An estimated 2,900 Canadians die on the roads each year, and nearly 200,000 are injured. (iStock)Engineers at the University of British Columbia are building what they hope will be a reliable system for predicting the likelihood of car crashes in a specific area, a development that could help save lives and cut pollution.

Despite the efforts of governments, about 2,900 Canadians die on the roads each year, and nearly 200,000 are injured. Worldwide, the World Health Organisation predicts that by 2030, road accidents will be the fifth-leading causing of death, beating diarrheal diseases, HIV/AIDS and tuberculosis.

A laboratory being built at UBC's Kelowna campus is designed to estimate how many collisions will happen in a neighbourhood, given specific road layouts, and help engineers improve their plans.

The lab can't tell who will crash or how many accidents will happen on any particular day, but the designers say the tool is a leap forward for road safety, because it doesn't rely on often-unreliable predictions of traffic congestion. Instead, it takes a range of factors into account, such as the road layout, population numbers, census details and other data.

Civil engineering professor Gord Lovegrove will lead the road safety laboratory when it opens in September 2011. Already there are suggestions that the system could be used to encourage authorities to carry out remedial work in neighbourhoods seen as dangerous — a prospect he says causes "some nervousness" on the part of city engineers.

Despite that, Ottawa, Vancouver, Victoria and Kelowna have been supporting the project by providing collision data, geographic information system mapping and other information to help build the models that calculate the likelihood of accidents.

The Canadian Foundation for Innovation and the British Columbia Knowledge Development Fund each contributed $92,000 toward the $240,000 cost of building the lab. The remainder involves donations of money and equipment from a variety of smaller contributors.

Buying a house?

Lovegrove wants eventually to build a user-friendly version that anyone can use over the internet.

Civil engineering professor Gord Lovegrove will lead the road safety laboratory when it opens in September 2011. (University of British Columbia)Those who fancy comparing their neighbourhood to one in another part of town would be able to go online and plug in the relevant details to reveal the estimated crash rate. The lab can make predictions for any suburb in Ottawa, Vancouver, Victoria and Kelowna. Lovegrove hopes that within a few years there will be enough information to estimate the crash rate anywhere in Canada.

Initially, he expects the lab will be used by city engineers and developers who are interested in safety. People designing new neighbourhoods will be able to use the tool to compare the safety of different designs.

"We're going to make it so you can put in an existing neighbourhood or a whole city of neighbourhoods … and it will tell you which are the least safe," Lovegrove, a former city engineer, says. That could help cities prioritise spending.

"I think the sky is the limit for testing existing neighbourhoods and saying, 'You know what? On average, these neighbourhoods need help.'"

Saving lives

'If you can get [a quarter of] any population doing anything, social marketing tells you that is basically the critical mass.'—Gord Lovegrove, professor

The UBC system is aimed at helping people figure out how to design roads in ways that can reduce the carnage. For example, many cities are spending millions altering roads for cycling — frustrating gridlocked motorists. One of the challenges facing cities that want to promote cycling is that cyclists tend to come off worst in collisions with cars.

Yet if roads and bike lanes are designed properly, Lovegrove says, both safety and congestion problems can be addressed simultaneously.

"As you get more people on a bike, the reduction in car crashes is greater — far, far greater — than the increase in bicycle crashes," he says.

Lovegrove believes there is a tipping point — maybe about 25 per cent of people cycling — beyond which the roads become safer for cyclists and drivers. Holland — where about a third of people cycle to work — has one of the lowest road death tolls in the world, for example.

"If you can get [a quarter of] any population doing anything, social marketing tells you that is basically the critical mass," he says.

Once that happens, drivers start taking more care, and cities becoming willing to spend even more money on safer cycling routes, he says. "It is a quantum leap for safety."

Up and running in 2011

One of the new laboratory's first tasks is to figure out at what point that quantum leap happens.

The laboratory is looking for research partners in China, Scandinavia and elsewhere to help figure out what happens to safety when cities have more cyclists than Canadian cities do today.

Better yet, Lovegrove wants to find a real-life neighbourhood willing to be built in accordance with a road-safety blueprint, so he can test and refine the models.

As for why the models work, he says the tools used before were inaccurate because they relied on long-term predictions of traffic volume, which were often wrong by about 30 per cent.

To get around that, he studied crashes in Ottawa, Victoria and Kelowna and found he could estimate collisions more accurately using land-use details, population, road lengths and other information.

Initial results suggest that traditional grid-based neighbourhoods could avoid 60 per cent of crashes by following safer design rules. Simple innovations such as roadblocks (to stop drivers short-cutting down residential streets) and three-way intersections (so drivers have to look fewer ways) could cut the number of crashes enormously, he says.

With more cycling, walking and public transport use, the results could be even better.

"Now I've just got to get somebody convinced to build a neighbourhood based on those (road-safety) principles," laughs Lovegrove.

The laboratory will be up and running in September 2011, with spaces for 15 post-graduate students.