Physics of 'green waves' could make city traffic flow more smoothly

May 22, 2013 by Lisa Zyga, feature

Understanding why green wave breakdown occurs may improve the green wave synchronization strategy and improve traffic flow in urban areas.
( —If you've been lucky enough to catch all the green lights as you drive down a busy street, you may have been benefiting from intentional synchronization called a "green wave." The green wave concept has been around in the US since the 1920s, but it doesn't always work as it should. When traffic gets backed up for some reason, "green wave breakdown" occurs. In a new paper, physicist Boris Kerner at the University of Duisburg-Essen in Essen, Germany, has modeled and analyzed the causes of green wave breakdown, and the results may lead to better coordinated green waves and more efficient traffic flow.

Kerner's paper, "The physics of green-wave breakdown in a city," is published in a recent issue of EPL.

Many large cities around the world, especially in Europe and the US, synchronize on the busiest streets to create green waves. When a green wave works as intended, all vehicles within the wave can drive through a sequence of green lights at a certain speed without having to stop at the signals. The timing of the lights can be controlled either by sensors or timers, and can be set up for traffic in one direction or both directions. Green waves have several benefits, such as allowing for higher traffic loads, reducing , controlling , reducing fuel consumption and emissions, and facilitating bicycle and .

The biggest disadvantage of green waves is that, when the wave is disturbed, the disturbance can cause traffic problems that can be exacerbated by the . In such cases, the queue of vehicles in a green wave grows in size until it becomes too large and some of the vehicles cannot reach the green lights in time and must stop. This is called over-saturation. As more and more vehicles stop, the traffic can cause a gridlock where vehicles can't move forward even when the light turns green because vehicles are backed up at the light ahead, which may still be red or turning green at the same time.

The physics of this green wave breakdown has not been thoroughly studied until now. In his paper, Kerner used two kinds of models to investigate the underlying mechanisms. He identified several general features of green wave breakdown that are independent of the model used, and also discovered that the physical characteristics of green wave breakdown depend crucially on which model is used.

In one model, called a three-phase model, green wave breakdown occurs due to an initial speed disturbance (for example, a car turning onto the main road from a side street) that causes a moving synchronized flow pattern (MSP). In an MSP, vehicles move slower than in the initial free flow of green wave, causing delays that destroy the green wave synchronization with the traffic lights. The result is oversaturated traffic at the traffic signals.

In the other model, called a two-phase model, an initial speed disturbance does not cause an MSP and the delays associated with MSPs. However, if the initial speed disturbance is large enough, and there is a large number of vehicles in the wave, then oversaturation and green wave breakdown can still occur.

In both models, Kerner found that one or more phase transitions are involved in the breakdown process. Also, breakdown occurs with a certain probability of less than 1; in simulations runs with the same parameters, breakdown occurs at some times but not at others.

While the models offer a theoretical explanation of how green wave breakdown may occur, experimental tests are needed to gather data and determine which model best fits the data. Kerner hopes that an empirical test will be performed in which the vehicle speed and flow rate are measured both upstream and downstream of a synchronized traffic signal. By studying speed disturbances and the possible emergence of MSPs, researchers could then determine the initial disturbances and how they lead to green wave breakdown.

"This would be a very interesting test for an EU project," Kerner told "However, as far as I know, there are no EU programs for such empirical studies of signal control. The problem is that my theory contradicts all classical theories of urban traffic, which, as is well known, do not work in the real world."

Explore further: What causes traffic gridlock?

More information: Boris S. Kerner. "The physics of green-wave breakdown in a city." EPL, 102 (2013) 28010. DOI: 10.1209/0295-5075/102/28010

Related Stories

What causes traffic gridlock?

September 7, 2011

Everyday life enters a different phase on the Tuesday after Labor Day, the unofficial start of autumn in the United States. As students and employees return from vacation, and vehicles fully flood roadways once again, drivers ...

Smart traffic lights reduce fuel usage and lower emissions

October 27, 2010

( -- Denso Corp. has designed the next version of 'the smart traffic light system'. By using messaging between vehicles and the traffic-light controller, better decisions about when to change signaling will help ...

Cell phone signals help manage traffic

September 26, 2011

In a pilot project in Texas, Siemens is developing intelligent transportation technology for the fast and orderly evacuation of citizens. In this project, traffic light timing systems register traffic flow and adjust the ...

Two new studies suggest ways of improving traffic flow

September 21, 2010

( -- Two studies have just been published that may help improve traffic flows. The first, presented as a Santa Fe Institute working paper for September, concluded that if traffic lights responded to measured traffic ...

IBM wants traffic lights to stop your car

May 26, 2010

( -- IBM has filed a patent application for a traffic light system that can remotely stop and start the engines of vehicles, with the aim of increasing fuel consumption efficiency at busy intersections.

120 intelligent cars warn each other in the test

August 6, 2012

One of the largest fleet tests in the world was launched today in Germany. 120 cars hit the road to test a system known as simTD, a technology that enables vehicles to communicate with each other and their environment. Cars ...

Recommended for you

Researchers study interactions in molecules using AI

October 19, 2018

Researchers from the University of Luxembourg, Technische Universität Berlin, and the Fritz Haber Institute of the Max Planck Society have combined machine learning and quantum mechanics to predict the dynamics and atomic ...

Pushing the extra cold frontiers of superconducting science

October 18, 2018

Measuring the properties of superconducting materials in magnetic fields at close to absolute zero temperatures is difficult, but necessary to understand their quantum properties. How cold? Lower than 0.05 Kelvin (-272°C).

The big problem of small data: A new approach

October 18, 2018

Big Data is all the rage today, but Small Data matters too! Drawing reliable conclusions from small datasets, like those from clinical trials for rare diseases or in studies of endangered species, remains one of the trickiest ...


Adjust slider to filter visible comments by rank

Display comments: newest first

3.7 / 5 (3) May 22, 2013
In one of the cities I've lived in, main street works just fine if everyone travels at about 34mph.
Of course, all it takes is one wannabe speed demon rushing to the red to mess up the works.
Some people just don't get it that speeding through that area slows you and everyone else down.
3 / 5 (3) May 22, 2013
In one of the cities I've lived in, main street works just fine if everyone travels at about 34mph.
Of course, all it takes is one wannabe speed demon rushing to the red to mess up the works.
Some people just don't get it that speeding through that area slows you and everyone else down.

Usually coordinated signals wouldn't be affected by such activities. They would probably make that speed demon wait. It all depends on where the sensors are placed too. If you have them between signals, you can measure the speed of the group of vehicles and see that only one is out in front and adjust your timing. If all you have is the sensors at the signals, then you'll detect the arrival of a car, but you wouldn't know if it was from the previous signal or if it pulled out of a nearby street or driveway and the timing between signals is probably based on the speed limit and traffic density. Actually, both scenarios are bad for the speeder.
3.7 / 5 (3) May 23, 2013
Usually coordinated signals wouldn't be affected by such activities.

It breaks the flow of traffic because they end up at the next light "early" and have to fully stop as it is still red. Then when the wave of traffic behind them would normally just pass through the green light, they have to stop or slow down instead while the speeder accelerates from their dead stop, which would not have happened if the speeder had just gone with the traffic flow instead of racing ahead.

I am not sure how sensors would affect things, my experience is with what I perceive to be fixed timing "green wave" configurations.
1.2 / 5 (9) May 23, 2013
Part of the problem is that even if a few dozen cars are working in perfect unison and all is going perfectly, it only takes one person to screw it all up. One person to pull out onto the street in front of the wave and take their merry little time getting up to speed. One person to decide to take their merry little time turning off the street. One person cutting someone off. One person deciding to smoke/text/eat/drink/whatever and drive at the same time. One person.

If we can figure out how to make it so one person can't bring it all crashing down, then I think things would be far better. Unfortunately, the only way I personally see that happening is when cars start driving themselves.
not rated yet May 23, 2013
Not knowing much about it, but optimizing traffic flow in a city always sounded like one of those really challenging problems. It's hard to account for all the inputs. How independent are they? [likely not as much as we would wish] What are the boundaries? [larger making things exponentially more complex] As if you could pour increasingly greater resources into developing a better solution, for relatively minuscule returns. Then, to have to re-solve the problem over and over as things change [more traffic, road repair, intersections change, businesses open and close, etc.]. And possibly end up with a solution that is optimal under some circumstances and unstable under others.
5 / 5 (1) May 23, 2013
We have a couple of cities (e.g. Hannover) where the traffic lights have additional indicators (automatically updated via sensor-driven data) that will tell you how fast you need to go to catch a green wave.
It works rather well in adjusting for changes in traffic flow.

And always remember:
Lights synchronized to 50km/h are also synchronized for 100km/h ;-)
1 / 5 (6) May 25, 2013

We have these here as well in Holland. That extra info makes it work well almost everytime. A big time and a big fuel saver for us.

Please sign in to add a comment. Registration is free, and takes less than a minute. Read more

Click here to reset your password.
Sign in to get notified via email when new comments are made.