Sidewalk Labs’s Material Co-Design

In 2014 Google debuted Material Design, a set of user-interface design guidelines “inspired by the physical world and its textures, including how they reflect light and cast shadows. Material surfaces reimagine the mediums of paper and ink” by presenting layers of content as if they’re solid surfaces of equal thickness, stacked atop one another. Unifying the user experience across all Google products, Mark Wilson wrote in Fast Company, “Google has become a second reality inside touch-screen devices – complete with its own rules of logic and physics – and if Google has its way, it will eventually break free of touch screens to quite literally shape the world around us.”

Well, Google has been having its way. The following year, before it metamorphosed into Alphabet, Inc., Google launched Sidewalk Labs, an “urban innovation” division dedicated to solving urban problems with technology. Sidewalk adopted its own form of Material Design for the “physical world,” but their materials were concrete and cameras rather than paper and ink. Sidewalk’s leaders broadcast their intention to reimagine cities “from the Internet up.” I speculated back in 2016 that Sidewalk might be involved in the master-planning of New York’s new Hudson Yards development, home to the company’s headquarters, but the following year, Sidewalk shifted its attention to Toronto, where they’re currently working with Waterfront Toronto, a government organization, to build a new district infused with Alphabet tech (from tall timber buildings to sensor-embedded streets).

That effort has faced significant resistance. Critics have challenged the terms of the public-private partnership, the opacity of the development process, and the lack of clarity over data governance. Given the obfuscatory nature of Sidewalk’s closed-door deals and proprietary technologies, it might be surprising that they’ve adopted relatively transparent, intelligible, accessible tools of self-defense: paper and ink. Material Design as rhetorical redoubt.

Cards and Chipboard

Consider Sidewalk’s Toronto HQ 307, where the company tests out and solicits ideas for the smart neighborhood taking shape nearby. This “experimental workspace” features an assortment of unassuming analog media – post-its, index cards, markers (not to mention face-to-face communication between trained liaisons and visitors) – amidst scaffoldings of unfinished plywood, chipboard, and cork. Paper signage is affixed to walls, tacked to bulletin boards, or hung from movable metal dividers. These layered sheets of white and pastel – which bear resemblance to a Material Design interface – mix simple stencil and san serif typefaces to tell us about everything from the project concept, job creation, and Ontario’s timber industry, to Sidewalk’s proposals for data use, environmental sustainability, accessibility, and social infrastructures – including myriad hypothetical health and wellness spaces.

All 307 photos by me

Embedded among the printed text and still images we find the occasional screen, where we can explore maps, animations, and Sidewalk’s design tools. Humble sawhorse tables host an array of annotated cards, each offering feedback or questions from fellow 307 “co-designers.” A nearby stack of blank cards and pile of markers call upon us to add our own voices: “A question I have about facial recognition is: how will sensors and cameras be used wrt safety / security in data collection,” one visitor noted. An “efficient unit prototype,” a sample apartment made of cardboard, invites visitors to annotate its features with colored post-its: “corridor too narrow for wheelchairs,” “I like looking at something while I cook.”

Designers Daily tous les jours says they’ve created a “set of colorful, rolling, modular, stackable, playful interactive tools” that turn 307 into a place for “input + discussion + experimentation + new ideas + action.” Not everything is interactive, however. Some features simply announce their presence and expect us to deal with it. Perched high in a corner is a pair of gadgets, a white cylinder and a grey box, which, a stencil-and-san-serif sign-on-a-stand informs us, is a Numina sensor: “Hello,” it says. “A low-resolution image of you may be taken in this area. This image is immediately de-identified and is only used to calibrate and validate pedestrian and vehicle movement operating in the area.” No opt-in or out. No solicitation of your opinion or request for consent. Numina simply is, as-is, and it may have already taken your photo – before you even knew of its existence.

The open room is painted primer-white, and at the center of its concrete floor is a platform composed of tessellated chipboard hexagons, each with an embedded LED light. Those hexagons reappear outside, as pavers made from concrete infused with powerderized post-consumer glass (thus, the signs say, keeping glass waste out of landfills, and decreasing the CO2 emissions generated through concrete production). Inspired by the work of the French Institute of Science and Technology for Transport, Development and Networks, these pavers deliver “four key features”: modularity, which allows for easy removal and replacement; heating, which hastens the melting of snow and ice; lighting, which can be programmed to offer real-time cues to regulate traffic and communicate street use; and permeability, which aids with stormwater management. A sign affixed to the side of a shipping container – the modern model of modularity – identifies these hexagons as the “holy grail of pavement.” It’d be sacrilegious not to exploit their virtues.

307’s humble material objects – concrete and chipboard, post-its and powderized glass – are rhetorical tools suggesting that, here, cities are “workshopped,” urban futures and data plans are co-created from a mix of proprietary platforms and public knowledge. Except for that Numina sensor: that’s registering your presence and taking your picture, whether you like it or not.

Cells and Sensors

“Sensors have become a part of our daily lives. CCTVs. Traffic cameras. Transit card readers. Bike lane counters. Wi-Fi access points. Occupancy sensors that activate lights or open doors. These are all examples of how digital technologies can integrate into our physical world to help make our public spaces more comfortable, responsive, and efficient.” So acknowledged Sidewalk Labs in its Designing for Digital Transparency in the Public Realm project, developed through Spring of 2019 and formally released on April 19.

Open government advocate Bianca Wylie decried Sidewalk’s “passive language, without agency.” They speak “as though the sensors and cameras just sprouted into the world, without creators or purchasers, without contracts or decisions.” Of course Sidewalk, its parent company Alphabet, and partners like Numina have put – and are putting – many of those sensors in place. This act of “subtle [sensor] normalization,” Wylie says, “works to “guarantee that a quantified city becomes a social norm.” It’s already happening, Sidewalk reminds us; we might as well accept the facts: “Look around you as you go about your day, and you’ll start to see how much sensing and data collection infrastructure is already all around you — but there is very little transparency around what data is being collected, by whom, and for what purpose.”

Think back to that Numina privacy notice inside 307. “Hello! A low-resolution image of you may be taken in this area.” Or it may already have been taken by the time you read this sign. What if we applied a similar tell-don’t-ask / declaration-without-consent strategy to the thousands of outdoor sensors around Quayside – and all across Toronto and other cities and towns. Boston and London are already posting signage that alerts people to the presence of digital technologies in the public realm, but those signs are often text-heavy, and they require visiting a website to gather more information. As Sidewalk team members Jacqueline Lu, Patrick Keenan and Chelsey Colbert explain, user research shows that “few users read long, jargon-filled privacy and data collection policies.” SL asks: “What if you could quickly communicate what technology was in use in the public realm in a way that was transparent and clear without being overwhelming?”

[April 8 – DTPR Release Shareout / Creative Commons Attribution 4.0 International (CC BY 4.0). All images are linked to their sources.]
Sidewalk’s team (which contains lots of talented folks with impressive design, civic tech, and government experience) responded to this question by, first, examining existing labels and signage – from nutrition labels to street signs to surveillance camera notices – and the relevant International Organization of Standardization guidelines. They then convened co-design sessions in Toronto, New York, San Francisco, and London, which made use of cards and forms like those available in 307 to assess participants’ questions and concerns about existing technology (as much recent research has shown, those tech concerns vary widely in relation to users’ race and class, which raises questions about the demographic diversity of workshop participants). Sidewalk’s facilitators first asked participants what questions they had about existing technologies, like Bluetooth beacons and infrared sensors, then identified concerns shared within the group. They then asked participants to imagine signage systems and icons that would provide useful information about those technologies. Sidewalk polished those user proposals into their own set of icons, which they further refined through user-testing. They held periodic online “shareouts,” too, to present the evolving design.

Through this iterative user research process (undertaken in collaboration with the Soofa signage team), Sidewalk learned that the questions most important to people are: the purpose of various digital infrastructures (e.g., emergency, mobility, accessibility, enforcement, energy efficiency, and so forth; it’s not the most ontologically consistent classification system), the entity accountable for each technology, and whether that technology collects identifiable information. The co-design process seemed to begin by assuming that the presence of these technologies is a given; [edit:] while some resistance was apparently permitted, workshops seemed to be structured such that little time was spent debating whether particular digital infrastructures belonged in the public realm in the first place. The co-design process ultimately frames sensing technologies as a necessity, an inevitability. Might as well normalize them by getting to know them.

Sidewalk developed a set of icons (derived from Google’s Material icons) for each variable and decided to set those icons within a hexagon badge. “We chose the hexagon,” they explained, “because this perfect shape that occurs naturally” – from honeycombs to snowflakes – “is the most efficient way to fill a space with the least amount of material.” Plus, it’s “currently unused in our vocabulary of signage shapes and slightly resembles a stop sign – giving users a slight ‘Hey, check this out’ without forcing a stop.” (edit: Alex Gekker also noted hexagonal resonances in sci-fi gaming.) And as with the hexagonal concrete pavers planned for use on Toronto’s streets, these hexagonal icons tessellate easily, in standard patterns (again, in accordance with Material Design principles, and as James Birch notes, in a fashion quite similar to that of the NFPA 704 hazard placard).

The “purpose” and “accountable entity” cells are black and white, while the “identifiability” cell is in color: if its captured data is de-identified, (e.g., by blurring recognizable information in photos) the cell is blue; if users are identifiable (as in surveillance camera footage), the cell is yellow. If a particular technology collects no identifiable data, there’s no colored cell. I must admit: I’m not quite sure what it means to be “air-” or “light-de-identified” (a breeze could dissipate any identifiable chemicals?).

A fourth cell offers a QR code and URL through which users can access, via their mobile devices, a “digital channel” that offers more detail. This transition from physical signage to digital interface again exploits the cross-platform continuity of Material Design. Online, the information is organized linearly, via a set of chained icons identifying, first, the responsible organization, the technology’s purpose, and the tech type – the same information embedded in hexagons on public signage. The next set of icons, enclosed in circles, addresses the data and its processing: the type of data, whether it’s identifiable, how it’s processed. The final set, framed in squares, pertains to data storage and access: how long, where, and how it’s stored, and who has access to it. Plus, as the Sidewalk Team explained in their April 8, 2019, shareout, “You’ll see in the digital channel [tha]
there’s always an area to give feedback, … a way for people to express their opinions about the technology. And hopefully that feedback actually goes to someone who can make a difference.” Yes, hopefully the platform facilitates the registration of concerns to an accountable entity, and that that feedback prompts real dialogue – rather than merely enabling the performance of “co-design.”

“We’re not planning this just for Quayside,” the team says. “The intention from the beginning was that this would be implemented well before Quayside, and that this can really forward provocations about digital literacy and understanding the public realm at large.” They intend for their Digital Transparency in the Public Realm (DTPR) project designs – and the whole design process – to extend well beyond Toronto, too. All workshop materials, icons, and design standards are available on Github under a Creative Commons license. Sidewalk wants folks in other towns and cities to adopt and adapt them – and thus further advance the cause of “digital literacy.” “Our goal is to co-develop meaningful design patterns that will eventually be adopted by cities, private partners, and other institutions that are interested in improving digital transparency in public spaces,” they noted in their March 4, 2019, shareout.

Transparency implies disclosing the presence of urban technologies and the harvesting of what Sidewalk calls “urban data,” which encompass data collected in public spaces, private spaces accessible to the public, and private spaces that aren’t controlled by their occupants (what a conveniently wide jurisdiction). By “providing transparency,” Sidewalk helps people “increase their awareness” – their “literacy” – “of how digital technology in the public realm works.” That’s not an ignoble goal. Transparency, Lu, Keenan, and Colbert explain, “can empower users to meaningfully engage in what is fast-becoming a critical conversation of our time. And, equally important, transparency can nudge both users and data collectors towards best practices.” Yet I can’t help but wonder in whose interest “best” is defined here.

DTPR, an acronym likely to evoke GDPR – a recent landmark EU regulation on data protection and privacy – isn’t really about privacy or consent, or about fundamental questions regarding the presence of extractive and surveillant technologies in the public realm. Instead, DTPR, like the 307 experience, marshals Material Co-Design – its stacked colored cards and tessellated hexagons, minimalist icons and ISO standards – to aestheticize and rationalize coercion (co-optation? hegemonic persuasion? I can’t find the right word!), to frame “performative ethics” as political action. As Rob Kitchin argues, “citizen-centric” design, or co-design, “has largely acted as an empty signifier, designed to silence detractors or bring them into the fold while now altering the technocratic workings, profit-driven orientation, or ethos of stewardship … and civic paternalism … of smart city schemes.”

By becoming literate in design methods, by framing deliberation within a totalizing Material Design system, by learning how to spot urban technology and celebrating our “awareness” of its operations (undoubtedly valuable skills!), we can lose sight of the bigger questions: about whether we want cities built in the image of a corporate internet, whether all digital infrastructures are necessarily in the public interest, and whether a public’s acts of citizenship should be reduced to filling out co-design cards and training Alphabet’s algorithms.

Leave a Reply

Your email address will not be published. Required fields are marked *