The technologies that will empower the smart cities of the future are refreshingly free of controversy. The concept itself is an easy sell: after all, who doesn’t want buildings to be greener, energy networks more efficient, traffic congestion and pollution reduced? So, while smart cities will require billions of pounds of investment in new technologies over the coming decades, there is an almost unanimous consensus among politicians, businesses and citizens about the benefits that it will bring. But while money, political will and public appetite are no barriers to the development of smarter, healthier cities, there is one fundamental issue that threatens to delay our progress towards this future utopia – IT infrastructures that were designed for a “dumber”, less data-intensive age.
Data demands of the smart city
Cities are built by people, but not necessarily for them. Business, industry and profit were the main forces that drove the organic, sprawling growth of the world’s great metropolises – not its citizens’ health, happiness, and convenience.
Smart technologies, including the internet of things (IoT), promise to solve many of the eternal problems that city-dwellers have been forced to put up with, from toxic air to gridlocked streets to public safety. New services will transform the way that municipalities manage public transport, shared infrastructure, city planning, waste and recycling, lighting, smart grids, and a host of other benefits such as access to healthcare, education or local government services.
But these new services all depend on the ability to generate, process and analyse previously inconceivable volumes of data. From sensors measuring air quality around the city, to the thousands of CCTV cameras monitoring transportation systems; from smart energy grids to “intelligent” bins that tell refuse collectors when they need emptying – the smart city will be a network of computers that is constantly creating huge amount of information.
This information is the raw intelligence that goes into making the key decisions on which smart services depend. Everything from transportation systems to smart grid and industry-related applications depend on instant, high-speed, ultra-reliable connectivity between the device collecting or generating this data, and the systems which process and analyse the information. This, however, is where traditional technology infrastructure threatens to delay – or even prevent – us achieving the full benefits promised by smart cities.
Living on the Edge
The traditional cloud model made a great deal of sense. Concentrating storage and processing power in hyperscale facilities enables users to take advantage of huge economies of scale to manage intense, high-volume workloads at a manageable cost.
Unfortunately, this model works far less well for the demands of modern smart city technologies. When storage, compute and analytics are located at a centralised hub – which may not even be in the same country, let alone the same city that the data is generated – it necessarily adds a lengthy lag.
This time delay might only be measured in milliseconds, but it can still have significant knock-on effects on services and applications that rely on instant communications, such as the machine-to-machine applications that are so central to many smart city initiatives. Consider, for example, the demands of an integrated transport system which simultaneously controls the movements of buses and trams, pedestrian crossings, traffic lights and much else. In applications like these, the need for speed goes beyond mere convenience and becomes a serious issue of public safety.
Clearly, connectivity and data transfer must be fast, stable and consistent at all times, in order to maintain operational runtime and reduce accidents or disruptions. That’s why smart cities need to take advantage of the recent trend towards “edge computing”.
When applications and data are moved from centralised hubs to the edges of a network, the distance between users and that data inevitably narrows, while speed, reliability and efficiency increases. This solves many of the challenges encountered by connected devices, by enabling data processing and other tasks to be performed at the edge of the network, rather than passing the workload back to a central data centre.
Edge computing is clearly one of the most important enabling technologies for smart city applications – but traditional hyperscale infrastructure still has a vital role to play.
Technologies such as drones, driverless cars, and integrated traffic monitoring systems are perfect examples of latency-intensive services which are well suited to “edge” locations. While these devices or services produce too much time-sensitive data for it to be dispatched to a remote location for processing and analysis, there are plenty of workloads that have much less rigorous connectivity demands.
For example, data required for citizen services such as taxation, welfare, healthcare, travel – while important – is not sensitive to nanosecond delays in the same way. It makes much more sense to host this data in the traditional server farms which can easily meet data-intensive demands, and scale quickly to changing demands.
For all the focus on exciting smart city applications like driverless cars, the role of low-level processing, backup and storage are just as important as they are in other areas of life. That is why successful smart cities will be built on hybrid edge-hyperscale infrastructure.
These hybrid models pose their own management headaches, however, requiring dynamic resource management to ensure that services are provisioned most effectively both at the edge and in the data centre. Technology providers such as hosting and networking firms are already working hard to provide solutions to these challenges; however, we also need to continue working closely with governments and municipalities to map out how a variety of new city services can be deployed on the best possible infrastructure model.
So great is the appetite for smart cities, and so universal the consensus that nothing need stand in the way of these transformational technologies. As long as we are all aware of the data management challenges that they represent, and are committed to solving them together, nothing will hold us back from building cities that are not just designed by people, but also for them.