Skip to content

A Mess on the Sidewalk

Inside the great digital mis-engineering of the Toronto waterfront

When Sidewalk Labs, a “smart cities” start-up launched by Google, made its Toronto debut in October 2017, a torrent of global adulation followed. Over the next few weeks, Sidewalk choreographed events featuring luminaries such as former Google chairman Eric Schmidt and Sidewalk founder and CEO Dan Doctoroff. There was even a cameo appearance from Canada’s hipster prime minister Justin Trudeau, whose Liberal government is investing heavily to attract innovation-driven companies to Canada.

For Toronto, which anchors a fast-growing metropolitan region that has become a hub for information technology companies, Google’s heavily hyped expression of interest showcased the city’s aspirations and promise. What that promise is, exactly, remains difficult to discern. Sidewalk Toronto’s pitch was to build a high-tech community on a mostly vacant twelve-acre patch of industrial waterfront site called Quayside, just east of the city’s core. The company’s initial proposal hinted at the urban revolution Google aspires to create: “a neighbourhood built from the internet up,” as Sidewalk Labs likes to put it. Quayside would be fitted out with affordable housing, green buildings, and Google’s Canadian headquarters. The waterfront locale would also become a test bed for high-tech infrastructure, such as automated mini-buses and self-navigating underground delivery carts. The new district will eventually provide homes and workplaces for thousands of people.

But Sidewalk also had its eye on a larger prize—a mostly derelict brownfield precinct just east of Quayside, known as “the Portlands,” an area that’s roughly the size of Central Park in Manhattan. The Portlands was built more than a century ago from landfill atop a fetid swamp at the mouth of a flood-prone river known as the Don. Home to works yards, contaminated industrial (and post-industrial) sites, and a clutch of fortress-like film studios, the Portlands’ scrubby real estate is worth untold billions but requires extensive remediation and costly flood-plain protection. Eventually, it would seem, Sidewalk wants to build an entirely new district, akin to London’s finance-and-culture hub, Canary Wharf, but fitted out with all manner of cutting-edge technology.

Doctoroff, New York’s former deputy mayor and later the executive who managed Michael Bloomberg’s information empire, assured Torontonians they’d be seeing early results with an ambitious timetable. He even had a pot sweetener: a $50 million (U.S.) kitty for joint planning and pilot projects, which Sidewalk would fund regardless of whether local authorities would ultimately sign off on the Sidewalk Labs project. While $50 million is a robust revenue stream for any Canadian city, it’s worth remembering that the sum is little more than a rounding error in Google’s world: Alphabet, the parent company of Google and Sidewalk, reported cash reserves alone in excess of $100 billion in 2018.

Toronto is no stranger to innovation, of both the urban-deal-making and digital- branding varieties. Yet Canada’s great capital of commerce and culture is still, well, very Canadian—which is to say, temperamentally averse to much of the fevered Silicon Valley boosterism that engulfs big-money tech launches like Sidewalk Labs.

Still, the seduction of a massive infusion of cash and digital technology to propel an industrial moonscape to the vanguard of digital urbanism has been intoxicating for the city’s political elite, though it may seem like an affront to basic urban virtues. Cities, after all, shouldn’t be reducible to a set of layered functional systems to be de-bugged, monitored, and optimized. But in the global race for jobs, investment, and the next big technological revolution, it may be that such schemes express the tradeoffs that come with any kind of twenty-first-century urban redevelopment plan.

They Cover the Waterfront

A mere year and half after the launch of Sidewalk Labs, the implications of those expedient tradeoffs are coming into clearer focus, and it’s a bracing picture. As a result, Sidewalk’s ambitious scheme for Toronto’s waterfront areas has devolved into a tangled mess.

From the beginning, the firm’s ambiguous goals, its patronizing style, and its penchant for secrecy triggered a wide public backlash. Among its most energetic critics is Jim Balsillie, the Waterloo, Ontario, tech entrepreneur who co-founded Blackberry and views Sidewalk’s gambit as an effort to commercialize Google artificial intelligence technology with little benefit for Canadian firms. Some fear Sidewalk’s stated desire to end-run municipal land-use planning authority. Others question whether Toronto was wise to offer up its valuable waterfront for an early-adopter-style pilot project. The Atlantic dubbed the scheme “Google’s guinea-pig city”—and a stint in a laboratory rarely ends well for guinea pigs.

Meantime, Waterfront Toronto (WT)—the government agency that invited Sidewalk to bid on an opportunity to establish a beachhead in Toronto and retains the authority to approve or reject the company’s “Master Development and Innovation Plan” later this year—has become mired in an existential crisis sparked in no small part by its early mishandling of the Sidewalk relationship.

Facing mounting criticism about its inability to adequately scrutinize a tech plan concocted by one of the world’s most powerful data companies, WT officials last spring set up an independent “digital strategy advisory panel”—made up mostly of law, intellectual property, and planning professors, as well as IT industry insiders—to pose hard questions about Sidewalk’s plans, especially those involving privacy, accountability, and ownership of the data. Despite the introduction of nascent oversight protocols, the list of key players involved in the whole messy process who have resigned, been fired, or otherwise opted out grows by the month.

The seduction of a massive infusion of cash and digital technology to propel an industrial moonscape to the vanguard of digital urbanism has been intoxicating for Toronto’s political elite.

The reasons behind the skepticism aren’t hard to fathom. Sidewalk’s plan turns on an extensive deployment of networked digital sensors embedded in the public realm—including civic buildings, municipal infrastructure, and traffic systems. (Google has made huge investments in automated vehicle technology through its Waymo subsidiary.) A growing chorus of tech and privacy experts warn that such surveillance-like devices must be rigorously monitored. But no one really knows exactly what “data governance” entails in an era defined by Facebook’s fake news crisis, the insidious manipulations of behavioral profiles, and consumer-cum-political algorithms coming under the fearsome reach of monopolies like Google and Amazon.

“While everyone is flailing around and trying to keep a straight face that Waterfront Toronto and its digital strategy advisory panel is in control here, along with all levels of government, this is fundamentally untrue and structurally impossible,” Bianca Wylie, an open government advocate who has been one of the deal’s most trenchant critics, cautioned in a November column in Medium.

Certainly, Sidewalk has become a screen upon which some Torontonians have projected a wide range of fears about the seemingly boundless expansion of tech’s reach into their private and civic lives. Yet the idea embedded at the heart of Sidewalk’s pitch invites serious reflection about the complicated relationship between digital technology and urban spaces.

City on the Make

In Sidewalk’s vision of the future, the twenty-first-century city—and particularly its infrastructure—can theoretically be “optimized” to improve quality of life. Sidewalk is hardly alone in this conviction. In recent years, dozens of tech companies, including IBM, Cisco, Samsung, and Siemens, have raced to sell hundreds of billions of dollars in smart-city-branded hardware and software to local and regional governments around the world.

The stakes, clearly, are huge. “The intent is to create a super-efficient city using real-time data,” says Lucie Greene, the author of the 2018 book Silicon States: The Power and Politics of Big Tech and What It Means for Our Future. Among major concerns, she cites the potential to leverage data culled from infrastructure and the scramble to control or profit from all that information.

But unlike conventional merchants of smart-city wares, Sidewalk sees data, rather than equipment, as its stock in trade. Using AI algorithms tethered to a vast array of sensors that effectively listen to cities going about their daily routines, Sidewalk proposes to create a “platform” that dynamically adjusts civic systems—everything from energy grids and waste management to road allowances, street furniture, and parking spaces—as urban conditions dictate.

Along the way, the company claims, city-dwellers, municipalities, and enterprising tech firms will all have an opportunity to make use of bits and pieces of this immense trove of semi-digested data for applications ranging from the prosaic (e.g., apps for reserving picnic tables or finding parking spaces) to the futuristic (self-driving mini-buses plying routes that vary depending on who’s waiting, where).

Under intense pressure from Canadian privacy advocates, Sidewalk has promised that all the sensor information it gathers will be “de-identified.” The company has also said that future commercial value lies not in the measurable details of any individual’s urban experiences but rather in aggregated but anonymous data tracking thousands of individuals living in or moving through this new district. These storehouses of data will furnish the raw material for algorithms designed to process various categories of urban signals, from the movement of vehicles along city streets to the perambulations of pedestrians along sidewalks, in public parks, and through retail zones.

Yet it doesn’t take much effort to conjure up ways that such modes of behavioral surveillance can lend themselves to ethical or legal lapses in domains such as law enforcement or the regulation of activity in public spaces. Another Google AI subsidiary, a health data firm called DeepMind Health, set off controversy in 2017 after its data-sharing agreement with the National Health Service was found to be in breach of the Data Protection Act. (Sidewalk’s initial proposal to Waterfront Toronto referenced community health clinics that could use data in novel ways but didn’t mention the work done by DeepMind.) “The potential of these technologies scares the hell out of some people and, justifiably so,” remarks Zack Taylor, a Western University political scientist who studies cities.

Cities, by virtue of their very complexity, can be particularly vulnerable to the law of unintended consequences. And this core feature of the urban experience likely explains why Sidewalk’s breezy promises for a better urban future have produced such blowback. As Ryerson University planning professor Pamela Robinson, a member of the digital strategy advisory panel, points out, “The technology in the abstract isn’t the same as the technology in a real place.”

Laboratories of Technocracy

From the earliest days of the industrial revolution, the chaotic state of fast-growing cities has prompted reformers to conjure political or technocratic solutions to the dilemmas of city life. Some, like Ebenezer Howard’s Garden City movement, came in the guise of land-use planning initiatives that would place a buffer between families and noxious industry. Others were packaged as architectural ideologies, such as Le Corbusier’s belief that “towers in the park” and mega-blocks would resolve the distasteful crowding and disorder of urban slums. Post-war cities and contemporary planning reflect both of these ideals, with varying success.

Cities at various points also become the staging ground for new technologies meant to confront a range of persistent urban ills. Some produced game-changing benefits. The late nineteenth century, for example, saw civil engineering solutions, like water treatment and networks of septic sewers, that sought to remedy chronic social problems, such as infectious disease outbreaks in crowded slums. But the marriage of science and cities hasn’t always ended well. A handful of late nineteenth-century technologies—elevators, steel-frame construction, rapid transit, and the combustion engine—produced dramatic urban changes. Some delivered immense gains as they scaled up, to use the parlance of the tech industry; others, however, begat severe side-effects, such as unchecked sprawl.

Cities, by virtue of their very complexity, can be particularly vulnerable to the law of unintended consequences.

It’s far from knee-jerk Luddism to note that no one can really know how new technologies will work once introduced into the world. “Everything,” writes Greene, “starts innocently, and positively.” The mid-century fantasies of bucolic, traffic-free parkways and drives in the countryside yielded rapidly to chronic air pollution and congestion. Amazon, a generation ago, introduced an easier way to order books. Today, the torrent of technology that began with e-commerce threatens to eradicate important social and civic institutions: main streets and retail markets.

Ken Greenberg, a highly respected Toronto urban planner, counted himself among those anxious about the “uncritical embrace” of technologies that have distorted the social lives of cities. The American-born Greenberg belongs to a generation of reform-minded planners who forged careers in Toronto in the heady days of the early 1970s, just after Jane Jacobs moved to the city and set to work halting downtown expressways and block-busting urban redevelopment schemes.

In 2015, he heard through a mutual acquaintance that Doctoroff was setting up Sidewalk with the encouragement of Google co-founder Larry Page. The company is also a giant landowner and developer, with extensive campuses and office hubs in cities like San Jose, New York, and London.

Greenberg reached out to Doctoroff’s team and soon found himself participating in a series of roundtables focused on the potential role that big data technologies, such as the “Internet of Things,” could play in improving urban quality of life. Those conversations, it’s worth noting, predated the scandals that have inflicted enormous reputational damage on the tech industry, especially those like Facebook’s secret deal to sell vast quantities of personal data to Cambridge Analytica, that have involved brazen privacy breaches.

“This was really exploratory,” Greenberg recalls of those early planning sessions. “It became apparent that a place was needed.” Some participants suggested Denver or Detroit. Others argued that Sidewalk should test its nascent ideas on a greenfield urban community that could be wired up from the get-go. Greenberg disagreed: “I always felt, and that is my contribution to the discussion, that it didn’t make sense to start with a big tabula rasa site.”

Walk This Way

At about the same time, Waterfront Toronto (WT) had to hire a new CEO to replace John Campbell, the genial commercial real estate executive who had run the agency since 2003. Jointly owned by the City of Toronto and the governments of Ontario and Canada, WT had been established in the early 2000s to steer the redevelopment of an industrial waterfront long mired in scandal, conflict, and inertia. By 2015, the project had made significant progress, building popular new public spaces and a growing portfolio of private sector projects (mostly mixed-use condos).

The new CEO, a planner named Will Fleissig, came from Denver via San Francisco. He evinced a distinctly American brand of enthusiasm that masked a conspicuously thin resume—stints as a municipal official and later a private sector planning consultant, but with no experience reporting to a high-profile board of directors. Perhaps reflecting his time in one of the world’s leading tech cities, Fleissig in early 2016 decided WT should place a big bet on smart-city technology by putting out a global call for proposals for an innovation-driven development plan for Quayside and, possibly, the Portlands.

As big cities go, Toronto can have the feel of a small town—at least so far as the metropolitan planning elite goes—so it didn’t take long for Greenberg’s conversations with Sidewalk to get back to WT, whose executives were also pondering some kind of tech-inflected partnership. The agency’s officials encouraged Sidewalk, as well as numerous other tech and development companies, to submit bids. A handful did, and WT officials eventually selected Sidewalk, even though its experience, two years from its founding, was as thin as Fleissig’s CV. The company was a major investor in the firm that built a network of public WiFi stations in New York, a project known as LinkNYC, and had helped cities competing for a $40 million smart-city grant from the federal government. But as of October 2017, Sidewalk had never developed anything, much less a twelve-acre brownfield site pre-zoned for 3.3 million square feet of density.

Formally, WT only granted Sidewalk the exclusive right to negotiate an innovation and development plan that would still have to pass muster with the agency’s board. But Fleissig, Doctoroff, and other Sidewalk officials frequently acted like they already had a done deal; they proceeded as though the board’s approval were a mere formality. They appeared frequently in public together, and Sidewalk briefly took to describing the arrangement as a “joint venture”—a characterization that misleadingly implied a formal legal partnership, until it was called out in the media.

Unlike urban infrastructure, Google’s infrastructure is an ever-mutating, lavishly capitalized, revenue-generating service.

What’s more, the initial framework agreement between the two organizations was kept under lock and key—a detail that also drew intense criticism once it came to light. Just nine months after the Sidewalk Toronto launch, the WT board jettisoned Fleissig for his conspicuous mishandling of the Sidewalk project, replacing him, temporarily, with an unflappable Bay Street public-sector pension executive named Michael Nobrega. (That initial contract, which turned out to be highly favorable to Sidewalk, was made public not long after Fleissig’s removal and replaced with one that evened the scales somewhat.) The scope of the proposed arrangement was also ambiguous: on paper, Sidewalk had bid only for Quayside’s twelve acres, but officials on both sides have continued to stress that WT is willing to let Sidewalk propose ways to deploy its infrastructure-based technologies “at scale.” This typically vague formulation clearly referenced future development on the Portlands, once the district is cleaned up and readied for development.

Meanwhile, the infrastructure technologies to be tested in the fledgling Sidewalk project only make sense if envisioned on a geographic scale far larger than a twelve-acre site. Among them were new systems for sensor-enabled waste separation and anaerobic digestion for composting, automated vehicle navigation, and dynamic traffic signal controls, wastewater monitors, and district heating and cooling networks tethered to energy-use sensors in individual buildings. As WT’s chief development officer Meg Davis admits, Sidewalk’s plans extend well beyond the project’s original site. “Quayside is part of a larger context,” she says. “You’ve got to be able to connect everything.”

Not-So-Soft City

When we talk about urban infrastructure, we mainly envision solid, immovable structures—streets, sewers, parks, community centers, electrical grids, subway systems, and the like. These backdrop features of city life embody the fixity upon which the kinetic drama of the urban experience unfolds. With lifespans counted in decades and centuries, these sturdy objects were platforms long before that word was co-opted by the mavens of social media.

These days, Google, too, can be regarded as a kind of virtual essential infrastructure—albeit a commercial variation, and one that rests upon a global system of privately owned telecommunications networks. Unlike urban infrastructure, Google’s is an ever-mutating, lavishly capitalized, revenue-generating service—one that exists by learning as much as it can about the lives and buying habits of those who use it and then using that data to target them with advertisements.

What’s the connection? Imagine a city street or a public park that can “know” something about who was moving along it, together with subsidiary information about where and how these urbanites were using the space (e.g., was a particular set of sensor-equipped park benches especially popular during early evenings?). Imagine, further, that this brand of smart urban space also had the capacity to adjust itself according to the output of an algorithm that presented a customized solution to some kind of unmet civic need. This is the foundational arrangement by which the worlds of civic and information infrastructure might intersect—or collide.

Kicked to the Curb

Last November, Sidewalk released its first detailed draft plan for Quayside. Superficially, the twenty-eight-page document depicts a cluster of buildings of varying scales abutting an array of open spaces positioned around an old slip off Lake Ontario. The planned technology and infrastructure was further detailed at a public roundtable and a meeting of the digital strategy advisory panel both held that December, including: “dynamic curbs” that use an array of sensors to let vehicles, especially self-driving ones, know where they can park or drop passengers off; sensors that monitor storm water run-off levels and building energy use; and a new form of pavement made from removable hexagonal blocks fitted out with multi-colored LED lights linked to a system of public space sensors. These devices are engineered to detect different traffic and pedestrian flows, and then use the patterns they observe to continuously adjust the lighting on each hexagon so the pavers show how much road allowance is afforded to vehicles, and how much is set aside for pedestrians and cyclists.

Craig Nevill-Manning, a former Google engineer who is now the head of engineering at Sidewalk, describes another application the company is experimenting with: a sensing device trained on pedestrians—think older people—crossing at intersections. These sensors, which he insists won’t capture any compromising or identifying information, can estimate the time someone requires to cross safely, based on a calculation of their walking speed. That data, Nevill-Manning explains, can be fed into an extended network of signal controls that, in theory, will dynamically tweak the intervals of the lights at the intersection.

It doesn’t take much imagination to realize that the complexity of this scheme grows exponentially when there are many pedestrians crossing at an intersection at varying speeds, and doing so in a heavily trafficked part of the city. The calculus becomes exponentially more complex as you multiply this example throughout any given city, with many nearby intersections generating streams of highly variable data, according to speed and traveling capacities of each pedestrian under review.

The Smart Set

The models of digital traffic and pedestrian surveillance are evidently in varying states of gestation, but they reveal the core of Sidewalk’s strategy. Notwithstanding the company’s earnest public pronouncements to Torontonians about the goal of developing waterfront affordable housing and the green advantages of using so-called tall timber construction methods, the long-term play is to develop a new and interactive breed of urban infrastructure to supplement the traditional forms.

Again, the “Labs” in Sidewalk’s moniker is the operative word: Quayside, and perhaps the Portlands, will serve as a kind of giant real-world petri dish in which the company can not only work out the relevant technical glitches but also concoct a business model for selling these technologies elsewhere.

The “Labs” in Sidewalk’s moniker is the operative word: Toronto’s open space will serve as a kind of giant real-world Petri dish in which the company can work out glitches and concoct a new business model.

It’s worth pointing out in this connection that Sidewalk/Google is dipping its oar into a vast and fast-moving industry. (One recent estimate put the global smart-city market at nearly $530 billion in 2017, growing to almost $2 trillion within five years—a 20 percent compounded annual growth rate.) Global smart-city tech firms sell all sorts of hardware and software-based systems that do everything from improving traffic flow to learning how to detect unusual behavior captured on closed-circuit TVs trained on public spaces.

Cisco, for example, promotes a system that involves trash bins: inside each one is a sensor that can detect how full a given bin is on collection day and transmit that information back to a central system, which aggregates all that data and uses it to optimize the routes of garbage trucks. (Other projects in the early stages of development veer toward the bizarre. BMW’s website takes note of a smart cities project in South Korea in which homes are fitted out with cameras and sensors to enable better communication between neighbors. “They can ask to borrow some sugar without even leaving their house,” BMW claims, as if texting and the phone had yet to be invented.)

Sidewalk’s suite of smart-city innovations has yet to come into clear focus, but it’s apparent there are a few core ideas. First, there’s the proposed transformation of civic infrastructure such that it can constantly respond to the data it gathers from the public realm. Second—and critically, from a commercial perspective—there are the third-party firms that will be invited to design apps and other marketable solutions that make use of all that gathered data, much in the same way that both Facebook and Google sell user information to advertisers (and other less reputable actors, such as Cambridge Analytica and Balkan-based fake-news combines).

To head off privacy concerns, Sidewalk initially assured Waterfront Toronto that it had the security of user data under control because it had retained a former Ontario privacy commissioner, Ann Cavoukian, to provide advice. But tech-savvy critics like Bianca Wylie, who was recently described in CityLab as the Jane Jacobs of the smart city, have raised crucial questions about vulnerabilities within Sidewalk’s system: Where do residents go to complain if they run into problems relating to Sidewalk’s technology? Who owns the data Sidewalk collects in its networks of sensors? And, crucially, what constitutes democratic participation in smart-city design?

Ryerson’s Pamela Robinson adds that de-identifying data—i.e., blurring faces captured on a CCTV trained on public parks—isn’t sufficient. If you are entering a heavily monitored district like Quayside, she says, you should be notified somehow that information about your presence may be captured, and that you should also have the opportunity to opt out.

The privacy issues are far from resolved. Cavoukian ended up quitting Sidewalk, saying she no longer felt the company intended to follow her recommendations. In advance of her resignation, WT announced that it was retaining a team of independent privacy experts to bird-dog the development of Sidewalk’s plan. The governance concept here is that Sidewalk’s data-gathering activities shall be accountable to a so-called data trust—a privacy monitoring approach pioneered in Estonia, one of the only countries in the world where nearly all of its public services are available online. But because Sidewalk’s technically complex plan demands a level of accountability that’s virtually without precedent, Robinson says any governance body will be costly: “It’s going to be expensive for those experts to do their jobs well.”

This Space for Rent

Yet privacy is only half the story. It’s not clear just how Sidewalk’s leaders plan to scale up this brand of civic infrastructure in the future—and just who will own the systems it embeds in urban precincts like Quayside. Consider Sidewalk’s “dynamic curbs” idea. This project dovetails with an extensive “curb-mapping” app launched this fall by Coord—another Sidewalk/Google venture. Dynamic curbs will use elements of physical infrastructure (sensors, LED signs, etc.) alongside back-end technologies designed to detect what’s happening at any particular segment of curb. This technology might, for example, know which parking rules may kick in at what time and determine whether a nearby parking spot is occupied or not.

Clearly, there’s a cost associated with creating and maintaining this kind of data system beyond the expense of caring for random patches of asphalt. Cities have long charged drivers for street parking on a time basis. But is there monetary value that can be attached to the knowledge that a given spot is available? And how is the value of that shard of transient data affected by the concurrent availability of nearby parking spots—or by similar incoming signals from nearby drivers looking for spaces?

Uber’s founders, of course, created an empire with an app and a set of algorithms that digest huge volumes of disparate demand and supply signals and then apply a variable pricing formula to determine individual fares. It’s not difficult to imagine someone designing a similar system for parking spaces, and one that would be especially useful for the self-driving vehicle of the future, which will need to idle somewhere between runs.

Suddenly, the lowly curb parking spot—that most quotidian kind of urban infrastructure—is embedded in a highly sophisticated data application, its occasional condition of vacancy now assigned a financial value within the broader commercial framework of an emerging transportation technology.

Where the Sidewalk Ends

Who pays? Who profits? Where, in this kind of system, does municipal infrastructure end and commercial activity begin? And are there secondary commercial or advertising uses for the information gleaned about parking spots? And what happens if Sidewalk someday decides to pack up and move on? The questions, as Wylie has reminded Torontonians, tumble out indefinitely, and Sidewalk has offered scant explanation, largely because the company hasn’t got answers—yet.

What’s clear, however, is that such formulations require us to re-imagine the way cities value infrastructure, as well as the financial relationships they’ll create with companies like Sidewalk. This new breed of civic-commercial partnership aims to use infrastructure-dependent technology configured to unlock new value from the sort of mundane, functional assets we never thought much about before.

Sidewalk, evidently, has no intention of developing a one-off system that it will sell only in Toronto. Yes, the company has committed itself to sharing some of the licensing revenues generated by the intellectual property developed here. But no one knows its specific longer-term commercialization plans. Nor can anyone get a clear fix on the implications for the sorts of details that preoccupy municipal officials, such as the ownership and maintenance of new tech equipment—sensors and the like—installed in the public realm. And given that so much of Sidewalk’s thinking involves digital technologies that age rapidly, there are also questions about obsolescence, as Sidewalk’s Craig Nevill-Manning acknowledges. After all, a parking spot doesn’t require much in the way of technical upgrades beyond the occasional slather of asphalt and gravel. (Et voila: Parking Spot 2.0!) However, the technology that transforms it into a monetizable asset may be out of date within a decade. These systems may someday become the printer-ink cartridges of municipal government.

Pamela Robinson poses one more important problem: when urban spaces and infrastructure are fitted out with an ever-growing array of sensors capable of gathering specific sorts of data from their surroundings, future managers of these systems, or end-users of the data, may think up additional sorts of information that should be collected. If Sidewalk deploys street furniture sensors in parks or along city streets that can tell when benches are busy, ostensibly in order to determine those periods when more should be put out, civic authorities or property managers might also find themselves curious to know if those sensors are showing bench use in the middle of the night, and what that might suggest (e.g., a sleeping homeless person). “Once you build it,” she observes, “you want to feed it.”

In fact, it may well be that such mission creep represents the greatest unknowable consequence associated with the interactive municipal infrastructure Sidewalk has proposed. The experts retained by data trusts will be mandated to game out potential misuses and/or abuses. Watchdogs will also strain themselves to identify data that may prove susceptible to misuse. But human ingenuity is a formidable thing. Was it predictable that a seemingly modest service for budget travelers, Airbnb, would evolve into a significant cause of rental housing shortages?

Robinson cites the case of an emerging idea in epidemiology. Data engineers are now able to deploy specialized sensors in sewer mains that can detect opioid loadings in wastewater. With the opioid epidemic showing only some signs of slowing, the technology—developed by a Cambridge, Massachusetts, start-up called BioBot Analytics—offers public health agencies a new tool for mapping the scope of the problem. “Wastewater contains valuable information about the health of communities,” the firm’s website says. “We collect it. We analyze it. We tell you how to leverage it to make your city better.”

The technology, which is deployed in municipal infrastructure, raises an intriguing variation of the old police-drama riddle about whether investigators have the right to rifle through someone’s trash after they’ve taken it down to the curb. As Robinson says, “Do people think of their waste as a data point?”

The second, and related, detail is whether an unintended application of such technology is to identify neighborhoods that seem to produce higher readings of illicit drugs—a piece of recon that the police could certainly use. (There’s no evidence that Sidewalk wants to deploy these sensors, but the company’s engineers are building applications to monitor wastewater.)

The overarching point—and the one that sits at the very heart of what Sidewalk seems intent on testing alongside the development of Toronto’s forlorn industrial waterfront—has to do with our evolving understanding of what precisely we mean when we talk about urban spaces and publicly owned amenities. Are these assets there merely to be optimized, monetized or, possibly, securitized? Infrastructure, as Lucie Greene points out, “is not an opt-in. The point of civic environments is that they’re there for everybody, and that they’re free.”