WATERLOO, ONT. — A research project aimed at developing three-dimensional, high-definition mapping (3D HD) could give Ontario an edge in the race to develop autonomous vehicles.
The new technology would give self-driving cars functional capabilities to handle any type of weather, and would be marketed commercially without any proprietary rights.
“To our knowledge, nobody in the world is doing this,” Ross McKenzie, managing director of the University of Waterloo Centre of Automotive Research (WatCAR), told Automotive News Canada. “Nobody has taken our one stepback approach to fill the gap we have identified. If this unfolds as projected, it will be a catalyst for additional connectedand autonomous-vehicle activity in the province and further development across other centres within Ontario.”
The four-year project, funded in part by a $5 million Ontario government grant, involves WatCAR, Waterloo Region’s Innovation Centre and Waterloo Region Economic Development Corporation. Their proposal received funding approval in May as part of Ontario’s Autonomous Vehicle Innovation Network.
McKenzie said a key component of the project is collecting information to equip the car to handle seasonal changes.
“What you would see on a 3D HD map today in the summer with a full canopy of leaves on all the trees is quite different than what you see in the spring and fall with no leaves on those trees, and strikingly different from what you see in the winter when, at a minimum, there is snow piled along the side of the road covering the curb or filling in a ditch,” he said.
“The vast majority of autonomous-vehicle driving logged across North America has been in fair, dry weather conditions, mainly in California, Texas and Arizona. In Canada we have the benefit of being able to drive in four different seasons of weather.”
The researchers are using a Lincoln MKZ hybrid sedan, dubbed Autonomoose, which belongs to the university’s autonomous-vehicle research team.
Autonomoose is equipped with numerous computers and a variety of sensors that run a complete autonomous driving system. Currently, to create a 3D HD map, the car first has to be driven on any road it will eventually drive on autonomously. Only after the 3D HD map is created can the vehicle drive autonomously.
3D HD mapping gives autonomous vehicles advance information they will encounter along the roadway before reaching them. This allows each vehicle to understand what a human driver differentiates along the way, such as an open space on a street reserved for parking, traffic signals or crosswalks. 3D HD maps will combine with satellite information and enhanced vehicle operating systems.
“We’re going to establish the 3D HD map for Waterloo Region and template the process so we can readily repeat it, creating separate Waterloo maps for summer, fall, spring and winter, followed by maps for other regions,” McKenzie said.
While the platform will be open, he said, it will not be available to download for free. Prices will vary on the size of the corporation or company. Entrepreneurs and startups will be charged less, either a nominal amount or no fee to encourage innovation through trial and error of new ideas for what can be done with 3D HD maps.
A second phase of the project will create a 3D HD map of the City of Stratford, located 150 kilometres west of Toronto, as well as Highway 7, which conitself as an autonomous-vehicle test bed, now has the city partially mapped using light detection and ranging (lidar) images.
Lidar is a remote sensing technology that uses the pulse from a laser to collect measurements, which can then be used to create 3D models and maps of objects and environments.