Toyota Research Institute
The best teams of AI researchers and data scientists who are passionate about cars think of future mobility.
The Toyota Research Institute (TRI), Toyota’s key artificial intelligence R&D center, is actively working in collaboration with Silicon Valley innovators in implementing open innovation processes to expand the possibilities of artificial intelligence technology. One of the initiatives recently undertaken by the institute in line with this mission was the Prius Challenge held in March 2017.
A mobility festival bringing together innovators from Silicon Valley.
The main event is a unique car race put on by Toyota in which participants compete to achieve the best fuel economy possible employing strategies they have developed using machine learning and sophisticated data analysis tools.
Big data comprising Prius driving and fuel consumption records was shared with teams in advance. Teams employed machine learning and other processes to analyze the data in preparation for the event.
Based on data analysis, teams ran simulations to try and determine factors related to achieving peak fuel efficiency, such as the best way of operating the accelerator, how to acquire optimal course positioning, and the best timing for applying regenerative braking.
Ten teams simultaneously participated in the race based on strategies they had worked out in advance. Monitoring detailed driving data in real time, they constantly tweaked their strategies mid-race in an effort to achieve the absolute best fuel efficiency possible and win the challenge.
Known as a center of technological innovation, Silicon Valley is home to many of the world’s leading IT and venture technology companies. It is also a home of TRI, and the location of the recently held Prius Challenge. Participants in the challenge were AI researchers and data scientists working to develop artificial intelligence technology. Teams developed strategies prior to the race, however, during the race itself they had to deal with the movement of competitors’ cars, something that is difficult to predict and also a major challenge facing researchers seeking to realize autonomous driving. While being forced to alter strategies due to competing vehicles moving differently than predicted, participating teams worked cohesively in solving problems and realized high levels of fuel efficiency that were more than 1.5 higher than the figures provided in car marketing catalogs.
TRI's pursuit of AI technology is making life better for society.
TRI serves as a base for Toyota’s AI technology research and development efforts. With an ultimate goal of realizing safe autonomous driving to greatly enhance vehicle safety and increase mobility access, the institute is also engaged in research and development aimed at alleviating traffic congestion and improving transportation systems and fuel economy. Efforts go beyond the traditional mobility framework and include initiatives such as the development of robots to support people’s daily lives, and research and development in materials science that utilizes knowledge gained through artificial intelligence and machine learning. All of these efforts are conducted with an aim of seeking to enrich society and better the quality of human life. While maintaining ongoing collaborations with a wide range of talented people and companies, going forward, TRI will continue to exceed expectations and take on challenges aimed at generating mobility breakthroughs for the future.