Advanced Search

Please click here to take a brief survey

The Problem with Self-Organizing Traffic Lights
Alex Steffen, 28 Sep 10

It's geek bait, really: a new study purports to solve transportation problems by using network models to let traffic lights self-organize into an optimal pattern. Over the last week, the blogosphere's been buzzing a bit about the idea:

We're fixed on the idea that lights should cycle on and off in a regular and predictable way, but this idea, they say, is unnecessarily restrictive. And less orderly patterns could be far more efficient... making traffic jams far less frequent.

Jams can arise, obviously, if traffic entering a road overloads its capacity. To avoid this, Helbing and Lämmer gave each set of lights sensors that feed information about the traffic conditions at a given moment into a computer chip, which then calculates the flow of vehicles expected in the near future. It also works out how long the lights should stay green in order to clear the road and thereby relieve the pressure. In this way, each set of lights can estimate for itself how best to adapt to the conditions expected at the next moment.

The problem is, great streets have many jobs, and moving cars is only one of them. Nothing is said here about the effects of this system on pedestrians or bicyclists. Indeed, no mention is even made of the fact that in many urban settings, cars moving at slower speeds makes everyone much safer (since cars moving faster get in more accidents, while drivers hitting pedestrians are much more likely to kill them when driving 40 miles-per-hour than 20). What optimizes an urban landscape for drivers may in fact make it dangerous, unpleasant, even unworkable for other people using the streets. Indeed, I suspect it's likely that this system is the very opposite of deep walkability.

Seeing only nails, traffic engineers scramble to build a better hammer. The problem may be with the very job description traffic engineer, an outmoded concept which seems to define the responsibilities of those planning our roads only by the speed of the cars passing over them.

What we need are planners who see land use, housing, transportation and quality of life as interwoven aspects of the same system, and plan accordingly. We can't build bright green cities by hacking the stoplights.

Bookmark and Share


I thought of doing this ages ago!

Systems like this ought to be approached in terms of maximising happiness rather than throughput (being stuck in a traffic jam being the bottom end... as is being stuck under an unexpected car!)

Actually, some years ago, a local council trialled a system that displayed a recommended cruising speed to avoid red lights. It sort of worked, but was not popular; possibly because it often recommended something that seemed counterintuitively slow (or way too fast!!) Perhaps it's time for a re-visit.

Posted by: Tony Fisk on 28 Sep 10

I get the point, but couldn't you integrate humans and bikes into such a system? Eg for pedestrians, the request button at a pedestrian crossing could be an input. Image recognition should be good enough to pick up bikes as well. You could even prioritise such traffic. Stop-start car traffic is also less efficient and safe than slower moving, but flowing, traffic. Stopped cars, unless hybrids, are particularly fuel inefficient, and fill the air with exhaust that discourages walking and riding.

So it's an argument for making sure the system is sufficiently sophisticated and robust, not an argument against making these systems take better feedback. I would agree that it shouldn't be used to encourage cards. But I guess I would say if cars are one of the tools a city uses for transport, a tool worth using is worth using well.

Posted by: Adam on 28 Sep 10

Second Adam's comment. Pedestrians and bikes should feed data into the system to allow for the smoothest operation for all road users. Perhaps the algorithm could even respond differently and prioritize pedestrians and bikes at peak times.

Posted by: BlackSun on 29 Sep 10

If such a system were to truly promote deep walkability, it would be built for pedestrians and cyclists, and THEN have m.v. flow integrated into IT. Our systems will not progress by continuing to think of bike/ped modality as an afterthought or an add-on, they must be primary.

Posted by: Stella Bonaparte on 29 Sep 10

Shades of 'the Midas Plague': it occurred to me that a modern sign of affluence is the capacity to dispense with a car.

While we're on the topic, why doesn't someone come up with a mechanism for guiding cars to a vacant parking space?

(I know! Don't encourage the drivers!)

Posted by: Tony Fisk on 29 Sep 10

It's fine if the parking space it directs them to is at the edge of a deeply walkable city centre :)

Posted by: Adam on 1 Oct 10

I also agree with Adam's comment about pedestrian and bike data being able to be fed into the system. This doesn't have to just be about optimizing car traffic.

Another reason to not dismiss this story so quickly is that you're missing the bigger backstory here, which involves a form of biomimicry. If you talk these particular traffic engineers, you'll likely find that they're getting inspiration from self-organizing natural systems such as flocks of birds and schools of fish. These animals don't have a leader, but somehow they manage to self-organize into a somewhat coherent whole. These examples are part of the basic intro given to people studying Complex Adaptive Systems.

I'm not trying to claim that this particular engineering example actually addresses deeper issues about sustainability, but my point is that people concerned about sustainability should be aware that there is a lot of interesting work going on in trying to understand principles of self-organization in complex systems, whether they take the form of traffic systems, ecology, social networks, or economics. We see that there are serious problems with top-down control, and what these traffic engineers are doing is a small subset of a much larger field of study that could have interesting implications for sustainability.

Posted by: Chris Davis on 24 Oct 10

Post A Comment

Please note that comments will remain open for only 14 days after the article is posted. While previous comments will remain visible, attempts to post new comments after this period will fail. This helps stop comment spam, so your forebearance is appreciated.

The Worldchanging comments are meant to be used for further exploration and evaluation of the ideas covered in our posts. Please note that, while constructive disagreement is fine, insults and abuse are not, and will result in the comment being deleted and a likely ban from commenting. We will also delete at will and without warning comments we believe are designed to disrupt a conversation rather than contribute to it. In short, we'll kill troll posts.

Finally, please note that comments which simply repost copyrighted works or commercial messages will be summarily deleted.

Yes No







MESSAGE (optional):

Search Worldchanging

Worldchanging Newsletter Get good news for a change —
Click here to sign up!


Website Design by Eben Design | Logo Design by Egg Hosting | Hosted by Amazon AWS | Problems with the site? Send email to tech /at/
Architecture for Humanity - all rights reserved except where otherwise indicated.

Find_us_on_facebook_badge.gif twitter-logo.jpg