Motor Mouth: The 'social ballet' of autonomous driving

Article content

“The social ballet of driving.” That’s what Dr. Ryan Eustice, senior vice-president of automated driving at Toyota’s Ann Arbor-based Research Institute, calls one of the biggest problems for the self-driving cars of the future. Different societal norms are going to make the task of having robots drive on the same roads as we humans a huge challenge.  

Article content

“Driving in Boston is different from driving in Detroit,” says the director of the University of Michigan’s Perceptual Robotics Laboratory, “and having driven in Rome, I feel it’s more of a sport there.” That’s important, says Ryan, because every little bit of the art of safely controlling an automobile — the speeds driven, the attention given to traffic signals, and even the differing ways different cultures treat the always difficult pas de deux of merging traffic — will come into play when fully autonomous cars start sharing roads with humans.  

We apologize, but this video has failed to load.

Try refreshing your browser, or
tap here to see other videos from our team.

Motor Mouth: The 'social ballet' of autonomous driving Back to video

And they will be most definitely sharing roads with us. As Stefanie Bruinsma, manager of industry engagement for the University of Waterloo, points out, “Governments are not going to build two highways” just so humans don’t have to share roads with computer-driven cars. In other words, humans and robots are going to need to learn to get along.

Article content

As to when this all might happen, there is some contention. Elon Musk, of course, is adamant that fully autonomous Autopilot is right around the corner. But then Navigant Research — along with many other organizations — dismisses Tesla’s robotics, which famously don’t use advanced LiDAR sensors, as “inconsistent” and says its products “do not match its proposed mobility business model.”  

In most estimations, Level 5 autonomy — which Bruinsma says allows the driver “to lay back and take a nap” while the car drives itself — is still a long way off. That’s especially true here in Canada where our winters pose far more difficult conditions — snow banks, black ice, and absent lane markings, not to mention the salt and sand we throw on our roads — than Waymo might find in San Francisco, or to which Uber is exposed in Arizona. That’s why the former powertrain researcher and Mini mechanic expects it will be at least 10 — and perhaps even 20 —years until we see fully autonomous cars on our roads. Those are prophetic words considering that Waterloo is Canada’s foremost university hub for autonomous vehicle development.

Article content

That’s not to say that we Canadians will not enjoy some form of autonomy in the near future. As Bruinsma points out, Level 3, where “the driver can sometimes remove their hands from the wheel”; and Level 4, or high automation (which can do everything the driver can but only under specific conditions), are much easier to attain. And, unlike the current Level 2 cars presently available — Tesla Autopilot and Cadillac Super Cruise being the leading examples — they don’t require human supervision at all times.  

Technically, no matter what the claims from certain manufacturers, current semi-autonomous vehicles require that a driver keep their hands on the wheel at all times. Even most experimental vehicles — for instance, the ill-fated Uber-ized Volvo that was involved in a fatality in Phoenix — are supposed to have an “emergency driver” who is supposed to remain alert at all times just in case of computer failure.

Article content

That said, our inclement weather — not so apparent as I write this on a record-setting warm March 11 — does provide us with an opportunity to become a “global hub” of autonomous research. Raed Kadri, head of Ontario’s Autonomous Vehicle Innovation Network, claims that, with “our existing strengths in automotive and IT, coupled with our strong entrepreneurial cluster and our innovative startups,” we have everything necessary to remain leaders in the autonomous and mobility sectors. In other words, if we play our cards right, Ontario’s frozen winters — and our abundance of engineering talent, of course — could put us at the forefront of autonomous research.

Humans and robots are going to need to learn to get along

It’s a bit of a surprise, then, that one of the technologies that might get autonomous vehicles through a Canadian winter actually comes from sunny California. But even though his laser lights were originally “brought to market for illumination reasons,” Paul Rudy, CMO of Kyocera SLD Laser, says his ground-breaking technology is finding an even bigger audience for its range-finding abilities. Similar in its range-finding abilities to the radar sensors currently used in autonomous vehicles, SLD’s co-founder says his lasers can “see” objects in a car’s path further and more reliably. As he points out, unlike radar sensors, which are located in the front grille — and which Motor Mouth can attest can often fall prey to snow and ice, not to mention salt and sand — his lasers are, of course, housed in the headlights, which typically have their own wiper/washer cleaning mechanisms.

Article content

As for the longer-term future, Rudy says   his lasers could some day do the same job as the LiDAR systems — as in light , detection and ranging — that most manufacturers maintain is crucial to full autonomy at a fraction of the cost.

Whether this all plays out over 10 — or even 20 — years, the research, as Eustice points out, is already paying off dividends. Toyota is already equipping most of its cars — as are some other manufacturers — with an advanced suite of advanced driver’s aids. Called Toyota Safety Sense, it includes, at its most basic, adaptive cruise control (which uses autonomous-like radar sensors to maintain a safe distance to the car ahead), lane departure alerts (which will also steer the car back into its lane) and a pre-collisions system that will brake automatically if it thinks you’re going to hit something. It’s all part of Toyota’s “Guardian” plan to “pair AI with humans to make drivers superhuman.” By using many of the same technologies as autonomous vehicles, says Eustice, Toyota can give drivers similar “perception, prediction and planning abilities” as self-driving cars. In other words, inasmuch as full autonomy is the ultimate goal, the immediate focus should be to use the same technology to make us all better drivers.

In the end, that’s perhaps the shame of all this hype surrounding full autonomy being right around the corner. If there’s something on which almost all experts agree , it is that our immediate priority should be making sure that advanced driver assistance systems — that “Guardian” Eustice speaks of   — are rolled out in vehicles across more price ranges. That would save more lives in the here and now.  

LISTEN: How EVs have saved Canada’s automotive manufacturing sector.

Subscribe to Plugged In on Apple Podcasts, Spotify, Stitcher, and Google Podcasts.