wood05Interviewer: Ryan Huang ’16, THURJ Staff


Robert J. Wood is an associate professor at the Harvard School of Engineering and Applied Sciences, a faculty member of the Wyss Institute for Biologically Inspired Engineering, and founder of the Harvard Microrobotics Lab. In recognition of his research, he received the 2010 Presidential Early Career Award for Scientists and Engineers from President Barack Obama and was named one of the top 35 innovators under the age of 35 by Technology Review in 2008. Other accolades include the 2007 DARPA Young Faculty Award, 2008 NSF Career Award, 2008 Office of Naval Research Young Investigator Award, and 2009 Air Force Young Investigator Award. I sat down with Professor Wood to discuss his research and thoughts on current technological trends.


Could you introduce us to the type of work you do?

We do research in robotics, mostly in small-scale systems (i.e. micro robotics) and also soft systems (i.e. soft robotics). A certain unifying theme and one of the main emphases of our work is materials, manufacturing, [and] design as enablers for unprecedented performance in robotics. For example, we have a project called Robobees, which is creating autonomous robot bees, and one of the [objectives] there is solving the challenges for fabrication at a millimeter scale using high performance materials. Another example is creating robotics [that] naturally interface with humans, so we have a number of projects in soft robotics which are addressing the materials and design side of things that you would wear, for example, to assist in some way for rehabilitation or augmentation purposes. We put a lot of effort into physical design and the instantiation of the robot so that we can make [robot] control easier, manufacturability easier, production cheaper, give [the robots] more robust capabilities, etc.

Tell us more about the Robobees project. What are potential applications?

Search and rescue, hazardous environment exploration, assisted agriculture, but really the applications are the fundamental science and engineering [problems] that we are using Robobees to motivate. For example, the things that get us excited [are] the solutions for how do I build it, how do I power it, what are the fluid mechanics that allow [movement], how do I think about controlling them, what kind of sensors am I going to have to develop, etc. So those questions—the questions of basic science and engineering framework—are the real motivation for this. That’s the real application. Things like assisted agriculture [and] artificial crop pollination, those are nice guiding motivations; those are nice things to say. But in reality we’re so far away from that, in terms of having robots that are ready to do any of these things, that it’s not very relevant to talk about some of these things just yet.

It’s about pushing technologies like small-scale power, multi-scale and multi-material manufacturing, lower power-high performance computation, developing coordination algorithms and programming methods, etc.

You described the coordination of Robobees. Does that mean there’s some sort of “hive mind”?

So there’s, I suppose you would call it, a hive mind. It’s like the following: You have lots of individuals and the individuals themselves are dumb, let’s say. They have minimal capabilities of what they can do and what they can understand about the environment. How do you take these simple individuals and do something useful and aggregate with many of them together? Researchers on our team led by Rhadika Nagpaul are thinking: how do I think about programming given limited resources and take a lot of bio-inspiration from things like termite colonies and bee colonies as well as [similar systems] where no one individual has the sort of grand mapping for everything. Rhadika’s work has shown with termite mounds these animals are extremely simple yet they perform these fantastic feats together, forming complicated termite mounds. In that sense, it’s sort of taking a very distributed approach to a larger goal, where the whole must be much, much greater than the sum of the parts.

Is there a central controller then?

The idea, at least with Robobees, is you have a hive, which is sort of where the all the coordination happens.  So, say send a certain number of bees in this direction, and then that’s the output. The input would be a bee just landed here, and it just saw the following things during its flight from this direction, very simple information coordinated somehow by the hive, and then it distributes individuals based from this information. So, yes, there is a centralized aspect, in the sense that there is a hive, which you’re assuming has infinite resources in terms of power and computation relative to the bees themselves.

One of your other projects was assisting humans with soft-bodied robots?

So we have interest in soft robots, and that’s for a number of things. Let’s take a sort of typical vision of what a robot is. You think about it in assembly line [environment]—something that’s massive, that’s bulky, that’s metallic and very precise, very fast, but also very dangerous. It can do high-precision repeated tasks and can do that very well, but you wouldn’t want a human down on the assembly line floor while this thing’s operating. So, you can image a couple of scenarios where you do want humans and robots interacting together—in a house, for example. The challenge is realizing something like that. One of [the possible solutions] could be better control algorithms or sensors. I sense when a human is near and I don’t swing my massive robot arm. We’re taking the approach of thinking about having them be more compatible in a material sense. So, if you have a “squishy” robot, something as soft as your skin, interacting with them wouldn’t be as nearly as big of an issue. That’s one of the sort of motivations. Another one is, if I got something that’s more rigid and has motors and gears in its arms and legs, you can imagine that it’s not going to be terribly robust [when dropped out] of a window right? So, if I want something which is going to be useful in real environments (rougher terrain or squeezing through rocks trying to find survivors in collapsed buildings) and if I had a robot that was really soft, that [robot] would be an effective solution to [real environments].

Also, there’s a class of [soft-bodied robot] projects that we are working on: wearable systems.  So, for things like rehabilitation, [we’re] making wearable devices that impede your motion no more than your clothes would, basically [making them] comfortable to wear. [It will sense] the motion of the patient in this case and sense when, for example, a patient will have a bad gait due to a neuromuscular disorder. [It could] have soft actuators built in to this artificial clothing that push the patient towards a more natural gait and start to re-train the neuromuscular system towards a more proper gait. That’s a big collaboration with the group in the Wyss Institute that we are looking into.

Another program that we’re involved with, that Conor Walsh is leading, is for augmentation or risk mitigation for users that might wear a suit. [It can] sense when you’re about to twist your ankle and freeze up [to prevent injury]. Soldiers [using this technology] could minimize fatigue when carrying heavy loads.

With the growing capabilities of technology, there’s more and more potential for abuse of those capabilities. As a roboticist, how do you feel about these developments?

It is an essential role of any scientist or engineer to gage the long-term impact of their work. This is because research most often out-paces law, so we have an ethical obligation to consider societal impacts in particular. As with most scientific research, we understand how our work could be applied to a spectrum of applications. We believe that our work could have many applications, with the benevolent far outweighing the negative.  However, it is the fundamental science and technologies that drive us far more than the uses of the robots themselves.  Fundamental breakthroughs in materials, manufacturing, biology, sensors, circuit design, controls, and programming are just a few of the topics that motivate our team.  Furthermore, one of our roles is to educate the public about our work—both our peers and the general public—through our scientific publications as well as our outreach activities.

What kind of future do you envision 30 years from now with robotics and technology? Is there any sort of breakthrough that you envision, or is it more of a steady progression?

I don’t know if this is going to be as discreet of a jump as it may sound. In some ways this is already happening in a slow, continuous fashion. The sort of old school science fiction notions of robots—robots that are cleaning your floors or cooking your dinner—that will likely start to happen, and in some ways it has already happened. You have…self- driving cars [as well]. I don’t know if this is going to be very accelerated or just a slow pace. Thirty years from now you’ll have a lot more automated, robotic tools to automate your existence in a good way. I don’t really think any of this is bad; I don’t think anyone’s going to complain about a robot that does your laundry for you; I don’t think anyone is going to miss doing laundry. I think those are the types of things [that we’ll see].


What kind of hobbies do you pursue in your free time?

I am into rock guitar and cooking, I generally gravitate towards relaxing activities or activities, so exercising, guitar, [and] cooking are all sort of either energetically or creatively relaxing outlets.

Do you have a favorite band?

Yes, Pink Floyd.

Do you have a favorite dish to cook?

I tend to grill maybe on average every other day every month of the year. So, grilled food, I guess. I’m trying to now step up; I’m getting myself to try sous vide.

Do you have a favorite movie?

I’m a fan of Kubrick, so I would say my favorite movie is Dr. Strangelove. I also like 2001 and others.

So, in your perfect day, what would you be doing?

Free day—are we living in a fantasy? I have a two-and-a-half-year old, so I’d be spending time with him of course. Probably interacting with my students as well. The most fulfilling days are some combinations of family time and interacting with students in a lab setting. I bought myself and my son a drum kit, so getting him on the drums and me on the guitar [would be great]. In the lab setting, (one of my roles in the lab other than advising is actually being the expert on manufacturing robots) maybe building robots and testing them out with my students


Who was your greatest source of inspiration?

I would say my inspiration as an engineer comes primarily from two places.  My dad is a savvy engineer, both in terms of creativity as well as salesmanship.  My grandfather always stressed the importance of education.

How did you get into this line of work? What caught your interest?

I remember being the typical “future engineer”—tinkering with Legos, taking stuff apart and putting it back together, building remote control airplanes, etc.  My father (electrical engineer) and grandfather (civil engineer) had quite an influence on me as well.  When I got to graduate school at U.C. Berkeley, I began looking for research projects that caught my interests.  Although I was trained as an electrical engineer, my interests were varied.  At the time, a project called the “Micromechanical Flying Insect” was starting at Berkeley, and this caught my eye.  This would lay the groundwork for later projects in my lab at Harvard.

I read that you were one of the top 35 innovators under 35 in “Technology Review.” What does it feel like to have achieved so much in such a short time? I’m sure you’re with peers twice your age sometimes.

It is, of course, gratifying to be recognized for your hard work. But engineering is evolutionary in nature. We build off of the breakthroughs from other researchers. This is true in our case as well; we are certainly standing on the shoulders of giants.