The Gallery: Smart Cities Pose Challenges

state house gallery

Last Week, CivSource posted IDC’s recent predictions for smart cities in 2014. A reader and project manager Jack Winters, tweeted us his own list of 10 smart cities issues. We asked him to turn them into an article which is below.

(1) There will be increased surveillance

The current crop of smart environments by their very nature are about data collection, sharing, indexing and recording. If this were just about managing and maintaining infrastructure, there would be little to worry about but, the fundamental idea that smart cities vendors are selling is that the behaviour of people should be managed at a granular level by centralized administration or decentralized cybernetic systems.

Sensors monitor people and that data is sent to a centralized server. The data can be used for surveillance purposes in three ways. Firstly, directly by security agencies for managing crime, managing crowds, controlling demonstrations, controlling access, identifying behaviour perceived as abnormal or classifying people as likely to commit crimes. Second, through systems which track individuals for reasons of efficiency or planning. For example, tracking the mobile phones of people in order to build models of people’s travel patterns. Finally, though systems which seek to monitor and control behaviour through incentives or punishments. For example you might gain or lose points, money or ranking depending on whether your driving, fitness, energy usage, purchasing or commuting behaviour matches that desired by the system designers.

This creates a risk for democracy because notions of efficiency, safety & the good life are promoted by designed systems but the fundamental foundations of these concepts are never to be discussed and are hard to change by democratic means.

(2) Outsourcing contracts will be badly written

No industry has gone through greater outsourcing catastrophes than government IT. The examples are almost too numerous to cite and the reasons have not been addressed.

Governments and especially local governments typically lack in-house expertise in effectively writing and enforcing contracts. (See examples here and here.) They lack both contract management and technical expertise. They are generally outclassed by the lawyers working on behalf of the vendors and individuals are not personally incentivized to write good contracts. Revolving door practices can also lead to conflicts of interest.

(3) Cities will be locked in to vendors

Along with contractual lock-in, smart city infrastructure increases the risk of technical lock-in. A well designed system should ensure that critical infrastructure was not owned by vendors or based on proprietary technology, but rather based on open source and open standards. But even in this case specialist knowledge required to manage complex smart city systems is likely to be difficult to transfer. This will force cities to choose between a small number of large vendors.

(4) Marginalized people will be more excluded than they are now

Not only will profiling lead to new exclusions and privileges, it will make it more difficult for mobility. For example, security screening could be a pervasive function of the environment making it even more difficult for marginalised people to move and use facilities. But also the needs of people who slip below the radar of big data will be ignored by a doctrine of data based decision making. A world shaped by big data will take into account the habits and preferences of people with smart phones and clear identities. The needs of marginal people who are not sensed will not be taken into account.


(5) Participation will be gamified

Consider this scenario: you wake up, you put out the rubbish at the appropriate time (10 citizen points). You have breakfast, the fridge reports your healthy living choice to your insurer keeping you in the low premium category despite your missing the gym yesterday. Then you leave the house and walk to the bus stop racking up some green points for your decision not to use your car. It’s not enough for you to claim the title of ‘street champion’ that prize goes to the Patel’s next door for their participation in a crowd mapping exercise. They won big for reporting the O’Conner’s for messiness – relegating the O’Conner’s to bottom of the street league.

(6) Lots of consultants will get rich

It’s a buzz world laden area – it’s inevitable.

(7) Infrastructure will be hacked

A recent study by Proofpoint found that over 100,0000 internet of things devices have been hacked. Fridges sending spam email is probably not a primary concern for most people, but if infrastructure is hacked the consequences could be devastating.

(8) Complexity will decrease resilience

Increasing complexity and interdependency of systems increases the likelihood of unpredictable failure modes or cascading failures. Critical functions like power, water, transport, or emergency response may have unforeseen dependencies or interactions with cloud servers or IT systems. Maintenance and adaptation may also be more difficult with increasingly specialist skills being needed.

(9) There will be little democratic control over smart cities

One of the big promises of smart cities is that people can participate in the process of city management by reporting issues or voting on things. However, this is only limited participation within a designed system. It is democracy as Reddit is democracy – through the ability to like or vote up. That is not the ability to challenge the governing system itself. Because of the difficulty of changing vendors it will be very difficult to vote out the management company. Because of complexity and sunk investment it will be even more difficult to vote out the automated management systems.

(10) Cities will not be able to control vendors effectively

City authorities will lack the technical expertise to effectively monitor supplier actions and to manage complex projects. They will take on another tier of consultants to manage those suppliers. These companies may themselves be vendors and will be reluctant to upset suppliers they may need to work with elsewhere.

Further, while the objective of private sector is to maximize profits, the public sector has a much more complex set of objectives that involve the maximization of social welfare. These objectives are hard to define — let alone measure — thereby weakening the power of incentives to deliver. Despite this limitation, performance is likely to be measured using KPIs because there is a lack of contractually enforceable alternatives. Vendors will have a strong incentive to ensure KPIs are easy to achieve, rather than real measures of the underlying complex social objectives. Since they control the data, they will be in a strong position to manipulate statistics.


This is not to say that technological solutions and judicious outsourcing cannot be effective tools, but rather that care needs to be taken not only to ensure effective governance and avoid becoming locked into vendors. This also extends to making sure that technological solutions do not diminish democracy, and replace politics with purely managerial and technocratic systems which are difficult to debate, let alone vote out. We need to recognize that smart systems are a form of authority and we should subject them to the same democratic scrutiny, checks and balances as we do human authority.

Jack Winters in a project manager working in the areas of big data and embedded systems. He writes and works with the privacy advocacy group Stop The Cyborgs

The Gallery is a forum for ideas and examination of matters facing state and local government. Readers, members of the media, academics or the business community are invited to submit guest columns to bailey{at}civsourceonline{dot}com. Member of the public sector? We’re interested in hearing from you too. CivSource does not endorse the views presented in The Gallery, but offers them in an effort to present more diverse coverage. CivSource will review all submissions but does not guarantee publication of all works submitted.