Getting Smart About Traffic

Thanks to GPS, sensors, artificial intelligence and even algorithms based on the behavior of E. coli, it’s possible to imagine the end of commuting madness

The road more traveled.
The road more traveled. Photo courtesy of Flickr user K2D2vaca

Usually I walk to work, but earlier this week, after another apocalyptic forecast of torrential rains and head-twisting winds, I fell prey to weather dread and drove in.

In no time, I was reminded of why Washington D.C. has the worst drivers in the U.S.–Allstate verified it–and also why it’s among the Top 10 congested cities in the country. The latest estimate is that drivers here waste an average of 45 hours a year in traffic jams. I don’t know if anyone’s come up with a comparable analysis of how much time the stress of sitting in gridlock takes off your life, but I’m guessing I said goodbye to 15 minutes or so that morning.

The experience revived my interest in the science of traffic flow and how GPS, sensors, and algorithms have made it possible to imagine a day when the commuting madness will end.

Here are some of the ways we just may get there:

1) Follow the wisdom of E. coli: That’s the thinking of two Chinese engineers wrestling with the hideous traffic of Guangzhou, a city of 13 million in southern China. They are advocates of applying “swarm intelligence” to traffic lights in the city, or more specifically, something known as Bacterial Foraging Optimization. This is an algorithm based on the behavior of E. coli, which, while very basic, ultimately results in the optimal solution to problems. In this case, the algorithm would be applied to stop lights, adapting them to traffic flow instead of keeping them on a fixed loop.

2) Failing that, you can still learn a few things from humans: Scientists at the University of Southampton in the U.K. found that real humans are better traffic controllers than computerized systems. So now they’re focusing on developing artificial intelligence for traffic control systems so they can learn from experience as humans do.

3) Or feel the pulse of social chatter: IBM studied traffic jams in three Indian cities over the past year through the social network comments of people stuck in them. The company’s evaluation of tweets, Facebook updates and other social network discussions of people in Mumbai, Bangalore and New Delhi is designed to show how social data can be used to read public attitudes on big urban issues, such as traffic. Among its findings: Drivers in New Delhi talked more about public transportation, weather and the stress of commuting, while those in Bangalore vented about the overall driving experience, construction and parking. And in Mumbai, they tended to rant about accidents and pollution.

4) Twitter intelligence is not an oxymoron: And Twitter is also being used in real time to stay on top of traffic accidents and backups on British highways. A mobile app called Twitraffic analyzes what people are saying on Twitter about traffic and warns you about problems that have popped up. The company behind the app claims it lets people know about accidents an average of seven minutes before the government’s Highways Agency does. It hopes to launch a U.S. version next month.

5) Meanwhile, back in the U.S.: There’s already a pretty impressive mobile app available here for helping you avoid commuting nightmares. It’s called Waze and it not only gives you directions, but it also monitors what other drivers are saying about what’s happening on the streets around you. It’s a traffic report through crowdsourcing, and one that constantly updates with new directions if there’s bad news coming in about the road ahead.

6) Just let the cars work it out: Since last month, about 3,000 vehicles around Ann Arbor, Michigan have been able to talk to one another. As part of a joint project of the U.S.Department of Transportation and the University of Michigan, the cars and trucks have been adapted to be able to communicate wirelessly and warn each other of potential accidents or backups. For instance, one vehicle could tell another when it’s approaching an intersection or if it’s stopping on the road ahead. The Michigan researchers think these wireless systems, if they become a standard feature, could cut accidents by 80 percent.

7) Car Talk was taken: MIT scientists are heading down the same road, developing something they calls CarSpeak. It’s a communication system for driverless cars that lets them “see” through the data provided by other cars on the road. And that would allow a car to cruise right through an intersection because it would know no other cars were coming.

Down the road

Here are a few other developments designed to help us get around:

  • Not so mellow yellow: A researcher at Virginia Tech concludes that one of our big problems is yellow lights because they create what he calls a “dilemma zone” for drivers. He’s developing a system for giving drivers a few seconds notice when a light is about to turn yellow.
  • We don’t need no stinking stretch limo: The largest buses in the world, 98-foot-long vehicles capable of carrying more than 250 people, will be rolled out in Dresden, Germany next month.
  • Nothing makes an old man feel young like driving at night: According to a study at MIT, the most important car feature for drivers over 50 are smart headlights, which adjust the range and intensity of light based on the location of other cars. The idea is to reduce glare and improve visibility at night.
  • I’m sleepin’ here: A new study of traffic noise levels in and around Atlanta found that almost 10 percent of the area’s population is exposed to traffic noise at a level described as “annoying.” And more than 2 percent live where traffic noise was described as “highly disturbing to sleep.”

Video bonus: How maddening are phantom traffic jams, you know, when everything slows to a crawl for no apparent reason? Here are two explanations, one from scientists, the other more like what we imagine.

More from

When Cities Run Themselves

Cars With Benefits

Get the latest stories in your inbox every weekday.