Toyota Invests $1 Billion in Artificial Intelligence in U.S.
Credit: Yuya Shino/Reuters.
PALO ALTO, Calif. — Silicon Valley is diving into artificial intelligence technology, with start-ups sprouting up and Google and Facebook pouring vast sums into projects that would teach machines how to learn and make decisions. Now Toyota wants a piece of the action.
Toyota, the Japanese auto giant, on Friday announced a five-year, $1 billion research and development effort headquartered here. As planned, the compound would be one of the largest research laboratories in Silicon Valley.
Conceived as a research facility bridging basic science and commercial engineering, it will be organized as a new company to be named Toyota Research Institute. Toyota will initially have a laboratory adjacent to Stanford University and another near M.I.T. in Cambridge, Mass.
Toyota’s investment invites comparisons to earlier research initiatives, such as the Palo Alto Research Center, or PARC, created by Xerox in 1970 to help the company compete with IBM. Xerox was never able to find a strategy to make it a significant player in computing, but the technologies invented at PARC during the next decade were used by Apple and Microsoft to completely remake the computer industry.
The new effort by Toyota is also the latest indication of a changing of the guard in Silicon Valley’s basic technology research. Last year, for example, Microsoft closed a satellite laboratory of its Microsoft Research division in Silicon Valley and laid off about 75 researchers.
Corporate research done by Internet companies like Facebook and Google has generally focused on things that can be turned into a product or service, breaking with the traditions of industrial laboratories run by AT&T and IBM, which focused on basic science.
International corporations like General Electric; Baidu, the Chinese search engine; Samsung, the South Korean conglomerate; and all the major automakers have been establishing research outposts in or near the region to take advantage of its engineering talent.
Artificial intelligence technologies were disappointing for decades, but they have finally begun paying off, leading to systems such as Siri, the personal assistant from Apple, and rapid improvements in self-driving vehicle technology.
And in recent years, there has been a rush to recruit talented researchers in so-called machine learning, many of them produced by Stanford and the nearby University of California, Berkeley. Toyota plans to hire 200 scientists for its artificial intelligence research center.
“The density of people doing this kind of work in Silicon Valley is higher than any other place in the world,” said Gill Pratt, a roboticist and former official at the Defense Advanced Projects Research Agency, or Darpa, who will lead the new company.
The new center will initially focus on artificial intelligence and robotics technologies and will explore how humans move both outdoors and indoors, including technologies intended to help the elderly.
When the center begins operating in January, it will prioritize technologies that make driving safer for humans rather than completely replacing them. That approach is in stark contrast with existing research efforts being pursued by Google and Uber to create self-driving cars.
“We want to create cars that are both safer and incredibly fun to drive,” Dr. Pratt said. Rather than completely removing driving from the equation, he described a collection of sensors and software that will serve as a “guardian angel,” protecting human drivers.
In September, when Dr. Pratt joined Toyota, the company announced an initial artificial intelligence research effort committing $50 million in funding to the computer science departments of both Stanford and M.I.T. He said the initiative was intended to turn one of the world’s most successful carmakers into one of the world’s top software developers.
A similar challenge is now facing many of world’s largest noncomputing technology companies. In September, Jeffrey R. Immelt, G.E.’s chief executive, predicted that the company would be “a top 10 software company” by 2020.
In addition to the software engineers in each of its businesses that make jet engines, locomotives, power turbines, medical imaging equipment and other devices, the company now has more than 1,200 engineers at a specialized software center in San Ramon, Calif., just across San Francisco Bay from Silicon Valley.
“There is a software layer over everything now,” said John Zysman, a U.C. Berkeley political scientist and the co-director of the Berkeley Roundtable on the International Economy. And that is a powerful magnet that continues to draw companies.
By shifting its focus to include mobility for a rapidly aging population, Toyota is also acknowledging that demographic changes may soon affect traditional automotive markets.
“Toyota has been a reasonable, conservative car company, so it is intriguing that they are making this move,” said Jameson M. Wetmore, an associate professor at the School for the Future of Innovation in Society at Arizona State University. “Kids are getting their licenses later and the car companies are becoming concerned they don’t have the place in society they once had.”
In addition to focusing on navigation technologies, the new research corporation will also apply artificial intelligence technologies to Toyota’s factory automation systems, Dr. Pratt said.
The company describes its manufacturing system as the Toyota Production Systems.
“As extraordinary as the T.P.S. is, we believe it can be improved still further through the use of more data and more A.I.,” he said. “There may also be advances in robot perception, planning, collaboration, and electromechanical design from T.R.I. that will translate into improvements in manufacturing robotics.”
Dr. Pratt said he began his career as a research scientist at Bell Laboratories and was influenced by his time there.
Before coming to Darpa, where he was responsible for the agency’s most recent contest for building robots capable of working in hazardous environments, he was known for his earlier research at M.I.T., where he pioneered so-called collaborative technologies making it safer for humans to work near robots.