Getty Images
Industryweek 11617 Toyota Logo

Exclusive: What’s Ahead for the Toyota Research Institute

Aug. 3, 2016
The leader of Toyota’s efforts in robotics applications and cloud-based research talks about how the $1 billion effort is progressing.

James Kuffner, the chief technology officer at the Toyota Research Institute, was part of the original team that built Google’s Self-Driving Car. He led Google’s robotics team until last January, when Toyota hired him away to focus on cloud computing at their new artificial intelligence subsidiary.

TRI now has three facilities, one in Ann Arbor partnering with University of Michigan, one in Cambridge, Mass partnering with MIT, and one in Silicon Valley partnering with Stanford. TRI focuses on vehicle safety and autonomy features and also mobility beyond cars—looking at robotics applications, especially for an aging population in Japan, and using the modern cloud computing infrastructure to accelerate scientific discovery.

In an exclusive interview, Kuffner talked to IndustryWeek about what’s ahead for TRI.

What are your research priorities?

Safety and accessibility. If you think about all of the advancements that the industry has been investing in over the last several decades—seatbelts, airbags, forward collision warning systems, emergency assisted braking. All of those things are helping reduce the number of fatalities. Nevertheless, we still have tens of thousands of people who die in the U.S. alone due to automobile accidents. Sometimes it’s human inattention or human factors.

So we are investing not only in the technology to make the vehicles safer and mitigate injury and death, but also make it better and easier for a human to drive and not be distracted. How can we build a complete system that takes into consideration human machine interfaces in a new way as this transportation evolves?

We think that all of the tech companies, Google, Facebook, most of them have open sourced their advanced machine learning and artificial intelligence software. And there are other open source alternatives like Caffe and Torch.

So now it isn’t who has the best algorithms. The algorithms are there. It’s who can have quality data. The most data. And for me, one of the reasons to join Toyota in their effort, and why I think they have a chance at having a leadership position is they have lots of good data. And its value needs to be unlocked. And so I’m really excited about the prospect of harnessing that value in the data, sharing it and being able to train and build these systems that will be much safer and improve overall the capabilities of the system.

Is that data coming from the connected vehicles?

There’s connected cars. There’s telematics data that we can get from basically all the testing of the vehicles. Now of course, it opens all of these really interesting issues about privacy, about security, about integrity and accuracy. We want to get in front of all these issues and start the conversation early. About how can we bring this technology faster to market to save lives earlier, and how can we do so in a responsible way that respects privacy and respects security and not have security as an afterthought. I’m proud that Toyota is really trying to take a proactive approach in thinking about it.

I just hired Jim Adler, who’s our head of data. He used to work in other companies, at Intelius, on privacy and security. And he’s trying to help us establish TRI as a thought leader in promoting "what does it mean when you have connected cars? What does it mean when you have data that you can share but it’s anonymized so that the entire society will benefit from it but without compromising the privacy of users and customers?" So really trying to think out all those safety issues in advance is important to us.

How big is your team over there?

We’ve just started. It’s been six months. But we have over 100 people now across our sites.

What’s the culture at TRI like?

We have a mix of artificial intelligence experts, we have academics, we have researchers, we have experienced technology veterans from Silicon Valley. We have people that we’re working with who’ve been longtime career Toyota engineers. They’ve actually transferred from Japan and are sitting with us side by side, helping us make sure that whatever technology we develop has the best chance of getting into Toyota’s products as soon as possible.

Sponsored Recommendations

Voice your opinion!

To join the conversation, and become an exclusive member of IndustryWeek, create an account today!