CITY-DWELLERS WORLDWIDE BY 2050
In 2008, for the first time in the history of mankind, the number of people living in cities drew level with the total rural population. The number of city-dwellers is now expected to reach 6.5 billion by 2050, i.e. 66% of the planet’s inhabitants. By way of comparison, back in 1800, people living in cities accounted for scarcely 3% of the world’s total population. This explosion in the urban population has led to the rise of megacities. Some studies predict that in 2050 there will be no fewer than fifty cities across the world with over ten million residents. This change poses a mammoth challenge: how to optimise mobility around these cities.
Large cities are now faced with huge congestion problems. In the United States the average daily commute to work is now thirty minutes longer than it was in 1990. Some cities such as Los Angeles, where the lack of public transport results in the car being the only viable means of getting around, are well-known for their interminable traffic jams, prompting Elon Musk to dig underground tunnels in order to relieve traffic congestion. The problem does not only affect cities in the United States; it impacts most large cities worldwide. In Sao Paulo, some residents spend up to four hours a day getting to and from work.
alleviating city CONGESTION
the annual cost of traffic jams, ACTING AS A BRAKE ON the us economy
Jim Hackett, President and CEO of Ford, a stalwart of the US automobile industry and an icon of the American dream, recently spoke on this subject on the Recode channel: “Today, the transport systems of most global cities have reached capacity. And yet, more and more of us seek the benefits of great urban centres. Faced with this rapid urbanisation and the pollution and congestion that comes with it, we have to admit that the model of the past is no longer tenable. It’s clear that we need to update cities to more efficiently move people and goods. In the process, we will improve the quality of life for all.”
Urban congestion is not only a nuisance for city-dwellers, it also has a substantial economic and environmental cost. A report from UK economics consultancy the Centre for Economics and Business Research (Cebr), entitled ‘The future economic and environmental costs of gridlock in 2030’, reveals that traffic jams cost the US economy $120 billion a year. The US Environmental Protection Agency reckons that they also generate 27% of the country’s greenhouse gases, a figure that could rise to 50% by 2030, according to other estimates. Jim Hackett argues that the new technologies could provide part of the solution. “With the power of AI and the rise of autonomous and connected vehicles, we have technology capable of a complete disruption and redesign of the surface transportation system for the first time in a century,” he writes.
DATA HELPING TO COMBAT TRAFFIC JAMS
Public/private partnerships for data
on the road to USING DATA
Artificial intelligence makes it easier to collect, process and analyse large quantities of data on urban mobility, which can then provide highly useful information to help improve existing transport systems. This is one of the main areas that Smart City projects are working on. Data gathering can identify the main trends in travel flows, and which routes and public transport are most in use. City authorities can then use these trend insights to guide their action on a range of variables – traffic lights, bus stops, parking facilities, etc – in order to make it easier for people to get around. In addition, once local authorities have sufficient amounts of data, they can use machine learning to automatically monitor traffic in real time. It thus becomes possible to direct vehicles efficiently and mobilise different resources according to the variations in traffic flow.
These days city halls are able to obtain some of the local mobility data – basically data generated by public transportation companies – but a substantial percentage of mobility data is in private hands. On-demand transport providers such as Uber, Lyft and Chariot and other navigation apps such as Google Maps, Swiftly, CityMapper and Waze all nowadays possess valuable traffic flow data. So cities will have to find ways to work with these private players if they want to optimise mobility within their perimeter.
at L'Atelier BNP Paribas
Collaborating with services such as Waze, Uber and Google Maps enables local authorities to carry out statistical analyses on the number of people who travel a certain route, and the average time it takes to get to their destination.
“Collaborating with services such as Waze, Uber and Google Maps enables local authorities to carry out statistical analyses on the number of people who travel a certain route and the average time it takes to get to their destination”, underlines Stéphane Leguet, a Strategic Analyst at L’Atelier BNP Paribas. These companies wish to be seen as facilitators of urban mobility rather than as troublemakers and they have started to play the game. Uber has set up a platform called Uber Movement, which analyses mobility data and makes it available to the general public. Uber has also forged closer collaborative efforts with some cities, including Seattle and Paris. In Paris the goal of their collaboration is to predict traffic disruptions arising from roadworks. Leguet explains: “The aim is to measure the impact that closing off a road because of roadworks will have, by measuring the traffic that normally uses that road and then transposing that traffic volume to neighbouring streets”. Cities can also combine mobility data obtained from private companies with their own public datasets so as to obtain a consistent, holistic overview of their transport ecosystem.
Avoiding negative side-effects
BRIDGING GAPS IN PUBLIC TRANSPORT
The next step is then to set up partnerships with these players so as to bridge gaps created by inadequacies in the public transport system. Working with Lyft and Uber helps users to make the connection between a bus stop and the nearest train station, to provide rides late into the night and/or in suburbs poorly served by public transport, and to ensure a means of getting around for elderly people, persons with reduced mobility or those with other disabilities.
Teaming up with private companies working in the transport field is also a good way to avoid friction and the sort of unwanted side-effects that may arise when local authorities and private sector firms each work in their own patch. Explains Stéphane Leguet: “Let’s suppose for instance that a navigation app identifies the fastest way of driving from point A to point B. That means the car driver can save time, but it might also cause upset in some way to the local area and people living in a given street. The fact is that traffic signs are specifically intended to take people along the main roads, which are able to cope with large volumes of traffic. Navigation apps might well undermine this whole approach by bringing large numbers of vehicles on to narrow roads. This could damage the entire structure of the locality, creating safety issues if, for example, there’s a school nearby.”
public- prIvate partnerships
Working together and sharing data should help to avoid this kind of clash. Local authorities could, for example, inform private companies of any at-risk areas, so that the private players could build this information into the algorithms used to calculate the best route. Timeframes could also be taken into account, for example in order to avoid sending car drivers past a school at times when schoolchildren are arriving or leaving. The city of Boston is an example of successful collaboration in this sphere. The city authorities have recently worked with researchers at MIT to make an analysis of available data in order to help optimise the planning of school buses. The MIT researchers combined mobility data provided by Google Maps with schoolchildren’s addresses provided by Boston’s state schools. By combining the different datasets, they were able to design the most effective bus routes, decide where the bus stops should be placed, and calculate the average number of students per bus. The whole calculation only took thirty minutes, compared with an average of seven weeks using the old manual methodology. The Boston state school network estimates that the new system – which was rolled out in September last year – will save $5 million a year and reduce carbon emissions by around 9,000 kilograms per day.
New Orleans is now planning to introduce a similar system to optimise the way its ambulances are deployed.
datA ENABLING PERSONALised SERVICES
Towards ultra-personalised recommendations
This is a very optimistic vision, but it shows that we need an overall view if we want to optimise transport times
Meanwhile the city of Pittsburgh has made some improvements to its traffic light system as a result of collaboration with Carnegie Mellon University spinoff Rapid Flow Technologies. The company has designed a system called Surtrac, which makes traffic lights ‘smart’. Sensors embedded in the lights calculate the speed and exact position of approaching vehicles. Surtrac then runs the data through AI algorithms to determine how long the lights must stay green in order to optimise traffic flows. This is a fully distributed system, which means that each intersection has its own computer, which stores data, makes calculations and communicates information to the computers at neighbouring intersections, which then in turn incorporate it into their own calculations. A recent pilot project resulted in a 40% reduction in the average waiting time at traffic lights.
In the long term, Stéphane Leguet expects to see a system that goes much further than simple transport flow management, incorporating all kinds of data so as to create a holistic, multimodal system. He predicts: “Today, when you want to go to the Fnac [a large French retail chain selling cultural and electronic products], Google Maps can also give you the opening hours and the estimated waiting time at the checkout at any given moment. Google Maps can also warn you if there are a lot of people in the store right now. In the future, if I want to go to the theatre, say, the app will also probably be able to calculate in real time the waiting time at the entrance and advise me to set out ten minutes later, might in addition provide me with a list of available parking spaces in the surrounding area, and so on. This is a very optimistic vision, but it shows that we need an overall view if we want to optimise transport times.”