By Edmund Sandoval
“When will the first autonomous vehicles be deployed in the real world?”
It’s a good question, and one we asked here more than two and a half years ago. More specifically, we asked, “When will ordinary people be able to operate automobiles on public highways with no driver intervention?” With 2018 in full swing, and autonomous vehicle technology having taken several bounds, we thought it time to revisit that question.
Is there any more consensus today as to when the driving (and non-driving) masses can expect fleets of autonomous vehicles to deliver them to and from work, play, and everything in between?
The answer is yes … with the usual caveats and qualifiers thrown in for good measure. So what can we say with certainty?
For starters, we can report that autonomous and semi-autonomous vehicles are already on the road, and they’re not just circling the block. Testing of fully autonomous vehicles has kicked off in California, Texas, Pennsylvania, Arizona, Michigan, Washington, and a covert-sounding former Air Force base somewhere on the West Coast.
Waymo vehicles, the self-driving arm of Google, have already logged more than 3 million miles on public roads. Since 2014, Tesla’s semi-autonomous vehicles—all newly manufactured cars contain Autopilot hardware—have banked more than 1 billion real-world miles courtesy of its customers. Impressive, right?
Yet, you can’t simply stroll down the block to the local Waymo dealer (any driving so far has been conducted by Waymo itself), and Tesla’s vehicles require drivers to be present behind the wheel when Autopilot is engaged. That means hands on the steering wheel, sensors to detect whether you’re on a mobile device or your attention is wandering to the backseat, and warnings that flash when any of those things occur. It’s also worth mentioning that Tesla itself does not consider its Autopilot technology to be self-driving technology; rather, as reported by Vox, it deems the option as a kind of advanced cruise control.
So, yes, there are autonomous and semi-autonomous vehicles on the roads in America, but you can only purchase the one that requires you to, you know, drive it.
That’s because the developers of these vehicles and the technologies required to make them work are playing a long game. It’s inevitable, but it’s not going to be quick. Rather, the transition will be gradual.
For instance, at the Davos World Economic Forum this year, Uber CEO Dara Khosrowshahi said the company expects to have self-driving cabs on the roads by mid-2019. But what he’s really saying, as reported by thenextweb.com, is that he expects a very small number of autonomous cabs to be on the road by mid-2019. Those self-driving cabs will only handle a very small number of rides—along routes Uber is confident they can handle. And riders will always have the choice of opting for a self-driving cab or person-driven cab. Any increase in self-driving cars will occur over the course of years. (Uber is also involved in a lawsuit with Waymo, which means that forecast could be optimistic.)
Where does that leave the other players? Well, depending on whom you talk to, the spectrum is broad, and that’s putting it lightly. Here’s a list that balances outlooks compiled by QuartzMedia.com and driverless-future.com, ranging from the confident to the measured:
- Tesla — 2018: It shouldn’t surprise that this frequent boaster is out front. Tesla and its visionary founder Elon Musk do, however, acknowledge that regulatory obstacles may impact this timeline. Musk has also claimed that, “a Tesla will drive in fully autonomous mode from LA to New York City.”
- Toyota and Nissan — 2020: A couple of years out, these two plan for varying levels of autonomous abilities. Toyota will get you on the highway; Nissan will take you to your favorite ramen joint during rush hour.
- BMW, Ford, and Lyft — 2021: BMW expects to take you on a highway trip and a spin around the city. Ford will do it all (as long as the routes occur on roads that have been mapped). Lyft plans for nearly all rides to be autonomous, but they’ll drive slowly.
- Otto (Uber’s trucking subsidiary) — 2027: We could see fully autonomous broad-scale deployment in about ten years. (Iron Skillet, now is the time to panic.)
- Kia — 2030: An autonomous Kia will pick you up and drop you off. All you’ll do is buckle and unbuckle your seatbelt. Whether this involves giant hamsters is unclear.
This is all to say that the consensus is still wobbly, if more dialed-in than just a few years ago.
What is apparent, though, is that all options have a ways to go. Every fender bender will be painstakingly investigated, and accidents will make national news, in spite of the tens of thousands of annual deaths related to human-caused collisions. Autonomous cars today are imperfect, about as effective as drivers as their human counterparts. Waymo, the highest-performing of the bunch, required human intervention approximately once every 5,000 miles; Mercedes-Benz, on the other hand, required intervention at a hair-raising rate of once every 1.8 miles, as noted by Wired.
Additionally, the regulation required to get these vehicles on the roads is likely to be as deep a morass as any, considering the numerous agencies and legislators that would weigh in.
Finally, as reported by Forbes, economic considerations will surely short-circuit some of these plans. Upstart companies may challenge established brands in new and emerging markets, and attendant industries — like trucking and delivery, auto insurance providers, energy companies, dealerships, parking lots, and ride-sharing services — may struggle to adapt to a post-driver reality.
Time will tell. Sooner rather than later. Maybe.
Please note that this article expresses the opinions of the author and does not reflect the views of Move Forward.