Toyota is making some serious moves in the race towards autonomous vehicles. It recently announced that it will be spending US $1 billion on an artificial intelligence and robotics technology research facility geared towards making discoveries that could be used for better self-driving automobiles. Toyota plans to spend an additional US $50 million into research centers to be co-established with Stanford and MIT.
The Toyota Research Institute will technically be a separate company headed by Toyota’s executive technical advisor Bill Pratt. Before coming to Toyota, Pratt acted as the research manager for the Defense Advanced Research Project Agency and led the DARPA Robotics Challenge. He will act as head of the Toyota research initiative at Stanford and MIT.
Pratt explained his plan at a press conference in Japan. The institute hopes to make developments that improve automobile safety, that make driving in personal cars accessible to everyone (regardless of age, disability status, etc.), and more specifically that make it easier for senior citizens to live on their own.
“TRI will focus first on collaborative autonomy and artificial intelligence- that way people and machines can work together, particularly in the area of mobility,” Pratt explained to the conference. “The purpose of TRI is to bridge the gap between fundamental research and product development of life-saving and life-improving technologies.”
Toyota aims to create a vehicle that can steer, change lanes, and merge into traffic on its own by 2020. It would be released initially in Japan.
Google announced earlier this year that it has begun testing its autonomous vehicles on public roads. So far all accidents involving Google cars have been the fault of other drivers.
Tesla released its autopilot software a few weeks back, but the software is meant to assist in parking, changing lanes, and reacting to accidents. It is not supposed to be fully autonomous and Tesla has recommended that drivers keep their hand on the wheel and remain alert while autopilot is activated.
These warnings didn’t stop one trio from using the feature to drive from Redondo Beach, CA to Manhattan in just under 58 hours. The group claimed to have generally driven at speeds around and over 90 mph and confirmed that they operated autopilot for 96% of the time.
Toyota’s research goals are not limited to the automotive world.
“While the initial focus of the investment may be on how to apply AI and robotics to making safer cars, Toyota’s larger aim is to expand the use of these technologies beyond the automobile,” explained Jeremy Carlson, senior analyst at IHS Automotive. “The goal is not only to make society safer, but to improve the living conditions of the larger society- particularly the elderly.”
Stanford and MIT will be there to help; Stanford plans to focus on sensing and perception, learning in all conditions (specifically deep learning and machine learning), human-centric interaction, and decision making and reasoning. MIT will specialize in creating tools to collect and analyze navigation data to learn to drive, precision systems for navigation, systems to handle different driving situations, predictive models to anticipate car and driver behavior, and a more intelligent human-machine interface.