Complex Choreographies of Mobile Mapping

Somewhere lies a mythical city where the cars drive themselves, and the people walk around with wearable guides that know what they want and where to find it. To make this city work requires the most detailed map ever created.
Google wants that map. So do Apple and Uber. And a consortium of German luxury carmakers just bought their way into Silicon Valley’s simmering map-making race, announcing Monday they will spend $3.1 billion to acquire Nokia’s mapping division, dubbed HERE. — San Jose Mercury News (8/3/2015)

I.

Maps, as representational media, are abstractions derived from something, or a reaction to something, but are not the thing itself. A map is a useful way-finding tool that may describe a territory by reproducing a similar structure, but as all are aware, a map is not the territory — which is to say that our perception of the map can never quite equal the territory.[1] Nonetheless, the semanticist Alfred Korzybski held that many people do indeed confuse maps with territories, by confusing models of reality with reality itself.

With that in mind, today’s digital cartographers continue their quest to develop a rigorously detailed intelligent one-to-one map of the world. Borges’ “On Exactitude in Science” describes the dilemma of such a perfectly accurate, yet imprecise, map: “In time, those Unconscionable Maps no longer satisfied, and the Cartographers Guild drew a Map of the Empire whose size was that of the Empire, coinciding point for point with it.” In other words, a map that is too exact becomes the thing it maps, endangering both.[2]

Further, a one-to-one map of the world, no matter how perfect, would still contain biases or flaws because every map is an interpretation that reflects the mapmaker’s point of view. And while to err is human, it is also a trait shared with highly sophisticated lidar scanners whose mapping errors relate to algorithmic misreadings of ambiguous data.[3] Admittedly all mapped information comes to us secondhand (it is a symbolic reconstruction), and even a one-to-one point cloud can only attempt to measure the built environment, not the qualitative components: the qualia — the unmappable elements — have, until recently, been left off the map.[4]

Yet Silicon Valley’s rush to digitally map the city causes one to pause and wonder if, by scribbling on the map, might we actually change the territory?[5] For instance, in The Image of the City (1960), Kevin Lynch explains how residents create social imaginaries of their cities, as a means of asking What do we want our cities to be? This key aspect of cartographic interpretation emphasizes the establishment of clearly legible urban environments to facilitate the inclusion of all residents. Along with Lynch’s social imaginaries, the 1960s simultaneously marked a theoretical transition away from the notion of urban space as a neutral container and toward an understanding of space as a conductive medium for the movement of people, information, and objects. Spatial theorists of the time envisioned relational structures as topologies that could be projected onto physical social space.

In particular, the architectural critic Reyner Banham’s “autopia”[6] and urban planner Melvin Webber’s notion of “city as a communication system”[7] contributed to an increased understanding of an urban condition preconditioned by information flows. Webber theorized that communication technologies would begin to define an urban realm that “is neither urban settlement nor territory, but heterogeneous groups of people communicating with each other through space.”[8] Moreover, he argued, “A city is not described by the buildings, but by the social relations which bind the city together.” Indeed, a community is defined by its social overlay — the everyday interactions and informational exchanges that weave it together.[9] Those theoretical frameworks positioned space as inherently caught up in social relations, thus both producing and consuming them.

II.

Imaginaries are intricately linked to the material; conceptual forms play a significant part in the mores by which daily lives are lived and decisions, policies, and actions are rationalized and legitimized.[10] Imaginaries play an important role in understanding place and in influencing the decisions that either enable or limit possible futures for urban habitants. Indeed, to the extent that maps are visualizations of imaginaries they precede the territory — and ultimately engender the territory.[11]

What might this mean for locative media and digital mapping? In competition with the Google X lab, Nokia has been coordinating an international fleet of vehicles that capture highly detailed street-level information. Atop each mapping car is a 5-foot rig with accelerometers and gyroscopes that record both vertical and horizontal deviations; a panoramic camera that captures images; a global positioning system tracking satellite-based coordinates; and most importantly, a fast-rotating laser scanner that mathematically pinpoints the edges of lanes, curbs, and tunnels.[12] This information is later uploaded into an image-processing program where point clouds of data are extrapolated into highly accurate 3D digital models. Now imagine this 3D digital map loaded into a robotic car. Visualize the on board computer relying on a continuous machine-to-machine exchange of environmental information to navigate streets and the infinite number of calculations required to synchronize the robotic car’s movements with the digital map — and you begin to realize the immensity of the task.[13]

“They know this [mapping] is going to be the brains of our cars,” said Eric Gundersen, CEO of San Francisco startup Mapbox, one of many firms riding a market wave of interest in next-generation cartography.[14] The competition is fierce because “mapping is core to everything mobile and everything about being able to move around a city.”[15] For an augmented city to function, it’s necessary to construct the most highly accurate, numerically defined urban map possible, update it in real time, and control who uses it. The team that solves the digital mapping model first will have an advantage in future industries from robotic taxis, drones, and package delivery fleets to personalized shopping apps.[16] This highly precise numerically defined one-to-one map of the world–essentially a massive database–is what Nokia has been developing with its “reality capture and processing” digital cartography project — aptly called HERE.

Digital maps such as HERE serve as examples of hybrid alliances between online systems and offline spaces. These maps are produced through a variety of technologies including laser scanning and radar (lidar), digital modeling, graphic software, GPS, and mobile telephony — all of which are connected via cellular networks, the Internet, and short-range signals such as wi-fi and Bluetooth. William Mitchell described this complex assemblage as ‘the mobility network.’ Variously termed the Internet of Things, the Internet of Everything, and the Cloud of Things, it denotes adaptive systems — for example, wireless mobile communication devices used to organize people and practices through sensors in real time and on the go. While for the most part invisible, an information and communication infrastructure — from the autonomous car to networked traffic signals that can be adjusted from afar to networked resource sharing collectively known as mobility-on-demand, including Lyft, Uber, and others, as well as the electronic allotment of parking spaces — coordinate options dynamically, in real time and on the fly. However, the vast number of computational interactions that would need to be transmitted wirelessly in real time by the mobility network in tandem with the autonomous car has caused Cisco to estimate that the Internet would need to expand by three to five times.[17]

More intriguing is that the hybrid alliance of online mapping and physical space is creating a secondary urbanism of data storage, such that the physical city and its data “doppelganger” exist and evolve simultaneously. While the mobility network is changing and will continue to change, it is wholly dependent upon a data doppelganger to orchestrate movements within the city. And as objects in the world — including people, places, and things — are increasingly integrated into the model, there is a possibility that distinctions between the physical world and its informational model may slip away.[18]

III.

While most scholars agree that soft infrastructure — wireless technologies, the Internet, and social media — may hold the potential to produce new kinds of space and enable new social practices,[19] the emphasis on technologies has resulted in uneven accessibility.[20] Recent studies suggest that informal modes of transit such as Uber, which promise inclusivity, are actually spatially restricted, privileging upper income neighborhoods while avoiding lower income areas.[21] Ideally, maps are designed to encourage access. With that goal in mind, might it be possible to re-embed the mapping project into a broader or different set of institutions? Through the adoption of universally accessible open source applications, maps might make a city more inclusive.

One proposal from architect Carlo Ratti of MIT Media Lab calls for the establishment of an “open source city” that is based on collaboration and knowledge sharing. Platforms such as Waze already rely on crowdsourced information to optimize traffic flow. Even Google Maps made its Keyhole Markup Language (KLM) available to the public, thereby democratizing geographic information systems (GIS) for amateurs and developers alike. There are also numerous platforms for crowdsourced crisis mapping, one of the best known is Ushahidi, which was utilized during the earthquake in Haiti and many other crisis areas.[22] While technology and automobile companies are fighting to control the commercial mapping arena, another project, called OpenStreetMap — a collaborative, open-source project originating in the UK — is freely downloadable for both public and private concerns. Paticipatory practices take up the mapping challenge by working to strengthen information access as an element of the public domain by enriching urban experience, cultivating environmental sensibilities, and renewing responsibility to some idea of a commons.[23]

IV.

Crowdsourced possibilities notwithstanding, we cannot presume that digital mapping is a wholesale solution to the mobile city. In addition to debate over the proliferation of sensors and telecommunication devices, a larger and more important discussion involves economics, policymaking, and governance of this technology. While the proliferation of locative media and other GPS-enabled maps has permanently changed the way we understand the urban landscape, what is driving (as it were) the connected car initiative?[24] The recent acquisition of Nokia’s 6,000-person maps team by the consortium of automobile makers Audi, BMW, and Daimler AG signaled the companies’ refusal to allow either Google or Apple (a newcomer to the field) to control location-based technologies. “High-precision digital maps are a crucial component of the mobility of the future,” said Dieter Zetsche, chairman of Daimler AG, in a statement about the Nokia deal. “With the joint acquisition of HERE, we want to secure the independence of this central service for all vehicle manufacturers, suppliers and customers in other industries.”[25]

While Zetsche’s claims may appear altruistic, Google, Apple, Ford, and GM, as well as many others such as GE, Cisco, Siemens, and IBM, are less interested in robotic cars per se than in capturing data from cars and selling that data to third parties. Market forces embed software opportunistically, to serve both commercial and political objectives, as applications and services create increasingly sophisticated ways to collect and monitor personal data.[26] At present, the government can monitor mobile phone conversations (they are considered public); in the future, law enforcement officials will not only be able to track how fast people are driving, but whether or not traffic signals are obeyed. What deserves additional attention is that both cars and location-based mobile applications record in­for­ma­tion about everyday so­cial­ity, as the metadata collected through user-generated content running on proprietary applications can be commercially lucrative.[27]

The ephemeral datascapes that defined initial encounters with the Internet are giving rise to a new media of geospatial information. What has become a matter of concern is that this future assemblage of wireless sensor networks and urban space has the capacity to instantiate an extensive applied control topology that weaves sensors with data, social media, and mapping — in other words, context. Today context implies a ubiquity of networked information intimately aligned with the perceptual and cultural realities of everyday life.[28] In this way, we have come full circle, from the placelessness of the early Internet to the present, where every nodal point can be located by longitude and latitude, interconnected, and aware. Just as de­mo­graphic, ge­o­graphic, so­cial, and even biometric data form the eco­nomic base of fixed internet conglomerates, such as Facebook and Google, the additional geospatial data retrieved from mobile devices and autonomous automobiles is as­so­ci­ated with sig­nif­i­cant mar­ket value.

Moreover, these initiatives all have consequences to our individual privacy.[29] What is presented as an emphasis on customization at the user end actually veils the commercial practice of personal data mining on the provider end. Users perceive a gain in control without being aware that they are constantly being monitored. “The extent, precision, and speed of this data gathering is unprecedented,” according to Internet theorist Felix Stalder.[30] Concern is also warranted over the surveillance of individuals by the National Security Agency and other international organizations.[31] Hence, digital mapping innovation is no longer about designing the best turn-by-turn navigation software or the most artful autonomous interface. Rather, maps are predicting and influencing behavior, evolving into disciplinary instruments, monitoring biometrics, and ultimately contributing to a commodified model of the city — its data doppelganger.

V.

While policymakers express guarded optimism that information and communications technologies, including large-scale data analysis, can bring increased understanding and order to urban processes, a critical realization is that digital mapping, automated transportation, social equity, urban infrastructure, and privacy can no longer be understood in isolation. Each of these is connected to the others via networked systems and wireless infrastructural integration. In that respect, the integration of locative technologies into everyday social practices compels us to reflect deeply on their protocols, platforms, and interfaces. The production of space is increasingly dependent on code, and code is written to produce space.[32] Digital cartography as a form of code is thus actively shaping socio-spatial organization, processes, and economies, along with discursive and material cultures. Those effects are becoming increasingly pervasive as more and more everyday practices are threaded through digital mapping platforms.[33] Moreover, as maps become more accurate in locating everything (and everyone) in space, and expand their usage parameters to include search mechanisms, we may not only come to mistake the model of reality for reality itself, we may also conclude that only what is searchable is worth searching for.

The larger questions here surround society. Who will control the future of location-based technologies? If maps reveal their makers’ biases, then whose city is actually being represented? Locative media is not outside the physical world; on the contrary, it is designed by, and entangled in, physical world social practices. Because locative media organizes people and practices in the physical world, the platforms and their applications have increasingly critical implications — to such a degree, in fact, that locative media as interactive maps effectively perform as a novel form of architecture.

Considering infrastructure — hard and soft — in this way creates opportunities for social interaction, because ultimately data spaces and physical spaces are intertwined. Thus, it is critical to understand how digital maps are produced and disseminated so that we can both envision alternatives and avoid reproducing actual inequities and inequalities. Cultural imaginaries and the collective discussion that emerges from such imaginaries, whether informed by political or design aspirations (and ultimately, all design is political), have significant implications. While the need for navigational tools is vital to cities, the urge to map is a populist passion. Our democratic institutions emphasize the value of active participation in the shaping of one’s society. If maps, as informational infrastructure, are open and decentralized, are egalitarian and support diversity, then maps can create numerous opportunities for many types of civic participation that do not exist today.[34] An egalitarian information infrastructure has the potential to open up engagement to all citizens — not only as a means for residents to represent their city, but also in allowing them to have a say in shaping their city’s future.

Notes

[1] According to Jim Walker, the phrase “A map is not the territory” means that a map can describe a territory in such a manner that allows for navigation, but that our perception of the map can never equal the territory, only our perceptual version of it, our map. It is Walker’s belief, along with cognitive scientists, that similar to other symbolic systems, maps provide a twice removed image of the world. Human perception receives multiple sense streams of data, which is then delayed as the brain processes it into a symbolic model. Further, our interpretations can present difficulties since any information, even if imagined or dreamt, can seem just as real to us as those that come from the outside world. Refer to Walker, Confusing the Map for the Territory Originated: 10 Feb. 2001 http://www.nobeliefs.com/MapandTerritory.htm

[2] Casey N. Cep, “ The Allure of the Map,” New Yorker (Jan 22, 2014) http://www.newyorker.com/books/page-turner/the-allure-of-the-map

[3] Kathryn Nave explains that many people do not question the results of laser scans. ScanLAB founders, Matthew Shaw and William Trossell, exploit this algorithmic aspect in their lidar experiments with materials such as glass, fog, mist, ice flows, even movement. “Why you can’t trust 3D scans of cities” Wired Magazine (5/9/2014), http://www.wired.co.uk/magazine/archive/2014/06/start/scanlabArchitects Shaw and Trossell

[4] UC Berkeley psychologist Eleanor Rosch along with professor of philosophy Alva Noë disavow the Cartesian paradigm, contending that you are not your brain — rather, that “consciousness is an achievement of the whole animal in its environmental context” https://philosophynow.org/issues/80/Out_Of_Our_Heads_Why_You_Are_Not_Your_Brain_by_Alva_Noe

[5] Future of Humanity Institute, University of Oxford, “Lesswrong,” March 2010, <http://wiki.lesswrong.com/wiki/The_map_is_not_the_territory>

[6] Reyner Banham. Los Angeles: The Four Ecologies (New York: Harper & Row, 1971.

[7] Melvin Webber. “Order in Diversity: Community Without Propinquity,” L. Wingo, Jr., ed., Cities and Space. (Baltimore: Johns Hopkins Press, 1963), 23–54.

[8] Webber, “The Urban Place and the Non Urban Realm,” in Explorations into Urban Structure (London: Oxford University Press, 1964), 116.

[9] Refer to “Order in Diversity: Community Without Propinquity,” in Cities and Space: The Future of Urban Land, ed. Lowdon Wingo (Baltimore: Johns Hopkins Press, 1963), 29.

[10] Rob Shields, The Virtual (London: Routledge, 2002), 47.

[11] According to Baudrillard “It is nevertheless the map that precedes the territory — precession of simulacra — that engenders the territory.”

[12] Laser scanning involves the controlled steering of laser (light) beams followed by a distance measurement at every point. This method, often called 3D object scanning or 3D laser scanning, is used to rapidly capture shapes of objects, buildings and landscapes. This remote sensing technology is also known as LiDAR.

[13] Google’s (and others) version of mobile mapping is simultaneously utilizing artificial intelligence (AI). The onboard computer is structured around robotic learning by sharing data with other vehicles and continually building up a networked library of environmental features. Each must be assessed, recognized, and ultimately understood in order for AVs to navigate safely through the streets.

[14] While Gundersen contends this map is “the brains of autonomous vehicles,” it is perhaps more accurately the automobile’s perceptual mechanism by offering a parallel to sight. As an example of instrumental realism, it calls to mind Descartes’ analogy of the blind man with a cane, only in this instance, instead of a cane performing as a prosthetic for the blind man, with the robotic car, a digital cane (Nokia’s HERE) is only in contact with a virtual data construction of the ground.

[15] Rex Santus. “Nokia Sells HERE to German Automakers for $3 Billion.” Forbes (8/03/2015) http://www.forbes.com/sites/rexsantus/2015/08/03/nokia-sells-here-to-german-automakers-for-3-billion/

[16] Matt O’Brien. “Silicon Valley Map Wars Heat Up.” San Jose Mercury News (8/5/2015)

[17] “New Cisco Internet of Things (IoT) System Provides a Foundation for the Transformation of Industries” Cisco News Site (6/29/2015) http://newsroom.cisco.com/press-release-content?type=webcontent&articleId=1667560

[18] Refer to Gordon, Eric and Adriana de Souza e Silva. Net Locality:Why Location Matters in a Networked World. (Hoboken, NJ: Wiley-Blackwell , 2011), 174

[19] The CFC Media Lab in Toronto developed Murmur as a documentary oral history experiment using the cellular telephone as an ubiquitous computing device. The interactive installation enabled people in the city to record personal stories about specific places, buildings, neighborhoods, and experiences for other visitors to access. Each account added additional layers of meaning thereby enriching the conception of the city beyond that of the officially endorsed versions perpetuated by tourism boards and commercial enterprises. http://murmurtoronto.ca/about.php

[20] Marvin and Graham’s Splintered Urbanism extensively documents uneven economic and technological development through hard infrastructure — transportation systems, electrical grids, fiber optic networks — based on a pattern of differential access to public services. Splintering Urbanism (New York: Routledge, 2001), 43–45.

[21] Jacob Thebault-Spieker; Loren Terveen; Brent Hecht. “Avoiding the South Side and the Suburbs: The Geography of Mobile Crowdsourcing Market” Urban Environments CSCW 2015, March 14–18, 2015, Vancouver, BC, Canada. http://journalistsresource.org/studies/economics/business/airbnb-lyft-uber-bike-share-sharing-economy-research-roundup#sthash.R1OSBx1I.1GEjoSHP.dpuf

[22] Also see Patrick Meier work on crisis mapping. http://irevolution.net/2013/11/11/humanitarian-technology-haiti-to-yolanda/ (Accessed 12.1.2015)

[23] In conversation with Malcolm McCullough. Ann Arbor MI, #Micities Symposium: Technology & Innovation in Michigan Municipalities and Beyond, Oct. 4, 2014.

[24] Refer to O’Brien. Op. Cit. Google’s Maps is now perfecting more detailed, internal maps for its self-driving cars. In 2015, Apple began deploying image-collecting mapping minivans throughout U.S. cities. Other players are TomTom, a Dutch company, known for its in-car navigation devices, who has partnered with Germany’s Bosch to map freeways for self-driving cars — its first project was Silicon Valley’s Interstate 280. Because mapping is integral to Uber maintaining an efficient ride-hailing service, Uber bought San Jose-based deCarta in March and is pursuing self-driving car technology.

[25] “You have to invest a lot of money in very expensive sensors and cars and tools to build the maps, and right now there’s no market for it. It’s a 2020 market,” Prioleau said. “It’s a market that takes a lot of investment today for something out in the future. I think that’s something they want to make sure they controlled.” Though the telecommunication company is based in Finland, Nokia’s mapping division has its roots in a Sunnyvale firm, Karlin & Collins, that was founded 30 years ago after one of its co-founders struggled to read an unwieldy paper map while driving around the Bay Area. The company helped launch the turn-by-turn computer navigation craze that took off in the 1990s. It moved to Chicago and changed its name to Navteq, which Nokia bought for $8.1 billion in 2007. The Finnish firm added Ristevski’s Earthmine, a Berkeley-based firm, in 2012; John Ristevski now heads up Nokia “REAlity Capture and Processing.”

[26] Another concern relates to anonymity, according to Dan Work, PhD, Transportation Engineer, University of Illinois Urbana-Champaign, is that individuals can be identified through any two repetitive location points, even if mobility data is anonymized.

[27] Information gleaned from mobile wireless networks include: whom we come into con­tact with and for how long, what value we, as in­di­vid­u­als, offer as a node in the net­work, and broad mo­bil­ity dy­nam­ics con­cern­ing our move­ment as a group are all im­por­tant data for de­ter­min­ing the re­con­fig­urable topol­ogy and rout­ing pro­to­cols im­ple­mented by the net­work, its ef­fi­ciency, and over­all per­for­mance.

[28] Op. Cit. Gordon, Eric and Adriana de Souza e Silva. Net Locality, 2

[29] Rachel O’Dwyer, “Network Media: Exploring the Sociotechnical Relations Between Mobile Networks & Media Publics” Istanbul, ISEA Conference, 2011.

[30] Felix Stalder, “Between Democracy and Spectacle: The Front-end and Back-end of the Social Web,” The Social Media Reader, (New York: NYU Press, 2012), 250

[31] Leo Kelion, “UN Internet Regulation Treaty Talks begin in Dubai” http://www.bbc.co.uk/news/technology-20575844 (Accessed 12.1.2015)

[32] Rob Kitchin, Martin Dodge, Code/Space: Software and Everyday Life. (Cambridge, MA: MIT Press, 2011) x.

[33] As social geographers Rob Kitchen and Martin Dodge argue, urban policy-making requires an interdisciplinary approach. Today GIS software, digital modeling programs, and wireless sensor information make it possible to build up a model of the city from user interaction and understand movement and circulation patterns in new ways. This enables designers and planners to study the city from the bottom up — from actual everyday social practices of urban residents.

[34] Celeste Pagano, “DIY Urbanism: Property and Process in Grassroots City Building,” Marquette Law Review. 97:2 336–389.