Paradoxes of efficient transportation

Some fascinating and counter-intuitive insights about traffic management from an article in Vancouver magazine. I have tried to de-Vancouverize it somewhat to distill the key insights that are more broadly applicable, but it is ultimately an article about Vancouver, with its somewhat contrary starting point of not having freeways going into the heart of the city:

… In 1968, the German mathematician, Dietrich Braess, was modelling the response of traffic to different road networks. Braess assumed that drivers keep adjusting their trips until they find the quickest commute possible. This is one of the premises of EMME/2, the computer program used to predict traffic flows in every big city around the world. The EMME/2 models assume that, like water molecules in braided streams, cars will distribute themselves until all roads are pretty much flowing equally. This is why the Alex Fraser Bridge was jammed within a few months of opening. It’s also why your shortcut to work never stays a shortcut for long. Braess discovered a curious wrinkle in the traffic universe. His math showed that adding new capacity to an existing network of roads can actually lengthen peoples’ commute time. This fact is fueled by human selfishness: We all try to choose the fastest route home. But when ten thousand of us make that same choice in isolation, we all just might arrive home later.

… [T]he streets of our city do not behave as we expect them to. New pavement will not always move more cars. Bus lanes will not always move buses more quickly. A straight line is not always the shortest distance between A and B. These quirks of physics and geometry have a special urgency in the Lower Mainland, where the car inventory is growing faster than the population. There are 36,000 new vehicles registered here every year: one every 15 minutes. Line ’em all up and you’d have a traffic jam from Kitsilano to Hope.

The problem for folks who spend their lives trying to move us around is that human psychology is as important to traffic flow as ample road space. “Inside every car or truck is a human, and you know the problem with humans, right?” says Clark Lim, the manager for transportation research at TransLink, the Greater Vancouver Transportation Authority. “We’re irrational.”

… Road scholars at the University of Toronto’s [Intelligent Transportation Systems] Centre point out that traffic accidents and stalls cause 60 percent of congestion in Toronto. … [but] we don’t really need [ITS] here in Vancouver. Here’s why: imagine you’re headed south on Cambie. A few blocks ahead, a dump truck tips over. Whump. Traffic quickly clots around the scene. If this were Toronto and you were on the Don Valley Parkway, you’d be screwed. Same thing on Highway 1 here.

“Freeways create a system that’s prone to failure—not just a little failure, but catastrophic failure,” explains [Strategic Transportation planner for Vancouver] LaClaire. “When freeways stop moving, the whole system shuts down. But here, if one of our major roads stops working, we can keep moving.”

Vancouver never succumbed to the freeway trap. For every blocked arterial, drivers have two or three alternatives a few blocks away. Maybe you hear about a crash on the radio. Maybe you see the clot up ahead. If you’re a selfish driver, and you do as the models predict, you won’t wait patiently. You’ll zip over to Oak or Main. Your impatience combines with Vancouver’s tight road grid to absorb the effects of accidents.

You probably think your traffic-beating maneuvers are cunning, but LaClaire and his team predicted them long before you got behind the wheel. Traffic planners expect you to be selfish, expect you to attempt the most efficient journey possible, to ditch Broadway for the Fourth/Sixth/Second route (which, by the way, has five fewer lights between Burrard and Main than Broadway has).

Lately, the planners have also been toying with the streetscape in order to manipulate you, to give the invisible hand of driver selfishness a little push. For example, the city adds 15 to 20 new traffic signals every year. Most are new pedestrian-or bike-activated crossings. It’s just one way that Vancouver has gone out of its way to make life easier for pedestrians, cyclists and bus riders, and thus harder for drivers.

… Commuters are not ruled by logic. Our journeys are about feelings as much as about getting places. We yearn to keep moving, even if it means going farther to get to our destination. The road is not just a medium for travel. It’s a symbol of a certain kind of freedom. Always has been. That’s why folks in the suburbs plead for more freeways, even though they’ll never recover the time they lose in congestion caused by years of freeway construction. It’s why stoplights seem to take lifetimes to change, even though the longest cycle in the city, that Zen master at Main and Terminal, lasts only 170 seconds from green to green during rush hour.

We crave the feeling of the unobstructed journey almost as much as we seek the destination. This is why Braess’ Paradox matters. It contains a coded message from the traffic cosmos, a commuter’s version of the Zen adage: if you really think the road is freedom, then you must kill it. That’s exactly what the citizens of Seoul did. Back in 1999, Seoul shut down a major road tunnel under the city. The closure didn’t lead to chaos. In fact, car volumes fell on other roads. The flow improved: Braess’ Paradox. Civic leaders then did the unthinkable: they tore down eight kilometres of perpetually jammed elevated highway in the heart of the city and replaced it with meadows and pathways. The experiment is now being replicated around the world.

Every summer in Paris, they convert a highway along the Seine into a sandy beach, complete with palm trees. Londoners have combined congestion charges with street closures to reclaim their West End. In Bogota, Colombia, a civic War on Cars has handed road space over to bikes and busses. The result is positively Braess-ing: much faster commutes for drivers.

Leave a Reply

Your email address will not be published. Required fields are marked *