GM’s driverless-car unit Cruise delays robot-taxi service
CEO says startup needs to log more test miles before launch, doesn’t offer new timeline
General Motors Co.’s driverlesscar business, Cruise, will delay the roll out of an autonomous ride-hailing service scheduled for late this year, underscoring the difficulty of safely deploying robotic vehicles on public roads.
Cruise Chief Executive Dan Ammann said the company needs to log more test miles on its driverless electric cars, which have been plying San Francisco streets for more than three years with safety drivers at the wheel. He declined to give a new timeline for beginning a commercial service, which he said would be offered first in San Francisco.
GM in November 2017 set a target of deploying an autonomous-vehicle service “in large scale” sometime in 2019, though executives have said that timing would depend on whether it can be done safely.
The delay by GM’s Cruise is the latest setback for a still-emerging technology that auto and tech executives have touted for years as having the potential to bring about a new era of on-demand transportation.
Car companies are racing to invest in new autonomous-driving capabilities, worried the rise of robot-taxi fleets would reduce the need for personal-car ownership and hurt their core automanufacturing business over the long term.
They also are vying to crack a potential market for autonomous-vehicle services that some analysts and industry executives say could generate trillions of dollars in revenue annually.
But the budding driverless-car industry has over the past year tempered its collective expectations as it faces nagging technical hurdles and an uncertain regulatory picture.
Other big players, such as Waymo LLC and Uber Technologies Inc., have also let their time lines slip or run into other problems.
Mr. Ammann said Cruise has made good progress refining the technology but declined to discuss specifics of the cars’ technical performance.
GM paid around $1 billion for Cruise in early 2016, when the San Francisco-based startup had 40 employees and was testing a self-driving system bolted atop Audi cars. It now employs nearly 1,500 people and has attracted more than $6 billion in outside funding, putting its value at around $19 billion, GM said in May.
Cruise is working to offer an on-demand, self-driving ridehailing service that would operate in pre-determined areas of San Francisco. Kyle Vogt, Cruise’s co-founder and president, has said robot taxis would offer riders a primary form of transportation and compete with Uber and Lyft.
Cruise has also explored plugging its cars into an outside ridehailing firm’s network, so they would be available through the Uber app, for example.
GM executives have said they can envision a day when autonomous-vehicle services could be a bigger portion of its business than manufacturing cars. They have cited profit margins of 20% to 30%, far greater than the 8% margin it earned last year.
Over the past two years, GM shares have rallied following news of major investments into Cruise and upbeat analyst reports about the division’s prospects as a future growth driver.
Shares were down less than 1% Wednesday.
In a research note, RBC Capital analyst Joseph Spak said GM’s delay wasn’t a surprise. He said autonomous-vehicle investors have been pushing back their time lines for commercialization of the technology.
Pushing back the commercialdeployment target puts at risk more than $1 billion that Japan’s SoftBank Group Corp. committed to Cruise last year, according to terms of the investment detailed in GM public filings. A provision of SoftBank’s $2.25 billion investment said $1.34 billion would be paid out once Cruise begins commercial service.
A SoftBank spokesman declined to comment.
“Having a positive first experience with this technology is going to be critically important,” Mr. Ammann said in an interview.
“Any time you’re working on something that’s never been done before, the timeline is likely to move around a little bit,” Mr. Ammann said. At an investor conference in January, he called deploying driverless cars “the biggest engineering challenge of our generation.”
Waymo, Alphabet Inc.’s selfdriving car unit, is viewed by many analysts as furthest along in developing driverless technology. It said in late 2017 that it would soon offer customers autonomous-vehicle rides without anyone at the wheel. However, the firm continues to use safety drivers in its commercial service in suburban Phoenix.
Ride-hailing firm Uber idled its self-driving-car program for several months after one if its test cars struck and killed a pedestrian in March 2018.
Ford Motor Co. also has set targets for rolling out an autonomous-vehicle business, with a plan to deploy driverless cars in limited geographic areas by 2021.
The Dearborn, Mich., car company took a stake in Pittsburghbased robotics startup Argo AI three years ago to help build its expertise in this technology. This month, Volkswagen AG agreed to invest $2.6 billion in Argo as part of a broader alliance between the two auto makers to develop new vehicle technologies.
“Ford has been different than others and very balanced about the arrival timing, in terms of development of the technology,” Jim Farley, Ford’s president of new businesses, technology and strategy, said in an interview this month.
Meanwhile, driverless-car developers are awaiting clearer direction from the Trump administration and Congress on how the technology will be regulated.
The industry unsuccessfully lobbied for a Senate bill last year that would have created nationwide standards for self-driving cars. The National Highway Traffic Safety Administration has yet to rule on a request from GM to allow public use of vehicles without manual controls like a steering wheel and brake pedals, a closely watched proposal among driverless-car developers.
Beyond improving the technology, Cruise is working on other pieces of its business plan necessary before robot cars hit the roads, Mr. Ammann said. For example, the company plans to build a facility in San Francisco that can recharge a future fleet of driverless electric vehicles, which would operate within pre-defined boundaries.
GM also wants to work more closely with San Francisco officials ahead of any public launch to ensure smoother operation of the service and win residents’ trust, Mr. Ammann said.
“The historical Silicon Valley approach ‘move fast and break things’ just doesn’t work for the kind of technology we’re working on here,” Mr. Amman said.