NATIONAL

Congestion relief

Calgary’s working on that age-old question: drive through or detour?

NICHOLAS KÖHLER August 6 2007
NATIONAL

Congestion relief

Calgary’s working on that age-old question: drive through or detour?

NICHOLAS KÖHLER August 6 2007

Congestion relief

NATIONAL

Calgary’s working on that age-old question: drive through or detour?

NICHOLAS KÖHLER

The GE5 Interchange, a confluence of the Glenmore Trail, Elbow Drive and 5th Street intersections now under construction, is the largest transportation project in Calgary history—30 years in the making, at a cost of $110 million. The Glenmore, which flatlines from east to west before meandering in a graceful arc towards the north-south 53rd Street corridor, carries 80,000 vehicles a day, making it one of Calgary’s most important arteries south of the downtown core. And for almost two years— until reopening last month as a 10-km long expressway five minutes quicker than its former iteration—it was a source of some consternation among Calgary drivers, already beset by the labyrinth of shut lanes, cranes and dusty detours the city’s become amid the Albert oil boom.

In 2004, shortly before the work began, Alex de Barros, a bespectacled civil engineer with the University of Calgary, asked the city for its Glenmore-Elbow traffic stats. De Barros, of the Intelligent Transportation Systems school of civil engineering—which holds as its Holy Grail the automated, driverless car—believed the numbers could help answer a question that vexes drivers everywhere on a daily basis: is it quicker to bypass upcoming construction by making a little detour, or should I grit my teeth and just drive through it?

Solving that dilemma, de Barros believed, would demonstrate the power of a computer-

ized traffic forecasting system he is developing that will treat congestion as a predictable thing, giving drivers real-time information on where delays will be 15,20 or 40 minutes into the future. The forecaster, which motorists will one day access via cellphone or GPS, works something like a meteorological model, with traffic-flow measurements and historical con-

gestion patterns replacing temperature and barometric-pressure readings in making predictions. Like other advances in transportation engineering that drivers should expect soon—shifting speed limits on highways designed to encourage fluidity or near-constant tolling via on-board GPS systems—the traffic forecaster shows that, as University of Toronto transportation engineer Eric Miller puts it, “a highway is not just a piece of pavement anymore—it’s an information system.”

To solve the construction dilemma, de Barros took traffic readings gauging speed, flow and density—numbers gathered from induction loops buried under key intersections—as well as origin and destination info gleaned from driver surveys, and pumped those data

into a computer simulation model tracking cars and trucks through a digital map of the Glenmore area. The result, a year in the making and eerily similar to a medical model demonstrating the flow of blood through a pumping heart, allowed de Barros to measure the baseline delay caused by the Glenmore construction when no drivers detoured. Little by little, de Barros then channelled cars into the arteries of smaller streets penetrating Glenmore’s surroundings. His findings were clear: with five per cent of drivers opting for alternative routes, Glenmore did indeed become less congested. But those benefits lasted only so long as less than 10 per cent of vehicles left the scene for smoother avenues. At the 10 per cent mark, Glenmore returned to the same levels of congestion as when no drivers detoured. More than 10 per cent jammed the whole network, leaving Glenmore as clogged as ever but seizing up its available escape routes too. “Overall,” says de Barros, “everyone loses.”

It’s not the sort of answer drivers can really act upon—as de Barros is the first to admit. What to do depends on what everyone else is doing, after all. That only underscores the benefits of a forecaster, which apart from reducing driver frustration could also improve road safety and lessen emissions. Still, the gizmo—a working model will likely be completed in five years—depends on the inclusion of a heady variable: the reaction of drivers to the very information the system itself provides. “Each individual driver is in some sense altering the system’s state through his second-bysecond decisions,” says U of T’s Miller. Fed into algorithms computing what’s most likely to be the smoothest route in the next hour must be a parallel analysis of how that information will affect those routes.

No one would benefit more from such divination than drivers in Calgary, where the GE5 interchange is just one of five multi-million-dollar road projects due for completion this year. Sans forecaster, Calgarians must still drive on instinct. So the question remains: to detour or not to detour? “Traffic always tends to get to a point where it doesn’t matter which route you take,” says de Barros, “the delay is going to be more or less the same. You’re better off just staying on Glenmore.” M

DRIVERS WILL KNOW WHERE DELAYS ARE GOING TO BE 15, 20 OR 40 MINUTES INTO THE FUTURE