Quebec City / INRS, Quebec City / Next Stormwater,Victoria, BC / University of Victoria
A slippery road or sidewalk is a dangerous thing.
To keep our walkways and roadways safe during Canada’s frigid, unforgiving winters, cities throughout the country have had to turn to road salt to increase tire traction. The trouble is, that salt doesn’t stay confined to the roads where it’s applied.
Eventually, the salt turns ice and snow into a briny slush that will find its way from the asphalt to nearby waterways, where it affects the fresh water it enters.
Enter team Clean Nature with a solution to reduce excess road salt.
Salt is extremely difficult to remove from water, team member Anne Carabin explained. The most common process involves reverse osmosis filters, which are very costly to set up and maintain. So the best way to keep waterways from over-salting is to go to the source and limit the salt applied to roads.
“We apply a lot of salt, a lot more than more than is necessarily required,” Carabin said.
What she and her team are developing is a computer model that uses real-time data to determine the amount of salt that needs to be applied to properly de-ice the roads, without compromising driver or pedestrian safety.
“If, for example, anything changes in the weather, like the temperature, if it begins to rain, the model will optimize the salt according to those different conditions,” Carabin said.
They’ve taken data provided to them to train their artificial intelligence model to look for patterns and find the optimal salting application. It will also keep track of where the salt-truck drivers are and compare that to traffic data, and adjust salt application and routing accordingly.
The computer model is just the first step. After perfecting that side of the operation, Carabin said they want to turn to producing equipment that would make these changes with minimal driver intervention. But the focus for now is on perfecting the computer model, Carabin said. Once that falls into place, perfecting the system will follow.