Who’s Really Driving Those Driverless Cars?

Fully-automated self-driving cars are still years away, but the hype for the new technology is already racing down the highway.

Ford, BMW, Volvo and Tesla are predicting fully autonomous vehicles on the roads by 2021 or sooner. They conjure up an almost-here, utopian future of an end to traffic congestion, vastly reduced pollution and safe roads.

Greener still: With so many fewer cars being driven, fewer parking spaces will be needed. You can start clearing away parking lots and putting in parks. The ultimate statistic of wasted space is that the average vehicle is used only 4 percent of the time and parked the other 96 percent! (We got that from Uber, which is already fielding a fleet of “self-driving” cars in Pittsburgh.)

But if the technology is indisputably real, the timeline is arguably fantasy.

“These statements are aspirations; they’re not really reality,” says Raj Rajkumar, a professor of engineering at Carnegie Mellon University, who advises General Motors. “The technology just isn’t there. … There’s still a long way to go before we can take the driver away from the driver’s seat.”

Mary Cummings, a professor of engineering at Duke University, maintains that, by definition, a fully autonomous car “operates by itself under all conditions, period. We’re a good 15 to 20 years out from that.”

In fact, the carmakers’ claims of a self-driving car by 2021 are not as sweeping as they sound. As The Wall Street Journal ascertained after a little digging, these guys are hedging. Ford’s, for example, will only be self-driving “in the portion of major cities where the company can create and regularly update extremely detailed 3-D street maps.” And who knows when that will be.

The recent bust-up between Mobileye and Tesla in the aftermath of a fatality involving the technology was quite revealing. Mobileye insists that its sensory system works fine — as long as you always remember to keep your hands on the wheel.

In other words, it goes by itself — but you can’t let go. You have to remain vigilant in case the sensors miss something, like a semi-trailer coming at you.

Mobileye’s seemingly self-contradictory statement stems from the fact that we have entered a transitional period. The new technology has arrived — but we aren’t there yet.

The publication a few days ago by the Department of Transportation of a 15-point safety checklist for self-driving carmakers was another indication of this in-between stage. The checklist on each model, which is to be made available to the public, addresses the overriding concern for safety. The car buyer will be able, they say, to see at a glance just how far the maker has gone to make the thing safe.

However, these are not legal requirements; they are voluntary guidelines for the industry, regulation with a light hand.

Transportation Secretary Anthony Foxx said that the checklist got a warm welcome at the Group of Seven nations on Sunday.

It’s the Obama administration’s way of giving in to industry pressure to allow them to move ahead without regulatory encumbrances, but at the same time letting the manufacturers know they are being watched very carefully.

Or, as President Obama put it, “Regulation can go too far. Government sometimes gets it wrong when it comes to rapidly changing technologies. That’s why this new policy is flexible and designed to evolve with new advances…”

But, “make no mistake: If a self-driving car isn’t safe, we have the authority to pull it off the road,” Obama said. “We won’t hesitate to protect the American public’s safety.”

All the talk about safety is a little disingenuous, though. The reason the technology is barreling along with that familiar, you-can’t-stop-progress inevitability is that deep down, no matter the homage they pay to safety, they’re not really so concerned. They’re confident they can satisfy the regulators and the public.

It’s not like driving today is perfectly safe. Nearly 100 people die every day on U.S. roads. The old adage, “If it ain’t broke, don’t fix it” doesn’t (or don’t) apply.

Foxx articulated the safety issue very well when he said: “One of the things I think that autonomous vehicles suffer from is that they get compared to perfection, and not to the 94 percent of car crashes that are attributable to human factors. We have to make the right comparisons,” Foxx said.

“These vehicles will not be absolutely perfect in terms of having no accidents, but by comparison they can be markedly better,” he said.

Going forward, the question will be how to define “markedly better.” How high a rate of malfunctions and accidents will be acceptable?

We have to be realistic. We can’t demand perfection. The desire for safety and stability are ever in conflict with innovation, which by definition creates change and entails risks, both economic and physical.

In the meantime, it might be a good idea for the auto companies to tone down their hyperbole, and to focus more on that checklist. Their idea of an acceptable level of risk may not be the same as ours.

Finally, let us hope and daven that by the time 2021 arrives, Moshiach will long be here, and we will have other things on our minds than driverless cars.