Jae C. Hong/Associated Press
Tesla CEO Elon Musk.
- In recent years, many of the companies developing self-driving cars have backed off from the aggressive timelines they had previously set to introduce fully driverless vehicles for personal or commercial use.
- And they are increasingly using investments and partnerships with competitors to share the immense research-and-development costs driverless cars require, as did General Motors with its investment in Cruise Automation.
- But Tesla has bucked both of those trends, saying that fully driverless cars will be available in 2020, powered by technology the electric-car maker has largely developed in-house.
- Tesla CEO Elon Musk's prediction that self-driving cars are imminent has been met with skepticism from experts and the electric-car maker has a history of making counterintuitive bets.
- Some have paid off, some have not.
- Sign up for Business Insider's transportation newsletter, Shifting Gears, to get more stories like this in your inbox.
- Visit Business Insider's homepage for more stories.
In recent years, the autonomous-vehicle industry has confronted a reality check. The promise of cars that can drive anywhere under a variety of conditions without human supervision has shifted from a near-term goal to the distant endpoint of a long, incremental process.
Additionally, the companies developing self-driving technology are increasingly using investments and partnerships with competitors and legacy carmakers to share the immense research-and-development costs that driverless cars require.
"We're not in an automation revolution. We're in an evolution of automation," said Bryan Reimer, a research scientist at the MIT Center for Transportation and Logistics.
Deadlines have become more conservative
Ford
A Ford self-driving test vehicle.
Earlier this decade, the companies developing self-driving vehicles took a more forceful approach to promoting themselves in an effort to attract attention and talent, said Brian Collie, the global lead for the automotive and mobility at Boston Consulting Group. Now, they are more measured in the way they talk about their goals.
In 2016, Ford said it intended to make a vehicle without a steering wheel or pedals in 2021. The next year, General Motors said Cruise, the autonomous-driving company it acquired in 2016, would launch a self-driving taxi service by the end of 2019. As recently as early 2018, Uber said it would give customers rides in self-driving vehicles in 2019.
All three companies have since backed off of those predictions.
"We overestimated the arrival of autonomous vehicles," Ford CEO Jim Hackett said in April, though the automaker still plans to start using self-driving vehicles for rides and deliveries in a handful of cities in 2021.
The more automotive and tech companies have tested their autonomous-driving technology, the more they have realized how much work is ahead of them, said Sam Abuelsamid, a research analyst at Navigant who focuses on mobility.
"There's a rule of thumb in engineering that the first 90% of a problem takes you 10% of the time, and then the last 10% takes you 90% of the time" he said.
Failing to set realistic expectations would anger investors and erode trust among potential customers, he added.
Tesla begs to differ
This year, Tesla has said that its vehicles will achieve the highest level of self-driving capability, called "Level 5" by the Society of Automotive Engineers, in 2020, while unveiling a computer chip designed for autonomous driving it says is the world's best. The moves were just the latest in a long series of bold bets that have defined Tesla's unconventional approach to making and selling cars.
"It's a company that operates in very different ways from other automakers," said Ed Kim, the vice president of industry analysis at AutoPacific.
Tesla did not respond to a request for comment on this story.
CEO Elon Musk has never shied away from making aggressive predictions about Tesla's autonomous-driving technology. In 2015, Musk said the company's vehicles would be able to drive themselves by 2017. He also said Tesla would send a self-driving vehicle across the US by the end of 2017, and when the company missed that deadline, he pushed it to the first half of 2018, a target Tesla also failed to hit.
But rather than adopt the more conservative tone taken by his rivals, Musk has continued to indicate that fully-driverless vehicles are more or less around the corner. In April, he said Tesla vehicles would have Level 5 capabilities by the end of 2020.
"I could be wrong, but it appears to be the case that Tesla is vastly ahead of everyone," Musk said during an April interview.
If he's right, Tesla will claim a sizable advantage over other automakers, whose vehicles are not likely to have self-driving capabilities anywhere close to Tesla's. But autonomous-vehicle experts have cast doubt on Musk's claims, saying it's unlikely they will materialize in the near term.
"None of that is going to happen," said Raj Rajkumar, a co-director of the General Motors-Carnegie Mellon Connected and Autonomous Driving Collaborative Research Lab, of Musk's prediction for 2020. "In 2021, not going to happen either."
There are many difficult situations that a Level 5 system would have to be able to handle, like black ice and dirt roads, the MIT research scientist Reimer said. Unless Musk is aware of a technological breakthrough that no one else has discovered, his timeline for the readiness of Tesla's Level 5 technology is unrealistic.
"A lot of these conditions are very complex and are not something that Elon or anybody else is going to solve in the next decade," Reimer said.
Already, Tesla is coming close to missing another deadline, that its vehicles will be able to control braking, acceleration, and steering in any environment, while still requiring driver supervision, by the end of this year.
"I think we will be feature-complete full self-driving this year, meaning the car will be able to find you in a parking lot, pick you up, take you all the way to your destination without an intervention - this year. I would say that I am certain of that. That is not a question mark," Musk said in a February interview.
Tesla has given no indication that will happen between now and the last day of 2019, just 10 days away, though Musk suggested on Thursday that an upcoming software update will bring Tesla vehicles closer to the vision he outlined during the interview.
Meanwhile, Tesla's rivals are forming partnerships
Waymo
A Chrysler Pacifica outfitted with Waymo's autonomous-driving technology.
Tesla has paired its bet that it is the leader in the self-driving vehicle industry with another, that it needs little help from other companies. Developing autonomous-driving software and hardware, as well as the vehicles that will use them, in-house gives Tesla the potential to reap all of the financial benefits from its investments.
But there's a reason other companies aren't taking that approach: It's expensive, and the payoff is uncertain. According to a June report from AlixPartners, total spending on autonomous vehicles will reach $85 billion by the end of 2025, a sum greater than the combined profits earned by Volkswagen, Toyota, General Motors, Nissan, Ford, Honda, BMW, and Daimler (the parent company of Mercedes-Benz) in 2018.
While automakers have been able to see some of the fruits of their labor in the real world - things like driver-assistance features that can brake to avoid an accident or keep a vehicle from drifting out of its lane - it's unclear when comprehensive self-driving systems will be ready for the road.
Many auto or tech companies developing autonomous-vehicle technology have either invested in or formed a partnership with another company. The number of partnerships related to autonomous vehicles increased by 122% from 2017 to 2018, according to AlixParnters.
In some cases, the deals have aligned companies with different skill sets. The Google spinoff Waymo, which at one point built a prototype self-driving vehicle, has in recent years bought vehicles from auto companies like Fiat Chrysler and Jaguar Land Rover as it has shifted its focus entirely to making the software and hardware necessary for autonomous driving. GM has said it will build a self-driving vehicle without a steering wheel or pedals that will use Cruise's technology.
But in other cases, companies have seemed to admit that they need help as global auto sales cool off slightly and government regulations require significant spending on electric vehicles. This year, Ford and Volkswagen announced they would expand their alliance to include electric-vehicle and autonomous-driving technology. Months earlier, BMW and Daimler said they would jointly develop self-driving cars.
Making vehicles for the present while investing in the future is expensive, and automakers have finite resources to spend on either, the Boston Consulting Group automotive and mobility lead Collie said. Partnering with competitors allows them to share research-and-development costs.
"People are just being stretched exceptionally thin, from a resource standpoint," Collie said.
Tesla prefers to work alone
Tesla once used software and hardware from Mobileye and Nvidia for its semi-autonomous Autopilot system, but has since shifted toward using more proprietary technology. Tesla has said its autonomous-driving computer chip has more processing power than a rival chip from Nvidia, though Nvidia disputed Tesla's claim, saying the electric-car maker's comparison was inaccurate.
Tesla bought a small computer-vision startup this year, but the electric-car maker has largely worked alone. Doing so allows Tesla to make decisions faster than if it were collaborating with another company, said David Whiston, a Morningstar analyst who covers Tesla and other auto companies.
"They don't want to be bogged down in bureaucracy and seeking the approval of one or two alliance partners," he said.
It helps that Tesla does not face the same pressure to produce short-term profits that more-established rivals like GM, Volkswagen, and Toyota do, Whiston said, which gives the electric-car maker the freedom to more aggressively invest in autonomous-driving technology.
While Tesla has never made an annual profit in its 16-year history, it has been able to repeatedly raise money from investors and maintain a stock price well above those of Ford, GM, and Fiat Chrysler. Some investors and analysts have indicated that their optimism about Tesla's future is based in part on their perception that the company's driverless-vehicle technology is ahead of its rivals'. Taking a more conservative approach would potentially diminish some of that enthusiasm.
Tesla has a history of making bold bets
Rick Eglinton/Toronto Star via Getty Images
The Tesla Roadster.
Autonomy is not the first area in which Tesla has made a counterintuitive bet. The company was started on one: That customers would be interested in high-end, performance and tech-oriented electric vehicles, a segment that was previously defined by slow cars with drab styling.
Other bets, like giving its vehicles the ability to receive updates wirelessly, and using an iPad-like touchscreen to control most vehicle settings, have also paid off, helping Tesla build an enthusiastic fanbase that has driven the company's sales growth.
But Tesla has also taken risks that have backfired, or have yet to produce clear benefits. An effort in 2017 to sharply increase the amount of automation at its vehicle-assembly plant in California brought the company to the verge of bankruptcy, and its 2016 acquisition of the solar-panel company SolarCity - and its $3 billion-plus debt load - have been followed by declining panel sales and the troubled rollout of solar roof tiles.
The widening gap between Tesla and the rest of the self-driving-vehicle industry has created two distinct visions for the future of the technology, one defined by practicality, the other by a brash ambition. The next year could have much to say about which will prevail.
Are you a current or former Tesla employee? Do you have an opinion about what it's like to work there? Contact this reporter at mmatousek@businessinsider.com. You can also reach out on Signal at 646-768-4712 or email this reporter's encrypted address at mmatousek@protonmail.com.
Read more:
