Roads may not be the sexiest elements of a city, but they’re among the most critical – and promise some of the biggest gains in deploying smart technologies and philosophies. Barnaby Page reports.
Roads, like sewers, are used by all and loved by few – which makes the challenge of running them more intelligently a double one. Not only do you have to improve the way the city’s (or indeed nation’s) roads and the vehicles upon them operate, but you also have to persuade everyone from regulators and budget holders to consumers that it’s worth doing.
As the Smart Transportation Alliance puts it, “there is an increased association of highways with exclusively negative attributes such as noise, pollution, acceleration of climate change, accident rates and external costs. The efforts deployed by the road infrastructure sector are largely ineffective; at the same time, the quality of service provided by these transport infrastructures in developed countries is currently decreasing as it suffers from massive spending cuts.
“Meanwhile, alternative transport modes (whether aeronautic, rail or other) have been subject to systematic improvements and heavy public investments.”
However, the benefits of smarter roads are difficult to deny, even if they don’t have the obvious and immediate progressive kudos of public transportation. Many aspects of smart traffic management apply especially to the more complex and congested environment of city-centre roads, but others – for instance variable speed limits – can have value on longer-distance highways too.
One of the biggest gains is in time. A research team at Carnegie Mellon University in Pittsburgh, Pennsylvania, found when it experimented with smart traffic signals that simply making them responsive to traffic conditions cut the time drivers spent stopping by 40 percent, and overall travel times by 26 percent – significant dents, indeed, in two of the biggest reasons people don’t love their roads.
There were green benefits too, with exhaust emissions down 21 percent, and while those don’t directly affect traffic itself they’re clearly a crucial fit with the smart city philosophy. Juniper Research estimates that by 2019 smart traffic management and parking will be reducing annual global emissions by an amount roughly equivalent to that produced by five million cars.
Gains that size, and even smaller ones, can accumulate into huge figures because of the sheer volume of road usage, especially in an increasingly urbanised world. Recent calculations by Juniper, for example, concluded that by the year 2021 smart traffic management and smart parking projects would save more than four billion man hours every year.
And that, of course, could have significant impacts on both quality of life and the efficiency of business.
With these benefits in mind, it’s not surprising that cities around the world are embracing smart road philosophies ranging from improved traffic control to encouragement of electric cars (or that vendors like IBM, SAP and Verizon are getting involved). And as in so many areas of the smart city, information is usually at the heart of these diverse projects.
Typically, data is obtained from sensors which may include in-road devices, video cameras, and (perhaps less often) devices installed within the vehicles themselves. The scale of information generated by these networks can be huge – Nanjing in China, for example, produces about 20 billion data points a year.
This data is then analysed, either locally for relatively simple applications such as deciding whether to change a traffic light’s colour, or centrally for more complex decisions affecting citywide traffic. The Really Big Data produced in the process can also be used retrospectively for analysing traffic behaviour and informing traffic planning.
The objectives are as varied as the cities and the roads themselves, but they can include smooth running of traffic – not only for the sake of drivers, but also because stop-start progress is wasteful and polluting; giving priority to public transport; reducing queues and jams; and responding to accidents, for example by rerouting traffic flows around them. Accidents account for around a quarter of all traffic congestion, according to the U.S. Federal Highway Administration, so that alone can make a substantial difference.
Smart road destinations
Los Angeles, famed for its smog and congestion, is also feted in smart road circles for its automated traffic control. Its $400 million system draws in information from magnetic road sensors and cameras to analyse traffic flow and control 4500 traffic signals accordingly; it’s credited with reducing delays at major intersections by 12 percent, and increasing citywide traffic speeds by 16 percent.
A similar approach using video cameras capable of performing their own analytics (as well as contributing to the central system) is likely to be rolled out across major Malaysian cities after a pilot in Cyberjaya.
Eindhoven in the Netherlands is taking the information-centric approach a step further. A traffic management system that it has piloted with IBM benefits from acceleration, braking and location information generated by sensors within vehicles themselves, as well as the more common on-road devices. Managers can thus respond to potential jams as well as accidents in nearly-real time, while drivers can also be given information on traffic problems via their smartphones or navigation devices – not only aiding them, but also helping to prevent the traffic situation worsening.
Also in the Netherlands, the experience of the Amsterdam region highlights the importance of collaborative working in traffic management. Three different bodies are responsible for roads here – the Amsterdam municipality itself, the province of North Holland and the national government – and sometimes their well-intended management measures conflicted. So they are now linking their systems.
To provide its traffic systems with the data on which to base their decisions, Singapore has turned to taxis – which of course travel the city more widely, and for much longer periods, than most cars. GPS technology installed in city cabs reports traffic conditions back to the Intelligent Transport Systems (ITS) project.
Parking, as well as driving, is also being digitised. In the Spanish cities of Santander and Barcelona, parking management systems that point drivers to the most suitable location for leaving their cars can increase revenue and diminish congestion (no more driving around looking for a space), as well as saving citizens the hassle of the search.
In a few cases, even the physical roads themselves are being made smarter, for example by installing lights controlled by motion sensors, or – as in one Nissan/Foster + Partners venture – embedding inductive charging equipment in the street so that electric cars can be recharged without parking at a special charging point.
Not just digital
Lower-tech approaches have their place too. In Stockholm, for example, drivers are charged more to travel during morning or evening peak hours than during the rest of the day. Similarly, variable parking charges can be used to encourage drivers to use a particular lot.
Some U.S. jurisdictions have for years employed high occupancy vehicle lanes, only available for use by cars carrying more than one passenger, which again are effective by imposing a cost on the drivers who collectively contribute most to congestion – though in this case the cost is felt in terms of delay, because the single-driver lanes tend to be busier, rather
Still, the vast majority of smart road projects do now focus on ways to exploit digital information. A sign of how these might develop comes from a project on the A14, the British road linking the port of Felixstowe on the east coast with the city of Birmingham 165 miles to the west. Because of the large volumes of container shipping coming into (and leaving) Felixstowe, this road linking it to the major population centres of the Midlands and northern England is a crammed one, and a pilot involving BT, the government’s Department for Transport and the Cambridge firm Neul demonstrates how the headaches of such highways could be eased through technology.
As Ofcom, the telecoms regulator which has an interest in the project because it regulates the wireless airwaves, describes it: “Sensors in cars and on the roads monitor the build-up of congestions and wirelessly send this information to a central traffic control system, which automatically imposes variable speed limits that smooth the flow of traffic. This system could also communicate directly with cars, directing them along diverted routes to avoid the congestion and even managing their speed.”
Smart road projects are not limited to privately-owned vehicles. Bikes can also be part of the mix (the Dutch firm Heijmans, for example, markets a product specifically designed to warn drivers when cyclists are nearing an intersection). Car sharing, a kind of digitally-empowered version of car pooling (or perhaps a drive-yourself version of Uber), is becoming increasingly popular. And many cities, for example Helsinki, are looking to package all these options together via a single transportation interface for consumers which will let them use a device such as a smartphone to choose from a menu encompassing everything from buses, to self-driving cars, to bike sharing.
The objective of many such schemes, of course, is to minimise car ownership – because reducing ownership inevitably reduces usage, and therefore pollution and congestion. But all the same, no medium-term smart road project would seriously envisage a world without private cars.
Enter the driverless car
A logical extension to smart road management is to build intelligence into the vehicles themselves as well as into the roads. At its most basic level this might simply be the capability to receive traffic and safety alerts, but more advanced functionality could include the ability to receive and understand electronically-transmitted “signage”, as well as the fully autonomous, autopilot-like driving modes which have occupied so many headlines.
Underpinning all this will be telematics capabilities, which Navigant Research expects to be ubiquitous in new cars for developed countries by 2025 (and which Juniper points out is being driven by the desire for in-car infotainment services such as Apple CarPlay and Android Auto rather than by more virtuous traffic management goals — so far, at least).
Tesla has received an enormous amount of attention for pioneering self-driving cars, and inevitably some of it was unwelcome when earlier this year the first known fatal accident involving an autonomous driving mode occurred with a Tesla vehicle in Florida. The driver of the Tesla Model S was killed after the vehicle was unable to pick out a white 18-wheel truck and trailer against a bright sky; the car drove directly under the trailer as a result.
The incident came only a few months after a more minor incident involving a Google self-driving car hitting a bus — in, embarrassingly, the tech heartland of Mountain View, California. But it doesn’t seem to have seriously damaged consumer confidence; a July survey in the U.S. by AlixPartners found that interest in self-driving had only dropped by about three percentage points since the Tesla crash, while awareness had risen sharply by around ten percent in just a few weeks, to top 80 percent.
Moreover, a similar survey by Boston Consulting Group in the U.S., China and Germany around the same time found that the Germans were the least willing to travel in a self-driving vehicle – yet even so, 41 percent would get in, compared with 48 percent of Americans. The Chinese were the most enthusiastic (around 81 percent, up from the previous year).
It seems that while consumers might well grow worried by continual accidents, the public has largely seen the Tesla incident in the light of the large number of miles travelled by self-driving cars and, of course, the well-known risks of conventional ones.
Getting there from here
Driverless vehicles, while real and here, still belong a little more to the realms of futuristic ambition than everyday experience. But while much discussion of driverless and other aspects of smart roads looks into the medium term – if not actually the distant future – to discern the benefits that widespread deployment might bring, one UK government agency is taking a much more pragmatic view in analysing the practical and regulatory impacts of smart road techniques that can be applied today.
Britain’s Government Centre for Connected and Autonomous Vehicles, which has recently launched a consultation on advanced driver assistance systems and automated vehicle technologies, says: “Organisations are taking notably different approaches in developing this technology. Some are taking an evolutionary approach in developing new advanced driver assistance systems (ADAS) – with incremental developments on existing technology such as automated braking systems. Others are taking a revolutionary approach in developing fully automated cars from the ground up, which can drive themselves for the entire journey.
“It is not clear when either approach will deliver a truly driverless car that people can purchase, or use. Experts think this could be any time from the mid-2020s onwards. What is clear is that vehicles, which can be parked within line of sight by remote control, or pilot themselves with human oversight on high speed roads such as motorways, will be available for sale in the next two to four years.”
Among the first steps, to be taken by the end of the decade, it identifies :
The Centre performed a regulatory review last year, looking at the implications for insurance among other issues, and then developed a code of practice for those testing these technologies. It is also funding R&D.
But it is realistic about the obstacles that will be faced on the route to widespread deployment: “It is a challenge to identify the right solutions for unknown future technologies. Indeed, because there will be a transitional period — where we will have a mixture of conventional cars, cars with increasingly sophisticated ADAS and ultimately, fully automated vehicles — the solutions for the distant future might not work now.”
That feeling is shared by automotive technology research specialist SBD and the mapping firm HERE — which is now owned by Audi, BMW and Daimler, having been sold to them by Nokia. The two produced a recent white paper on driverless vehicles and, as HERE’s Carrie Cox puts it, “we believe traffic will be reduced long term — that is, a generation from now — but the traffic situation will get worse during the complex deployment process which must come first”.
Among other reasons, HERE and SBD believe that the total number of cars on the road could increase while driverless and conventional vehicles share the space. It’s also possible that the novelty of driverless cars might cause drivers of normal ones to be more cautious, producing delays in itself.
To ease that pain by minimising complexities as well as congestion, HERE and SBD urge auto makers to collaborate on producing technologies that work with each other, rather than confining themselves to proprietary systems.
It seems, if these forecasts are right, that despite the many benefits of smart roads we may not be quite loving our intelligent streets and highways anytime soon — yet we’re not going to stop relying on them.