It is now widely accepted as fact that autonomous vehicles are the way of the future. The only really big question is, “how soon?”
Some proponents, both within and outside the auto industry, have predicted, even promised, fully-autonomous Level 5 vehicles by 2020, if not before. But, more and more, those optimistic projections are being challenged by reality.
Key players, including Ford, have back-pedaled on earlier predictions, acknowledging that they underestimated the massive challenges involved, which extend far beyond the technical to the regulatory and the ethical.
Still, two recent announcements – one by Tesla, the other by Waymo – suggest that full-autonomy, at least within closely-defined parameters, could soon be available to some.
How realistic are those prospects? Here’s our take on them both.
Tesla: no stranger to overstatement
At a recent invitation-only event for investors, Tesla’s controversial CEO, Elon Musk, proclaimed that, “A year from now, we’ll have over a million cars with full self-driving,” a task that would involve upgrading the software, and in many cases the hardware, in virtually every Tesla ever built.
Lest one confuse the term “full self-driving” with the latest software option released by the company, which unfortunately uses that very name, Musk made it clear that in this case he meant full Level 5 autonomy with no geofencing or need for human oversight. In other words, cars capable of driving themselves anywhere on the planet, under all possible conditions, with no limitations.
To put that prediction in perspective, there are no cars on the road anywhere today that achieve Level 5. Not from Tesla, nor anyone else. So, even for Musk, who is no stranger to overstatement, it’s an audacious claim, bordering on the preposterous.
Beyond the sheer magnitude of that prediction, there is the question of Tesla’s ability to achieve Level 5, or even Level 4, capability on a production-ready basis.
At the core of wide-spread skepticism in that regard is Tesla’s reliance on ultrasonic, camera, and radar-based technology for its “autonomous” features, without the use of Lidar – a reliance that many believe may have already been complicit in fatalities involving the use of Tesla’s so-called AutoPilot system.
Every other serious attempt at full autonomy – including Ford, GM Cruise, Uber and Waymo – has accepted the need for Lidar to ensure the necessary levels of redundancy and safety.
There’s also the matter of Tesla’s acknowledged willingness to treat its customers as beta testers, apparently following the Silicon Valley mantra of “move fast and break things,” rather than the more conservative auto industry practice of testing, testing, testing, before releasing for production.
That’s why, along with Tesla’s established history of hyperbole, we’re not taking this latest prognostication seriously.
Waymo: a decade of intense experience
Waymo, which began as Google’s self-driving car project in 2009 and is now a separate subsidiary of Google parent, Alphabet, recently announced plans to build Level 4 autonomous vehicles in Detroit, beginning later this year.
It will do so in partnership with Canadian tier-1 supplier, Magna, at a plant owned by American Axle and Manufacturing.
Unlike Tesla, it won’t be building entire cars but modifying and completing vehicles supplied by Fiat Chrysler and Jaguar Land Rover to install necessary self-driving hardware and software.
With a decade of intense experience in the development of self-driving technologies, Waymo is widely regarded by many as the leader in the field. According to the company, its self-driving vehicles have already accumulated more than 16 million km of testing on public roads.
It ranked number one in a 2018 study by Navigant Research, which evaluated potential autonomous vehicle providers on the basis of 10 criteria: vision; go-to market strategy; partners; production strategy; technology; sales, marketing, and distribution; product capability; product quality and reliability; product portfolio; and staying power. (Tesla ranked 19th out of 20.)
Not only has the company employed many, if not most, of the leading researchers in the field, it has balanced its perspective with solid experience from within the auto industry. Larry Burns, a former vice president of research, development and planning at General Motors, came on board as an advisor in 2011, and industry veteran John Krafcik, most recently CTO of Hyundai Motor America, has been Waymo’s CEO since 2015.
(Burns is co-author of the book “Autonomy: The Quest to Build the Driverless Car and How it Will Reshape Our World.”)
Waymo has taken a measured, step-by-step approach to the development of self-driving technologies, most recently announcing establishment of a limited, commercial Level 4 robotaxi service in the Phoenix area, called Waymo One, in conjunction with Lyft.
The Waymo One self-driving car service still has Waymo-trained test drivers behind the wheel as backups although it will be fully self-driving within a geo-fenced area. The safety driver will eventually be removed from the vehicle.
It’s the responsible approach in our opinion.