Back in September 2014, Seigo Kuzumaki, who was then Toyota’s deputy chief safety technology officer, said Toyota didn’t plan to develop a fully autonomous vehicle because “Toyota’s main objective is safety.” Ken Kobushi, head of Toyota’s intelligent vehicle division, said infrastructure improvements would have to be made to ensure the safety of self-driving vehicles.
But now, Akio Toyoda, Toyota’s CEO, will spend $1 billion to create a new Toyota Research Institute with a focus on robots and self-driving cars, George Anders wrote for MIT Technology Review.
Robotics Researcher to Lead Effort
Toyota named Gill Pratt to head the institute, which will have offices in Michigan, Silicon Valley, and Cambridge, Massachusetts. The noted robotics researcher is authorized to hire hundreds of scientists and engineers.
Pratt is charged with developing semi-autonomous systems using artificial intelligence to boost driving safety. Driver-assistance systems such as lane-change warning and collision avoidance are steps on the way towards fully autonomous vehicles, according to Business Insider.
Toyota’s goal is to have a car that can drive itself on highways by 2020, which will coincide with the Japan Olympics in Toyota’s home country.
Artificial Intelligence and Self-Driving Cars
Artificial intelligence (AI) would help cars react better to possible road dangers than technology now in use, such as image sensors. Pratt explained the benefits of AI:
The intelligence of the car would figure out a plan for evasive action […] Essentially [it would] be like a guardian angel, pushing on the accelerators, pushing on the steering wheel, pushing on the brake in parallel with you.
Pratt gave as an example an AI-equipped Toyota knowing to move itself into another lane to avoid a car accident if another vehicle enters its lane.
Pundits say Toyota’s decision to invest in research on robotics and autonomous vehicles is an indication that the carmaker does not want to be left behind when so many other companies, such as Google and Tesla, have been working on self-driving vehicles for a longer time.
Toyota has teamed up with Stanford, the University of Michigan, and MIT to look to the future, even if it might be a decade or more before self-driving vehicles are on the market. At Stanford, Fei-Fei Li, a computer science professor leading her department’s partnership with Toyota, told Technology Review that her team’s work might be relevant as soon as 2018, or as late as 2028, or any time in between.
In one Stanford project, John Duchi, an assistant professor of statistics and electrical engineering, is working on approaches to defensive driving. For example, he is looking to build software making it possible for an autonomous vehicle to make good decisions if a bicyclist should suddenly appear.
And Li’s team is employing 3D vision and pattern recognition that would allow a vehicle to recognize children playing ball near a road, or pedestrians walking while looking at smartphones. Li said: “It’s a very difficult research problem” to make software and sensors equally discerning.
George Anders opines:
It’s possible that a generation from now everything from roadway design to driver certification will be radically reshaped by the ubiquity of semi- or fully autonomous vehicles, and carmakers without the requisite technology will be as imperiled as the sellers of silver-halide film in the age of the digital selfie.
Image by Tokumeigakarinoaoshima, via Wikimedia Commons.