After two years serving in an interim capacity, Professor Srinivasa Narasimhan will be stepping aside as Carnegie Mellon welcomes the sixth director of its Robotics Institute. A graduate of CMU’s School of Computer Science in 2005, Matthew Johnson-Roberson arrives at the school after stints as an associate professor of engineering at University of Michigan’s Department of Naval Architecture and Marine Engineering and the Department of Electrical Engineering and Computer Science.
Johnson-Roberson also served as a co-director of the UM Ford Center for Autonomous Vehicles, from whose office he fielded our call to discuss his plans for his new position, as well as what he believes the future will look like for robotic research.
TC: You’re currently in Michigan, working out of that new Ford wing?
MJR: Yeah, I’m here and a few of the students are here. Just doing some robot stuff. Having a good time.
What has your focus been over there?
A number of different things, but it’s long-term, blue sky research for Ford. They’re doing lots of work, as is Argo, on things that will come out on the roads, hopefully in the next six months to three years. We’ve been looking much more at things that are five-10 years off of getting to the road. That’s one of the nice things about universities is we can take this blue sky approach to things, everything from weird new sensors to thinking a lot about human prediction and safety guarantees.
I see the Ford/U of M setup being a model universities increasingly look to. Universities — especially those like CMU — have a long history of wealthy benefactors. Do you see this kind of partnership being the model for university research going forward?
It reflects the sort of transformation that robotics has undergone in the past 20 years. So many of the technologies that were developed in the 90s and 00s are now reaching a level of maturity where they are getting rolled out in commercial products and making a big difference to the future of a lot of industries. I think it’s a natural extension of that. You begin to see relationships between universities and companies. Even if you look at Pittsburgh as a city, the transformation that it’s undergone from heavy industry focused around natural resources and steel, that transformation is only going to accelerate.
Part of my goal is to continue relations and build new ones. Beyond just industries, thinking about government and policy and all the other things that are now going to become more relevant for robotics — making sure we make those relationships and build on the strengths of the technical work that already goes on at the institute. That is something I’m particularly excited about.
In Pittsburgh, you have a number of these homegrown startups, but also big companies like Google coming in to be closer to the research and court graduates. What can CMU do to further foster a relationship like that?
They’re opening a Waymo office to work with one of the professors at CMU, so you can see see that these relationships — not just the faculty but the students. The lifeblood of all of these companies is new, highly trained employees, and anything you can do to get a leg up on recruiting and helping to build a culture that people want to come to is a huge advantage for these companies. You see them co-locating, sponsoring research, doing all of these things that are helping them to both develop new projects but also form new relationships with incoming students. The richest part of the university is that every year you get a new crop of some of the smartest people in the world.
Is part of your job going to be helping grow some of these startups natively within sort of the broader context of the university?
Yeah. Having had the opportunity to do a startup myself, I think knowing that there’s this huge knowledge gap, that there are so many students that are so smart and have such big ambitions for the world, I think figuring out ways of helping them fulfill that is my role. One way that you just highlighted is startups. People talk about ecosystem a lot. Part of that is that there are other startups in the area, but secondly, you have a community where you can find likeminded people to do stuff with and build things.
You’re at the U of M, so you’ve seen some of the transformation that’s happening in Detroit. Detroit may not be as far along as Pittsburgh, when it comes to fostering the startup community, but there’s a lot of opportunity there. What role can the school play in keeping the talent that CMU attracts to the city?
There are a couple of things. One thing that I see as increasingly important is making sure that you acknowledge that the opportunities are out there. The speed and size of the robotics industry is accelerating at a pace I don’t think any of us could have anticipated. An important part of this is acknowledging that and not trying to stay stationary. The industry is changing, the ecosystem around robotics is changing and the size and scale around these companies is changing. Let’s figure out how to make this happen.
Robotics hasn’t historically been the most inclusive field. What role can CMU play in that? I assume by the time most people enroll at a place like CMU, robotics has already been their focus for some time?
I want to leave as my mark on CMU in this time two things. One is to increase opportunities, make sure we are broadening participation and we are seeing representation in the field. Secondly, I think this is perhaps more important, is that universities are good at shaping the minds of young people. I couldn’t imagine a better position to try and affect change on bringing more diversity and inclusion into robotics than being at the premier research institution for robotic research. That means you’re at the genesis point for the next generation of roboticists.
You’re a good example of that — you didn’t start at CMU with robotics as a focus.
Exactly right. And I’ll go a step further. When I got to CMU, I struggled. It was the first place I’d ever been where everyone was smarter than me. That is what I think is really special about that place. Whatever happened, it didn’t result in me quitting robotics, leaving and never coming back. And I think that’s a testament to the people that were there then and are there now.
What excites you most in robotics, these days?
We’re really at this inflection point when it comes to big, deployed robotic field systems out there in the world. One day, I want it to be that you look out your window, wherever you are in the U.S. and the world, and you see a robot doing something useful. That’s not the world we live in right now. If you go to a factory floor or a few other places, you can see a robot. Maybe you have a robot vacuum, but I want it to be at a point where you look out your window and see a robot.