25 June 2019

Philip Low

Addressing the needs of their citizens The United Nations predicts that by 2050, 68% of the world population will live in urban developments; 60% in megacities of over 10 million people. So, over 6.5 billion people living in cities and it is only headed in one direction. The dystopian view has tended in fiction to future cities that will become grid locked, cess pits of pollution, with residents struggling to breathe and avoid contagion of one sort or another. Optimists in the genre hope for a better outcome; some putting their faith in the discovery of cheap, non-polluting power, others in off-loading population to other planets. None of this seems likely in the foreseeable future, in the meantime, the best hope is probably smart cities. Smart cities are being created at the confluence of interlocking new technologies:

The Internet of Things (IoT), Edge Compute, fibre optics, artificial intelligence (AI)/machine learning (ML), 5G and Flash-memory. All of these things may be coming together in the nick of time to ensure cities become liveable spaces not wastelands. Between them, these technologies have the potential to enable us to manage everything from traffic flow to street lighting, parking to security and waste management to messaging. They can also help manage the atmosphere inside and outside buildings and make our energy production, distribution and usage more efficient. So, we come to the other dystopian vision – big brother is watching you. Smart cities depend on data but how much are the citizens willing to trade off loss of personal data security for the potential benefits.

The key to smart cities is data IoT will provide the data, with sensors everywhere inside and outside buildings, to monitor traffic hot spots; bins that need emptying; parking metres that are available or not working; CO2 levels and other pollutants; demand on utilities and public transport and always more. There are still problems to iron out: as always with new technology, we have to develop standards that will ensure systems can communicate with each other, to pull data out of the silos. This in turn raises the risks for cybersecurity to address. There is also the minor problem of whether we can produce and power all the sensors envisaged. By all accounts we are talking in billions and into trillions – that is an awful lot of batteries or power connections to manufacture. Scepticism perhaps should rise in proportion to the number of zeros being bandied about.

The backbone of any smart city will be, and in some cases already is, a fibre optic spine that provides connectivity: public service Wi-Fi, data collection and dissemination, communications, transportation, education and other public services. This has to be regarded as infrastructure, a shared facility, just like a road, along which anyone can transport their data and communications. This raises the perennial problems for technology, standardised connectivity and partnership funding. 5G, which is still a decade away in development and full deployment even in major cities, will, provided enough investment is found, eventually create capacity to handle huge volumes of connections operating in a smart city. Connection density will increase from the current 4G of 2,000 to over 100,000 active users/sensors per square kilometre, alongside the much-hyped increase in speeds to gigabits per second and much reduced latency. Edge Computing will not replace data centres but will enable data to be processed locally – some predict around 75% of all data by 2025 - and the development of local applications.

Take, for example, video analytics for traffic management. It makes sense to avoid the bulk movement of this kind of data and process it locally, which will also match the latency requirements for micro-management of adjustments in traffic handling. There are security issues that require local storage and processing of data, sometimes for legal compliance which, for example, may not allow data to be moved out of a legal jurisdiction. Networking a number of edge sub-stations together creates a local distributed computing resource and Wi-Fi network. The savings on backhaul costs could be significant, coupled with reduced latency and greater resilience. The introduction of Artificial Intelligence (AI) and Machine Learning (ML) plus, now deep learning, will surely mean that the enormous increase in data can be handled and processed ever more efficiently and productively, in time.

Other technologies are also emerging to help things along.Flash-memory and NVRAM technologies are finally reversing the stupendous growth in cooling and drive wattage costs, as well as floor space for racks, in both data centres and edge sub-stations. Revolutions So, the technologies for smart cities are exciting but we are already seeing the results of a total focus on technology which forgets the people who have to live in our smart cities. Examples exist from earlier industrial revolutions where steam and water power first brought people together in grim, polluted cities. The second industrial revolution gave us mass production and the third automation and electronics. None of these were unqualified successes for the growing city populations. Digitalisation and data are being heralded as the fourth industrial revolution and need to genuinely produce quality of life as well as efficiency in production.

To get smart cities right we need to clarify who is responsible not just for the infrastructure but also for the data: creation, management and security. That means addressing the regulatory issues from city planning to sharing infrastructure and controlling polluters/pollution to the imposition of cyber security standards. “Smart Cities” need to decide what they are for. They may be looking to drive economic growth by attracting certain types of businesses through the services they can provide access to or the integration of physical transport to attract manufacturing or distribution. Perhaps first they need to address keeping data secure and anonymised. IoT is all very well but it is a melange of personal, corporate and city data with many permutations of which data is useful to different participants but paramount is what data citizens are willing to have shared. Leading companies in the digital age are beginning to realise, as some of the industrialists in the 19th century did, that all their workers and those that service them, have to live somewhere. So, when hi-tech jobs drive prices of housing upwards, somehow provision has to be made for the cleaners and shop staff, drivers and teachers that will enable the high earners to work productively. Sheer numbers mean we have to rethink how cities operate – better transport solutions, better atmosphere control, inside and outside buildings, better planning to meet housing needs and leisure needs.

Data for all This year's Datacloud Europe, Global Congress & Exhibition, taking place in Monaco 4-6 June, like its predecessors, brings together many of those who will be making decisions now that will dictate how smart cities evolve. This means both existing major cities that are being ‘retrofitted’ with fibre optic spines and IoT and the new megacities that will blossom, particularly in Africa and Asia. Data, its generation, collection, processing and application has to become leaner, less wasteful of resources in operation and cooling. It also has to become better at identifying the needs of all those who will live in the smart cities of the future. 

By Philip Low, chairman, Broadgroup