local chapters began to organize as far afield as Sao Paulo, Copenhagen, Portland, and Kuala Lumpur. By the dawn of 2009, thousands of Web developers, urban planners, environmental designers, students, and government employees had enlisted in the effort. With the help of software developer Sean Savage in San Francisco, Geraci organized a bicoastal pair of meet-ups, held on January 14, 2009 (the same day Google announced it would shut down Dodgeball). His goal was to bring together coders and urban planners for the first time to brainstorm an agenda for the nascent movement. Helped along by some free publicity on the popular geek blog BoingBoing, both meetings were packed.

But DIYcity wasn’t only about talking. Geraci wanted the movement to build “a suite of tools that residents of any city, anywhere, can plug into and use to make their area better.” He had his eye on Washington, DC, where Apps for Democracy, the first city-sponsored apps contest, had run during the preceding autumn. Geraci had concluded that apps contests were an inspired idea but too open-ended and too driven by government data and the programmers’ own desires instead of the problems of citizens. So he devised a series of DIYcity Challenges that started with problems—ride sharing, bus tracking, tracking the spread of communicable diseases. To accelerate the process, and keep the focus on users, not tools, he even dictated key parts of the design solution—for instance, a Twitter bot to crowdsource traffic reports. And rather than inviting competition, Geraci’s approach was for the entire community to collaborate on a single solution. It was the collaborative culture of Red Burns’s Interactive Telecommunications Program reemerging at an opportune moment. He recruited developers and even worked on the teams himself as they built solutions to the challenges.

The immediate goal was to log a couple of quick wins that showed the DIYcity approach could work. The results were impressive, given the pace of the challenges— which lasted just a few weeks—and the lack of prize money. The first challenge produced DIYtraffic, a service for creating personalized text-message alerts based on a feed of traffic-speed data Yahoo provided at the time, culled from roadway sensors and anonymous tracking of mobile phones by wireless carriers. Presaging the popularity of crowdsourced traffic apps like Waze that would arrive a few years later, DIYtraffic also allowed users to add their own reports to the official feed. In keeping with Geraci’s emphasis on reusable tools, a kind of “write once, run anywhere” approach to local software, DIYtraffic was skinnable, meaning that anyone could set up the same service for their own city by simply customizing the outermost layer of the underlying software.

Another challenge focused on public health led to the creation of SickCity, a tool inspired by Google Flu Trends. Both tools sought to map epidemics by mining Internet activity. Flu Trends relied on searches for terms related to flu symptoms and treatments, which Google could geographically tag based on the users IP address. SickCity was more crude, simply scanning the Twitter stream for keywords like “flu” and “fever.” But while it lacked the sophisticated automated methods Google uses to build its list of terms that might indicate illness and had significantly fewer data points, SickCity did have several advantages over Flu Trends. First, people were likely to start reporting symptoms to a social network before the illness was full-blown and they began searching Google for treatments. Second, SickCity offered the ability to see trends at smaller scales—Google didn’t start publishing city-level slices until January 2010, almost a year after the release of SickCity. Finally, by changing the filter key words, the tool could be applied to any variety of public health concern, from food poisoning to anxiety.

Created in an all-night marathon of collaborative coding, SickCity was DIYcity s most successful challenge and spread widely in a frenzy of open-source replication. According to Geraci, over one hundred local instances were set up within seventy-two hours. While not scientifically validated like Google’s project (a collaboration with the Centers for Disease Control), and flooded with spurious data by the emergence of swine flu (which polluted Twitter with discussions of the disease by people who were not themselves ill), SickCity showed the viral potential of lightweight Web apps that feed off social interactions to address urban problems.

And then, just as fast as it had blown up, DIYcity was gone. After just a single meet-up and five challenges, Geraci made a difficult choice. In 2011, over coffee in Manhattan’s Little Italy two years after the end of the DIYcity Challenges, he laughed as he recalled it. “I had a new baby, no job, and wasn’t prepared for the success of DIYcity.” And as any social entrepreneur will tell you, conceptual success doesn’t always translate to financial success. “How do you pay your rent?” he wondered; “It is a question that still hangs over the entire DIY movement, not just DIYcity.” As Geraci struggled to find a business model for the project, its early energy was dissipating. Local groups that had formed on the DIYcity site began carrying on their discussions in other forums. “People didn’t see a need to stay united,” he concluded. Geraci returned to the start-up world. For him DIYcity “lived out its natural cycle. It didn’t outlive its usefulness.”

But DIYcity did live long enough to become an inspiration, catalyst, and blueprint for organizing civic hacking groups for years to come. It was a People’s Computer Company for a generation weaned not on PCs but social media, mobile computers, and open data. It’s no coincidence that present in the crowd at that sole DIYcity meet-up in Manhattan was a cadre of civic hackers who would go on to shape the grassroots smart-city movement: Crowley and Selvadurai launched Foursquare a few months later; Nick Grossman and Philip Ashlock of Open Plans would write open-source software for online 311 systems as well as start Civic Commons, a repository for open-source cityware; Nate Gilbertson, a policy advisor to the director of the Metropolitan Transit Agency, would push an open-data initiative through a creaking bureaucracy; and his colleague Sarah Kaufman would see it through.

As Geraci described it, “DIYcity was a totally bottom up organization ... there was nobody giving orders ... it was driven by people showing up, looking at what needed to be done, and doing it.” Like ITP, “it was loose and collaborative and open and that’s what made it work.” What Geraci provided was a lens to focus their energy and a well-crafted moniker under which to carry it forward.

Sociability: The Smart City’s Killer App

“Use the Internet to get off the Internet,” commanded the new marketing slogan for Meetup.com in 2011. Launched in 2002, Meetup was an early pioneer of the hybrid social networks that are commonplace today, bridging online and offline lives to help people congregate face-to-face around shared interests and hobbies. In less than ten years, more than 10 million people had joined over a hundred thousand Meetup groups all over the world. To mark the accomplishment, founder Scott Heiferman reminisced, “I was the kind of person who thought local community doesn’t matter much if we’ve got the internet and TV. The only time I thought about my neighbors was when I hoped they wouldn’t bother me. When the towers fell [on September 11, 2001], I found myself talking to more neighbors in the days after 9/11 than ever