See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/262008155

The real-time city? Big data and smart urbanism

Article in GeoJournal · February 2013

DOI: 10.1007/s10708-013-9516-8

CITATIONS READS

2,101 6,053

1 author:

Rob Kitchin
National University of Ireland, Maynooth
346 PUBLICATIONS 22,720 CITATIONS

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

The Programmable City View project

ICLRD: Spatial Planning and Cross-border Cooperation on the Island of Ireland View project

All content following this page was uploaded by Rob Kitchin on 21 October 2015.

The user has requested enhancement of the downloaded file.

GeoJournal (2014) 79:1–14
DOI 10.1007/s10708-013-9516-8

The real-time city? Big data and smart urbanism

Rob Kitchin

Published online: 29 November 2013

 Springer Science+Business Media Dordrecht 2013

Abstract ‘Smart cities’ is a term that has gained Introduction

traction in academia, business and government to
describe cities that, on the one hand, are increasingly For the past two decades, urban analysts and theorists
composed of and monitored by pervasive and ubiqui- have been charting the evolution of cities during an era
tous computing and, on the other, whose economy and where information and communication technologies
governance is being driven by innovation, creativity (ICTs) have been exerting a growing and pervasive
and entrepreneurship, enacted by smart people. This influence on the nature, structure and enactment of
paper focuses on the former and, drawing on a number urban infrastructure, management, economic activity
of examples, details how cities are being instrumented and everyday life. Cities which have embraced ICT as
with digital devices and infrastructure that produce a development strategy, being pioneers in embedding
‘big data’. Such data, smart city advocates argue digital infrastructure and systems into their urban
enables real-time analysis of city life, new modes of fabric and utilising them for entrepreneurial and
urban governance, and provides the raw material for regulatory effect, have been variously labelled as
envisioning and enacting more efficient, sustainable, ‘wired cities’ (Dutton et al. 1987), ‘cyber cities’
competitive, productive, open and transparent cities. (Graham and Marvin 1999), ‘digital cities’ (Ishida and
The final section of the paper provides a critical Isbister 2000), ‘intelligent cities’ (Komninos 2002),
reflection on the implications of big data and smart ‘smart cities’ (Hollands 2008) or ‘sentient cities’
urbanism, examining five emerging concerns: the (Shepard 2011). Whilst each of these terms is used in a
politics of big urban data, technocratic governance and particular way to conceptualise the relationship
city development, corporatisation of city governance between ICT and contemporary urbanism, they share
and technological lock-ins, buggy, brittle and hack- a focus on the effects of ICT on urban form, processes
able cities, and the panoptic city. and modes of living, and in recent years have been
largely subsumed within the label ‘smart cities’, a term
Keywords Big data
Smart cities
Urbanism
which has gained traction in business and government,
Real-time analysis
Data analytics
Ubiquitous as well as academia.
computing
Governance The term ‘smart city’ has been variously defined
within the literature, but can broadly be divided into
two distinct but related understandings as to what
R. Kitchin (&)
makes a city ‘smart’. On the one hand, the notion of a
NIRSA, National University of Ireland Maynooth, County
Kildare, Ireland ‘smart city’ refers to the increasing extent to which
e-mail: [email protected] urban places are composed of ‘everyware’ (Greenfield

123
2 GeoJournal (2014) 79:1–14

2006); that is, pervasive and ubiquitous computing and development and governance and how they can be
digitally instrumented devices built into the very enhanced by ICT. In this scenario, networked infra-
fabric of urban environments (e.g., fixed and wireless structures are enabling technologies, the undergirding
telecom networks, digitally controlled utility services platform for innovation and creativity, that facilitates
and transport infrastructure, sensor and camera net- social, environmental, economic, and cultural devel-
works, building management systems, and so on) that opment (Allwinkle and Cruickshank 2011).
are used to monitor, manage and regulate city flows What unites these two visions of a smart city is an
and processes, often in real-time, and mobile comput- underlying neoliberal ethos that prioritises market-led
ing (e.g., smart phones) used by many urban citizens to and technological solutions to city governance and
engage with and navigate the city which themselves development, and it is perhaps no surprise that some of
produce data about their users (such as location and the strongest advocates for smart city development are
activity). Connecting up, integrating and analysing the big business (e.g., IBM, CISCO, Microsoft, Intel,
information produced by these various forms of Siemens, Oracle, SAP) that, on the one hand, are
everyware, it is argued, provides a more cohesive pushing for the adoption of their new technologies and
and smart understanding of the city that enhances services by cities and states and, on the other, are
efficiency and sustainability (Hancke et al. 2013, seeking deregulation, privatisation and more open
Townsend 2013) and provides rich seams of data that economies that enable more efficient capital accumu-
can used to better depict, model and predict urban lation. For city officials, national governments and
processes and simulate the likely outcomes of future supra-national states such as the European Union,
urban development (Schaffers et al. 2011; Batty et al. smart cities offer the enticing potential of socio-
2012). Everyware thus works to make a city knowable economic progress—more liveable, secure, func-
and controllable in new, more fine-grained, dynamic tional, competitive and sustainable cities, and the
and interconnected ways that ‘‘improve[s] the perfor- renewal of urban centres as hubs of innovation and
mance and delivery of public services while support- work (Kourtit et al. 2012; Townsend 2013). Hollands
ing access and participation’’ (Allwinkle and (2008) thus identifies five main characteristics of a
Cruickshank 2011: 2). It also provides the supporting smart city as evidenced by industry and government
infrastructure for business activity and growth and literature: widespread embedding of ICT into the
stimulates new forms of entrepreneurship, especially urban fabric; business-led urban development and a
with respect to the service and knowledge economy. neoliberal approach to governance; a focus on social
On the other hand, the notion of a ‘smart city’ is and human dimensions of the city from a creative city
seen to refer more broadly to the development of a perspective (alia Florida 2004); the adoption of a
knowledge economy within a city-region (Kourtit smarter communities agenda with programmes aimed
et al. 2012). From this perspective, a smart city is one at social learning, education and social capital; and a
whose economy and governance is being driven by focus on social and environmental sustainability.
innovation, creativity and entrepreneurship, enacted These five characteristics, Hollands (2008) suggests,
by smart people. Here, ICT is seen as being of central leads to an inevitable tension within smart cities
importance as the platform for mobilising and realis- between: serving global, mobile capital and stationary
ing ideas and innovations, especially with respect to ordinary citizens; attracting and retaining an elite
professional services. In and of itself, however, the creative class and serving other classes; and top-down,
embedding of ICT in urban infrastructure is not seen to corporatized, centralized development and bottom-up,
make a city smart (Hollands 2008). In other words, it is grassroots, decentralised and diffuse approaches.
how ICT, in conjunction with human and social capital Another vital conjoin between these two visions of
and wider economic policy, is used to leverage growth a smart city is the prioritisation of data capture and
and manage urban development that makes a city analysis as a means for underpinning evidence-
smart (Caragliu et al. 2009). Whereas the first vision of informed policy development, enacting new modes
a smart city focuses on ICT and its use in managing of technocratic governance, empowering citizens
and regulating the city from a largely technocratic and through open, transparent information, and stimulat-
technological perspective, the second encompasses ing economic innovation and growth. Data are thus
policies related to human capital, education, economic viewed as essential constituent material to realising a

123
GeoJournal (2014) 79:1–14 3

smart city vision. Such data are seen as providing studies that are characterised by data scarcity (Miller
objective, neutral measures that are free of political 2010).
ideology as to what is occurring in a city, with the The hype and hope of big data is a transformation in
weight of data speaking an inherent truth about social the knowledge and governance of cities through the
and economic relations and thus providing robust creation of a data deluge that seeks to provide much
empirical evidence for policy and practice (Mayer- more sophisticated, wider-scale, finer-grained, real-
Schonberger and Cukier 2013). And yet, there has time understanding and control of urbanity. There is
been to date been little critical focus on the new forms no agreed academic or industry definition of big data,
of data being produced (or not produced), how they are but a survey of the emerging literature denotes a
being mobilised by business, government and citizens, number of key features. Big data are:
and the implications of real-time data analytics.
• huge in volume, consisting of terabytes or peta-
In this paper, the data explosion that has occurred
bytes of data;
over the past decade, the role of cities as key sites in
• high in velocity, being created in or near real-time;
the production of such data, and how these data are
• diverse in variety, being structured and unstruc-
being used to re-imagine and regulate the urban life
tured in nature, and often temporally and spatially
are examined. In particular, the analysis concentrates
referenced;
on the new phenomena of ‘big data’ and the generation
• exhaustive in scope, striving to capture entire
of enormous, varied, dynamic, and interconnected
populations or systems (n = all), or at least much
datasets that hold the promise of what some see as a
larger sample sizes than would be employed in
truly smart city—one that can be known and managed
traditional, small data studies;
in real-time and is sentient to some degree (Batty et al.
• fine-grained in resolution, aiming to be as detailed
2012; Townsend 2013). I detail a number of projects
as possible, and uniquely indexical in
that aim to produce a real-time overview and analysis
identification;
of the city, and provide a critical reflection on big data
• relational in nature, containing common fields that
and smart urbanism.
enable the conjoining of different data sets;
• flexible, holding the traits of extensionality (can
add new fields easily) and scaleability (can expand
Big data and cities
in size rapidly).
There has long been the production of very large (see Boyd and Crawford 2012; Dodge and Kitchin
datasets, such as national censuses, government 2005; Laney 2001; Marz and Warren 2012; Mayer-
records and geomatic surveys, that provide informa- Schonberger and Cukier 2013; Zikopoulos et al.
tion about cities and their citizens. Likewise, busi- 2012).
nesses have collated significant amounts of data about In other words, big data consists of massive,
their operations, markets and customers. However, dynamic, varied, detailed, inter-related, low cost
these datasets often rely on samples, are generated on a datasets that can be connected and utilised in diverse
non-continuous basis, the number of variables are ways, thus offering the possibility of studies shifting
quite small, are aggregated to a relatively coarse from: data-scarce to data-rich; static snapshots to
spatial scale, and are often limited in access. As a dynamic unfoldings; coarse aggregation to high res-
result, these large datasets have been complemented olution; relatively simple hypotheses and models to
by what might be termed ‘small data’ studies— more complex, sophisticated simulations and theories
questionnaire surveys, case studies, city audits, inter- (Kitchin 2013).
views and focus groups, and ethnographies—that There is little doubt that since the early 2000s there
capture a relatively limited sample of data that are has been a transformation in the volume of data
tightly focused, time and space specific, restricted in generated. Zikopoulos et al. (2012) detail that in 2000
scope and scale, and relatively expensive to generate c.800,000 petabytes (250 bytes) of data were stored in
and analyze, to provide additional depth and insight the world. By 2010, MGI (cited in Manyika et al.
with respect to specific phenomena. Much of what we 2011: 3) ‘‘estimated that enterprises globally stored
know about cities to date then has been gleaned from more than 7 exabytes [260 bytes] of new data on disk

123
4 GeoJournal (2014) 79:1–14

drives… while consumers stored more than six against various databases in real-time, and new data
exabytes of new data on devices such as PCs and are generated such as CCTV, photographs, finger-
notebooks.’’ They further estimated that in ‘‘2009, prints or iris scans; or spatial video, LiDAR, thermal or
nearly all sectors in the US economy had at least an other kinds of electromagnetic scans of environments
average of 200 terabytes [240 bytes] of stored data… that enables mobile and real-time 2D and 3D mapping.
per company with more than 1,000 employees. Many In the case of automated data, data are generated as an
sectors had more than one petabyte in mean stored data inherent, automatic function of the device or system.
per company.’’ Based on their review of data volume There are a number of different means by which
growth, Manyika et al. (2011) projected a growth of automated data are produced, including: capture
40 % in data generated globally per year. Such is the systems, in which the means of performing a task
phenomenal growth in data production, Hal Varian, captures data about that task (such as scanning items at
Chief Economist at Google (cited in Smolan and a check-out till being used to monitor the till-operators
Erwitt 2012), estimates that more data are being performance, as well as collecting information with
produced every 2 days at present than in all of history regards to the items purchased and who purchased
prior to 2003 and Zikopoulos et al. (2012) expects data them); digital devices, such as mobile phones, that
volumes to reach 35 zetabytes [270 bytes] by 2020. In record and communicate the history of their own use;
2013, EU commissioner for Digital Agenda, Neelie transactions and interactions across digital networks
Kroes, stated that 1.7 million billion bytes of data per that not only transfer information, but generate data
minute were being generated globally (Rial 2013). about the transactions and interactions themselves
Such explosive growth in data is due to a number of (such as indexical logs of payments or bank transfers
different enabling and driving technologies, infrastruc- or email); clickstream data that records how people
tures, techniques and processes, and their rapid navigate through a website or app; sensed data
embedding into everyday practices and spaces. These generated by a variety of sensors and actuators
include the widespread roll-out of fixed and mobile embedded into objects or environments that regularly
internet; the development of ubiquitous computing and communicate their measurements; the scanning of
the ability to access networks and computation in many machine-readable objects such as travel passes, pass-
environments and on the move; the embedding of ports, or barcodes on parcels that register payment and
software into all kinds of machines transforming them movement through a system; and machine to machine
from ‘dumb’ to ‘smart’ and the creation of a plethora of interactions across the internet of things (Kitchin and
purely digital devices; the roll-out of social media and Dodge 2011). In contrast, volunteered data are gifted
Web 2.0 applications; advances in database design and by users. These include: interactions across social
systems of information management; the distributed media such as the posting of comments, observations
storage of data at affordable costs; and new forms of and the uploading of photos to social networking sites
data analytics designed to cope with data abundance such as Facebook or Twitter; and the crowdsourcing of
(Dodge and Kitchin 2005; Greenfield 2006; Kitchin data wherein users generate data and then contribute
and Dodge 2011). These developments not only enable them to a common system, such as the generation of
the accessing and sharing of data, but are also the GPS-traces uploaded into OpenStreetMap to create a
means by which much big data are generated. For common, open mapping system (Kitchin and Dodge
example, mobile devices such as smartphones allow 2011; Sui et al. 2012).
their users to access information at the same time as Whilst directed and volunteered data can provide
they record the information accessed, and when and useful insights into urban systems and city lives, it is
where it was requested and how it was used. automated forms of data generation that have most
The sources of big data can be broadly divided into caught the imagination of those concerned with
three categories: directed, automated and volunteered. understanding and managing cities. In particular, there
Directed data are generated by traditional forms of has been an interest in automated forms of surveil-
surveillance, wherein the gaze of the technology is lance, sensor networks and the internet of things, and
focused on a person or place by a human operator. the tracking and tracing of people and objects. Here,
Such systems include immigration passport control the city is envisaged as ‘‘constellations of instruments
where passenger details are collected and checked across many scales that are connected through

123
GeoJournal (2014) 79:1–14 5

multiple networks which provide continuous data empty spaces in a car park, and track the progress of
regarding the movements of people and materials’’ buses and trains along a route, and smart tickets, such as
and the status of various structures and systems (Batty the Oyster card on the London Underground, can be
et al. 2012: 482). As such, the instrumented city offers used to trace passenger travel. All of these forms of data
the promise of an objectively measured, real-time are growing rapidly (by 2013 over ten billion objects
analysis of urban life and infrastructure. were connected to the in internet of things, with this set
Automated forms of surveillance include: anony- to rise to over 50 billion by 2020; Farber 2013).
mous paper tickets being replaced with automatically Some of these data are generated by local govern-
trackable ‘smart cards’; automatic number plate ments and state agencies, and some by private
recognition (ANPR) systems that use digital cameras companies, and by no means are they all open in
to scan license plates and pattern match the details to nature. Nevertheless for urban managers these forms
owner details and can be used to trace vehicles as they of instrumentation provide abundant, systematic,
cross a city and provide inputs into intelligent dynamic, well-defined, resolute, relatively cheap data
transportation systems (ITS); automatic meter reading about city activities and processes, enabling the
(AMR) that communicates utility usage without the possibility of real-time analytics and adaptive forms
need for manual reading and can do so on a continuous of management and governance (Kloeckl et al. 2012).
basis; and automated monitoring of public service
provision, such as RFID chips attached to rubbish bins
detecting whether they have been collected (Dodge The real-time city
and Kitchin 2007a; Hancke et al. 2013). Sensor
networks consist of an array of very small, inexpensive Many city governments now use real-time analytics to
sensors or actuators that can be embedded or placed on manage aspects of how a city functions and is
different structures to measure specific outputs such as regulated. Perhaps the most common example relates
levels of light, humidity, temperature, gas, electrical to movement of vehicles around a transportation
resistivity, acoustics, air pressure, movement, speed, network, where data from a network of cameras and
and so on. Sensors can be passive and read by transponders are fed back to a central control hub to
scanners, or can be active, broadcasting data at regular monitor the flow of traffic and to adjust traffic light
intervals over local or wide area networks, or they sequences and speed limits and to automatically
might have near field communication (NFC) capabil- administer penalties for traffic violations (Dodge and
ities that enables two-way communication (Hancke Kitchin 2007a). Similarly, the police might monitor a
et al. 2013). Sensors networks can be used to monitor suite of cameras and live incident logs in order to
the use and condition of public infrastructures, such as efficiently and reactively direct appropriate resources
bridges, roads, buildings, and utility provision, as well to particular locations. Data relating to environmental
as general environmental conditions within a city. conditions might be collated from a sensor network
Urban places are also now full of objects and distributed throughout the city, for example measuring
machines that are uniquely indexical that conduct air pollution, water levels or seismic activity. Many
automatic work and are part of the internet of things, local governments use management systems to log
communicating about their use and traceable if they are public engagement with their services and to monitor
mobile. These include automatic doors, lighting and whether staff have dealt with any issues. In nearly all
heating systems, security alarms, wifi router boxes, cases, these are isolated systems dealing with a single
entertainment gadgets, television recorders, and so on. issue and are controlled by a single agency.
Many of these devices transfer data between each other, More recently there has been an attempt to draw all
in turn leading to new derived data. Devices such as of these kinds of surveillance and analytics into a
mobile phones can be traced through space by triangu- single hub, supplemented by broader public and open
lation across phone masts and others with built-in GPS data analytics. For example, the Centro De Operacoes
receivers, such as mobile phones, tablets, and satnavs, Prefeitura Do Rio1 in Rio de Janeiro, Brazil, a
can record and transmit their own trails. Transponders
can be used to monitor throughput at toll-booths,
1
measuring vehicle flow along a road or the number of http://www.centrodeoperacoes.rio.gov.br/.

123
6 GeoJournal (2014) 79:1–14

Fig. 1 The Centro De Operacoes Prefeitura Do Rio in Rio de Janeiro, Brazil. Source George Magaraia, http://ultimosegundo.ig.com.
br/brasil/rj/2012-05-03/ig-visita-o-centro-de-operacoes-do-rio-de-janeiro.html

partnership between the city government and IBM, one-stop data analytic hub to weave together data from
have created a citywide instrumented system that a diverse set of city agencies in order to try and
draws together data streams from thirty agencies, manage, regulate and plan the city more efficiently and
including traffic and public transport, municipal and effectively. Terabytes of data stream through the office
utility services, emergency services, weather feeds, on a daily basis enabling the analysts to cross-
and information sent in by employees and the public reference data, spot patterns and identify and solve
via phone, internet and radio, into a single data city problems (Feuer 2013). They have also started to
analytics centre (see Fig. 1). Here, algorithms and a make some of the data available in open form,2
team of analysts process, visualize, analyze and enabling developers to build apps that take the data
monitor a vast amount of live service data, alongside and rework and repackage it for daily consumption by
data aggregated over time and huge volumes of city dwellers. Likewise, Dublinked,3 provides opera-
administration data that are released on a more tional data from Dublin’s four local authorities in an
periodic basis, often mashing the datasets together to open format, and many other municipal governments
investigate particular aspects of city life and change around the world have started to release various kinds
over time, and to build predictive models with respect of administrative and operational data using various
to everyday city development and management and kinds of open data models (see Ferro and Osella 2013
disaster situations such as flooding. This is comple- for an overview of eight different models). An
mented by a virtual operations platform that enables example of an app using such open municipal data is
city officials to log-in from the field to access real-time SmartSantanderRA an augmented reality app that
information. For example, police at an accident scene provides information on about 2,700 places in the city
can use the platform to see how many ambulances of Santander (beaches, park and gardens, monuments,
have been dispatched and when, and to upload points of interest (POI), tourism offices, shops,
additional information (Singer 2012). The stated aim galleries, museums, libraries, public buses, taxis,
of the city’s mayor, Eduardo Paes, was ‘‘to knock bikes, parking places, and so on), along with real time
down silos… [between] departments and combine access to traffic and beaches cameras, weather reports
each one’s data to help the whole enterprise’’ (Singer
2012).
2
Similarly, the Office of Policy and Strategic https://nycopendata.socrata.com/.
3
Planning for New York city has sought to create a http://www.dublinked.ie/.

123
GeoJournal (2014) 79:1–14 7

Fig. 2 SmartSantanderRA augmented reality app. Source http://www.smartsantander.eu/index.php/blog/item/174-smartsantanderra-

santander-augmented-reality-application?template=retro

and forecast, public buses information and bike-rental fire and rescue, communities, housing, health, and
service4 (see Fig. 2). tourism—though these data are more administrative in
In other cities, such as London, live feeds of real- nature and not in real-time. Rather than simply
time data are being communicated to citizens through providing the raw data, these sites produce visualisa-
what have been termed ‘city dashboards’. For exam- tions that aid the interpretation and analysis, especially
ple, in the London case5 (Fig. 3), developed by CASA for non-expert users, and allow citizens to monitor the
at UCL, citizens can find out real-time information city for themselves and for their own ends.
about the weather, air pollution, public transport For those developing and using integrated, real-
delays, public bike availability, river level, electricity time city data analytics, such centres, apps and
demand, the stock market, twitter trends in the city, dashboards provide a powerful means for making
look at traffic camera feeds, and even the happiness sense of, managing and living in the city in the here-
level. These data can also be mapped. This is and-now, and for envisioning and predicting future
complemented by the London Dashboard,6 a data scenarios. Rather than basing decisions on anecdote or
visualisation site that tracks the performance of the intuition or clientelist politics or periodic/partial
city with respect to twelve key areas—jobs and evidence, it is possible to assess what is happening
economy, transport, environment, policing and crime, at any one time and to react and plan appropriately.
Moreover, the use of large samples and the linking of
4
http://www.smartsantander.eu/index.php/blog/item/174-smarts
diverse forms of data provide a deeper, more holistic
antanderra-santander-augmented-reality-application. and robust analysis. For advocates of such systems it
5
http://citydashboard.org/london/. thus becomes possible to develop, run, regulate and
6
http://data.london.gov.uk/london-dashboard. live in the city on the basis of strong, rationale

123
8 GeoJournal (2014) 79:1–14

Fig. 3 CASA’s London City Dashboard. Source http://citydashboard.org/london/

evidence rather weak, selective evidence and political about the world. Data can be taken at face value; they
ideology. As such, it is argued, the use of such big data are pre-analytic and rhetorical (Rosenberg 2013).
provides the basis for a more efficient, sustainable, Likewise, the algorithms used to process these data are
competitive, productive, open and transparent city. neutral and non-ideological in their formulation and
But just as smart urbanism underpinned by big data operation, grounded in scientific objectivity (Kitchin
offers a seemingly attractive vision of future cities, it and Dodge 2011). Such a framing of data and
also raises a number of concerns, five of which will algorithms enable smart city projects themselves to
now be examined in brief. present an image of being politically benign and
commonsensical; that big data urbanism is inherently
a good thing, seeking to make a city safer, more
Five concerns about a real-time city secure, efficient, productive, sustainable and so on by
employing rigorous, technical practices that capture,
The politics of big urban data process and analyze vast quantities of transparent,
neutral, objective data. Data, however, are more
Data within smart city initiatives are portrayed as complicated than that. Data do not exist independently
being benign and lacking in political ideology. Data of the ideas, techniques, technologies, people and
are simply data: natural and essential elements that are contexts that conceive, produce, process, manage,
abstracted from the world in neutral and objective analyze and store them (Bowker and Star 1999;
ways subject to technical constraints. Sensors and Lauriault 2012; Ribes and Jackson 2013). As Gitelman
cameras have no politics or agenda. They simply and Jackson (2013: 2, following Bowker) put it, ‘‘raw
measure light or heat or humidity, and so on— data is an oxymoron’’; ‘‘data are always already
producing readings and pictures that reflect the truth ‘cooked’ and never entirely ‘raw’.’’ As such, no data

123
GeoJournal (2014) 79:1–14 9

are pre-analytic, or are objective, value-free, and Technocratic governance and city development
benign. What data are generated is the product of
choices and constraints, shaped by a system of The drive towards managing and regulating the city
thought, technical know-how, public and political via information and analytic systems promotes a
opinion, ethical considerations, the regulatory envi- technocratic mode of urban governance which pre-
ronment, and funding and resourcing. Data then are sumes that all aspects of a city can be measured and
situated, contingent, relational, and framed and used monitored and treated as technical problems which
contextually to try and achieve certain aims and goals. can be addressed through technical solutions; display-
It is no different with big data used to underpin ing what Mattern (2013) terms ‘instrumental rational-
smart urbanism. Whilst big data may seek to be all- ity’ and Morozov (2013) calls ‘solutionism’, wherein
encompassing, exhaustive and politically benign, as complex social situations can be disassembled into
with all data they are a selective sample and are neatly defined problems that can be solved or
framed within a thought system. What data are optimized through computation. Here, there is a
captured is shaped by: the field of view/sampling reification of big data; they can provide the answer
frame (where data capture devices are deployed, what to all problems (Mattern 2013). By capturing a
their settings/parameters are, who uses a space or phenomena as real-time data it seemingly becomes
media); the technology and platform used (different possible to model, understand, manage and fix a
surveys, sensors, lens, textual prompts, layout, etc. all situation as it unfolds. As Hill (2013) puts it: ‘‘[smart
produce variances and biases in what data are city thinking] betrays a technocratic view that the city
generated); the context in which data are generated is something we might understand in detail, if only we
(unfolding events mean data are always situated and had enough data—like an engine or a nuclear power
contextualised with respect to circumstance); the data station—and thus master it through the brute force
ontology employed (how the data are calibrated and science and engineering.’’ Indeed, Mattern (2013)
classified); and the regulatory environment with suggests that big data urbanism suffers from ‘‘datafi-
respect to privacy, data protection and security cation, the presumption that all meaningful flows and
(Kitchin 2013). Big data generally captures what is activity can be sensed and measured.’’ Within such
easy to ensnare—data that are openly expressed (what thinking there is ‘‘an often-explicit assumption that the
is typed, swiped, scanned, sensed; people’s actions universe is formed with knowable and definable
and behaviours; the movement of things)—as well as parameters [that] assures us that if we were only able
data that are the ‘exhaust’, a by-product, of the to measure them all, we would be able to predict and
primary task/output. It takes these data at face-value, respond with perfection accordingly’’ (Haque 2012).
despite the fact that they may not have been designed Employing an evidence-based, algorithmic processed
to answer specific questions and the data produced approach to city governance thus seemingly ensures
might be messy, dirty, full of occlusions and biases. It rational, logical, and impartial decisions. Moreover, it
is less well suited to contextualising such data or provides city managers with a defence against deci-
revealing the complex contingent and relational inner sions that raise ethical and accountability concerns by
lifeworlds of people and places. Moreover, the data enabling them to say, ‘It’s not me, it’s the data!’
are generated within systems designed to enact a (Haque 2012).
particular political and policy vision. The result are However, technocratic forms of governance are
data that are inflected by social privilege and social highly narrow in scope and reductionist and function-
values, especially within domains that function as alist in approach, based on a limited set of particular
disciplinary systems (such as law enforcement) kinds of data and failing to take account of the wider
(Johnson 2013). There is no doubt that big data effects of culture, politics, policy, governance and
initiatives do produce data that are useful for under- capital that shape city life and how it unfolds.
standing and managing cities, but the politics and Technological solutions on their own are not going
limitations of such data and the methods used to to solve the deep rooted structural problems in cities as
produce and analyze them need to be teased apart and they do not address their root causes. Rather they only
examined as to the values and agendas underpinning enable the more efficient management of the mani-
them and whose interests they serve. festations of those problems. As such, whilst smart

123
10 GeoJournal (2014) 79:1–14

city technologies, such as real time analytics are that beholden cities to particular technological plat-
promoted as the panacea for tackling urban gover- forms and vendors over a long period of time creating
nance issues, they largely paper over the cracks rather monopoly positions (Hill 2013). The danger here is the
than fixing them, unless coupled with a range of other creation of a corporate path dependency that cannot
policies. Further, control and command systems easily be undone or diverted (Bates 2012). As Hill
centralise power and decision making into a select (2013) details, the strategy adopted by IT corporations
set of offices, at the same time that they make elements mirrors that of US car manufacturers in the mid-
of the data publicly available. There is clearly a twentieth century in creating a form of technology-led
delicate balance to maintained as new forms of urbanism centred on car transportation. Here, public
technologically rooted monitoring and management transport networks were closed down to be replaced by
are rolled out. On the one hand, such technologies a vast road building programme that then shaped
enable aspects of the city to managed more efficiently patterns of urban development in the following
and effectively on a dynamic basis rooted in a strong decades. Haque (2012) thus wonders ‘‘what the smart
evidence-base. On the other, these data and technol- city equivalents might be of Robert Moses’ tangled,
ogies need to be complemented with a range of other congested and polluted freeways or the failures of the
instruments, policies and practices that are sensitive to Pruitt Igoe housing complex.’’ Third, that it leads to
the diverse ways in which cities are structured and ‘one size fits all smart city in a box’ solutions that take
function. little account of the uniqueness of places, peoples and
cultures and straightjackets city administrations into a
The corporatisation of city governance narrowly visioned technocratic mode of governance
and a technological lock-in (Townsend et al. nd). Indeed, IBM is now selling a
product called ‘IBM Intelligent Operations Center’,
Alongside the critique that smart city governance is which combines a number of the systems that were
too technocratic in nature is a concern that it is being designed for Rio into a single product that can be
captured and overtly shaped by corporate interests for applied to any city (Singer 2012). Given these
their own gain (Greenfield 2013; Townsend 2013). concerns, Hill (2013) thus warns that ‘‘[l]iterally
The smart city agenda and associated technologies are hardwiring urban services to a particular device, a
being heavily promoted by a number of the world’s particular operating system, is a recipe for disaster, not
largest software services and hardware companies efficiency… Put simply, city fabric changes slowly yet
who view city governance as a large, long-term technology changes rapidly… There is a worrying lack
potential market for their products. Either through of thought about adaptation in this desire to install the
being major partners in building cities from the ground consumer tech layer as if it were core building
up (e.g., Songdo or Masdar City), or partnering with services.’’ That’s not say that such a corporate lock-
established cities to retrofit their infrastructure with in is inevitable, but it is clear that is the desire of a
digital technology and data solutions, these companies number of very large corporate players.
have been seeking to make their wares a core,
indispensible part of how various aspects of city life Buggy, brittle and hackable cities
are monitored and regulated. As such, as Schaffers
et al. (2011: 437) note, ‘‘smart city solutions are The embedding and use of ubiquitous and pervasive
currently more vendor push than city government pull computing in city environments is creating city
based’’, with companies working to build working services and spaces that are dependent on software
relationships, put in place favourable market condi- to function. Dodge and Kitchin (2004) term these
tions, divert funding streams and create public–private environments code/spaces, wherein software and the
partnerships. spatiality of everyday life become mutually consti-
The concern around such a move is three-fold. First, tuted, so that if the software fails a space is not
that it actively promotes a neoliberal political economy produced as intended as the old analogue system and
and the marketisation of public services wherein city associated tacit knowledge has been entirely replaced.
functions are administered for private profit (Hollands For example, if the software used to control a subway
2008). Second, that it creates a technological lock-in system crashes, then the trains do not run (as has

123
GeoJournal (2014) 79:1–14 11

happened in many cities in the past few years; see all highly vulnerable to direct attacks that force objects
Townsend 2013); or if a supermarket’s checkout tills to exceed their design parameters or operate in
crash, shoppers cannot make purchases as goods dangerous ways, misdirection through distorting read-
cannot be scanned or payments made, with the shop ings leading to user error and damage, and the theft of
effectively becoming a warehouse. As such, whilst information. Whilst the nascent deployment of smart
potentially solving a diverse set of urban problems, the city technologies have had some teething issues,
creation of code/spaces through smart city projects contra to the concerns expressed above they have been
leaves cities vulnerable to other issues. In particular, it relatively robust despite their vulnerabilities. None-
has the potential to create buggy, brittle and exposed theless, as more systems are deployed, Townsend
city services and spaces that are prone to viruses, concludes: ‘‘The only questions will be when smart
glitches, crashes, and security hacks that can cause cities fail, and how much damage they cause when
havoc (Kitchin and Dodge 2011; Townsend 2013). As they crash.’’
Kitchin and Dodge (2011) detail, software is an
unusual product because it is sold in full knowledge The panoptic city?
that it is inherently partial, provisional, porous and
open to failure. Software-enabled technologies, espe- Over the past couple of decades, with the development
cially those that are networked and distributed, of various forms of directed, automated and networked
routinely have to be patched and updated to cope with digital technologies, there have been increasing con-
new contingencies. And as systems become ever more cerns over the rising level of surveillance in societies,
complicated, interconnected and dependent on soft- explicitly acknowledging the politics of data. It is now
ware, the challenge of producing stable, robust and possible to track and trace individuals and their
secure devices and infrastructures increases (Town- actions, interactions and transactions in minute detail
send 2013). across a number of domains (work, travel, consump-
What are the implications then of creating exten- tion, etc.). This level of monitoring has been driven by
sive city systems that are reliant on software to a growing ‘culture of control’ that desires ‘security,
function? Of taking two highly complex, contingent orderliness, risk management and the taming of
and open systems—cities and information systems— chance’ (Garland 2001, cited in Lyon 2007: 12).
and interweaving them together? Or as Townsend However, despite systems becoming more wide-
(2013) puts it: ‘‘What if the seeds of smart cities’ own spread, fine-grained and sophisticated, they have
destruction are already built into their DNA? What if largely operated as independent systems and the
the smart cities of the future are buggy and brittle? notion of a panopticon (an all-seeing vantage point)
What are we getting ourselves into?’’ He suggests that has remained open to vertical (within an activity) and
as more and more systems are layered on top of ICT horizontal (across activities) fragmentation due to
networks, each in a relatively brittle state, the risks of agencies communicating imperfectly or being unable
critical failures and ‘normal accidents’ (everyday or unwilling to exchange or compare information
glitches) become compounded. At the same time (Hannah 1997). Governance has thus consisted of a set
analogue alternatives are disappearing. Moreover, of oligopticons—partial vantage points from fixed
ever more systems are becoming open to malicious positions with limited view sheds (Amin and Thrift
forms of attack. For example, the Israeli government 2002).
acknowledges that its essential services such as water, Big data and data control centres, such as the Centro
electricity, banking, rail and road infrastructure is the De Operacoes Prefeitura Do Rio, that integrate and
target of numerous cyber attacks, with Israel Electric bind data streams together, work to move the various
Corp reporting that it receives 6,000 attempted hacks oligopticon systems into a single, panoptic vantage
every second (Paganini 2013). And in October 2012, point and raise the spectre of a Big Brother society
the traffic management system for a major artery in based on a combination of surveillance (gazing at the
Haifa was hacked causing traffic chaos for hours world) and dataveillance (trawling through and inter-
(Paganini 2013). And beyond critical systems, Mims connecting datasets), and a world in which all aspects
(2013) reports that smartphones, tablets, and the of a citizen’s life are captured and potentially never
various devices making up the internet of things are forgotten (Dodge and Kitchin 2007b). There is an

123
12 GeoJournal (2014) 79:1–14

inherent tension then in the creation of systems that opportunities, they also raise a number of concerns
seek to enable more effective modes of governance with respect to the politics of such data, technocratic
which also threaten to stifle rights to privacy, confi- governance, the corporatisation and further neoliber-
dentiality, and freedom of expression. As more and alisation of city management, the possibilities of a
more aspects of urban life are captured as data in technological lock-in, system vulnerabilities, ethical
dynamic ways at finer resolutions, this tension is set to issues with respect to surveillance, dataveillance and
grow and it will be important to balance the benefits of control, as well as other concerns relating to data
data analytics with individual and societal rights in quality, fidelity, security, the validity of analytics that
order to maintain democracy and trust in government, utilise data dredging techniques, and how data are
especially when so much of the data will be processed interpreted and acted upon. Given the role that such
by corporate systems. Without regulated oversight and systems are likely to play in shaping urban governance
enforcement concerning abuses of data, then there is there is a pressing need to interrogate the nature and
likely to significant resistance and push-back against production of urban big data, the composition and
real-time analytics by citizens. functioning of urban analytics and control centres, and
the implications of technocratic, corporatised and real-
time forms of governance. This paper has provided
Conclusion some initial entry points, but wider synoptic overviews
and in-depth empirical studies are required to examine
The notion of smart cities has gained much traction in existing and potential smart urbanism. As Greenfield
recent years as a vision for stimulating and supporting (2013) and Townsend (2013) argue, without such
innovation and economic growth, and providing critical interrogations the smart cities of the future will
sustainable and efficient urban management and likely reflect narrow corporate and state visions, rather
development. One significant aspect of the smart than the desires of wider society.
cities concept is the production of sophisticated data
analytics for understanding, monitoring, regulating Acknowledgments An early version of this paper was
originally presented at the ‘Smart Urbanism: Utopian Vision
and planning the city. As cities have become increas-
or False Dawn’ workshop at the University of Durham, 20–21
ingly embedded with all kinds of digital infrastructure June 2013. Many thanks to the organisers and attendees for
and networks, devices, sensors and actuators, the constructive feedback.
volume of data produced about them has grown
exponentially, providing rich streams of information
about cities and their citizens. Such big data are varied,
fine-grained, indexical, dynamic and relational References
enabling real-time analysis of different systems and
to interconnect data across systems to provide detailed Allwinkle, S., & Cruickshank, P. (2011). Creating smart-er
cities: An overview. Journal of Urban Technology, 18(2),
views of the relationships between data. For citizens 1–16.
such data and its analysis offers insights into city life, Amin, A., & Thrift, N. (2002). Cities: Reimagining the Urban.
aids everyday living and decision-making, and London: Polity.
empowers alternative visions for city development. Bates, J. (2012). ‘‘This is what modern deregulation looks like’’:
Co-optation and contestation in the shaping of the UK’s
For governments, big data and integrated analysis and Open Government Data Initiative. The Journal of Com-
control centres offer more efficient and effective city munity Informatics, 8(2). http://www.ci-journal.net/index.
management and regulation. For corporations, big php/ciej/article/view/845/916. Accessed 6 Feb 2013.
data analytics offers new, long term business oppor- Batty, M., Axhausen, K. W., Giannotti, F., Pozdnoukhov, A.,
Bazzani, A., Wachowicz, M., et al. (2012). Smart cities of
tunities as key players in city governance. the future. European Physical Journal Special Topics,
Over the next decade, the real-time city is likely to 214(1), 481–518.
become a reality in many cities as urban administra- Bowker, G., & Star, L. (1999). Sorting things out: Classification
tions seek to capitalise on new data streams and new and Its consequences. Cambridge: MIT Press.
Boyd, D., & Crawford, K. (2012). Critical questions for big data.
commercial products are bought to market that help Information, Communication and Society, 15(5), 662–679.
governments and citizens make sense of the city. Caragliu, A., Del Bo, C., Nijkamp, P. ( 2009). Smart Cities in
Whilst such big data analytics offers a number of Europe. Series Research Memoranda 0048. VU University

123
GeoJournal (2014) 79:1–14 13

Amsterdam, Faculty of Economics, Business Administra- Illinois. http://papers.ssrn.com/abstract=2241092. Acces-

tion and Econometrics. sed 16 Aug 2013.
Dodge, M., & Kitchin, R. (2004). Flying through code/space: Kitchin, R. (2013). Big data and human geography: Opportuni-
The real virtuality of air travel. Environment and Planning ties, challenges and risks. Dialogues in Human Geography.
A, 36(2), 195–211. Kitchin, R., & Dodge, M. (2011). Code/space: Software and
Dodge, M., & Kitchin, R. (2005). Codes of life: Identification everyday life. Cambridge, MA: MIT Press.
codes and the machine-readable world. Environment and Kloeckl, K., Senn, O., & Ratti, C. (2012). Enabling the real-time
Planning D: Society and Space, 23(6), 851–881. city: LIVE Singapore! Journal of Urban Technology,
Dodge, M., & Kitchin, R. (2007a). The automatic management 19(2), 89–112.
of drivers and driving spaces. Geoforum, 38(2), 264–275. Komninos, N. (2002). Intelligent cities: Innovation. Routledge:
Dodge, M., & Kitchin, R. (2007b). Outlines of a world coming Knowledge Systems and Digital Spaces.
in existence’: Pervasive computing and the ethics of for- Kourtit, K., Nijkamp, P., & Arribas-Bel, D. (2012). Smart cities
getting. Environment and Planning B, 34(3), 431–445. perspective—A comparative European study by means of
Dutton, W. H., Blumler, J. G., & Kraemer, K. L. (1987). Wired self-organizing maps. Innovation, 25(2), 229–246.
cities: Shaping future communication. New York: Laney, D. (2001). 3D Data management: Controlling data
Macmillan. volume, velocity and variety. Meta Group. http://blogs.
Farber, D. (2013). Counting the internet of things in real time. gartner.com/doug-laney/files/2012/01/ad949-3D-Data-
C|Net, July 30th. http://news.cnet.com/8301-11386_3- Management-Controlling-Data-Volume-Velocity-and-Var
57596162-76/counting-the-internet-of-things-in-real-time/. iety.pdf. Accessed 16 Jan 2013.
Accessed 18 Sep 2013. Lauriault, T.P. (2012). Data, Infrastructures and Geographical
Ferro, E. & Osella, M. (2013). Eight business model archetypes Imaginations: Mapping Data Access Discourses in Can-
for PSI re-use. Open Data on the Web workshop. http:// ada. PhD Thesis, Carleton University, Ottawa.
www.w3.org/2013/04/odw/odw13_submission_27.pdf. Lyon, D. (2007). Surveillance studies: An overview. Cambridge:
Accessed 10 May 2013. Polity.
Feuer, A. (2013). The Mayor’s Geek Squad. New York Times, Manyika, J., Chiu, M., Brown, B., Bughin, J., Dobbs, R., Rox-
March 23rd. http://www.nytimes.com/2013/03/24/nyregion/ burgh, C., et al. (2011). Big data: The next frontier for
mayor-bloombergs-geek-squad.html. Accessed 9 May 2013. innovation, competition, and productivity. McKinsey
Florida, R. (2004). The rise of the creative class. New York: Global Institute.
Basic Books. Marz, N., & Warren, J. (2012). Big data: Principles and best
Gitelman, L., & Jackson, V. (2013). Introduction. In L. Gitel- practices of scalable realtime data systems. Manning:
man (Ed.), ‘‘Raw data’’ is an oxymoron (pp. 1–14). Cam- MEAP edition.
bridge: MIT Press. Mattern, S. (2013). Methodolatry and the art of measure: The
Graham, S., & Marvin, S. (1999). Planning cybercities: Inte- new wave of urban data science. Design Observer: Places.
grating telecommunications into urban planning. Town 5th November 2013. http://designobserver.com/places/
Planning Review, 70(1), 89–114. feature/0/38174/. Accessed 15 Nov 2013.
Greenfield, A. (2006). Everyware: The dawning age of ubiqui- Mayer-Schonberger, V., & Cukier, K. (2013). Big data: A rev-
tous computing. Boston: New Riders. olution that will change how we live, work and think.
Greenfield, A. (2013). Against the Smart City (The City is Here London: John Murray.
for You to Use). New York: Do Projects. Miller, H. J. (2010). The data avalanche is here. Shouldn’t we be
Hancke, G. P., de Carvalho e Silva, B., & Hancke, G. P, Jr. digging? Journal of Regional Science, 50(1), 181–201.
(2013). The role of advanced sensing in smart cities. Sen- Mims, C. (2013). Coming soon: the cybercrime of things. The
sors, 13(1), 393–425. Atlantic, August 6th. http://www.theatlantic.com/technology/
Hannah, M. (1997). Imperfect panopticism: Envisioning the archive/2013/08/coming-soon-the-cybercrime-of-things/
construction of normal lives. In G. Benko & U. Strohmayer 278409/. Accessed 15 Nov 2013.
(Eds.), Space and social theory (pp. 344–360). Oxford: Morozov, E. (2013). To save everything, click here: Technology,
Blackwell. solutionism, and the urge to fix problems that don’t exist.
Haque, U. (2012). What Is a City that It Would Be ‘Smart’? New York: Allen Lane.
Volume #34: City in a Box. http://volumeproject.org/blog/ Paganini, P. (2013). Israeli road control system hacked, causes
2012/12/21/volume-34-city-in-a-box/. traffic jam on Haifa highway. The Hacker News, October
Hill, D. (2013). On the smart city: Or, a ‘manifesto’ for smart 28th. http://thehackernews.com/2013/10/israeli-road-
citizens instead. City of Sound, 1st Feb 2013. http://www. control-system-hacked.html. Accessed 13 Nov 2013.
cityofsound.com/blog/2013/02/on-the-smart-city-a-call- Rial, N. (2013). The power of big data in Europe. New Europe,
for-smart-citizens-instead.html. Accessed 5 Feb 2013. May 24th. http://www.neurope.eu/article/power-big-data-
Hollands, R. G. (2008). Will the real smart city please stand up? europe. Accessed 27 May 2013.
City, 12(3), 303–320. Ribes, D., & Jackson, S. J. (2013). Data bite man: The work of
Ishida, T., & Isbister, K. (2000). Digital cities: Technologies, sustaining long-term study. In L. Gitelman (Ed.), ‘‘Raw
experiences, and future perspectives. LNCS: Springer. data’’ is an oxymoron (pp. 147–166). Cambridge: MIT
1765. Press.
Johnson, J.A. (2013). From open data to information justice. Rosenberg, D. (2013). Data before the fact. In L. Gitelman (Ed.),
Paper presented at the Annual Conference of the Midwest ‘‘Raw data’’ is an oxymoron (pp. 15–40). Cambridge: MIT
Political Science Association, April 13, 2013, Chicago, Press.

123
14 GeoJournal (2014) 79:1–14

Schaffers, H., Komninos, N., Pallot, M., Trousse, B., Nilsson, Sui, D., Elwood, S., & Goodhild, M. (Eds.). (2012). Crowd-
M., & Oliveira, A. (2011). Smart cities and the future sourced geographic knowledge: Volunteered geographic
internet: Towards cooperation frameworks for open inno- information in theory and practice. Dordrecht: Springer.
vation. In J. Domingue, et al. (Eds.), Future Internet Townsend, A. (2013). Smart cities: Big data, civic hackers, and
Assembly (pp. 431–446). LNCS: Springer. the quest for a new utopia. New York: W.W. Norton & Co.
Shepard, M. (2011). Sentient city: Ubiquitous computing, Townsend, A., Maguire, R., Liebhold, M. & Crawford, M. (no
architecture, and the future of urban space. Cambridge, date) A planet of civic laboratories: The future of cities,
Mass: MIT Press. information and inclusion. Institute for the Future, Palo
Singer, N. (2012). Mission control, built for cities: I.B.M. Takes Alto.
‘Smarter Cities’ Concept to Rio de Janeiro. New York Zikopoulos, P. C., Eaton, C., deRoos, D., Deutsch, T., & Lapis,
Times, 3 March 2012. http://www.nytimes.com/2012/03/ G. (2012). Understanding big data. New York: McGraw
04/business/ibm-takes-smarter-cities-concept-to-rio-de- Hill.
janeiro.html. Accessed 9 May 2013.
Smolan, R. & Erwitt, J. (2012). The Human Face of Big Data.
New York: Sterling.

123

View publication stats