Canada Rolls Out Own Testing of Autonomous Cars

OTTAWA – A pilot test of automated vehicles is under way in the Canadian province of Ontario, aided by a groundbreaking regulatory system designed to encourage technology and automotive companies to increase R&D.

The experimental program, which began Nov. 28, takes advantage of a provincial law enacted Jan. 1 that allows companies to test self-driving vehicles on Ontario roads. Other Canadian provinces, which have jurisdiction over transport issues in the Canadian confederation, still are developing such rules.

Several U.S. states also are angling to become centers for autonomous-vehicle development in North America. Eight states, as well as the District of Columbia, have adopted legislation that to some extent allows autonomous vehicles to travel on public roads.

Participants in the Ontario pilot test include the University of Waterloo, Canadian communications giant BlackBerry and Germany-based motor-home specialist Erwin Hymer Group.

 The three vehicles being used in the project are a Lincoln MKZ hybrid sedan dubbed the “Autonomoose” by its Waterloo University operators; a Roadtrek E-trek recreational vehicle equipped with different levels of automation operated by Erwin Hymer; and a ʼ17 Lincoln with automated features being tested by BlackBerry, which is developing software in association with the project.

To gain access to Ontario’s roads, the participants had to demonstrate the vehicles complied with the J3016 standard for automated driving systems, released in 2014 by SAE International, a U.S.-based automotive-research association.

It has six levels ranging from 0 (no automation) to 5 (full automation), based on the control or components of the vehicle, and monitoring of the driving environment, whether by a human or the vehicle.

In a briefing note, SAE says under the standard the biggest difference occurs between level 2, where the human driver performs part of the dynamic driving task, and level 3, where the automated driving system performs the entire dynamic driving task.

Experimental programs in Ontario must involve vehicles with an automated driving system that operates at levels 3 – where the system can request a human driver to take over under certain conditions – or 4 or 5, the latter indicating the autonomous-driving technology completely controls all driving functions of the vehicle.

In addition to adhering to the SAE standard, participants must comply with Ontario’s Highway Traffic Act, have human drivers holding licenses appropriate to their vehicle who are ready to take over at any point, and have insurance of at least CN$5 million ($3.75 million).

The province’s pilot framework does not limit autonomous vehicles to select roads or highways, and no special permits, plates or identifiers are required.

Ontario's Manufacturing, Tech Infrastructure Touted

Ontario is “the most well-positioned region in North America” to deliver innovation in automated vehicles, Transportation Minister Steven Del Duca tells WardsAuto in an interview, citing the province’s existing large automotive manufacturing industry and its technology center in the Kitchener-Waterloo area, home to BlackBerry and other tech companies.

The transportation ministry, along with key stakeholders such as auto and tech companies, universities and infrastructure companies, worked together to develop both the regulatory framework and the project itself. While the regulations are to span 10 years, Del Duca says that is enough time for regulators and industry to maintain “the ability to be nimble and to respond to what’s happening in the marketplace.”

David Adams, president of Global Automakers of Canada, says: “We will likely see fully automated vehicles in closed areas on closed regular routes by 2025. However, it will be at least 2040 to 2050 before fully automated vehicles are mainstream.”

Adams notes the approach to vehicle automation is “bifurcated” between traditional automakers who will introduce increasing levels of automation with each new generation of vehicles, and newer players such as Google, who would want to release a fully automated vehicle “right out of the gate”.

Barrie Kirk, executive director-Canadian Automated Vehicles Centre of Excellence, suggests it would be easier to roll out more autonomous vehicles than those that require land-based infrastructure to guide their driving. The idea is to put as much artificial intelligence as possible into the vehicles, “as the cost to add special infrastructure to all the highways (and) urban and rural roads across the country is impossibly expensive.

 “Once (automated vehicles) are launched with significant (market) penetration, then there are ways we can optimize the infrastructure to make it better for AVs,” Kirk says.

On the governmental side, Del Duca says investment is less important than “an attitude or cultural shift to be open and nimble, and to want to support technological advancement.

“What I want here in Ontario is not only for us to take advantage of the technological lead that we have – and that’s globally recognized – but also to say to our major manufacturers in the auto sector, ‘How can we make sure jobs continue to stay in Ontario, and have more investment come to Ontario?’ Because this is the place you want to be,” Del Duca says, “especially in the North American context – for automated, connected vehicles of the future.”

Set the Vehicle Controls, Then Go to Sleep

Kirk notes the current project is particularly useful because it includes autonomous technology for mobile homes.

“In the future, being able to get into a mobile home, get into bed and have the (the mobile home) drive you someplace while you’re sleeping sounds wonderful as a vision.”

He predicts the next few decades primarily will see the advent of self-driving taxis while human drivers eventually will be rendered obsolete.

What experts need to consider for the future, Kirk says, is how the shift from human-driven to fully automated cars should be structured. He questions whether partial automation is even a phase to be considered, as previous tests by Google and Tesla have shown humans are not consistently alert monitors of vehicles of which they are not fully in control.

“Let’s fast-forward to 2030 when the data is in,” he says. “If computers really are much safer drivers than silly, stupid humans, is it ethical to allow humans to keep driving and cause all that death and carnage on the road, when there is a better and safer technology alternative?

“At what point in time will we have to ban human drivers?”