Interview with Manuela Veloso
If the first word you come up with when you think, 'Manuela Veloso', is RoboCup,
then you're missing a whole lot! Come find out her reasons for coming
to Carnegie Mellon, her passions, and her favorite place in the world—all
in this exclusive mini-interview. Plus, learn more about the history of
So RoboCup started in about 1996, and it was a result of some people in Japan, Yuraki Kitano and myself, and some others getting interested in the problem of multi-robot systems—systems that involve more than one robot accomplishing tasks. I had been doing research in planning—in very classical AI [Artificial Intelligence] planning algorithms. But then I became interested in planning and execution and I had been working with some of my students, Garren Haden, my first student, on planning and execution in a real robot. Then in 1994, my student Peter Stone saw a little demo of a one-on-one playing soccer at a major AI conference, and this was a little demo set up by Michael Sohota who was student of British Columbia University in Canada, and he’s advisor was Allen Macworth. Peter was a big soccer fan and he was all interested in this little demo—so he asked me if he could do his thesis on robot soccer. So the combination of my interest on planning and execution in real robots, my own growing interest in multi-robot problem, and my student Peter Stone’s having seen the demo and his love for soccer all kind of made it happen. Peter then pursued his thesis from 1996-1999 on simulated robot soccer, which was what mainly available then—at the same time we were developing little prototypical small robots playing. But then things grew much more here at Carnegie Mellon, and now there are all these robots we have here today!
From my point of view, I have to say that my aspirations are mostly at the technical level. I don’t aim at always winning—we are trying to understand very deep technical problems: increasing the robustness of single robots, removing their brittleness in terms of sensors and behaviors, and also since this is the soccer problem, we are very curious about the opponent-modeling and how the robots respond to opponents, how to create team-work, and how to make team strategies—all in real time constraints.
So even if people think that I really want to win the competition—which in some sense is what we want—it’s not what drives our research. Our research is really at the level of trying to understand how these multi-robot systems work. So what I want for Carnegie Mellon is to have the greatest impact in bringing to the world, from the scientific point of view, these multi-robot teams.
Where are you originally from?
I am from Lisbon in Portugal. I was born in Portugal and I have been in the States since 1984.
How did you came from Lisbon to becoming involved in robotics at Carnegie Mellon? Why robotics? Why Carnegie Mellon?
I’m an Electrical Engineer. I did my Electrical Engineering degree in Portugal, and then we came to the States for personal reasons. I did a Master’s in Computer Science at Boston University, and then I applied to a PhD program. In those days, the real thing that attracted me to Carnegie Mellon was Jaime Carbonell’s work on analogy. When I came to CMU, I was completely overwhelmed and completely amazed at Allen Newell and Herb Simon’s contributions to AI, and I developed a passion for this concept of learning, planning, and having the full cycle of perception, cognition and action. So I went all the way from Electrical Engineering to soccer on multi-robot systems through a path that was a slow transition in interests. The AI is still my passion now.
What do you like best about your job?
Meeting with my students. Definitely. So the best part of it is when my students and I engage in really deep technical discussions. That’s when I understand really how we can advance the scientific state of the art of where we are, what are they really struggling with—that’s where we hypothesize what can be done, we generate experiments—so that’s where I see the scientific method being carried ahead. I bring back what I thought, they bring back where they are, and we try to push our research forward. That’s by far the best thing. I also like teaching. But teaching is something more of like you teach, you try to explain to the students what’s going on, and you come out of the class and its over. You can teach again, but its not continuous. The research with my Grad students is continuous—in the sense that I keep thinking—what are the results that Pat is getting, what are the results that Scott is getting; did Scott already figure this out—so I’m constantly thinking about the problems that I have with each of the students. All of these things go on in my mind all the time.
What’s your favorite thing about CMU?
I really think that it is these amazing students that you interact with. The environment is very rich from an intellectual point of view, and very friendly also in the sense that people are willing to talk to each other along very different dimensions. I have experienced a lot of having people to talk with that make the effort to explain what they are doing in a completely different area. When my colleagues and I talk, it requires an effort to explain what you are doing to someone who is not in your area. And Carnegie Mellon people are willing to make that effort. That’s something unique. I don’t know—people by osmosis, or the air—enter into this willingness to share with each other what their passions are at the research level, and what their technical difficulties and advances are. That makes conversations and interactions quite interesting.
Are you married and do you have children?
Yes, I’m married and I have two sons. 16-years and 21-years old. My 21-year old son just finished his undergraduate degree in Computer Science at Harvard University. And my younger son has done robotics camps in the summer. Both of them have distant interest in robotics, but they understand what’s going on. It’s a part of their lives.
What's your favorite vacation spot?
Oh Portugal! There’s a place in Portugal called Praia da Luz—that’s in the Southern part of Portugal and it’s the most beautiful spot in the world. I’ve been there since I was young and I always go back there.
If you didn’t go into robotics or academics in general, what would you have done in another life?
I would have liked to be a doctor. I really would have liked to go into the problem of studying diseases of human beings and trying to cure them. I don’t know why I never did it—it’s not like it didn’t exist before—but I guess that since my father is an engineer and the whole family is math-based, there was not as much motivation to do medicine. Plus I didn’t like blood and all of that…but now, looking back, I think I would have really liked being a doctor. Definitely.
What hobbies or passions do you have outside of your work?
That’s a good question. With a life with children, sometimes your hobbies and passions are really driving your kids around and taking them to school, etc.! But mainly I like swimming, reading, playing cards, playing squash—all these things are done with very little depth though. And I like to travel. I enjoy very much traveling with my family—we go around to Brazil, Europe, Portugal—we travel a lot, and so I enjoy that.
What motto/philosophy do you live life by?
I try to live in the present moment. The past is gone; the future is not here yet. So I try to take the best advantage that I can from the present moment. I believe that when you pass away what remains is the personal relationships you create in personal moments. In some sense that’s what really matters—all the big things will be forgotten. What you remember is the little personal relationships. I tend to care a lot, in terms of my life, to try and build these personal relationships with my students, whom I interact with so much. My Grad students have really almost become like family. Those kinds of things make you more personal rather than just technical—I’m very technical, yes, but I really try to live the present moment and try to see the people that are around.