Chinese Internet search company Baidu is easing into the passing lane for driverless cars. The Beijing-based firm plans to mass-produce its own autonomous “Apollo” car by 2021 with help from such heavyweights as Daimler, Ford and nearly a dozen Chinese carmakers. A prototype is due next year and Baidu is testing driverless vehicles in California.
Baidu, which is at the fore of developing artificial intelligence (AI) technology in China, has also made its self-driving-car software available for any company to download. The idea is to promote the use of Baidu’s platform (which already works “out of the box” on some vehicles) by allowing free access and data collection. More than 50 partners, including Microsoft, have signed on, allowing Baidu to close the gap with Waymo, the Google spin-off that’s considered the current leader in autonomous-vehicle technology.
Some say this positions Baidu to become a dominant force in the driverless space. However, as with all disruptive technologies, this is easier said than done – even for an innovative and well-invested player like Baidu. Despite a strong push, analysts say, Baidu needs to prove that it can build a global brand for driverless cars, and this entails more than just developing the software or creating a prototype vehicle with outside help.
“It’s the whole car that matters, and Baidu will have to commercialize it and build a global brand,” said David Garrity, a technology and investment analyst who heads New York-based GVA Research. Whatever vehicle Baidu and its partners develop, according to Garrity, needs be an integrated product that appeals to consumers from a performance, comfort and design point of view.
Baidu’s challenge, Garrity says, is the same one that faced Japanese and, later, Korean automakers such as Toyota and Hyundai in the 1980s and 1990s: They must build a high-end global brand that resonates with consumers outside home markets.
But unlike Toyota, Baidu is an Internet company with no in-house expertise in automobiles. Even with a hand from big carmakers like Daimler and other allies, some say it’s unclear if Baidu can successfully fuse its driverless software with a vehicle built by other companies.
Apple and Google, on the US side, have shelved plans to develop their own self-driving cars – underscoring the challenges involved. Baidu also faces regulatory headaches in China, where driverless cars are still banned – though there’s no reason to believe, at this point, that the government will block the technology.
Whatever the challenges, Baidu has no choice but to move ahead. Its entry into autonomous tech and AI generally is aimed at offsetting a slowing of its core online-ad business in China, where the government is placing new curbs on Internet advertising.
Norman Anderson, a Washington, DC-based infrastructure strategist, says any company developing a driverless car, especially those in the US, faces a clash in corporate cultures. Marrying Detroit with Silicon Valley, he says, isn’t a great idea because Old Economy auto executives may not work well with New Economy geeks. He says the former tend to be narrowly focused problem solvers, while the latter are visionaries eyeing a much larger picture.
“Why mate Detroit with Silicon Valley? Why go back to the past?” asked Anderson, the president and chief executive of CG/LA Infrastructure, a consulting firm that advises global infrastructure projects.
Project Apollo, Baidu’s partnership program to create a driverless car, also seems to be having a few skids. It originally hoped to begin testing self-driving cars on urban roads by 2018. This was to be followed by testing on highways and open city roads by 2020.
But Baidu has now shifted its timeline to “fully autonomous driving on urban roads and highways by the end of 2020”, a noticeable postponement, according to a company statement. Its ability to fulfill this deadline will also be an acid test of its driverless-car ambitions.
Baidu, moreover, exited a much-vaunted partnership with BMW in late 2016 that was supposed to test driverless vehicles in Chinese cities. The two companies reportedly disagreed on research priorities and project scheduling.
Cost, performance, design and other details on Baidu’s Apollo car are also scant, raising questions about its international appeal.
Baidu didn’t respond to questions from Asia Times on the challenges it faces in driverless cars.
Still, some major Chinese players have proved that they can manage and promote international brands. Geely Automotive on the carmaker side is successfully managing its Volvo and Lotus brands, while Lenovo has excelled in personal computers after acquiring IBM’s PC business.
With the right help, Baidu could come up with a driverless car that is a hit in international markets. Baidu’s ace in the hole is its ability to gather the vast amounts of data necessary in creating the “deep learning” programs that enable autonomous driving software.
Driverless cars may be leapfrogged
Technology, however, is in constant flux. Anderson notes that the current approach to autonomous vehicles involves adding software and sensors on a 120-year-old invention – the automobile.
Converging technologies make it likely that electric vehicles will soon merge with autonomous tech to create a revolutionary new product. Any gasoline-powered driverless cars developed by Baidu could be leapfrogged.
Designers are also mulling driverless vehicles that look more like offices or living rooms on wheels than cars. Flying-drone technology raises other possibilities.
Companies like Baidu will have a tough time keeping up. “The winners will be those who redefine mobility,” Anderson said. “A lot more is going to happen. Software allows for an infinite number of things.”
Winn Schwartau, a noted US cybersecurity expert, sees a bigger problem in perfecting self-driving cars with autonomous decision-making capability.
“How do you build ‘moral authority’ into algorithms?” Schwartau said, alluding to the so-called runaway “trolley problem” in which a human operator in a hypothetical accident must decide whether to kill one person to save five.
“I predict that driverless cars are going to be thrown off the streets,” Schwartau said.