A few weeks ago, a white Ford Motor Co. vehicle sporting an Uber Technologies Inc. logo and a lot of self-driving gear on the roof drove down my street in San Francisco. There was clearly a driver at the wheel and a passenger next to him, and it was hard to tell if it was in self-driving mode, since the driver’s hands were on the steering wheel.
I was the only one staring at the car on the street in a possibly jaded San Francisco Bay Area, a hotbed of serious research and hype on autonomous cars and how they will change our lives. Many in Silicon Valley are convinced that robot cars are literally around the corner, as this one was for me.
The recent barrage of news—from agreements between big auto makers and tech upstarts, to trials by nuTonomy in Singapore and Uber in Pittsburgh, to new set of federal recommendations and guidelines—seems to back up that view. But there is still a lot of work to do to make driverless car technology ready for prime time: get to affordable pricing, convince the leery about safety and most important, strong federal rules for the road.
The disconnect between a slow-moving reality and the Formula One pace in Silicon Valley minds is similar to the gradual approach from car makers and a hard push from tech companies, especially Alphabet Inc. GOOG, +0.39% GOOGL, +0.27% The resulting confusion from consumers was demonstrated at a recent panel on driverless cars, when speakers at the Commonwealth Club of California were asked whether people should wait before buying a new car, as they tried to get a sense of how close self driving cars are to reality.
“We have to think about the fact that the laws are not going to move that quickly,” said Linsey Willis, a spokeswoman for the Contra Costa Transportation Authority, home of Concord, California’s GoMentum Station, the largest secure self-driving test center in the U.S. “If you are not able to take advantage of a shared network, my advice is to buy a new car.”
That is music to the ears of anyone in the automotive industry. For all the partnerships and R&D efforts at auto makers around the world, they all are now in the classic innovator’s dilemma situation, where they can see a technology looming in the future that could disrupt their entire business. Ford F, -0.25% readily admits that the industry is indeed at a serious inflection point and is one of the big auto makers forging partnerships, working with and buying startups.
Car makers, though, are taking a different approach to the self-driving car than tech companies like Google parent Alphabet, which is the farthest along in the autonomous-car race. Car makers are working within the existing automobile design and are adding more and more computing functions, while Alphabet‘s Self-Driving Car Project is starting from the ground up and redesigning the car, without brakes or a steering wheel, aimed at reaching full autonomy faster.
Car makers’ evolutionary approach—with add-ons enabled by more sensors for semiautonomous driving features like adaptive cruise control, blind spot detection and parking assistance—is a way to ease consumers into the idea of more automated functions. This should increase consumer acceptance for the potentially scary but also convenient idea of a car that eventually drives itself, pushing through scares like a recent panic about Tesla Motors Inc.’s TSLA, +0.70% semi-autonomous features .
Car-hailing companies offer more promise for early autonomous ambitions. Like Uber, Lyft hopes to lower its operating costs by eventually eliminating many of its human drivers and operating a service populated with many autonomous cars.
“This is coming to our society,” said Emily Castor, director of transportation policy at Lyft, another panelist at the Commonwealth Club. “We’d rather be the Netflix than the Blockbuster in this situation.”
Even the car-hailing vision, though, will likely play out gradually, first in limited-service vehicles like taxis with specific routes, such as to an airport, and partial truck routes on stretches of freeways—Uber recently acquired a self-driving truck company. That type of usage would allow the technology to learn and improve accuracy in vision and reactions.
“I bet in 10 years, there are autonomous cars that you will be able to buy.”Alex Lidow, chief executive of Efficient Power Conversion Corp.For consumers, a gradual approach is likely because the cars are initially going to be extremely expensive. Alex Lidow, co-founder and chief executive of Efficient Power Conversion Corp., a privately held chip company in San Jose, said he expected autonomous driving will gradually evolve like the once costly options such as anti-lock brakes and air bags, as the cost starts to come down.
“I bet in 10 years, there are autonomous cars that you will be able to buy,” said Lidow, whose company makes chips for LIDAR systems that create three-dimensional images or maps of the surrounding area, a core component of self-driving vision systems. “They will be expensive, but you will be able to buy them.”
Lidow, 61, believes it will be possible to buy a self-driving car before he has to give up his driver’s license due to old age. He estimates that autonomy functions will cost an additional $10,000 or more and first arrive for consumers in high-end models with luxurious amenities.
The wild card in the gradual approach is Alphabet, the closest to actually producing an autonomous car. Mike Ramsey, now an analyst with Gartner Inc., said it is easy to see that the Google parent company “is way ahead of everybody” based on data companies file with the Department of Motor Vehicles, so-called disengagement reports that tell how many times the vehicle operator had to intervene and take over the controls. In its most recent monthly report on its self-driving car project, Google said its cars have driven more than two million fully-autonomous miles on public roads.
“It can get very confusing when you see all these announcements,” Ramsey admitted. “Uber is not ahead, they are not even close.…The lead in capability for Google is substantial, it is not insurmountable forever, but right now it is big.”
So far, none of these deals has actually occurred, and Apple has never confirmed that it is working on any kind of a car, though Chief Executive Tim Cook has waxed enthusiastically about the car market. At the company’s shareholder meeting this year, he hinted that there are exciting things in development when he told a shareholder that investors would have to wait like a kid at Christmas, and that it would be “Christmas Eve for a while.”
Despite excitement from investors, Apple does not seem to be a real player yet. And Google desperately needs to find a partner to complete its vision, Ramsey said.
“They are now struggling to find someone to build a purpose built vehicle just for them,” Ramsey said. “They would love to sell that. They have the data from the car, they have you in the center of their ecosystem.”
But car makers don’t want to be contract manufacturers for Google, especially because Google wants to own the whole system. Google has a deal with Fiat Chrysler Automobiles NV FCAU, -2.15% to test self-driving Pacifica minivans in a Detroit suburb, as part of a new research center it is establishing in the Motor City area but those vehicles were also recently spotted in a garage in Mountain View, Calif., according to the Verge.
Every company has a big stake in this future, however it plays out. Apple needs a big product hit, Google wants to own all of your data, the car makers need to keep selling cars. Startups are trying to jump into a field with computing and robotics expertise, where their best hope is to be acquired for talent. And the government is trying to move faster, so it does not fall behind in making new rules for robots on the road.
All of this competition and uncertainty means that we should probably hit the brakes on hopes for autonomous cars driving down our streets in mass numbers anytime soon, instead of the single oddball I spotted. When it does eventually happen, Alphabet and traditional auto makers will likely be the main players, but expect many bumps and collisions along the way.