Interview Transcript
Q1: If we could start with your name, where you were born, and some of your early education, how did that unfold and get into your life?
Hugo Rodrigue: Okay, so I’m Hugo Rodrigue. I’m Canadian. I was born in Montreal. I did my undergrad at McGill in mechanical engineering. I did my master’s in 3D manufacturing at the University of Montreal. Then I took some time off and moved to Korea in 2011 to do my PhD—at that time in manufacturing.
Then I changed to soft robotics and graduated in 2015, and in 2016 became a professor at Sungkyunkwan University, which is very hard to pronounce. We are affiliated with Samsung, so let’s say it’s pretty nice. And I’ve been a professor there since 2016.
Q2: How did you decide to go from Canada to Korea? What inspired that?
Hugo Rodrigue: So I thought at that time I had many options. My advisor was moving back to France, and I kind of wanted to move fields at the same time. So I looked at all options—some were just for, “Okay, maybe this sounds more fun.” I kind of wanted to learn a third language (I had failed to learn German previously), as I speak French and English and wanted to add a third language.
The Korean government scholarship offers a program where you learn Korean for one year—and only learn Korean—and you get paid, and then you can do your PhD. I thought that was really attractive. So I moved to Korea to learn Korean and then did my PhD, and I’ve just stayed there since.
Q3: How did you get interested in robotics?
Hugo Rodrigue: I mean, I think everyone likes robotics. My family has a lot of engineers, so engineering was always around. The most interesting applications of engineering in movies when I was young were robots. I mean, I know there are cars also, but I was never interested in cars, so I liked robots.
When I was in high school, I thought robotics were cool and I would like to do that with my life. But then when I started my undergrad, I had a class about robotics, and it was really just a bunch of mathematics. I thought, “Okay, maybe I don’t like robotics.” So that’s why I did my master’s not in robotics.
When I moved to Korea, I wanted to do manufacturing, but my lab also started doing soft robotics, and I saw that it was more design-based. That was quite interesting for me, and that is when I moved back to robotics—specifically soft robotics.
Q4: When you mentioned seeing robots in popular media as a university student, what kinds of things came to mind?
Hugo Rodrigue: Things like Terminator. I mean, Terminator is nice. There’s also a soft robot there, right? The enemy made from liquid metal is actually maybe a soft robot. So that’s the main one that pops into mind. And there’s Star Wars with the different robots with different shapes, which was interesting. I think those are the main two ones that I remember right now.
Q5: When you got interested in soft robotics during your PhD, what kind of problems was your team working on?
Hugo Rodrigue: At that time, we were trying to build shape memory alloy-based soft robots. That’s what they were working on, and I was told that’s what we work on, so that’s how I ended up on that specific topic. What we were trying to do was to create new types of motion using shape memory alloys.
At that time, soft robotics was a pretty new field—I mean, it’s a pretty old field, but at that time it was kind of a renewal of the field in a way. Most people could only do very basic motions with non-pneumatic actuators, and bending was difficult using shape memory alloy; twisting was not possible.
So I thought, “Okay, twisting is not possible, so I’m going to focus on using shape memory alloy and polymers, combining them to create twisting motions.” And that was my PhD topic.
Q6: You mentioned that soft robotics is both new and old. Can you give a little background on that history?
Hugo Rodrigue: In the late 1950s, they invented the McKibben actuator, which is a pneumatic actuator. That was pretty much the same time that rigid robots started to exist. But rigid robots have been very useful for industrialization and making things; they’re everywhere in factories. So at that time, there was not as much interest in soft robots.
You can find some examples of research every decade in soft robotics, but it was not a very large field. Then in the late 2000s and early 2010s, we saw the research from George Whitesides at Harvard, who was actually a chemical engineer presenting soft robots as an application of polymer processing techniques and polymeric materials. That brought it out and created a big boom—saying, “Oh, we can make robots from soft things, it looks great, and it might be useful.” I think that is when our field really started becoming bigger and bigger very quickly.
Q7: Do you see the field driven by particular applications, or by technical and material-based solutions that provided a push at the time you entered?
Hugo Rodrigue: I think a lot of it is a desire to do more basic research. I feel soft robotics is a field that mixes basic research—like, “How about we just study the motion of animals through robots?” like a worm or other kinds of soft animals, or even animals with bones, and replicate them with soft materials or combine them with rigid materials. The application of smart and intelligent materials is driving the field forward without directly having an immediate application in the same way rigid robots do.
Q8: What were some of the challenging technical things you had to solve to get your twisting actuators working? Where did you go from there?
Hugo Rodrigue: We had a manufacturing issue at first that we were trying to solve—not actually 3D printers yet. 3D printers really helped our field. If you have polymers, you need to mold them. If you have a 3D printer, you can easily make a mold; whereas before, you needed to machine a mold, and machining costs a lot of money, machine time, and people. 3D printing made it very easy: you get the mold, you can make it. That’s why it started getting more popular.
Then, the issue was how to design the interaction between the shape memory alloy and the polymers, how to assemble them, and how to make it into a working actuator. There is an interaction between the shape memory alloy and the polymer that is not always easy to resolve—balancing the dimensions and properties—since shape memory alloy is actually a metal and very rigid, while the polymer is very soft. They don’t necessarily interact well together naturally.
Q9: What kinds of things did you pursue after completing your PhD?
Hugo Rodrigue: After my PhD, I did a very brief postdoc of around six months in the same lab, and then I became a professor directly—a pretty brief path compared to most. Once I started my own lab, I decided I didn’t want to work with shape memory alloys anymore, so I immediately worked on pneumatic actuators, which is a more mainstream direction in soft robotics.
Within this, I focused on vacuum-based actuators. Instead of adding air, you remove it, and that changes the design a lot.
Q10: What are some of the main research questions you explore with vacuum actuators?
Hugo Rodrigue: I focus on the mechanical design of soft robotic elements rather than overall soft robotics applications. My main question is: How much work and how much power can we get out of a soft actuator? I try to make actuators that can produce more work, produce more displacement, and move faster to enable more dynamic motions. I don’t focus as much on applications compared to most people.
Q11: Where do you get your inspiration, or how do you decide where to go next in your research?
Hugo Rodrigue: That’s a good question. I feel I wake up and I’m like, “This is the direction.” I don’t have a single directing axis. I kind of just try to evolve from my own technology and say, “Okay, these are the current issues that we have, so let’s find different ways of solving them.” Then we just evolve from there. That’s how I’ve done it most of the time.
Q12: Can you tell us about your lab? What kind of expertise do the people you work with have, and how is your science workflow?
Hugo Rodrigue: When I started my lab, my PhD background was in manufacturing, so I had never attended robotics conferences until I became a professor and didn’t know many people at the time. I was quite lonely by myself, so I just worked on my own things without many collaborations. I still don’t have so many; I have collaborated mainly with other Korean professors, though I’m trying to set up some collaborations with people in Europe at the moment.
My lab is pretty small by Korean standards—I have around 12 graduate students and one postdoc. Every single one of them is a mechanical engineer. I would like to have different expertise, but the Korean system tends to draw students from the same major, and we don’t get a broad pool of postdoctoral applicants.
Q13: What are some of the projects in your lab that you are most proud of or find most interesting?
Hugo Rodrigue: The project I’m working on now involves building wearable robots to help nurses in hospitals handle patients. I used to work only on actuation, but for this project, I’m working on sensing, and that is something I’m quite happy to be working on.
Normally, if you use a sensor, you use something like an IMU to detect orientation in the world. But I’m working on stretchable sensors, trying to build sensors that can stretch more using less force. Rather than making it purely a materials issue, I’m studying the physical design layout of the sensors to achieve these properties. I found a lot of common themes with my actuator design, which has been a great new research axis for my lab.
Q14: How did you end up taking on an application-based project like the nursing assistant?
Hugo Rodrigue: This is a collaboration where someone else is handling the broader application. I just focus on my sensor, integrating it, and doing the process of calibration—not as much as the application itself. Most of my individual projects focus heavily on the actuator and components, with the application taking a backseat.
If you work on an application, at some point you have to shift from developing the component to developing the application. Whereas if you focus on the component, you can work on improving it from day one to the last day. I work on components because a lot of people work on applications, but not as many work on fundamental components. If we can make better components, everyone can use them to improve their own applications.
Q15: Why do you think fewer people focus on the basic component side?
Hugo Rodrigue: I think it depends on personal interest and what appeals to funding agencies. If you work on solving a real-world problem—like wearable suits for nurses or shipping robots for manufacturing—that seems like a very good thing to fund. Whereas saying, “I want to make an actuator that is 20% stronger,” is not as attractive to fund in comparison.
Q16: Where do you get your research funding, and how do you manage that?
Hugo Rodrigue: Mainly from the National Research Foundation of Korea, which funds most basic research in Korea, and from the Korean Evaluation Institute of Technology, which focuses more on applied research where I work with collaborators toward a given application.
Q17: Are there other applied areas your lab has worked on?
Hugo Rodrigue: I’ve worked on logistics and shipping with CJ in Korea, focusing on unloading trucks and palletizing boxes and packages. I’ve done two projects with CJ on this. Most of my other projects focus on either specialized robots, robotic arms, or soft robotic arms—which, to me, is still essentially an assembly of components to build an arm.
Q18: Can you say a little more about your work on soft robotic arms?
Hugo Rodrigue: Most robotic arms are rigid; that’s what is installed in every factory in the world. We used to have very large ones, and now we have cobots that are smaller but still rigid. If you want to build a soft robotic arm, the issue is that most soft actuators are not strong enough and don’t produce enough deformation to support a good payload with a wide range of motion.
My goal is to use stronger actuators with a bigger range of motion to build a soft robotic arm that can be controlled and carry a good payload while remaining compliant. Compliance makes it inherently safe. You wouldn’t use a soft arm for very precise work like soldering a car part, but if you want to handle a patient—towel them off or give them food—you want inherent safety in case they move or the robot miscalculates.
You can create compliance using motors, but that is generated through software control, and control can potentially fail. If it is physically compliant in every scenario, it is always safe. I’ve been thinking a lot about whether a soft arm needs to be soft everywhere, or if a combination of rigid and soft components—where the soft components provide compliance in every direction—is enough.
Q19: How do you test compliance and safety in your lab?
Hugo Rodrigue: On my students! I always tell my students, “Stop trying to measure things right away—just play with it and see if it matches what you expect to happen.” Human hands are great sensors. If an actuator makes 0 to 10 kilos of force, you can pull it with your hands to feel it. If you put the arm on a table and it pushes against it, you can observe how it deforms and adapts its shape using its compliance. Just look at it and observe it.
Q20: What parameters besides compliance and safety do you take into account when making the robot arm?
Hugo Rodrigue: For me, it’s really the force, the payload, the deformation, and accuracy. In soft robotics, a gripper uses compliance to grab objects of any shape or size, so accuracy isn’t as important. But with a robotic arm, accuracy is critical because you need to place the gripper in the right region. Implementing sensors to achieve that accuracy is a critical design parameter.
Q21: Do you develop specific sensors for the arm, or use off-the-shelf solutions?
Hugo Rodrigue: I try to use off-the-shelf rigid components where appropriate. By using rigid components for joints, you can use traditional sensing solutions on those joints to get good accuracy.
Q22: What performance targets are you aiming for with your soft robots? Are you trying to match human capability or surpass it?
Hugo Rodrigue: Everything has a different goal. A human lifting weights can lift maybe 10 to 15 kilos. If a soft arm can achieve a similar payload, it can accomplish a similar set of tasks as humans. If a human lifts boxes all day and a robot can’t lift that same box, it doesn’t have the same utility.
In some cases, I try to surpass humans in speed. With pneumatic actuators, I’ve worked on jumping robots. People ask why I make jumping robots, and I say it’s just for fun! We have robots that can jump 2.5 meters high. I can’t jump two meters, so in that respect, the robot surpasses humans.
Q23: Is there anything you would have liked to pursue in your research career but haven’t had the chance to yet?
Hugo Rodrigue: I really would have liked to work on biological tissue-based actuators. But that requires more approvals, paperwork, specialized equipment, and space that I don’t see myself being able to assemble right away. Doing that paperwork in Korean is also more difficult for me, so I tend to stay with what is easier for me to execute.
Q24: What do you see as your most important contributions to robotics?
Hugo Rodrigue: Since I work on components, my biggest contributions are in making better actuators. Specifically, my work on vacuum actuators has achieved some of the highest payloads and contraction ratios demonstrated for any soft actuator.
Second is my stretchable sensor—it is very cheap, reliable, and highly stretchable with very little force. I think both will have a lot of utility in the future.
Q25: Looking at how robotics has evolved since you started, what are the biggest developments that have influenced your work?
Hugo Rodrigue: The biggest overall advancement is AI, which is huge right now and enabling humanoids to transition from academic pursuits to major industry focus. But I try not to use AI because I don’t focus on software-related solutions. If I worked on AI, I would be doing work that isn’t as good as what other experts are doing. By restricting my focus, I can stay dedicated to hardware design.
The technology that has changed my work the most is 3D printers. My lab has around 15 3D printers—I think I have more 3D printers than students! I buy multiple units of the same printer so students can print and prototype faster. It’s like a wall of 3D printers.
Q26: Where do you see the field of robotics heading in the future?
Hugo Rodrigue: In countries like Korea and Japan, there are declining birth rates, aging populations, and shortages of workers for jobs people don’t want to do or in places people don’t want to live. We really have to find robots that can fulfill these niche jobs. The main direction forward is building robots to fill those remaining hard-to-staff labor gaps.
Q27: Do you see nursing robots as a particularly promising direction in healthcare?
Hugo Rodrigue: Yes, nursing robots would be extremely useful. In Korea, a family member often has to stay in the hospital next to the patient to assist them. As more people live alone, they may not have family available. A robot that can handle basic assistance tasks has tremendous potential to lower healthcare costs and raise care quality.
Nursing is demanding, and the nurse-to-patient ratio in Korea is quite tough. Patients often need simple things—a blanket change, a pillow adjusted, or temperature checked during the night. If a robot handles these basic tasks, the quality of the patient’s experience improves, and human nurses are freed up from running around so they can spend quality time giving the human touch where it matters most. Mixing the two yields the best outcome and highest acceptability.