Although the industry has made great strides over the last five years, we are a long way from the finish line of fully automated cars. When you look at what is currently being tested and developed in the field of autonomous vehicles, you will find that these systems can only handle certain speed ranges, certain weather conditions, certain street complexity, or certain traffic. Most of what has been collectively accomplished has been relatively easy because most driving is easy. Where we need autonomy to help us, is when the driving is difficult. And it’s this hard part that TRI intends to address. That means understanding difficult driving scenarios, and building AI systems that learn from, and evolve to predict, such scenarios. 

TRI will also apply AI technology to the challenge of home robotics. Here, fueled by our aging society and the remarkable progress in electronics and computer science, we see a need for machines to assist in mobility beyond the realm of what is currently possible. Home robots may become even more personally prized in our future than cars have been in our past. It is entirely possible that robots will become for today’s Toyota what the car industry was when Toyota made looms.

Finally, Toyota's quest for a new world of mobility also depends on novel materials. Robots in the home will need materials that are stronger, lighter, and less expensive than the machines on the factory floor. Electric vehicles need a new type of battery that stores energy as efficiently as gasoline, and is equally simple and fast to recharge as a gas tank. And fuel cell technology currently requires methods to store and process Hydrogen at high pressure, posing significant engineering challenges. TRI wants to apply the promise of machine learning and AI to discover, even design, new materials that will vault past these limitations.

Society tolerates a lot of human error. But we expect machines to be much better. We expect them to be bullet-proof… ever-ready… and nearly perfect. However, achieving this level of quality is difficult given the issues that exist in the current state-of-the-art in AI software. Presently, many of the most advanced AI systems use machine learning techniques where large datasets are used to train the software how to respond. A challenge with this approach is that it is hard to know what the system will do when faced with a novel input. 

It is this quest for safety, utility and reliability that drives TRI to understand the hard parts of driving, to understand how humans can benefit from robot assistants, and to understand how to apply AI to designing and creating new materials.