Sensing the City

It's called “the nursery.” A half-dozen white-domed machines lie on racks in a high-ceilinged room at Argonne National Laboratory, about half an hour west of the city. With the low, winter afternoon sun hitting them just right, it’s not a stretch to imagine them as eggs warming in an incubator.

These little plastic nodes are packed with sensors and backed by millions in federal funding. Eventually, the microwave-sized devices will make their way out to lampposts in Chicago or Detroit or Denver or beyond to quietly measure the world around them. They’ll look for traffic patterns, and they’ll measure sound. They’ll count particles in the air and note the amount of carbon monoxide, sulfur dioxide, and other pollutants present. They’ll measure vibration, magnetic fields, and light. And if all goes according to plan, they’ll send this information back to a database where scientists, city officials, hacktivists, and residents will be able to access and analyze the streams of hyperlocal data.

This is the vision of the Array of Things (AoT), a joint initiative between Argonne, a U.S. Department of Energy laboratory operated by a subsidiary of the University of Chicago, the University of Chicago, the City of Chicago, and various technology firms. The project expects to start publishing data from its preliminary nodes to the city’s open-data portal earlier this year, at which point they hope to have a hundred of them up around the city quietly quantifying the traffic, noise, and emissions that make city living unpleasant at least, and environmentally unjust at worst.

AoT is the brainchild of Charlie Catlett, the project lead and a senior computer scientist at Argonne. About six years ago, Catlett began exploring how computer science might help address the challenges of mass urbanization. He looked at geographic variation in life-expectancy data and other metrics, and became alarmed by how much a person’s address dictated their chances of suffering asthma or lead poisoning.

“Depending on what neighborhood you’re in, your experience living in the city is dramatically different than a neighborhood that’s a mile away,” Catlett said. “This sort of bothered me.” He wanted to help “cities move in the direction of ‘better’—however you want to define ‘better’—that could mean less congestion, better air quality, equitable services and opportunities, etc.”

Charlie Catlett, AoT's project lead

In June 2012, Catlett began several years of conversations with scientists and city officials about what a functional city-wide sensor network might look like—what it should measure, and how it might be deployed. In 2015, the National Science Foundation awarded the University of Chicago a $3.1 million grant to support AoT’s development. The effort adopted the name “Array of Things”—a sort of portmanteau of “array telescope” and the “internet of things”—to suggest a distributed but unified, discrete but interconnected, collection of measuring instruments aimed at understanding life in the city.

AoT is an early foray into an urban design revolution based on the principle that massive amounts of new data can help identify and repair urban woes. It places Chicago at the epicenter of a global “Smart-City” revolution. Metropolises across the globe seek to adopt AoT or technology like it, hoping that a smarter, more connected, and more analyzed city means a cleaner, safer, and less-congested one.

Cities are entering “a fourth stage of modern transformational change,” according to a February 2016 White House report on urban technology. The first came about with the steam engine, the second with the power grid, and the third was made possible by reliable mass transit. Now, the report says, “new physical and digital technologies” offer innovative solutions to twenty-first-century problems.

“Information and communication technologies, the proliferation of sensors through the Internet of Things, and converging data standards are...combining to provide new possibilities for the physical management and the socioeconomic development of cities,” the report reads. “Local governments are looking to data and analytics technologies for insight and are creating pilot projects to test ways to improve their services.”

There’s a lot of chatter around the idea of the smart city, but it can sometimes feel reductive, as if to suggest that something as complex as urban poverty can be understood as a sequence of ones and zeroes. Chicago has a long history of failed utopian schemes, from the City Beautiful movement of the 1890s to 1920s to the high-rise public-housing projects of the mid-twentieth century. On its face, AoT stirs up some of the same old urban-design tensions—while introducing the uniquely twenty-first-century concerns of data security and digital privacy. It might explain why Catlett is a little wary of the “smart city” label.

“Our idea of a ‘smart city’ was more: let’s see if we can equip people who are making decisions and operating the city to do it better,” Catlett said. “When I think about the Array of Things and ‘smart cities,’ I don’t think about the overall design of a smart city so much as: we can just pick off one problem at a time and, at the end, we’re going to be improving the city. If we never become ‘smart,’ at least we’ll be smarter than we are now.”

In other words, the smart city is not a silver bullet, even if it’s sometimes billed as one. But if AoT works out as Catlett hopes it will, then it will open the door for other opportunities: a developer might use the data to build an app to warn asthma sufferers of nearby irritants; local activists might use air-particle levels to make the case for restrictions on diesel traffic. Or the city—armed with reams of new data—might better understand how a bus route change downtown affects traffic flow.

“Once you have a number of these deployed across the city, you begin to measure it in a way it hasn’t been measured before,” Catlett said.

✶ ✶ ✶ ✶

If you stand on the corner of Damen Avenue and Cermak Road and squint up at the lamppost above the gas station, you can make out one of the twelve AoT nodes already deployed across Chicago. The Damen/Cermak node is part of a subset of nodes focused on “Urban Air Quality and Health Education,” and it’s pointed at an area that has a history of fighting for a cleaner environment.

Pilsen and Little Village have long been undergoing the slow process of deindustrialization. With it come the pains of job loss and a long-overdue reckoning with industry’s toxic legacy. The blocks surrounding the Damen/Cermak node, in other words, are an excellent place to test one of AoT’s core theses—namely, that when citizens and officials better understand their environment, they are better able to improve it.

In 2012, the nearby Fisk and Crawford coal power plants closed down after years of protests and community organizing by the Little Village Environmental Justice Organization (LVEJO), the Pilsen Environmental Rights and Reform Organization (PERRO), Pilsen Alliance, and others. The closings may have had as much to do with broader market forces and federal environmental regulations as they did with local activism, but they were considered an enormous environmental win nonetheless. Organizers leaned heavily on a 2002 Harvard School of Public Health study that literally put a number on how many deaths, emergency-room visits, and asthma attacks the two plants caused each year.

"We inherently knew something was wrong ... but we couldn't prove anything," says Kim Wasserman, LVEJO's executive director. "It was finally getting those numbers that we were able to tell the story of how we were being impacted, outside of: 'My kid has asthma.' And it was those numbers that moved people to want to take action. There's nothing like having 40 body bags in the hallway of City Hall to really accentuate how we were being impacted."

Jasmine Mithani

The AoT nodes themselves won’t go as far as to draw a link between poor air quality and asthma attacks or fatalities. In these early stages, the project-runners are mostly concerned with making sure the citywide sensor network—potentially the first of its kind—really works. But the data they produce could empower more scientists and activists to make those kinds of analyses. If successful, AoT might multiply exponentially that quantitative awareness, giving neighborhoods access to all kinds of data that could confirm or deny suspicions about the air they breathe. Instead of just a dozen or so air monitors scattered sporadically across the city, a projected 500 AoT nodes will be logging air-quality data in neighborhoods that have historically suffered from high levels of asthma but have lacked the mechanisms for connecting it to specific environmental factors.

“Data is power,” Wasserman-Nieto said. “When you are putting out things like the Array of Things you are crossing into very unknown territory because what’s going to happen when those results come out? Who is going to take responsibility for what is there?”

Wasserman-Nieto and LVEJO—along with other community groups on the South Side—have already taken the matter into their own hands. LVEJO is part of Shared Air, Shared Action, a 2016 grant from the Federal Environmental Protection Agency to explore how community groups and local residents might use increasingly affordable sensor technology to do their own air monitoring. For LVEJO, the current concern is diesel pollution, with high truck traffic in the area and proposals for new diesel-heavy industrial facilities.

Volunteers for Shared Air/Shared Action canvass the South Loop

Shared Air, Shared Action volunteers walk predetermined routes in their neighborhoods with boxy plastic air monitors strapped to their chests. They look for and note anything that might add context to a sudden aberration in data—the passing of a smog-belching truck, say, or a particularly odorous barbecue nearby. They send their observations and the data back to a centralized database for scientists at the University of Illinois at Chicago and Kansas State University to analyze. It’s all a bit like traditional, shoe-leather neighborhood canvassing, except that instead of drumming up votes, these engaged pedestrians are painting a portrait of their neighborhood’s environmental health.

For Wasserman-Nieto, getting community members directly involved like this creates a bottom-up approach to air monitoring in which those with the greatest human stake in the results are present in formulating and performing the experiment at hand. It’s also, Wasserman-Nieto notes, an opportunity to introduce questions of science, technology, and engineering to communities that have traditionally lacked access to STEM education.

“A lot of times universities will make assumptions about the problem—and a lot of times they’re right—but [they] will dictate how to solve that problem,” she said. “If you want true partnership with community, then the community leads the request. The community leads the process. The community is the principal investigator.”

✶ ✶ ✶ ✶

If the implicit benefit of cheap sensors and big data is greater knowledge, then the new technology’s inherent risk is to information security and personal privacy. Every day reams of information about our lives are beamed across networks in exchange for access or convenience. Not a month goes by without some new security breach or hack that briefly makes us reconsider our faith in the gospel of the Cloud.

It’s no wonder, then, that AoT has raised the eyebrows of privacy critics. After all, its aim, in the superficial sense, is to install hundreds of artificial ears, eyes and noses throughout public spaces. Initially, AoT had planned to include in the node a mechanism for counting nearby Bluetooth devices as a way to track pedestrian flow. When the Tribune published an article on it in 2014, it prompted concern over exactly what data was being collected and who would control it. The feature was ultimately dropped, and AoT eventually published an Operating Policies document outlining their approach to privacy, but the genie was already out of the bottle. Chicagoans had gotten a whiff of Big Brother, and it amplified a skepticism many already hold toward city government.

The final node design moved forward instead with a camera to count pedestrians and cars, and to detect street flooding. To guard against privacy intrusions, AoT relies on edge computing, in which images are processed and analyzed right at the point of collection. In other words, according to AoT, the node extracts whatever quantitative information it needs, destroys the image, and just sends the information (e.g. number of cars) back to Argonne’s database. For calibration purposes, less than one percent of images will randomly be stored. AoT insists they will not contain personally identifiable information, but they will be controlled and protected as if they do.

It begs a uniquely digital thought experiment: if a camera records a person in the city, but only a computer sees the image, is it a violation of personal privacy? Should Chicagoans have the ability to opt out of such collection, even if it is ultimately anonymous and limited to public spaces?