TOKYO — Six months into his new job, Toyota Motor Corp.’s top visionary for new mobility is spelling out priorities and ambitions, promising new safety technologies for cars within five years and envisioning a time when children are chauffeured to soccer practice by autonomous vehicles.
Gill Pratt, CEO of the newly created Toyota Research Institute Inc., also said he expects to be talking with many other companies, from technology giants to startups, about possible partnerships for self-driving cars and robots.
Pratt, a robotics expert who formerly worked with the Pentagon, predicted a takeoff point at which the performance of autonomous cars expands exponentially as vehicles fully harness the potential of wireless networking.
“It’s the wireless Internet that’s really going to make this thing explode,” Pratt said of the potential for self-driving cars and advanced robotics.
As head of Toyota’s $1 billion incubator for next-generation technologies, Pratt will have a big hand in guiding the world’s biggest automaker into a new era where the software counts as much as hardware and technology companies such as Google and Apple are putting down stakes.
In the six months since TRI’s startup, Pratt has been busy hiring people. The team now numbers about 100, including 30 from Toyota Motor in Japan.
TRI also opened a third r&d hub in Ann Arbor, Mich., following centers with the Massachusetts Institute of Technology in Cambridge, Mass., and with Stanford University in Silicon Valley.
But its efforts also are supported by separate r&d operations at Toyota Motor in Japan, including another division here dedicated to robotics.
Speaking to reporters June 17 during one of his monthly visits with r&d colleagues in Japan, Pratt said TRI is focusing on three main tasks: improving automotive safety, improving vehicle accessibility for people unable to drive and developing robot technology for use in everyday life.
As Pratt sees it, the safety effort will provide the quickest payoff. The breakthroughs, he said, likely will focus on evasive technologies that sense when a collision will happen and steer the car to safety. While such technologies are already on the market in rudimentary forms, Pratt suggested that next-generation versions may steer a car automatically around obstacles such as pedestrians darting into the street, even if it means steering out of the lane.
Pratt cautioned that the goal is not necessarily to have accident-free cars but to develop cars that aren’t responsible for causing an accident.
“If you are sitting in a car, and you are in between two other cars, and you can’t move, and another approaches you from 90 degrees and smashes into you, there is nothing you can do,” he said.
Toyota hopes to parlay TRI’s r&d into new technologies for autonomous vehicles and other products, such as household robots. More broadly, Toyota wants to be a leader in software, just as it is in automotive hardware.
Indeed, next-gen cars will overlap greatly with robots, Pratt said. “The car of the future is as much about software as it is about hardware,” he said. “When you put software and hardware together, what you get is a robot.”
The mandate there is more than just avoiding accidents. Pratt envisions creating cars that can drive people who are disabled, elderly or tired — or even drive children.
“Someday — I don’t know when it will occur — someday, these cars will be safe enough that children could go from one place to another without having to have their soccer mom, or their soccer dad, drive them,” he said.
Pratt said he is also on the lookout for possible partners and expects to meet soon with the autonomous-driving team from Uber. Last month, Toyota said it would invest in Uber to provide leasing options to the company’s drivers and explore new mobility businesses.
The quest to make computer-driven cars as safe as human-driven ones is just beginning, Pratt conceded. Humans have an impressive safety record behind the wheel. In both the U.S. and Japan, he said, road fatality rates are approximately one death for every 100 million miles driven.
“To my knowledge, there is no autonomous car system that comes even close to that,” Pratt said. “We still have a long way to go.”