Smart cities have been more of a concept than a reality until now. As government hands more budget and power to elected mayors, a new generation of senior analysts will be needed to support evidence-based decision making.
Have you visited a smart city? If you have been to London, New York or Lyon, then you have, but you might not have realised it. On the ground, the evidence of how data and analytics are being applied to the running of a metropolis can be relatively limited - the countdown system at London bus stops is probably one of the most visible manifestations.
Behind the scenes, however, significant efforts are being made to harness the power of data to the political shifts currently taking place. With government pursuing a policy of city devolution, placing responsibility for policies and budgets into the hands of elected mayors, comes a parallel policy of applying big data and analytics to the challenges of public service consumption, business growth, congestion and pollution. Seeing proof that taxpayers’ money is delivering real benefits is also embedded in that thinking.
For this move to be successful, however, some important developments are required. The first is a scaled-up, better-integrated suite of initiatives, especially where the internet of things is concerned. IoT projects are proliferating in cities, from car pools to parking sensors, often driven by technology firms keen to test their innovations, but without any common purpose.
The second requirement is one familiar to any private sector business - the need for a central analytical function to capture this data, align it with the business need and make sense of it for the organisation. While the public sector in the UK is a leading practitioner of open data, in many respects it lags behind on analytics.
For that reason, when Policy Exchange published its Smart Cities report in January, one of its key recommendations was that, “as part of the city devolution negotiations, cities should explain how they will put in place key leadership roles such as a chief analytics officer and chief technology officer.” New York City is so far the only example of a CAO being appointed by former Mayor Bloomberg, generating a raft of insights and policy changes as a result. London and Manchester may well follow suit as they respond to the Government’s challenge.
Eddie Copeland, head of unit at Policy Exchange and co-author of the report, explains: “Clearly, there has been a lot of excitement around the goal of smart cities and the application of technology to important social and business goals. A lot is happening at the moment that doesn’t seem to be gaining traction - cities are testing technologies, but there is no follow-on.”
He argues that the devolution agenda is giving the smart cities movement a new relevance as a way of prioritising what local government does for its citizens. Just as the last decade has seen the use of data to support direct service provisioning, through better identity validation and a unified view of each citizen, so IoT should become a core component of how cities function.
“At the moment, UK cities don’t have the basics right,” says Copeland. “If a smart city is hungry for data, it will spot the gaps in what is being provided and they will be solved. As new technology arises, the risk is that there will be even more data it doesn’t know how to use unless it has a CAO.”
Politically, this concept has now moved into the very centre of thinking about elected mayors. In London, Conservative candidate for the mayoralty, Zac Goldsmith, has adopted the recommendation for a CAO into his manifesto. A new initiative in Greater Manchester is aiming to put it into the forefront of data sharing and analysis to help improve public services. Greater Manchester Combined Authority (GMCA) is proposing to establish a data sharing authority, called GM-Connect, to help break down the barriers which stop public services sharing information.
Copeland believes that the handful of UK cities with sufficient scale to establish a centre of excellence and deploy a COA should not be the end point for this idea. “It is not enough for each local authority (LA) to use data and technology to make its own processes efficient. The scale of larger cities should encourage LAs to work in a co-ordinated fashion across their regions. That is a massive opportunity.”
He believes this is not dependent on party politics - in London, Labour’s candidate Sadiq Khan looks set to adopt the recommendation, too. “In the 21st Century, having the ability to drive change based on data in smart cities is no longer a ‘nice to have’, it is a fundamental competency on which government depends,” says Copeland.
A number of barriers still stand in the way of smart cities and CAOs. The first is a lack of suitable candidates - within the public sector, there are very skilled data and analytics practitioners, of course. But this new role will be political by its nature, despite the objectivity of data, so the CAO will need the right leadership skills to navigate such potentially choppy waters.
A second obstacle is the embedded culture within city organisations, combined with some long-term lack of investment into the necessary infrastructure. In this respect, at least, the fresh impetus this idea is getting comes at a good time since the cost-base of big data analytics has dramatically altered in recent years.
Simon Dennis, director of central government for SAS UK and Ireland, has seen this challenge up close. "I've witnessed many attempts at adopting analytics in government over the last two decades. I've been privileged to work on some great success stories and, in hindsight, thankful we didn't win some of the projects that we chased where, even with all the advantages of using SAS, we might have struggled to deliver delight and receive warm enthusiasm from the end users.”
He adds: "The status quo probably comprises analytically-underwhelmed customers with an entirely justified pre-disposition to low expectations from a low-budget government IT project with sketchily-defined outputs, variable input data quality and ridiculously short delivery timescales.”
In contrast to that historical experience, the new push for an Office of Data Analytics within local government should benefit from what Dennis describes as “an almost perfect storm of pre-conditions for success.” Leveraging Hadoop big data architectures and low-cost, in-memory analytics, local authorities can dodge the need for mainframe solutions that might cost millions and jump to low-cost delivery of self-service insights to decision makers.
"Local authorities can take demographic data, consultation surveys and intelligence from open source, such as the internet, to gain a deep understanding of population needs and priorities. The software can enable the selection of the correct analytical technique and can ensure any guidance is statistically valid, so that this virtual real-time analysis is not hijacked by vested interests, nor will it lead to mob rule,” he says.
Dennis also envisages a potential political pay-off: “Imagine being able to prove the tabloids are misrepresenting public opinion with hard facts. It's no longer lies, damn lies and statistics - this can be become the era of lies, damn lies or statistics!"
In line with what has been happening in the private sector, it seems highly likely that the public sector will embrace data and analytics with all the people, processes and technology that come with them. Part of that will be using external partners to build the infrastructure and competencies, but there will also be a rapid development path to create a sustainable, internal resource.
Jacqui Taylor, CEO of Flying Binary, points out that, “the UK was the first country to appoint a chief data officer at national level. Is a CAO the next step?” To ensure this happens, there will need to be some parallel developments that keep smart city initiatives on track and aligned with each other.
As one of the authors of the national information infrastructure policy for the Cabinet Office, she understands the context that will ensure success. “How do you make a city smart? I have just been mapping the PAS 182 smart cities data concept model against ISO. It builds on PAS 181 - the smart cities framework - and will lead to PAS 183, a data sharing model. If you understand it as a change to the landscape of how government does business at this level, it gives you a sense of the size of that change,” says Taylor.
Smart cities are complex systems of sub-systems which is why it is vital to ensure the right policies and common standards are established from the outset, since changes later on become near impossible. At the moment, the whole movement is at the bottom of the maturity curve, assembling the raw data which should eventually become an information engine for local government.
Most of that activity has been around the bulk publishing of open data drawn from planning and public services sources, points out Taylor. It is some distance from operating at this level to creating a centre of excellence. “Whether it is a CDO or CAO, smart cities need that capability, which is why they are starting to follow the New York City model,” she says.
After London’s mayoral election in May, it could be that the first local government CAO gets appointed and the whole smart cities movement will gain a domestic reference point. It is unarguable that data and analytics move into the centre of decision making in the public sector, both to improve efficiency in an era of budget cuts and also to provide the proof that devolved government is working for its citizens. So the next time you do visit London or Manchester, you might begin to notice it has become just that little bit smarter.