Crossing to safety
around the world, project teams are helping wildlife make it across roads alive
A road sign in Australia
Martin Magnusson's team needs kangaroos. In charge of Volvo's collision avoidance initiatives, Mr. Magnusson and his team have launched a research project to develop a kangaroo detection system. But the project requires data: lots and lots of recordings of kangaroos in various light and weather conditions, to train the vehicle's sensory system to distinguish between a marsupial and debris floating in the wind. That has required the team to slowly drive the winding roads of Tidbinbilla Nature Reserve near Canberra, Australia to encounter and record kangaroos about to cross.
41 kangaroo collisions happen every day in Australia, and 1 out of every 20 vehicle wrecks in the U.S. involves animals.
As vehicle-wildlife crashes continue to plague roads—41 kangaroo collisions happen every day in Australia, and 1 out of every 20 vehicle wrecks in the U.S. involves animals—projects like Volvo's aim to prevent these accidents. But getting enough data has been a struggle.
“It's quite hard to collect data from real animals just driving past them,” says Mr. Magnusson, of Gothenburg, Sweden, who also spearheaded a project to build the Volvo S90's vehicle detection system for larger mammals. “If you look at moose, elk, reindeer, they probably run away and try to escape [when a car drives by]. It's been a real challenge.”
In addition to using live animal data, Mr. Magnusson's project team created an augmented reality program that could help train the system to identify animals by overlaying virtual fauna on top of what the system sees in real life. “Augmented reality helps us to set up more-or-less impossible situations without any danger and is an important verification platform to us,” he says.
Darren Quinn, project manager for Parks Canada Agency, Radium Hot Springs, British Columbia, completed a CA$14.5 million project aimed at reducing large-animal collisions in his province's Kootenay National Park. The initiative, completed in two phases in 2013 and 2015, included installing 15 kilometers (9.3 miles) of fencing that prevents creatures from crossing the driving lanes of Highway 93 South and building nine underpasses that allow wildlife to safely access habitat on each side of the highway. Motion- and heat-sensing cameras monitor each of the underpasses to help the project team study how readily animals take to their new route and if there are any problems.
But convincing forest-dwellers to take a new path can be tricky, especially since adaptation times vary from species to species. The Kootenay wildlife crossings are largely based on a similar project from Banff National Park in Alberta, where 82 kilometers (51 miles) of fencing, 38 wildlife underpasses and six overpasses prevent roads from disrupting animal habitats. Data from the Banff project shows that new crossings have reduced wildlife-vehicle collisions by more than 80 percent, though it typically takes two to five years for animals to adapt to the new route—and they hardly agree on whether over or underpasses are better. Species like wolves and deer tend to prefer short, high overpasses, while black bears and cougars like long underpasses.
Despite the Kootenay wildlife crossings still being in the adaptation period, Mr. Quinn says he's been “pleasantly surprised” by the number of animals and array of species using them soon after construction.
The nearly two years of data gathered during the project has shown over 1,600 successful wildlife crossings in the first three underpasses by medium to large animals, Mr. Quinn says. “It's a huge success.”
Long before such benefits to wildlife can be realized, however, organizations sponsoring road crossing infrastructure projects must work with stakeholders to gather support for the initial outlay. Animal underpasses are expensive to construct, though generally easy to maintain, Mr. Quinn says.
And some organizations have had to cope with unexpected costs after project closing. Michael McVaugh, a traffic and safety engineer with the Colorado Department of Transportation, Durango, Colorado, USA, serves as project manager for the state's animal detection system. At three locations along U.S. Highways 160 and 550, his team has installed 2 miles (3.2 kilometers) of a specialized underground cable that detects changes in the Earth's magnetic field. When larger animals disrupt the field, the system illuminates roadside signs alerting drivers that an animal may be crossing. Unlike underpasses, the cable system doesn't require any behavioral changes for animals, so there's no adaptation period.
“It's quite hard to collect data from real animals just driving past them.”
—Martin Magnusson, Volvo, Gothenburg, Sweden
But that doesn't mean the detection system works flawlessly. When the project began in 2008, the team had a budget of over US$1 million for researching, testing and installing the system (which no state had created before). But the organization wasn't prepared for the high cost of protecting and repairing the cable.
“Burrowing species like moles, pocket gophers, prairie dogs, they like to chew on the detection cable,” Mr. McVaugh says. “We've had a problem with maintaining the system to this point, because of the ongoing impacts from burrowing species. Each time the cable is chewed on and compromised, the system is disrupted and detections are impacted.”
Colorado is currently investing in wildlife crossings similar to those used in Canada. Despite the cost of maintaining the detection system, Mr. McVaugh believes the project can provide data to make future infrastructure buildouts more effective. “It's been an ongoing learning process,” he says. —Christina Couch
PM NETWORK APRIL 2016 WWW.PMI.ORG
APRIL 2016 PM NETWORK