Article - Issue 50, March 2012

What's so clever about Smart Cities?

Rick Robinson and William Webb FREng

Download the article (601 KB)

The Royal Academy of Engineering published Smart infrastructure: the future in January 2012. A working definition of ‘smartness’ is outlined together with applications of smartness that have benefited different industries. The sectors specifically discussed are energy, water, land and maritime transport, communications and the built environment. The report is available at www.raeng.org.uk/smartinfrastructure

The Royal Academy of Engineering published Smart infrastructure: the future in January 2012. A working definition of ‘smartness’ is outlined together with applications of smartness that have benefited different industries. The sectors specifically discussed are energy, water, land and maritime transport, communications and the built environment. The report is available here

Rising traffic congestion, growing populations, resource constraints and environmental concerns such as climate change put ever more pressure on the smooth and efficient running of modern cities. The availability of cheaper sensors, mobile communications and data processing techniques have opened the door to smarter cities and infrastructure that can adapt to changing circumstances. Rick Robinson of IBM and William Webb FREng of Neul Ltd examine the smart city phenomenon.

Smart grids, smart cars, smart cities – it seems everything is becoming ‘smarter’. But what do we mean when we talk about smart systems? In its recent report, Smart infrastructure: the future, The Royal Academy of Engineering describes ‘smart infrastructure’ as infrastructure that uses a feedback loop to respond to changes in its environment, and to provide evidence for informed decision-making. The report describes different levels of smartness. A system may collect data to help designers produce the next, more efficient version, or to help a human operator take decisions. A smart system can even take action without human intervention.

Systems do so by gathering information on factors such as user demand and changes in the operating condition of infrastructure components to improve their performance. For example, smart traffic systems can detect congestion and inform drivers (see Traffic Prediction). Smart electricity grids can send information to people’s homes to help consumers reduce their energy use. In this way, a smart grid can control energy demand to better match it to supply, reducing the need for spare generating capacity. A similar approach can be applied to managing water supplies.

Cities without limits

Depending as they do on efficient operation of connected infrastructures – most notably transport, energy and water – cities are clear candidates for smart systems. The research company ABI Research has estimated that the global market for technologies that support and enable the development of smart city projects will grow from $8.1 billion in 2010 to more than $39 billion in 2016. (ABI Research Smart Cities Market Data 3rd Quarter 2011).

Some infrastructure systems have been smart for years. For example, operators of mobile phone networks optimise their systems to maximise the use of the scarce, expensive radio spectrum. Other infrastructures, particularly those with ageing physical systems such as water supply and sewerage, are evolving more slowly.

With increasing demand on resources and growing environmental concerns, technologies that enable smarter cities are now high on the agenda of city leaders as well as infrastructure operators. Even in a time of spending cuts in the public sector, there is an intention to stimulate economic growth and improve community vitality and well-being. There is also a need to reduce the consumption of natural resources. Every city is a net importer of water, energy and raw materials. The more of those materials that are required to create value, the more vulnerable the city economy is to supply disruptions.

In a smarter city, networks of sensors gather data from physical and information systems with the intention of changing the behaviour of those systems, or of inhabitants or city workers. Temperature sensors might detect areas with near-freezing conditions and send gritting lorries there first. Congestion monitors might change traffic light phasing. Air pollution monitors could restrict certain vehicles entering the city. Rubbish bins from commercial premises might detect when they are nearly full, enabling dynamic routing of refuse lorries. Sensors in parking spaces could allow drivers to be directed to free spaces rather than having drivers inefficiently search for them – some 20% of city traffic is estimated to consist of drivers searching for parking spaces.

Data collection

Smarter city IT systems collate data about key parameters. They then format and process the data to make it easy to use. Applications then generate insights from the data and take or advise action from that information.

A good example is how CCTV cameras, roadside sensors and location data from mobile phones and GPS devices in fleet vehicles are making it easier to make predictions based on real-time and historic measurements of traffic speed and volume. But the sheer scale of city systems can make the deployment and connection of such sensors a significant challenge. A sensor in every parking space, for example, could mean hundreds of thousands of sensors alone. Add in other sensors in rubbish bins, street lighting, every rented bicycle and the number could run into millions. Some estimate that this would add up to 10 sensors per person so that in a city like London there could be approaching 100 million sensors.

Fortunately, it is not always necessary to invest in new sensor networks. Data is already collected in many different ways:

  • Organisations such as Traffic Master can provide information on traffic flow gathered from sensors often deployed on bridges and other road platforms.
  • Mobile phone networks know where users are and can provide information on crowd density and deduce factors such as traffic speed.
  • Many vehicles, for example buses, have GPS sensors which track their location.

As data becomes more readily available at lower cost, the use of predictive technology to draw meaningful insights from it will surely spread. Indeed, the technology behind smart cities could trickle down to individual users. As new technology becomes available, many innovative and socially valuable applications will emerge. We might see the equivalent of a ‘smarter city app store’ where individuals and companies build inventive new applications that exploit the data gathered in smarter cities. Examples are already emerging, such as the many smartphone applications that identify the location of automated external defibrillator devices for use by first-aiders.

However, even when data is available, it can be a complicated task to turn it into useful real-time information – for example user density data from mobile phones into information on crowding. It is often simpler to install new sensors to monitor something that you want to measure than to use data gathered for other uses. So smarter cities are being greatly facilitated by the availability of low-cost sensors that can be deployed simply and cheaply.

Sensor Technologies

Micro-electro-mechanical systems (MEMS) or similar technologies can be used as sensors for a wide range of conditions, from temperature to motion at chip-scale for very low cost. However, some sensors, particularly for medical applications, are more expensive. Even in this area, technology is changing rapidly with the development of new sensors for humidity, carbon monoxide and other factors. Sensors are also being miniaturised and becoming cheaper, with individual sensors being integrated into multi-function units.

Historically, the problem with developing networks of sensors has been getting information from the devices. Generally this needs to be done via a wireless system but for example the cost of laying wires to every car parking space would be prohibitive. While cellular telephone networks are ubiquitous, devices have limited battery life, use relatively expensive wireless modules and sensors could be a load on networks designed for people rather than machines. Novel wireless solutions which bounce data from one sensor to another until they reach a point of internet connection (known as ‘mesh’ systems) have proven useful in some scenarios, but require complex algorithms to find efficient routing paths, fail when there are insufficient nearby nodes to form a mesh and offer limited coverage.

There are various possible options for bridging the communications gap. One solution being developed is a standard called ‘Weightless’ which operates in the gaps in the radio spectrum between digital TV transmissions. Compared to cellular technologies, Weightless is slower and has lower capacity, but it is an order of magnitude cheaper and meets requirements such as a 10-year battery life.

Weightless is specifically designed to facilitate the development of the increasingly popular domain of machine-to-machine communication (M2M). M2M uses low-cost, long-range base stations to track, trace and report on the status of sensors and other devices.

A growing range of businesses, including mobile telecoms companies, are developing new M2M technologies. M2M networks could support a wide variety of functions such as smart meters for electricity supply or smart automotive applications. One trial of M2M technology is taking place in Cambridge where the city council is looking specifically at smarter city applications for refuse collection in which commercial dustbins send back signals when they need emptying, allowing optimisation of the collection route and avoiding bins overflowing.

M2M technologies underpin the broader concept of the ‘internet of things’ – the idea of a world of internet-enabled devices that can network and communicate with each other and with other web-enabled gadgets. The general area is attracting growing interest from the R&D community to develop new ideas and technologies to exploit it. In January 2012, the Technology Strategy Board announced that it is investing £500,000 in research by 10 UK companies to improve our understanding of the challenges and opportunities associated with the internet of things. Later this year the Technology Strategy Board plans to invest up to £4 million on demonstration projects to study the merging applications and services through the internet of things.

Research Councils UK, through its Digital Economy Programme, is supporting research into smart cities. One multidisciplinary research programme, Digital City Exchange at Imperial College London, is investigating ways to digitally link resources and services within a city through harnessing next generation digital systems to combine and repurpose city data. The researchers are exploring the technical and business opportunities that could arise from integrating city-wide energy, transport, waste and utility resources. This promises to transform the planning and use of cities, generating significant technological, economic and societal impact.

Kuala Lumpur Stormwater Management and Road Tunnel (SMART). In its normal operational mode, the tunnel has two tiers of car traffic. In its second mode, when there are average floods, radial gates open diverting water into a holding pond. When this is full, the water spills into the tunnel and passes through the void under the lower road deck. In the once-a-year major flood in the region, the Motorway Control Centre will decide whether to use one or both road decks for flood flow. Traffic is then stopped from entering the tunnel, CCTV and a control car ensures that the roads are clear and then the tunnel is opened to receive the flood water © Sepakat Setia Perunding (Sdn) Bhd/Mott MacDonald

Kuala Lumpur Stormwater Management and Road Tunnel (SMART). In its normal operational mode, the tunnel has two tiers of car traffic. In its second mode, when there are average floods, radial gates open diverting water into a holding pond. When this is full, the water spills into the tunnel and passes through the void under the lower road deck. In the once-a-year major flood in the region, the Motorway Control Centre will decide whether to use one or both road decks for flood flow. Traffic is then stopped from entering the tunnel, CCTV and a control car ensures that the roads are clear and then the tunnel is opened to receive the flood water © Sepakat Setia Perunding (Sdn) Bhd/Mott MacDonald

The Royal Academy of Engineering has published a report on smart infrastructure that describes some of the applications that would form part of genuinely smart cities. For example, it suggests that consumers will be exposed to smart metering of water as well as electricity. Thames Water, which is currently running fixed-network trials in London and the Thames Valley, expects its new system to give more timely, frequent and accurate data to improve management of its water network. Frequent meter-read data will enable Thames Water to understand exactly where the water goes when it enters supply - be it customer usage, leaks in its own supply network or on the customers’ sites.

Future smart water systems could have meters read remotely and send messages by an M2M technology such as Weightless. Water and wastewater treatment plants could also be remotely controlled by an M2M network.

Smart systems can also encompass services that many cities deliver to their inhabitants, such as health care and support of an ageing population. One project, in Bolzano in Italy, provided 30 elderly people living in sheltered accommodation with a mixture of air quality monitoring, mobile tele-assistance and touch-screen access to self-service information and healthcare. The system can alert wardens’ smart-phones so that they can respond to unexpected changes in the monitored environment.

One example of smart infrastructure in the service of city management is the dual use of a tunnel in the central business district of Kuala Lumpur in Malaysia. The Stormwater Management and Road Tunnel (SMART), which opened in 2007 (see Ingenia 30), normally operates as a road tunnel, but during one of the city’s recurring floods it turns into a stormwater channel. Computer modelling of the traffic water flows and of the time needed to clear the tunnel justified the safety case to create a flexible road tunnel-cum-stormwater drainage system.

Who pays the bill?

Different parts of the smart city could well use much of the same technology, especially communications networks. Some sensors, such as CCTV cameras, may have multiple uses. This raises issues relating to who will pay for the installation and running costs. A key decision for each city is who will invest in deploying which sensors, where and for what purpose.

Some smarter city applications deliver social or environmental benefits rather than direct economic value; others require large up-front investment from one organisation to generate returns elsewhere. For example, a congestion reduction scheme funded by a city transport authority may generate benefits in the form of growth for the wider city economy, in improvements in environmental quality and in the general wellbeing of the city’s inhabitants.

Furthermore, while trials have proved that several kinds of smarter city schemes can deliver improved outcomes, many of them have included a substantial funding contribution from research. Few of them have developed the sustainable financial or commercial models that would justify the investment for full-scale deployment from the sponsoring organisation alone. For these reasons, smarter city investments tend to come from a mixture of public and private sources.

Another aspect of smart cities that needs to be addressed is the concern about privacy and commercial confidentiality which can restrict access to information. It is also important to ensure that intelligent cities are not vulnerable to severe weather conditions or to malicious attacks such as GPS jamming. For this reason, systems like Weightless are designed not to rely on GPS and to be able to use alternative location and timing sources where necessary.

Wider implementation

There are clearly many benefits that smarter cities can bring to their societies, economies and environment. The key problems have been gathering the required data from sensors, processing this data and funding the schemes. With new wireless technology becoming available, data collection promises to become much less expensive, and, with improvements in analytics software, the processing is simpler, cheaper, more flexible and more open.

There is now real opportunity for larger-scale implementation of the technology. There is much work to be done on the integration of such systems into society but it is already clear that cities that embrace a smarter future could deliver real benefits in resource efficiency, health and wellbeing for their citizens.

[Top of the page]