How Autobahns Use Tech to Fight Traffic

Sensors embedded in German roadways monitor traffic patterns and provide information to help avoid future jams

Turns out, all you have to do to avoid traffic jams is embed wire sensors in the roads. That, combined with data from the past, allows German traffic experts to forecast up to 90 percent of all traffic jams.

The data is all there: "When German train drivers held warning strikes in July, there was about 15 percent more traffic on the highways all around Frankfurt until 6:00 a.m. the next day," says Wolfgang Scherz, the head of the State Office for Roads and Transportation in the German state of Hesse. Afterwards, the traffic flow returned to normal.

Should the train drivers make good on their warning and begin a full-scale strike, the traffic scenario in heavily populated western Germany would be much worse. Tens of thousands of rail customers would switch to their cars from one day to the next. Still, Scherz and his colleagues have a pretty good idea of what the situation would look like. They have been watching Germany's autobahns for years.

"Our data pool on traffic flow is our most valuable treasure," Scherz says. "And that treasure grows every day." Already, he can predict most highway traffic jams days in advance as long as the congestion is not caused accidents, which bring about 20 percent of Hesse's traffic jams. Indeed, the state authorities feel the predictions are so reliable -- with a margin of error of only 10 or 15 minutes -- that Scherz's traffic forecasts will soon hit the radio waves. From September onwards, drivers will know of traffic jams a day before they happen.

Drivers Behave Pretty Much the Same

Seeing the traffic future is made possible by thousands of counters on German highways, which are most often just wires placed directly in the asphalt that send real-time info to a central collection point on how many cars are travelling in which direction and the speed at which they are moving. And since the vast majority of drivers behave pretty much the same, further calculations are simple. The more results gathered and analyzed, says Duisburg-based traffic expert Michael Schreckenberg, the more precise the prognoses end up being.

Along with other experts, the physicist has developed a mathematical model that constantly compares the present traffic density with stored data from the traffic past, calculating travel times on this basis and thereby allowing reasonably reliable congestion forecasts for the highway network of the German state of North Rhine-Westphalia. When it comes to predicting what will happen in the next 30 minutes, the calculations are 90 percent accurate, says Schreckenberg. One hour in advance, a respectable 80 percent of traffic jams are still accurately predicted.

The experts in Hesse want their predictions for the entire day to have the same level of accuracy. To this end, they have fed their computer with data about events that tend to result in snarls. For example, when Frankfurt's football club is playing against Munich's (Germany's soccer version of the New York Yankees), fans arrive in greater numbers and from different directions than when the away team is Energie Cottbus. Visitors to the Frankfurt Book Fair behave much more intelligently than those heading for the International Motor Show, who tend to jump into their cars all at once and drive into the same traffic jam at the same time.

'Self-Defeating Prognosis'

Weather forecasts are also to be part of the calculation. When roads are wet and visibility is limited, a highway's capacity decreases by as much as 15 percent, says Gerd Riegelhuth, a traffic expert from Hesse. And that figure is often precisely the difference between flowing traffic and an impassable bottleneck.

But there are risks to over-accurate forecasts. Traffic jam expert Schreckenberg warns about the hard-to-calculate "phenomenon of the self-defeating prognosis," which holds that if too many drivers react to the radio warning and head for alternate routes, the predicted congestion may never happen. Even so, drivers may still find themselves stuck in traffic on the route they chose instead.

Before it's here, it's on the Bloomberg Terminal.