On the way to intelligent traffic control

Transport models Mobility Transport behaviour Data analysis Telecommunication Computer calculations IT systems Satelittes

An important step towards a more efficient running of urban traffic is knowing how many cyclists, pedestrians, and drivers are at which places, when they are there, and how quickly they move. Researchers have now examined whether it is possible, via the road users’ smartphones, to count and identify them.

We are in central Copenhagen on H. C. Andersens Boulevard, one of the most heavily trafficked roads in the inner city. If traffic can be improved here, it will result in a number of advantages. Both economic and environmental.

The CITS (Copenhagen Intelligent Transport Systems) project will create a basis for ensuring that the right political decisions can be made when it comes to improving traffic in Copenhagen.

The project was carried out on a 130-metre stretch of H.C. Andersens Boulevard, where six lamp posts—three on each side of the road—were provided with Wi-Fi capability.

“If you want to test traffic solutions, H.C. Andersens Boulevard is a great location, as the technology is really put to the test,” says Søren Kvist, Senior Consultant at Copenhagen Solutions Lab—the City of Copenhagen’s incubator for smart city solutions.

He explains that this area is also one of the places with most bicycle traffic. And that is saying a lot in a city where up to 45 per cent of all trips to and from work and educational institutions are by bicycle. Every day, several thousands of bicycles and cars pass the small section of H.C. Andersens Boulevard, which became a laboratory for researchers and companies for testing whether it was possible to get a clear picture of traffic in the area using Wi-Fi data from mobile phones

.Photo: Mikal Schlosser
Morning rush hour traffic on H.C. Andersens Boulevard was analysed by means of data from road users’ smartphones.
Photo: Mikal Schlosser

Rush hour a challenge

“Copenhagen is experiencing major challenges with its rush hour traffic—among other things with traffic jams. Of course, you can solve them just by making the roads wider, but it is not necessarily the smartest approach. Some of the characteristics of the current situation is that the traffic lights change at fixed intervals, and that we lack good data on the different means of traffic,” says Søren Kvist and continues:

“Using data, we can learn to control traffic much better. If we can streamline traffic management, we can save time, because people don’t have to endure traffic congestion or spend too long looking for parking spaces. We can also reduce pollution and accidents, and we can use our resources on other things than building new roads,” he says.

The City of Copenhagen is currently replacing street lighting with more intelligent and eco-friendly lighting, and in this connection, Wi-Fi capability will be installed on the lamp posts, among other things.

This makes it possible to use them for many other purposes than lighting, including data collection, as it has been done in the CITS project,” says Professor Per Høeg from DTU Space.

DTU Space has, among other things, expertise within the field of GPS technology and satellite-based communication—knowledge which can be exploited to develop other positioning technologies, e.g. Wi-Fi positioning.

“We have tested whether it’s possible to use Wi-Fi signals from road users’ smartphones to identify their position and speed, and it is. In addition, it was possible to identify three types of road users: bicycles, pedestrians, and cars—based on how long their smartphones were in the area,” says Per Høeg.

The system, which was developed for the CITS project, used Wi-Fi signals to record the movements of road users with smartphones. The section, which was monitored, was only 130 metres, and as there is a big difference in how often different phones communicate with Wi-Fi access points, it is expected that many people—especially motorists—have passed through the short section without being registered. And only around 75 per cent of the Danish population have a smartphone. The remaining are not registered.


H.C. Andersens Boulevard on 2 February 2015. Tivoli on the left the City Hall Square, and Industriens Hus further ahead. 

Blue: Road users in the direction away from the City Hall Square.

Orange: Road users moving in the direction of the City Hall Square. 

Green: Crossing traffic (typically pedestrians).

Illustration: Jacob Senstius

 

Illustration: Jacob Senstius



Higher level of detail

For Thomas Alexander Sick Nielsen, transport researcher at DTU, the CITS project was a welcome opportunity to get a better insight into some of the road user types that we lack knowledge about.

Cars are currently counted in several places where, e.g., coils are installed in the road to register passing vehicles. But automatic counts are used in particular for cars, while soft traffic is counted less frequently. And this gives some uncertainty in the overall picture of traffic in Copenhagen.

Illustration: Jacob Senstius     

 

 

The lines are generated by the road users’ smartphones and consist of the same data as in the picture above, but before sorting by direction, speed, and road user type.

Illustration: Jacob Senstius.

“Bicycles are counted manually, and the same applies to the few places where you count pedestrians. By registering Wi-Fi units, we place cyclists and pedestrians on equal footing with other road users. Today, for example, it’s a problem that we don’t know how many pedestrians there are, and whether there are places people avoid because they feel insecure there. Here, you can use the level of detail and even register per minute. For road users, we also have very little knowledge of what is happening at night. And we have no ‘fluid’ data showing at short intervals how traffic builds up and breaks down,” says Thomas Alexander Sick Nielsen.

For Professor Per Høeg from DTU Space, the CITS project was a very successful proof of concept experiment providing the researchers with more information than previously seen in similar systems.

The project showed that Wi-Fi data can be collected and used to show the number of road users and their movement patterns.

“We have also attracted great interest from several cities in Europe and the USA. And, of course, also from the City of Copenhagen which uses CITS as background material in other projects,” he says.

This is confirmed by Søren Kvist from Copenhagen Solutions Lab. The testing of different technologies to create a smart transport system is now being tendered out:

“We are currently creating a more permanent test area around H.C. Andersens Boulevard where we test other sensor types, parking sensors, pollution sensors, and Wi-Fi for tourists—based on the smart city concept Copenhagen Connecting, which received the award as the best in the world in Barcelona in 2014. If we develop systems with precise data about traffic, we can optimize traffic by 11 – 32 per cent. The reason for this large span is that there are still many uncertainties associated with it. But the potential is enormous, if we can control the traffic intelligently,” he say