Under the cover of darkness, a hillside overlooking the banks of the Chilcotin River began to collapse. When officials from the B.C. government surveyed the site the next morning, the July 30 landslide had created a dam of rock and soil 600 metres wide and 30 metres deep.
A former helicopter pilot and now the chief operating officer at the Vancouver-based Spexi Geospatial Inc., Wilson’s team had flown a number of tailored drone missions across Canada, monitoring rail lines, inspecting dams and surveying construction sites. Today, that model has created a visual aerial database that spans 131 B.C. communities in what the company describes as "Google Streetview of the sky." Proving to Canada it works Founded in 2017, the start-up is one of several companies and government agencies looking to fill the gap between commercial satellite imaging and traditional aerial photography.
Spexi's drone network cannot feasibly monitor forests across Canada like a satellite. But it can be tasked to fly cities and in specific disasters. Another advantage of drones over satellites is they can often avoid high cloud cover, according to experts who have trialled the technology. "The quicker that we can get information to emergency services — highways, the railway people, even firefighters — the better."
When Lakeland launched Spexi seven years ago, many foresaw drones on the cusp of transforming everyday life. But that dream has faced a number of challenges in recent years. Amazon’s decade-long promise to launch drone-delivered parcels has yet to fully take off despite plans for limited expansion by the end of 2024.
Once pilots finish their flights, they can upload the images to the cloud database within hours, adding another piece to a growing aerial atlas. Outside of Canada, the company has expanded its visual database with flights across Mexico, the U.K. and a number of U.S. cities like San Francisco, Austin and Houston. In Galveston, Texas, captured aerial imagery shows detailed roof damage from a recent hurricane, something Lakeland says could be used for rapid insurance assessments. Spexi’s growing visual network is regularly updated to see change over time.
Under Spexi’s current contract with the Canadian government, the company is now working on training its algorithms so the computer models better recognize different tree species and whether they are alive and standing or knocked over. "They are extremely interested in what we're doing. We just need a critical mass of it to be able to make sense for them," he said.