Cities are getting “smarter.” But are they getting wiser?

By Brian Barth

“Oh, no. My phone is dead. Better head to the park.”

Walk past the basketball court down at Anita Stroud Park, toward the little creek below, and you might find a gaggle of teens clustered around a very modern-looking bench that would seem more at home outside a coffee shop in Soho than in a tiny neighborhood park next to I-77 on the north end of Charlotte, North Carolina.

A pair of USB ports on a console on the front of the bench provides juice from the solar panel mounted at lap level between the seats. Who wouldn’t want to hang out at a bench like this? It certainly catches the eye of passersby. What these kids might not realize, however, is that this bench is watching them back. Underneath that solar panel is a small Wi-Fi enabled sensor that sends data back to an office building in East Cambridge, Massachusetts. Anyone who passes within 150 feet of the bench with a Wi-Fi enabled mobile device in their pocket is picked up by the sensor and registered as a unique visitor to the park. The sensors can’t access personal information from your phone—rather, they’re designed to pick up the unique ID associated with any Wi-Fi enabled device—but still, if you come back the next day, it knows it’s you again, not a new visitor. It may make privacy advocates squirm, but such data is very handy for park planners.

“The idea that we can learn about how many people are using the space, when they are there, and how long they are there, without having to literally send someone out there to count people, is very valuable,” says Monica Carney Holmes, the planning coordinator for Charlotte’s urban design office.

Since bolting the bench into place last October, Holmes’s office has learned that 85 percent of visitors to Anita Stroud are repeat visitors and that Tuesday, Wednesday, and Thursday nights are the most hopping. Outdoor Zumba and Tai Chi classes were scheduled for Saturday mornings last summer, Holmes says, but this summer they’ll program more events for weeknights to capitalize on the higher traffic.

Melissa Gaston, the president of the homeowners association of the Park at Oaklawn, an affordable housing community adjacent to Anita Stroud Park, has used the bench and seen its data—“I think it’s cool,” she says—but acknowledges that some residents are wary. “People think it’s Big Brother watching them, a 1984, George Orwell sort of thing. The city says [the bench] doesn’t identify individual people, though I’m sure they could use it for that at some point. So that has given some residents pause.”

The bench in Anita Stroud Park is manufactured by Soofa, a company founded in 2014 by three graduates of the MIT Media Lab. It is the first tangible sign of a plan to revitalize an area of Charlotte branded the North End Smart District, which includes Brightwalk, the Park at Oaklawn, and six other adjacent neighborhoods. The 3.6-square-mile district houses approximately 9,000 residents—primarily low-income and predominantly African American—clustered around a dilapidated warehouse district between the University of North Carolina campus and the downtown core, and it is facing heavy development pressure, especially now that the city’s light rail network is being extended into the area.

The intent is to redevelop the area with the sort of cutting-edge digital infrastructure that will attract new employers to set up shop, particularly IT companies and other enterprises that tend to draw on a young, upwardly mobile, tech-savvy workforce. Exactly what technological infrastructure will end up in the North End Smart District is an open question at this early stage, but these days most cities of any size are touting “smart city” technology, and this is Charlotte’s pilot effort. This term of art encompasses things like gigabit public Wi-Fi, parking apps instead of parking meters, streetlights that dim to save energy when no one is around, automated buildings that heat and cool each space only when in use, and trash and recycling bins that alert sanitation personnel when full.

Data-collecting street furniture (Soofa isn’t the only company making benches with sensors) is one of the first areas where the smart city concept has spilled over from buildings and infrastructure into green space, though one could argue that features such as free Wi-Fi in parks, GPS-guided interpretive walks, interactive light and water installations, and computerized controls of lighting and irrigation systems—increasingly common in the urban environment—also fit the bill.

Beyond benches, urban data collection is advancing rapidly. The Center for Urban Science + Progress (CUSP) at New York University, established in 2012 as one of the few research centers in the world with degree-granting programs in urban informatics, is home to the Sounds of New York City (SONYC) project, an initiative to map noise pollution—and ultimately to inform policies and design strategies to reduce it—through a network of sensors, augmented by information on noise conditions provided by citizens through an app. At the Urban Observatory, CUSP’s research facility in Brooklyn, specialized cameras and remote sensing equipment are mounted on top of the building to collect data on things like the heat island effect and sources of emissions that are invisible to the naked eye. A hyperspectral camera pointed toward the skyline around Prospect Park allows researchers to view the infrared light emitted by plants to analyze the health of tree plantings in different areas (plant stress is perceptible in the infrared spectrum long before it is visible to the human eye). This information is then cross-referenced with highly localized weather and air quality data to look for cause-and-effect relationships.

Perhaps the most ambitious urban informatics project in the country is in Chicago, where the city has begun the installation of 500 nodes, each housing 16 different sensors, plus a microphone and a camera, on utility poles inside what looks like a rocket-powered backpack from The Jetsons. Known as the Array of Things, a joint initiative between the City of Chicago, Argonne National Laboratory, and the University of Chicago, the sensors detect basic weather and air quality information, along with magnetic fields (useful for assessing traffic flow) and physical vibrations (from machinery and heavy trucks), among other measurements. The camera is connected to software that can parse vehicular, bike, and pedestrian traffic, giving an accurate count for each location. An early concept for the project included a sensor to detect wireless devices, similar to the Soofa benches, though the idea was abandoned owing to public opposition.

Some 42 nodes have been installed in Chicago so far, with the full rollout scheduled for completion by the end of 2018. A small number of nodes will soon be delivered to nine other participating cities in the United States, Mexico, and Europe as well. (The first data sets will be available at www.plenar.io, a global hub for open data, this summer.)

In Charlotte, Holmes says the city is planning more benches like the one in Anita Stroud Park—at bus stops, perhaps, and in civic plazas—as the North End Smart District is developed. By amassing reams of data on urban life, the theory goes, future planning and design efforts should be more accountable to conditions on the ground.

“The bench is part of a much larger conversation,” says Ed Krafcik, a former landscape architect who now works for Soofa. “By embedding technology into public infrastructure, a city is better able to respond to the evolving needs of citizens. Companies like IBM and Cisco have been talking about this for years, but the technology comes full circle when the data gathered is actually used by a parks department, an economic development agency, a planner, or a designer in a way that responds to how citizens actually use the city.”

Although the upshots of the so-called smart city movement are fairly clear—better data equals better and more transparent decision making—some people believe its potential pitfalls merit greater attention. Chris Speed, the chair of the design informatics program at the University of Edinburgh in Scotland, says a number of issues stem from the fact that “most of the smart city agenda is top down, and based on economics rather than social values.”

Speed is concerned that the smart city movement may contribute to the privatization of the public realm, rather than its democratization, citing the recent installation of a Nintendo gaming station in a prominent public plaza in Edinburgh as an example. By providing a venue for the company to promote its latest product, the city was able to extract revenue out of a space that ordinarily only consumes taxpayer dollars. But hanging a corporate logo over public space always brings up questions about whose interests are being represented. “Does the public even realize what sort of information is coming out of sensors in parks or lampposts, or from a video game kiosk?” Speed asks. “Where is the opportunity for their input?”

One roadblock to such input is the public’s general lack of understanding about mass-scale data collection and its impact on matters that are important to them. Most people have at least a vague notion that using the Internet means surrendering personal information to the Googles and Amazons of the world—that’s what all those legal disclosures that most of us click our agreement to without actually reading are about, right?—but at least we are asked for our consent, even if we have only a foggy idea of what we’re agreeing to. We just hope that the government is somehow keeping these corporations in check, and not collaborating with them to snoop on citizens.

Transfer the Internet privacy debate from cyberspace to public space, however, and another set of ethical implications arises. For example, there are no signs marking a 150-foot perimeter around Soofa benches warning pedestrians to turn back if they don’t want their devices registered. There is no way to opt out of being photographed by the Array of Things or recorded by SONYC microphones. Even if one is to trust the individuals and organizations behind such projects, what happens down the road when they are acquired by a multinational tech corporation, which quietly changes the wording of the privacy policy to enable other, more lucrative types of data collection from the sensors? As with any sexy new technology, the potential for mission drift is hard to forecast amid the glare of the initial hubris.

According to Soofa’s privacy policy, its sensors “observe the information being sent by mobile devices, including the device’s MAC address [media access control address, a unique identifier used to communicate with a wireless network, similar to an IP address], manufacturer, and signal strength. After receiving the nonidentifiable data sent by mobile devices, Soofa applies a cryptographic function to the MAC addresses to further anonymize them. Soofa analyzes the data it observes and provides aggregate anonymous information to its customers.” The privacy policy does not address whether Soofa might sell the data, stating only that it may “share data with research organizations both for and not for profit…to advance our research platform.” Krafcik maintains that selling data is not part of Soofa’s business model, though it’s not hard to imagine that other companies in the growing industry of urban data providers would want to engage in such commerce.

Digital privacy experts have long held that anonymous data, such as a cell phone’s identification number, is not as anonymous as one might think, even if encrypted. Timothy Yim, the director of data and privacy at the San Francisco-based nonprofit Startup Policy Lab, says “de-identification,” the process of stripping “personally identifiable information” (or “PII” in techspeak) from a data set does not necessarily prevent “reidentification” at a later date. “It is very hard to guarantee that any de-identification process is 100 percent foolproof,” he says. “And the more data that we have sitting in private repositories, like data aggregators and data brokers, the greater the likelihood that reidentification is possible.” Few people realize, he adds, that cell phone-detecting sensors are widely deployed by owners of large retail environments to analyze pedestrian activity. Spatial DNA, a Toronto-based firm that provides those services, is now using its PeopleFlow system to track consumer activity not just in malls, airports, and sports venues, but on public streetscapes as well.

Yim consults with local governments on the privacy implications of drone use and other smart city initiatives, and has followed the progress of the Array of Things closely. While the project has been well received in many circles, he says it has met with stiff pushback from some community members and data privacy experts. During the public comment period for the project, concerns were voiced about the surveillance capabilities of the units, including how future changes to the software and privacy policy would be evaluated, who will have access to the sound recordings and images collected by the units, and what would happen if a subpoena were issued for access to images and recordings in the course of a criminal trial.

Responses to 80 questions from the public were posted on the project website, including to the question of how law enforcement requests would be handled, on which the project team demurred, saying only that “the University of Chicago, as copyright holder of the data, [will] be responsible for responding to law enforcement requests.”

Yim counsels his municipal clients to err on the side of extreme transparency when rolling out data collection initiatives, and to draft privacy policies with exhaustive detail, imagining every current and future scenario that might result in a breach of public trust. The Array of Things’s initial privacy policy was troublingly vague and thin on detail, he says, though it was beefed up considerably following a string of negative news stories—“Array of Things Sensor Policy Leaves Law-Enforcement Question Open,” exclaimed the Chicago Tribune—that chronicled a bout of public backlash last year.

But he applauds the project for one privacy practice that he says could have been better publicized in the initial rollout to help allay the public’s Big Brother fears: raw images and recordings are processed within each node, so that only numerical data—such as ambient noise levels, pedestrian and traffic counts, and the presence of standing water—is transmitted to the project server (the Array of Things privacy policy states that raw image and audio data will occasionally be collected to improve instrument calibration, but will only be reviewed by select personnel under “strict confidentiality obligations”).

“It is incredibly difficult to implement traditional privacy concepts like notice and consent in the context of drones or sensors in public space,” Yim says. Planners and landscape architects, he adds, can play a role in helping communities define “the balance between the benefits and the potential privacy risks, as well as how to mitigate some of that risk.”

The line between collecting data for a valid public purpose and the unreasonable surveillance of private citizens can be tough to tease out. Beyond clear dangers like hacking and data breaches, and underlying concerns about private corporations somehow benefiting from data collected on the taxpayer’s dime, are existential questions about privacy as a basic human right. Helen Nissenbaum, a fellow at NYU’s Information Law Institute and a privacy rights activist, says she is “horrified” at the lack of government oversight regarding the so-called Internet of Things as it expands into public space.

“The notion of public space where people can walk around and not be surveilled, not be held accountable for what they do, is really a diminishing idea—unless we pursue the idea of smart cities with our eyes open and a great understanding of what we’re doing,” Nissenbaum says. “We need better policies that control how data flows from one party to another and who can get access to what data under what conditions. I personally believe that if someone has a bunch of information about me, it’s like the sword of Damocles—they have power over you.”

According to Yim, very few laws, at any level of government, address what sort of data may be collected in public space, though he says this is likely to change as the legal system catches up with the technology. Reform at the political level notwithstanding, both he and Nissenbaum encourage anyone in the position of recommending or procuring such data collection systems to delve into the ways the systems work and what the ramifications of their use are. Cell phones, for example, transmit several different signals, which have different security implications, says Nissenbaum, who suggests a few basic questions to ask: What happens to the data once it is registered? How long is the information going to be stored? Will it ever be shared with anybody else? Can the purchaser of the system conduct an audit to make sure it is working as the company claims it is?

“People say things like, ‘I don’t care; I’ve got nothing to hide,’” Nissenbaum says. “But privacy is not about having some guilty secret that you don’t want other people to know. If people understood how the sharing, aggregating, and analysis of data works, it would throw a lot of those assumptions out the window. Hopefully it won’t be too late by the time we as a society realize this.”

Krafcik is well aware of the skepticism regarding “Big Data” and what corporations and governments might be doing with it. But he argues that Soofa benches are potentially a means to build public trust around urban data collection. “The goal is to humanize the Internet of Things in a civic context. When sensors are hidden in light poles and electrical boxes and other places that are not tangible for citizens, it can create a contentious relationship with the technology. Rather than thinking it’s all about creating a better experience for them in the city they live in, it’s like, why is the city spying on me? Taxpayers are financing a lot of these innovations, so it becomes hard to advance if it’s not in a form they can connect with.”

Whatever the merits of Soofa’s approach, it has drawn interest from scores of municipalities, along with a few technology giants. Verizon and Cisco signed on as partners for Soofa’s first pilot project, a dozen benches scattered about Boston. The benches have now been installed in 65 cities and five countries, and the company’s second product launch—solar-powered, Wi-Fi enabled E Ink signs (the same technology used in a Kindle) used for wayfinding and public service announcements—is under way in Boston. The signs, which do not include sensors at this time, are used to communicate everything from transit updates and job postings from nearby businesses to notices for town hall meetings and Boston Mayor Marty Walsh’s Twitter feed.

In Charlotte, the question of who benefits from smart city technology takes a social equity cast. Rachel Stark, ASLA, a landscape architect in the city’s urban design office who is spearheading community planning efforts for the North End Smart District, says it was quickly apparent that the latest whizbang technology is not exactly a high priority for many residents in this part of the city, where the median income is around $25,000, one in six properties is vacant, the violent crime rate is four times the Charlotte average, 28 percent of residents are unemployed, and 27 percent of households lack a car. The appeal of a phone-charging bench has its limits in a community where not everyone has a smartphone in their pocket.

“Is this going to be helpful, or is it just going to raise property values and drive people out?” Stark asks. “The most important question at this point is, how can we be smart about our community engagement? Language is so important. As landscape architects, it is very important to understand that, yeah, we may be the educated professional, but that doesn’t make us better. It makes the situation worse if we don’t actually communicate in a way that the people we’re talking to can understand. So we’ve come to local community groups since the beginning to find out the best ways to get their constituents engaged in the discussion.”

Melissa Gaston and her husband, Darryl Gaston, were among the first community members consulted. “I embrace the smart city concept, but I cautioned the folks from the city that because of things we have experienced historically in Charlotte, residents here are a little apprehensive about the government coming in and telling us what they want to happen,” says Darryl, a third-generation resident of Druid Hills—one of the eight neighborhoods within the North End Smart District—and the president of its neighborhood association (the couple, who met at a community organizing workshop, maintain residences in their respective neighborhoods). “We are already experiencing all sorts of growth pains—predators knocking on our doors, saying, ‘I know that your taxes are in arrears, but if you sell me your house you don’t have to worry about the city taking it.’”

Based on this feedback, city staff is taking an incremental, bottom-up approach with the Smart District plan. The installation of the Soofa bench last year was part of the No Barriers Project, a prelude to launching the Smart District that entailed a six-month series of events in Anita Stroud Park put on by the tech-centered design firm IDEO, the Parsons School of Design, and the Mecklenburg County Park and Recreation Department. Activities included whimsical evening events, like playing with “light boxes”—lightweight chair-size cubes with an illuminated dry erase surface for people to write out their thoughts about the community (or whatever was on their mind)—along with barbecues and dance parties.

Stark says the social capital generated by the No Barriers Project is now feeding into a series of community input sessions about how the Smart District plan could support the immediate needs of residents. One idea that’s arisen is to subsidize weatherization improvements and smart thermostats for the aging housing stock in the area. Another suggestion, which came up in discussions about impediments to mobility, is for the city to create a sort of banking app for residents who have a smartphone but not a bank account or credit cards. Uber and Lyft don’t accept cash, so the app would allow users to load money onto their account at a convenience store, for example, and then pay for ride share services or for municipal transit using a phone.

One Smart District project already under way is a workforce development program to train adults in technology skills to help bridge the digital divide, Stark says. A program for high school students will provide two weeks of training in a field of technology, followed by a four-week internship working in that particular industry. “The idea is that by having the training and providing localized jobs, people don’t have to travel as far, their incomes are hopefully a little higher, and they have a few more options in life,” she says.

The Gastons plan on holding them to it. They recently formed the North End Community Coalition to unite the various community groups in the area against the pitfalls of gentrification and ensure that the upcoming investments in technology are harnessed for the interests of current residents. Consensus in the community is in support of the Smart District plan, says Melissa Gaston, as long as it continues to be on the community’s terms. “There are some things that the city has come to us and said they wanted to do and we said, ‘No, we are not on board with that.’ But they’ve addressed some of our initial concerns. We’ve been left out of a lot of good things going on in the city in the past, so I recognize the value of being part of this.”

Brian Barth is a Toronto-based journalist with a background in urban planning and landscape design.