• Photo: Jose-Luis Olivares/MIT

    Full Screen
  • This figure shows two maps with colored lines that represent the main roads in Lausanne, Switzerland. The three colors represent how long it takes to commute: red is the longest commute, yellow is average, and green is the shortest commute. The left map, with conventional traffic light programming, has many red lines that represent long commutes. The right map, which uses the researcher's improved system, has many green lines that represent short commutes.

    This figure shows two maps with colored lines that represent the main roads in Lausanne, Switzerland. The three colors represent how long it takes to commute: red is the longest commute, yellow is average, and green is the shortest commute. The left map, with conventional traffic light programming, has many red lines that represent long commutes. The right map, which uses the researcher's improved system, has many green lines that represent short commutes.

    Image courtesy of the researchers

    Full Screen

Traffic lights: There’s a better way

MIT researchers develop an improved system for timing of urban lights to minimize commuting times.

Press Contact

Andrew Carleen
Email: acarleen@mit.edu
Phone: 617-253-1682
MIT News Office

Media Resources

2 images for download

Access Media

Media can only be downloaded from the desktop version of this website.

Anyone who has ever driven a city street and been frustrated by having to stop again and again for red lights has probably thought that there must be a better way. Now, researchers at MIT have developed a means of computing optimal timings for city stoplights that can significantly reduce drivers’ average travel times.

Existing software for timing traffic signals has several limitations, says Carolina Osorio, an assistant professor of civil and environmental engineering at MIT. She is lead author of a forthcoming paper in the journal Transportation Science that describes the new system, based on a study of traffic in Lausanne, Switzerland.

“Usually in practice, when you want to time traffic lights, traditionally it’s been done in a local way,” Osorio says. “You define one intersection, or maybe a set of intersections along an arterial, and you fine-tune or optimize the traffic lights there. What is less done, and is more difficult to do, is when you look at a broader scale, in this case the city of Lausanne, and you want to change signal times at intersections distributed across the entire city, with the objective of trying to improve conditions across the entire city.”

Such an expansive aim triggers complications, such as the ripple effect that a change at one intersection can produce across the surrounding area, or changes in driver behavior following changes in traffic-light patterns: For example, if wait times on a particular route increase, drivers may seek alternative routes that feature fewer red lights.

The new optimization process developed by Osorio and graduate student Linsen Chong can time traffic lights in large urban areas while accounting for the complex and diverse reactions of individual drivers. Their approach uses high-resolution traffic simulators that describe, in detail, the behavior of drivers in response to changes in travel conditions.

In detailed simulations of Lausanne’s traffic, they found that the timings produced by their approach reduced the average travel time for commuters by 22 percent, compared with timings generated by commercial traffic-light timing software.

Some cities currently make use of these high-resolution simulators, known as microscopic simulators: Behavior down to the level of individual drivers is simulated to estimate the impact of a given timing pattern. But the complexity of such models makes them computationally intensive. For instance, in the case of Lausanne, more than 12,000 individual drivers are simulated.

The new approach allows these models to be used in a practical and computationally efficient way. Other citywide models can be used to help determine proposed timings, but they treat traffic flow simplistically and homogeneously, rather than as a collection of individual travelers with distinct and complex behavior.

The new simulation-based optimization model proposed by Osorio and Chong aims to bridge these options, providing a detailed vehicle-level analysis but applying it to city-scale optimization.

The system, Osorio says, starts with a modest premise: “What if we combine information from these microscopic simulations with [citywide] information from these simple traffic models that are very computationally efficient and run instantly, but have very low resolution?” The approach combines the accuracy of high-resolution models with the computational efficiency of low-resolution traffic models.

The basic system, Osorio says, is also being applied toward different goals: Instead of just minimizing commuting times, it is also being used to minimize fuel consumption, and even to determine the optimal location for services such as vehicle-sharing hubs.

The work is currently being extended to help in the design of timing systems that can adapt to changing traffic conditions. Work on this topic is ongoing in collaboration with officials in New York City’s Department of Transportation, focusing on peak-period traffic in areas of Manhattan.

That agency’s Mohamad Talas, a deputy director of system engineering who was not involved in the research but is working with the MIT team on testing, says, “Such a model can validate our active traffic-management system in Manhattan, and allow us to fine-tune our processes and improve the network operation.”

Talas adds, “I believe that this approach is economically viable, with cost savings for any jurisdiction that needs to assess and improve traffic conditions for a large area of the transportation network.”

Topics: Cities, Urban studies and planning, Traffic management, Computer science and technology, Transportation, Automobiles, Civil and environmental engineering, School of Engineering, Research


The problem with such an approach for reducing car based commuting time in a city is that it will attract more traffic, until the very level that is prevalent right now (level of commuting time) is reached again. Cities do not have commuting times that are based on how well the traffic is guided. Cities have commuting times that are based on the local average commuter's individual will on how long to sit in a vehicle to commute. If it becomes easier, more people will fill in the resulting spaces because they will realize "hey this use to be 1 hour and now it's just 40 minutes, I'm ok with that" and they use their cars, buy houses/rent places further out, until all is back to the very same levels of wasting time (by commuting) as it is right now. It isn't the traffic light software. It's us.

Optimizing traffic flow can really only be done in one direction at any one time since the timing of the lights for north bound traffic directly interferes with the traffic flow south bound and the cross town traffic. and traffic does not have constant volumes. and distances between lights are not constant ( even in cities with primarily a grid layout )

What is so special with this work except a shiny re-branding with MIT's name? Using microsimulator as the black box performance
evaluator for some (or whatever) objective functions, then find out the "best" timing plans that optimizes the objective function. This is nothing but moot academic exercises. These type of ideas have been there for half century and tons of papers out there. As a seasoned traffic engineer I cannot help but just chuckle for this type of academic exercise that knows little about how real-traffic controller works and how to perform signal optimization in real-life, except just putting together some hype term and start playing numbers in paper.

interesting to watch streets all over taiwan at evening rush hour when police come and cycle lights manually depending on traffic

Optimization of signal timing for urban street networks is a more complex task than the simplified simulation models often used for academic research. For example, do these simulation models take into account that there are pedestrians that need adequate time to cross wide streets, there are buses and LRT routes on the same streets. Then there are emergency vehicles that come through and disrupt any kind of optimizaiton routine that is being run. Now, if someone has a areawide microscopic traffic simulation model that can realistically account for all that, and is also tied to a regional travel demand model I will take a look at that. Otherwise, this too far removed from reality for any traffic engineer to take it seriously. However, I think efforts like this are good simplified exercises for training future traffic engineers in graduate schools.

Robotic cloud connected cars are the ultimate solution and will make traffic lights obsolete in a few decades. Automated taxi service will be the solution to commuters, parkings and urban mobility.

The author assumes that city officials want to minimize commuting times. Quite to the contrary. In cities like Boston the goal is to reduce the number of cars in the city by making driving as uncomfortable as possible. To this end, many streets in downtown Boston have a light at the next intersection turn red as soon one light turns green. There are times when many seconds go by with not a pedestrian in sight and no car moving in any direction.

Only ordinary motorists are mentioned in the article, with the goal of minimizing commute times. One commenter suggested timing signals to discourage motorized travel -- though that might also, I'd suggest, increase pollution and
congestion. Pedestrians, buses and emergency vehicles are mentioned in the comments, so far. So, I'll comment about bicyclists, who are generally slower than motorists except during times of congestion.

Timing traffic signals on some streets for typical bicyclist speed (say, 15 mph, but dependent on slope) can slow and calm motor traffic and establish preferred bicycle routes, also desirable design goals. This is done in Portland, Oregon, among other cities. Installing a special bicycle signal so bicyclists can merge into position for through travel or a left turn at the next intersection when a street is clear of motor traffic also is a valid design goal. Placing bicyclists on a separate bikeway in the street corridor requires special signals to avoid conflict with turning motor traffic, and inherently results in reduced throughput for either or both, see for example the discussion here: http://john-s-allen.com/blog/?....

A sufficiently sophisticated analysis would account for the various design goals, some of which are in conflict with one another, and optimize accordingly. Determining what is considered optimal involves political as well as engineering decisions.

John S. Allen, MIT '75

Some areas already have systems that do this. For example, in southeast
Denver, East Arapahoe Rd, east of the freeway, and connecting roads
north thereof, towards Aurora. Visiting in July 2013, i found that EVERY
traffic signal in the area was timed to stop me just as I got to it.
For days straight, it NEVER failed. I had to wonder, is there a special
transmitter in my car, or do they do this to every car here? And Why?
Pull out fast or pull out slow, stop at the next signal. 10 mph below
the speed limit or 10 mph over the speed limit, stop at the next signal.
(Seeing this, they'll probably mail me a ticket.) See a red light
ahead, so slow down a bit to avoid stopping; it turns green; speed up,
and it turns red as I get to it - DAMN! This is the new, moneyed side of
town - is this a deterrent against outsiders even to tour or even visit
the area? In the tireder, poorer northwest of the city (demonstrated by
cheap motels and a rare closed McDonalds), the traffic lights did not
jam me up the same way. Yes I stopped, but it seemed about typical luck
basis, not Every Damn Time.

We have known of most of these possibilities since the 1950s. We have had computers capable of helping since the 1970s. Some cities and suburbs have major computerized traffic signal coordination.
The dream improved when wireless emerged: If cars can declare their intended routes, the system can use that information to improve timing and economy for everyone. If cars can "listen" for advice, the system can suggest small adjustments to speed and route, yielding even better improvement. Safeguards are needed: Jamming can remove benefits. Hacking can cause huge problems (sabotage). Adding optical signalling (line-of-sight infrared) in parallel with radio could add robustness. (Optical jamming range is shorter.) The system can detect faulty GPS readings or reports, by their proximity to actual base stations.
Additional benefits are possible if cars can talk to each other, but cars are more hackable than "the system", so vastly more caution is needed (unbreakable encryption and identification, and central authentication with verification against a reputation blacklist). Cars could form ad-hoc wireless networks for connectivity to the system, greatly reducing infrastructure cost. (But the relayed signals must be impervious to man-in-the-middle attacks.)
Car-to-car declarations of emergency stops can be especially dangerous if forged, and present an opposite danger if they are omitted when relied upon. A car should identify the cars in front of it and authenticate before being ready to trust an emergency declaration from it.

Something crazy is going on on this site. Every comment and every edit gets "moderated" (okay), but moderation seems to damage many comments by inserting hard line-breaks that I did not put there. (Look at some comments here.)
SOME kinds of editing can accidentally add hard line-breaks. It happens to me when I copy my text from Disqus and then paste it back in. Soft line-breaks that are there for display somehow become hard line-breaks in the entry. To avoid this, I paste the text into Notepad (actually Notepad2-Mod), Select-all, and then Cut, just to strip the clipboard text of formatting. Then, when I paste it into Disqus, spurious hard line-breaks don't happen.
I don't think you edit every article, but I can't be sure you have not made some tiny edit. Whatever viewing or editing you do in the review process, please find a way not to mangle every posting that you look at or every posting that you adjust.
When I edited one of my damaged comments on this page (to remove the spurious hard line-breaks), the edited comment went back to "Hold on, this is waiting to be approved by MIT News", treated like any new comment. (Even though a machine could easily have detected that I only made white-space changes, and let the existing approval apply to the changed version.)

it is so sad that this country can't do anything right
our bridges are on the cusp of collapse
roads are full of potholes
and so on
infrastructure everywhere is falling apart

Back to the top