Page 82 - ITU Journal - ICT Discoveries - Volume 1, No. 2, December 2018 - Second special issue on Data for Good

ITU JOURNAL: ICT Discoveries, Vol. 1(2), December 2018

depth or language, makes analytics and aggregation rality issues such as bad data contagion, the dissem-
difficult and more generally prevents the data from ination of erroneous information resulting in panic,
being used effectively. This challenge is accentuated and the dissemination of fake news [8]. Further-
by data ubiquity, which makes data governance more, the potential data dependencies generated by
very different from the governance of other re- data flows between systems raise the question of
sources to which they are often compared (oil, fi- the impact of a massive blackout (electricity, trans-
nancial resources, etc.). portation, banking, etc.) [9] or bankruptcy of a ma-
jor service provider. As described in the IRGC Guide-
Operational data governance is therefore essential, lines for the Governance of Systemic Risks [10],
to control, monitor and protect the ecosystem. This “systems prone to systemic risks are highly inter-
necessarily involves a cartographic understanding connected and intertwined with one another.”
of circulating and stored data: which data, who pro-
duces the data and why, where does the data come 4.2 Business resilience, future value and
from and where is it stored and secured? Building rising costs
information modeling (BIM) and city information
modeling (CIM) are examples of large-scale initia- Smart cities rely on IoT and on the implementation
tives whose objective is to foster governance and of innovative solutions and technologies such as
harmonize building and city data [3]. By extension, machine learning and blockchain, whose large-scale
operational data governance relies on the ability to use is still in the testing phase. The current eco-
capture the “big picture” of the ecosystem’s data nomic pressure on innovation could generate unex-
through data aggregation capabilities and end-to- pected future costs. As stated by Sculley, Holt,
end visibility of data processing [4]. Golovin, et al., “it is dangerous to think of these
quick wins as coming for free” [11]. This encourages
Smart cities are data-driven, but massive data col- us to consider technological debt as a risk. Moreo-
lection is not a panacea unless proper governance is ver, just like traditional businesses, players in the
in place [5]. In fact, in the long run, improving the data economy still have to prove their strength and
efficiency of data use may well involve minimizing resilience in this new era of technology, where a
data volumes. growing share of the value chain is based on intan-
gible capital [12].
4. NEW RISKS AND THREATS
4.3 Decision risk
The resilience and stability of smart sustainable cit-
ies requires active management of uncertainty. The Smart cities entail a growing number of data-based
emergence of new technologies and new business decisions related to a wide range of topics (energy,
models calls for faster implementation of an traffic, tax, safety, insurance, etc.) and stakeholders
adapted risk management approach, as safe growth (citizens, cities, companies, etc.) hoping for effec-
may be hampered by the many unknown unknowns tive, fair and unbiased decisions that will result in
generated by data management and processing. Ex- operational efficiency, sustainable economic
isting standards and methodologies [6] could be growth and social justice. The efficiency of the deci-
adapted to the specific needs, contexts and com- sion-making process depends on technical and non-
plexities of cities, communities and projects. But technical parameters such as algorithms, data qual-
whatever the chosen approach, the most important ity and governance, each of which could be a source
component is the ability to anticipate new risks and of bias or error. What, then, are the economic, social
threats. Data-related threats evolve as smart cities or environmental consequences of wrong decisions,
develop, so anticipating risks, cyber risks and bias or errors due to poor quality data, misinterpre-
threats is an ongoing process [7]. Below, we pro- tation or an inability to use the data effectively?
pose three specific risks that should be taken into
account. Projects under construction should be challenged in
the light of these risks and threats. But risk identifi-
4.1 Systemic risk cation is only the first step in implementing a risk
management approach. Other issues must also be
The smart cities paradigm in which systems and addressed by stakeholders, including the need to
data are interconnected raises questions about vi- improve risk assessment (likelihood and magni-
tude) through the collection and analysis of loss

60 © International Telecommunication Union, 2018