U-M physicist: Smart cruise control eliminates traffic jams

ANN ARBOR—We’ve all been there, wilted behind our steering wheels while our car inches and jerks forward like an ice chunk through a straw: Yes, you’re in a traffic jam, the bane of every major commuter city in the country.

But such infuriating backups could be avoided—at least on single lanes that don’t require merging—if just 20 percent of cars were equipped with adaptive cruise control (ACC), according to University of Michigan adjunct physics Professor L.C. Davis. Doing computer simulations on his laptop from programs he wrote himself, Davis studies the origins of traffic jams and the impact that vehicles equipped with ACC have on traffic flow.

Adaptive cruise control detects the presence of a preceding vehicle and measures the distance as well as relative speed using a forward-looking sensor, and automatically adjusts the vehicle speed to keep a proper range. ACC is an option on some high-end vehicles.

In a recent paper published in The American Physical Society’s Physical Review, Davis demonstrated a "spectacular" reduction in single lane jams if just 20 percent of the vehicles in a 600-car platoon were equipped with ACC.

Davis’ paper also revealed other interesting traffic jam behavior. For instance, with all other factors being equal, faster traffic is more ‘sensitive’ and will jam more quickly than slower vehicles. For example, at 30 meters per second (about 67 mph), when Davis changed the 188th car in the platoon from ACC to no ACC, a jam formed immediately. But when the same change was made at 15 meters per second (about 34 mph), traffic slowed more gradually and did not experience an abrupt jam.

"There is different behavior at lower speeds than at higher speeds," Davis said. "At 30 meters per second you get a discontinuous change. You either get a jam or you don’t. At lower speeds you get something in between."

The good news is that equipping even a small percentage of vehicles with ACC appears to greatly improve traffic flow. But the news is bad Davis said, when you add merging traffic to the mix.

Even though you can prevent traffic jams, the introduction of ACC doesn’t improve the average speed through an on-ramp area as much as Davis had hoped, he said. There are two kinds of congestion, he noted. A jam on a single lane road is one kind of congestion, he said, and the slow downs where two roads merge are another.

Much research has been done on traffic flow and jamming behavior with conflicting results. Davis said his modeling shows that even the best of those earlier results are optimistic. For example, a European study of the German Autobahn (East, Treiber and Helbing), suggested that if just 10 percent of vehicles had ACC, it would reduce travel time due to jams by 80 percent. But Davis’ paper disagreed with those findings, and showed that concentrations of 10 percent or less of ACC vehicles would not improve traffic flow.

Davis, of Plymouth, is a retired manager with Ford Motor Co. A seasoned rush-hour commuter, he started modeling traffic behavior in 2001 while still at Ford. He joined the University of Michigan shortly after he retired in 2002.

So, what does Davis’ research suggest commuters can do to reduce traffic jams now?

"I think what we should do is encourage drivers, when they start coming toward a place where one lane ends, to start moving over," Davis said, confirming that the person who zooms ahead of the pack to sneak abruptly into the line of merging traffic does indeed cause a traffic jam.

"If everyone is obeying and getting in early, one person can still ruin the entire thing."

For more information on the Physics Department, see: http://www.physics.lsa.umich.edu/nea/