I just got back from the two most important auto shows of the season. Kicking off the year in frigid January, the Consumer Electronics Show in Las Vegas and the North American International Auto Show in Detroit pretty much bookend where the auto industry has been, and more importantly, where it is going. One is ascendant, the other seemingly on its last legs. One debuts flying taxis, the other supercharged relics of pony cars past. One, it really can’t be overstated, represents the future of the auto industry; the other, unfortunately, its past. And I don’t think I have to tell you which is which.
There is one thing, however, both shows — or at least, the exhibitors at both shows — can agree on: There’s going to be a pause in our headlong rush into self-driving cars. Oh, we’re going to bump up the typical car’s autonomy — which, as I will explain in a minute, is not the same thing as self-driving — and there were plenty of cars running around without steering wheels, especially in Vegas. But, except for those Silicon Valley startups that literally have nothing to lose, the entire industry realizes it got ahead of itself in promising fully autonomous automobiles in the next few years.
Waymo-nee-Google, of course, upset the whole apple cart back in 2012 when it announced it would transition from human-driven cars to the full robot in one step, quite literally jumping from what the National Highway Traffic Safety Administration in the U.S. labels Level 0 (absolutely no driver’s aids) to Level 5 (no driver ever needed; you can have sex in the back seat for all Google cares) in one fell swoop. With sensationalism all but owning media coverage these days, automakers — whose engineers are trained almost from birth to adhere to the “go-slow” school of revolutionary advancements — were forced to abandon their traditional step-by-step incrementalism in order to keep up.
Ever since, automakers have fallen all over themselves, into a game of one-upmanship quickly developing as normally conservative industry leaders promised “self-driving” cars on ever-tighter time spans. (And let’s remember that, for the average consumer, autonomous means completely self-driving, not some partial version thereof.) Ford, Volvo, GM and others, not to mention various Silicon Valley interlopers, kept upping the ante until the public — again, prodded by the media — started expecting that we’d all be being chauffeured around in little steering wheel-less bots by 2020; 2021 at the latest.
Well, not that that was ever a reality, but the industry as a whole has almost completely backed away from this prediction, adopting a much more “humble attitude” says The Information.
Automakers previously trumpeting their self-driving programs have suddenly gone radio silent. High-tech firms delving into the mobility business for the first time — AEV, Udelv and Nuro — are now focusing on the autonomous transport of goods rather than people. Even Waymo is rolling back its much-trumpeted autonomous taxi service. “A couple years ago, everyone was like, ‘we’ll be done [developing self-driving cars] in three, four, five years,’” Heiko Kraft, director of analysis and testing for Mercedes-Benz self-driving car research, told author Cory Weinburg. “Everyone was promising a lot.”
There’s a simple reason for this back-pedalling. Actually, two: Tesla and Uber. Or more specifically, the two fatalities — Joshua Brown in a Model S and pedestrian Elaine Herzberg by a Volvo — involving Tesla and Uber’s robotic drivers. Though both companies have tried to downplay the significance of these accidents — Tesla by denying responsibility; Uber by simply stopping testing in Arizona — the consequences have been far-reaching, especially in the court of public opinion.
Indeed, a recent study by Cox Automotive reveals the depths to which the promise of self-driving has dropped in the public’s perception. According to Karl Brauer, executive publisher of Autotrader
and Kelley Blue Book
, while awareness of fully autonomous automobiles has increased by more than 50 per cent in the last two years, trust in Level 5 self-driving has dropped some 40 per cent.
The raw numbers are even more startling — while two-thirds of Americans now count themselves as familiar with self-driving technology, barely a quarter actually trust it, an incredible reversal from the 47 per cent who said they completely trusted autonomous automobiles just two years ago.
According to Brauer, the blame for the dramatic levels of distrust can be attributed directly to those two high-profile accidents, 61 per cent of respondents claiming knowledge of the Uber accident alone. So traumatized are they by the fatalities attributed to automotive autonomy that, while previous studies indicated that those in favour of self-driving wanted Level 4 and 5 abilities — full control ceded to the machine — the new Cox study shows now more skeptical consumers would prefer that automakers limit themselves to Level 2.
It marks a dramatic shift in our desire for automated driving. Nearly half of us would never buy a Level 5 autonomous vehicle (an increase of more than 50 per cent from just two years ago), fewer of us believe our roads would be safer if all cars were fully autonomous and, perhaps most telling of all, twice as many of us “would feel [more] uncomfortable riding in a vehicle fully driven by a computer … than in a vehicle driven by a complete stranger.”
Three short years ago, the lure of autonomy was the total, no-steering-wheels-needed autonomy promised by Waymo. In the wake of those devastating accidents, what we now want is the Advanced Driver Assistance Systems — adaptive cruise control, lane departure assist, et al — currently available on most luxury cars and, increasingly, even little econoboxes. (Ninety-five per cent of Toyotas sold in Canada, including base-model Corollas, are equipped with the company’s Safety Sense semi-autonomous driver’s aids.)
In other words, what we want is the exact level of automated driving we have now. We have had, if Cox’s research is to be believed, quite enough of progress.