Self-driving vehicles are constantly undergoing development tests, but, as Canada continues to experience its typical winter weather, a question has been raised—how do autonomous vehicles operate in snowy conditions?
Newly released data from a collaboration between the University of Toronto and the University of Waterloo will help train future self-driving cars to handle the challenges of winter driving.
Along with their teams, Steven Waslander, an associate professor at U of T, and Kryzysztof Czarnecki, a University of Waterloo professor, unveiled the Canadian Adverse Driving Conditions dataset earlier this week. Based on actual scans of icy, snow-covered Canadian roads, the dataset acts as a virtual training course for the computer algorithms that enable the cars to drive themselves.
“There are lots of great training datasets out there already, but they were collected on sunny, summer days,” said Waslander. “If you take algorithms trained on those datasets and try to use them in adverse conditions, they tend to get confused. They can misclassify objects—such as pedestrians and other vehicles—or even miss them entirely, all because of the changes in sensor data caused by snowfall.”
The dataset was created with Autonomoose, a Lincoln MKZ hybrid that has been equipped with a full suite of sensors, including eight onboard cameras, a lidar (light detection and ranging) scanner and a GPS tracker. Waslander and Czarnecki developed the vehicle as a testbed for self-driving software, but the Autonomoose also has a recording mode that captures data at a rate of 10 images or scans per second.
Over the past two winters, the teams have taken the Autonomoose around southwestern Ontario, recording data from more than 1,000 kilometres of driving. Of this, approximately 33 kilometres in harsh, snowy conditions were selected to form the basis of the dataset.
“Data is a critical bottleneck in current machine learning research,” said Alexandr Wang, founder and CEO of Scale AI, which collaborated with the universities for the project. “Without reliable, high-quality data that captures the reality of driving in winter, it simply won’t be possible to build self-driving systems that work safely in these environments.”
Ford’s Mustang Mach-E will arrive at dealerships across North America in late 2020, but the automaker is not waiting until then to train technicians to repair and service the all-electric SUV.
Ford has partnered with Bosch to develop a virtual reality training tool. Ford technicians will don a virtual reality (VR) headset to learn the ins-and-outs of repairing Mustang Mach-E, like removing and installing the battery pack.
Technicians will don an Oculus Rift VR headset to perform tasks on the Mustang Mach-E.
“Technicians will be immersed in a simulated and gamified world, meaning they won’t need to rely on actual Mustang Mach-E vehicles to learn about its components, including the electric SUV’s new high-voltage system,” said Dave Johnson, director of Ford’s service engineering operations. “This new virtual reality training tool allows technicians to understand the components and steps required to service these high-voltage systems, then confidently perform diagnostics and maintenance.”
Thanks to the VR training program, those dealerships don’t have to wait for the Mach-E and related equipment to arrive. Technicians can navigate through a set of training modules as if walking through classrooms. The process prepares the technicians for all the necessary repair tasks, including niche skills in Ford’s training program. The technicians can access the system from any location.
Bosch developed a proof-of-concept for the VR training in 2019. Ford now becomes the first automaker to pilot the application using repair courses designed for the Mach-E.
Eight-million for infrastructure
Last Thursday, Navdeep Bains, Minister of Innovation, Science and Industry announced that the government is investing $8 million to build 160 fast chargers at 73 locations to help Ontarians transition to a clean energy future.
Funding is provided through Natural Resources Canada’s Electric Vehicle and Alternative Fuel Infrastructure Deployment Initiative and will allow Hydro One and Ontario Power Generation to build one of the province’s largest EV networks.
While Ontario has been making strides toward increasing its EV charging infrastructure, Canada as a whole is still lacking in the charging network department. With consumers being accustomed to fuelling their cars almost everywhere, the nation has yet to build up enough capacity to provide similar convenience for electric cars.
There are currently roughly 7,700 public charging stations in Canada. According to the Kent Group’s 2018 National Retail Petroleum Site Census, Canada has 11,929 retail gasoline stations or around 3.2 stations for every 10,000 Canadians—and that’s just stations, not individual pumps.