Connect to UAB
School of Engineering
School of Natural Sciences and Mathematics
Last August two motorcyclists tried a new way to beat the heat: testing their machines in an early-morning race down Birmingham’s Interstate 459. The winner managed to hit 97 miles per hour—above the speed limit, that is.
A few hours after the bikers were clocked doing 167 (and arrested), tens of thousands of commuters on U.S. Highway 280 oozed by the scene in typical workday congestion at an average speed of less than 30 miles per hour. Trapped behind a wall of red lights, they could be forgiven for looking over at the wide-open freeway with a twinge of envy.
Highway 280 is Alabama’s second-busiest roadway, and to judge by the newspapers and local politicians, Birmingham’s biggest headache. Stand on the western slope of Oak Mountain on any weekday morning and you can trace an unbroken river of taillights flowing mile after idle mile, from the stalled SUVs at your shoulder to the minivans inching up Shades Mountain in the far distance. Drivers drowning in the middle of a 280 traffic jam have plenty of time to think up easy fixes for the congestion, and they do: add more lanes, ban the big rigs, build some public transport, and, best of all, turn this baby into one giant freeway.
“Everybody who has a driver’s license thinks they’re a traffic engineer,” says Steven L. Jones, Ph.D. Jones and fellow civil-engineering faculty member Andrew Sullivan really are traffic engineers. Between them, they have decades of experience handling the sticky mix of economics, politics, and pop psychology poured into every road built in America. And both are interested in using powerful new traffic simulation programs to explore how our cars affect our cities, our lives, and ultimately our planet.
Dell on Wheels
Until 1913, Alabama law required all citizens to spend 10 days each year helping to maintain state roadways. Could the century-gone legislators who passed that law have foreseen that by the year 2001 the average Alabamian would spend 30 hours a year at a dead standstill in traffic? And there’s no relief in sight; in fact, the number of vehicles on the urban highways in this state is expected to jump 40 percent from 2001 to 2015.
Last year the Regional Planning Commission of Greater Birmingham asked UAB to study several proposed solutions to the 280 gridlock, including the popular favorite, a series of so-called urban interchanges for the highway’s worst intersections. Urban interchanges send a main road hurtling up and over cross streets, eliminating the need for traffic signals. But they are also complex and costly, requiring years of work, a maze of ramps and access roads, and enough open land to build them all.
“It’s one thing to look at an interchange like that on paper,” says Jones. “But the planning commission wanted to know how these changes would make 280 function as a system, and the only way to figure that out was with a simulation.” So Jones and Sullivan rebuilt the highway’s most troublesome dozen miles on their computers, using government data and their own field observations to create a model of morning rush-hour traffic that is accurate to within half a minute.
The simulation looks like a video game, with thousands of tiny rectangles representing cars, trucks, and buses zipping back and forth. They force their way in from side streets, cut each other off, and run red lights with abandon—just like real Birmingham drivers. Each rectangle is free to follow its own agenda within certain rules, and each is endowed with one of 10 driver personality types, “ranging from very aggressive to little-old-lady,” says Sullivan.
“That’s what’s so interesting about traffic engineering—it’s largely a human problem,” Jones adds. “A structural or mechanical engineer making a widget faces questions of materials and properties and physics. But in traffic engineering you’re dealing with personalities.
“Driver behavior affects things like lane changing,” he explains. “Say I want to turn left across oncoming traffic to go to Target. Am I going to force myself into the smallest gap possible, or am I going to be extra careful and wait? Because if I’m extra careful, then folks are queuing up behind me. In real life, you have a mix of different drivers, and all of that gets captured in this model.”
Highway 280 Revisited
The results of the simulation disappointed Birmingham business and commuter lobbies, because Jones and Sullivan found that no one approach would “fix” 280 instantly. The network of urban interchanges did indeed cut the most time from an average commute, but only by a slim margin. “Our analysis showed that if you had the ability to write a massive check and the interchanges appeared instantly, you’d save about seven minutes,” Jones says.
Why such a small time savings from such a huge road-improvement project? The answer, to paraphrase John Donne, is that no highway is an island, entire unto itself—it’s just one part of a vast local-transportation web. While the urban interchanges increased traffic flow significantly on the eastern half of the model, those speeding cars were ground to a halt by the narrow transition from 280 to the Red Mountain Expressway close to downtown Birmingham.
“People have asked, what if you just expanded the on-ramps at the expressway?” Jones says. “But if you did that, you’d just be pushing the problem to the cut where the expressway goes through Red Mountain, and you’re not going to widen that. But what if you did? Then the issue becomes the ramps at University Boulevard, and then. . . .” he trails off. “Eventually, you get into a problem you can’t solve.”
I BRAKE 4 NOTHING
Normal congestion is bad enough, but nothing sends commuters over the edge like the “phantom jam”—a miles-long backup that appears and disappears again for no apparent reason. One minute you’re trapped solid, the next you’re back up to 65 miles per hour with no wreck, no police activity, and no roadwork to blame for it all.
The reason is actually pretty mundane, says Steven Jones. “It’s called shockwave theory, and it’s been around since the 1950s.” Picture a commuter racing in to work. He drops something on the floorboard and taps his brakes briefly to assess the damage. The driver behind him, who can only react once she sees and responds to the flashing taillights ahead, has to brake a little quicker and stronger. As the wave travels backward it multiplies, until at some point the entire stream comes to a halt. As long as more cars arrive at the back of the queue faster than they’re cleared from the front (a virtual guarantee during rush hour), the slowdown will continue.
“All of a sudden, you’re sitting in a solid traffic jam,” Jones says. “And it’s all because somebody spilled their coffee.”
So how do you solve a problem like 280? The answer may have as much to do with rain clouds and ice cubes as asphalt and traffic lights.
In the 1990s, the traffic world was invaded by physicists who had noticed that their equations for representing natural phenomena—such as the movement of gas molecules or complex weather patterns—could be applied to the flow of cars down a road.
Until that time, traffic-flow theory had been based on hydraulics. “Picture a stream of water moving through a culvert,” says Jones. When the traffic stream encounters an obstacle such as a jackknifed semi, it reacts like a creek meeting a fallen tree: the main flow is slowed while the water squeezes its way around. But newer theories abandon that wide-angle approach to focus more tightly on how “group” behavior is actually created by individual units—whether molecules or minivans.
“The analysis is getting smaller,” Jones explains. “Where we’ve been modeling large traffic streams, now we’re getting down to the molecular level. It’s giving us the ability to model very specific travel behaviors, rather than just general traffic attributes.”
Your Mileage May Vary
“Where things get really interesting,” Jones says, “is in true regional-transportation planning. Computer simulations will allow us to picture how many new vehicles a new subdivision might put on the road, for example, or how many trips a new Wal-Mart Supercenter might generate. As the models get more and more specific, it’s going to be a huge help for city planners in understanding how people actually travel.”
Jones thinks such simulations will eventually help make a strong case for public transit. “One of my primary interests is the environmental impact of transportation,” he says. “And I have this graph that shows that even as cars have gotten cleaner and cleaner over the years, net emissions are still going up.
“Why? Our cars are clean, but the thing is, we drive them more. And that’s important to think about, especially when we start considering spending zillions of dollars on hydrogen power. If I have a gas-guzzling Suburban without a catalytic converter, but I only drive it twice a week on grocery runs, I’m polluting less than the guy in the Toyota Prius driving 70 miles every day. The issue is not only how clean the car is, but also how much it gets driven.”