A Roundtable Discussion with Josie Hughes and Yiğit Mengüç

Abstract

Robotics ReportsVol. 1, No. 1 Roundtable DiscussionOpen AccessCreative Commons licenseA Roundtable Discussion with Josie Hughes and Yiğit MengüçModerator: Surya Nurzaman, Participants: Josie Hughes, and Yig˘it MengüçModerator: Surya NurzamanSurya Nurzaman, Monash University Malaysia, Bandar Sunway, Malaysia.Search for more papers by this author, Participants: Josie HughesJosie Hughes, École Polytechnique Fédérale de Lausanne, Lausanne, Switzerland.Search for more papers by this author, and Yig˘it MengüçYiğit Mengüç, Marble Wit Research, Pittsburgh, Pennsylvania, USA.Search for more papers by this authorPublished Online:19 Oct 2023https://doi.org/10.1089/rorep.2023.29004.snuAboutSectionsPDF/EPUB Permissions & CitationsPermissionsDownload CitationsTrack CitationsAdd to favorites Back To Publication ShareShare onFacebookTwitterLinked InRedditEmail Surya NurzamanJosie HughesYiğit MengüçSurya Nurzaman: Robotics Reports has invited two cochairs of the IEEE technical committee (TC) and soft robotics, which is a nonprofit organization that tries to push forward the field of soft robotics. With us today, we have Josie Hughes and Yiğit Mengüç to discuss the progress and challenges in the soft robotics field. My name is Surya Nurzaman, and I am the editor-in-chief of Robotics Reports, and I will be the moderator of the discussion. I am a senior lecturer at Monash University at the Malaysia campus, and my field is related to soft robotics and bioinspired robotics. Josie and Yiğit, please introduce yourselves and your fields.Josie Hughes: Hi, I am Josie Hughes, and I am an assistant professor at École Polytechnique Fédérale de Lausanne, where I started 2 years ago. My background is in manipulation, soft robotics, and my laboratory, CREATE Lab, looks at developing new design methodologies and fabrication approaches for soft robots that look to exploit interactions with the environment.Surya Nurzaman: Thank you Josie. Yiğit?Yiğit Mengüç: Thank you for organizing, Surya. My name is Yiğit Mengüç and I am currently an independent researcher, I guess I would say, and my background is in mechanical engineering. Formerly, as a professor of my laboratory, I used to work on materials, mechanics, and the manufacturing methods for creating soft robots. Then I worked at Facebook for a few years looking at virtual reality, so wearable soft robots, haptic gloves. Now, I am looking at the relationship between technology and the impact it has on us as individuals and on our communities.Surya Nurzaman: Thank you, Josie and Yiğit. To begin, can you explain how you define soft robots and why it is important?Josie Hughes: Good question to start with. Soft robotics is two things: one, it is a way of technically and from an engineering perspective designing robots that seeks to move away from a paradigm where we have very rigid and precise control to something where we exploit the environment, so this means moving more to soft materials, soft structures, and developing these technologies and controllers.However, I think it is also a mindset as much as anything else, it is this disciplinary way of approaching robotics that moves away from just sense, think, act, to more of a bioinspired approach where we try and develop and understand intelligence, and how we design robots and the emergence of behaviors and intelligence in more of an interdisciplinary way, which I think is quite different to the rest of the community. It gives us a huge strength in what we do, and that we can really approach the science of robotics in this unusual way.Surya Nurzaman: Thank you, Josie. Yiğit?Yiğit Mengüç: I think that is a great answer, I am going to feed on that because the thing that I heard you say, Josie, is this idea of the difference in the paradigm, I think it is almost like a deliberate approach to a paradigm shift. The hypothesis is almost not testable by any one of us, and that is why the paradigm feels appropriate. No one person, no one laboratory can successfully test the hypothesis that a soft robot is more effective at such and such task than a rigid robot, because every rigid robot has a niche that it has filled and it is better than the soft robot in some way.We have been talking about soft robots being better for manipulation, for search and rescue, for human interaction, but in every case, there is a rigid robot that is coming out of industrial process that has had 20 years of development by a thousand engineers, that will be better than the soft robot that we show, almost every single time.Yet we keep doing this, and our hypothesis is constantly falsified, so why do we keep working on it? I think it is because of this paradigm shift and the mindset difference. I think we believe that there is this different paradigm that is based on emergence, which is based on this interaction that you are describing as environmental interaction, that without which, we can already see there are limits of the current paradigm, and so that dissatisfaction leads us to say, “Okay, there must be some other one.”To me, that is the heart of soft robotics, I think the materials and other things are the avenues to test this hypothesis, because they are surprising. The softness means that they surprise us, so the paradigm of emergence is where we are satisfied by soft robotics and not satisfied by rigid robotics, and I think that is what binds the community together more than anything else.Josie Hughes: I think that is a very nice way of putting it.Surya Nurzaman: I think we agree with the shift of the paradigm. Yiğit, you have years of experience in different institutions, what has been your biggest challenge in exploring the soft robotics field so far? Is it the actuators, the fabrication, or something else?Yiğit Mengüç: In the same vein, it is not actuators, and control, and things like that. The thing about soft robotics that is interesting is the emergence and the fact that it interacts in surprising ways with the environment, which means that it is not modeled to begin with. Even further, it is not modelable at this moment. I am not saying that it is not modelable in principle, but the whole point of emergence is that you do not have a model to explain what to expect.Because of this, it is hard as a soft roboticist working with this mindset to be able to get resources from institutions, scientific institutions, or industrial institutions that expect a predictive modeled approach to engineering and science. The hardest part for me is that when you go to funders, and this is both in industry and in academia, and you say, “ I think that soft robots can benefit us in…” The haptic gloves is a good example. You say, “In haptic gloves, this is useful,” and then there is the question of, “How—how much better is it going to be?”My answer is, “I don't know, the current approach is not very good, so I'm going to try this thing and whatever comes out, I will tell you what I find out.” That is unsatisfactory, that is dissatisfying to the mindset of the old paradigm. This is the hardest part for me, I personally struggle within this kind of interpersonal conflict of the different paradigms. That is the biggest challenge in soft robotics.Surya Nurzaman: What do you think, Josie?Josie Hughes: I think you positioned it really nicely, I think we can think of these as top-down and bottom-up challenges. Surely, from the bottom-up, yes, there are engineering challenges in terms of fabrication, in terms of sensing and actuation, these can always be better. There is a pathway there, though, and we are making some progress. There are always going to be challenges there, but we can see the engineering steps toward making these better. Then there is this top-down philosophical approach that makes soft robotics really challenging; we need this interdisciplinarity, and as a student coming to soft robotics, it is not easy because you need to be able to think in these big terms and connect even discussions of consciousness in terms of philosophy, in terms of neuroscience, back down to a very tangible, “This is my robot and how does this interact?”I think that is what makes it so beautiful and just an amazing field to work in, but I think also for people new to the field, that is really hard, to not just get your head around these engineering principles, but also these really challenging concepts that drive us in soft robotics and as you said, to try and go beyond what can be achieved with this rigid robot paradigm, but what we are trying to achieve, maybe not now, but in the future. I think that is a big challenge.The other thing, I do not know whether it is a challenge or not, is showing what we can do with soft robotics. Is there an application or is there, maybe not an application, but something we can do that cannot be done with rigid robots? This is something we always get asked and we can say, “X, Y, Z is better,” but it is very hard to do. I think that we still as a field, we need something to put a flag on. There are many potential things, but as for drivers in terms of funding, in terms of everything through to the next 10, 20 years so soft robotics continues to have this wave of excitement and future direction.Surya Nurzaman: Thank you, Josie. Yiğit, you have worked within industry before. Do you think there is any flagship application that can convince the industry when soft robotics could be better?Yiğit Mengüç: You are saying, this idea of looking for the flag to plant, to show, okay, this is the new land we have discovered with the oldest exploration. I think even that mindset is the previous paradigm of science, this idea that you make, you invent, you find particle physics, you find energy mass equivalence, and then you can blow up cities, it is like, okay, that is a very strong marker.I think that the whole mindset of softness, this idea of emergence means that we need to be entirely soft in our mindset of where that flagship is, where that flag is, which means we can be entirely flexible, ourselves like our robots, how do we make our minds? And this is why I love, Josie, that you're bringing up this idea that it's the challenges that you're coming in, you have to make a thing, and at some point, you're going to take a photograph of this thing, but for the students, it requires them to understand or at least appreciate, or themselves go through all these connections, that consciousness, and philosophy, and all of it has to be there, baked in somehow to get a really good, beautiful, new thing.I was very frustrated trying to explain the flagship idea of soft robotics to these corporate institutional structures that go back decades. Now I am of the perspective of what would a soft robot do? They flop over, they go around, you grow around it. If you are trying to convince somebody, if you are antagonistic with someone, I think that that is a sign that you should flex and grow somewhere else, so grow toward where it is easy. Personally, my perspective is that soft robots are particularly well suited for artwork, for art-making craft, because it is accessible and artists are extremely good at learning techniques. I will pause there, and that is something I might talk about more later, but my point is that we can be very omnivorous, and I think that is my perspective now.Surya Nurzaman: Thank you. Josie, you had some projects in the field of agriculture. How did you convince the industry partner in that sense for this new paradigm?Josie Hughes: I think this omnivorous approach is very nice because I think there is no application in theory that you could not throw a soft robot at. The versatility in approach makes it very interesting, which is not only beneficial in terms of gaining funding and application, but also a bit challenging of where do you put your time and resources, and how can we deploy them when it is both a paradigm shift and a new technology? Playing devil's advocate a bit here, we have such big sustainability challenges, we have such big social challenges, that at the same time as focusing on this philosophy and driving perhaps the science of embodiment and emergence further, that I think we also need to look at how can we deploy our robots to address these sustainable development goals in the very short term?It is a very nice thing that we can make not only this very long-term philosophical and paradigm shifting approach, but also in the very short term, I can develop a robot that can help harvesting. I can develop a robot that can help with food science, or I can develop a robot that can help with wearables in health. I think that is a tricky balance to make, because how do you deal with these very big paradigm-shifting philosophical concepts as well as here and now? I want to make robots that you can take out of the laboratory and can really help people.It is great that we can work on these two different ends of the spectrum, but it is difficult operating in this niche, and I think it is very difficult for choosing a direction to go in. I think there is increasing acceptance from, for example, agriculture, because they need these solutions. There is a real problem, they do not have solutions to climate change, they do not have solutions to monitoring biodiversity, so there is an absolute pull factor that they need new technologies and they are willing to think out of the box to accept new solutions.This creates an opportunity for us as soft roboticists, and I think it is important that we dedicate our time and energy to trying to contribute toward these big global challenges. I think it is also important that we hold on to this paradigm-shifting philosophical approach, because in the long term, I think this is what is going to generate our new robots, 3.0, 4.0, that are going to drive us toward significant advancement in the future.Surya Nurzaman: Thank you. Yiğit?Yiğit Mengüç: There is a tension there, and we can get into that, and I appreciate you bringing this up. I will point out that even this approach that you are describing, I think it is already a different mindset to the process of the work. We are saying that, “Soft robotics needs to do both at once,” and I think that itself is a difference from the industrial rigid robot research path. They can say, “Look, we have a $10 billion industry in warehouse robots that slide under boxes,” or whatever, each little niche warehouse robot. My point is that already the mindset is different, and that tension is in soft robotics, and that is the thing that makes it different already.Josie Hughes: True. You approach it by having that longer term, bigger vision, even if it is an applied problem, you approach it in such a fundamentally different way.Surya Nurzaman: Let us also talk about the interdisciplinary nature of soft robots. Both of you are already the cochairs of the TC, and I believe you organize some activities. I do not know whether you are helping to organize the next RoboSoft, the next conference on soft robotics. If so, which people that you would like to invite maybe to give a keynote that you believe so far we are missing in this field and could contribute to the field?Josie Hughes: I think for soft robotics to continue to have the wave we have, we need to be interdisciplinary. I think we can always include more biologists; I think we need to have more on the philosophy side and also more of this understanding of neuroscience and also cognition. So I think having this more awareness in this area is what is going to drive soft robotics through the next 10, 15, 20 years, because I think there is a risk otherwise that we stagnate into what we are doing at the moment.The research fields work in a very different way, so I think soft robotics can do a lot in trying to break barriers and create a language with which we can communicate. It has already been done very well in material science. I think that is perhaps the strongest interaction between soft robotics or maybe biologists. Doing more of a job in terms of breaking barriers, in terms of cognitive sciences and philosophy, is an important direction. I think we can never be interdisciplinary enough for soft robotics, because that is really where our strengths are.Surya Nurzaman: Right. What do you think, Yiğit?Yiğit Mengüç: I like the way you said we can never be interdisciplinary enough. From that perspective, there are no disciplines, the whole, all of human experience falls within the idea of the research that we are doing.Josie Hughes: Right.Yiğit Mengüç: There was one workshop that a colleague from Facebook organized at the previous Robosoft in Edinburgh. He is a neuroscientist, and the workshop was titled something like “Stop Making Robots and Instead, Reinvent Life.” They had speakers from neuroscience talking about cognitive models for chat, talking robots and so on. This was before ChatGPT got really big, so that was interesting.They had a speaker who was talking about the evolution of brains, and this was an evolutionary biologist showing that the evolution of brain stems into our frontal lobe and what came next. That was interesting to me because we talk about biological inspiration in soft robotics, but we do not talk about biological process inspiration. Following the process of development in robots is an interesting new concept that I saw from there, so you cannot be interdisciplinary enough.What else I could add in the future is the philosophy of softness. Where I have seen it reflected most explicitly is actually in Taoism and Lao Tzu's writing, and where he says, “Be entirely soft, to follow a spiritual path, you must be completely soft.” That is the closest that I have seen, and this is a very old concept. If we cannot be interdisciplinary enough, that to me is the logical extreme at the moment, of who else to invite to the conversation.Josie Hughes: A good point. The art, and I think creativity and touching on even artists and that domain is interesting. This label of soft robotics is not always good because it is a more continuous sphere than that. I think it is also very important that we have a conversation with more typical roboticists, because I think we are approaching similar problems from the same direction and I think we need to understand how we can leverage what they are doing and also how they can leverage what we are doing.I think that understanding more of a conventional and a learning-based approach to complexity and intelligence is very important, especially as we see with developments, for example, in large language models (LLMs), you could argue it is really transformative, it is a step function in what we can do with machine learning, whether it is good or bad, but we need our equivalent for soft robotics and robotics in general. We need to be aware of this type of more perhaps conventional intelligence and how we can leverage it. We cannot live in our isolated worlds, I think we need to be aware of what other more conventional types of intelligence and robotics are also occurring and how we can leverage that or how we can augment it as well.Surya Nurzaman: That is a very interesting discussion. As you are also both the chair of TC, how do you drive this community in the next 10 years?Yiğit Mengüç: My intent is to move out of the committee. I have been on the committee through the pandemic, and I think we understand now that that is too long for the way that we want to keep the community fresh. I will just say that the way that I expect to drive it forward is to work entirely outside of the community. I will extend that to include that I am also not a professor right now, I am not managing researchers, so my goal is to try and find any other avenue to explain the perspectives of the value of soft robotics, I am exploring that in ways outside of the traditional modes of research.Josie Hughes: I think that the interdisciplinarity, as we discussed, is one very important thing, how can we grow this, and how can we attract other disciplines into our community, and how can we create a language in which we can understand each other. Another aspect, which is very practical, is students and education, because I think it is a challenging field, because we have assisted interdisciplinarity, and I think, as a technical community, as a soft robotics community, empowering the next generation of students and researchers, and also trying to make it more well rounded, make it more representative, and I think also make it more diverse, representing this interdisciplinary parity and also this diversity and thinking that we need.I think it is very important that we use soft robotics with this flexibility, this omnivoracity and thought to reflect that in the kind of community we have. I think it is very important that we can use this to create a more diverse community, and I think that will do wonders for the research we do as well as being, I think important socially and in terms of what we can do in terms of inclusivity and diversity.Another thing I think we can work toward addressing is can soft robotics, with this mindset, try and contribute to doing some good in society? Be it in terms of environment, be it in terms of society, be it in terms of aging and health care. Are there some sustained activities as well as this philosophical and larger point of view where we can try and contribute to humanity in society? I think that is great in terms of engaging the next generation who have such an awareness of sustainability and society that is perhaps not as prevalent in other generations, so I think that in turn will help feed and sustain the community.Surya Nurzaman: We talk about the direction that maybe we plan to go in in the next maybe 5 to 10 years, but I am also curious whether you think there is anything that is beyond soft robotics. Will we go in the same direction in the next 10 years or will things radically change in the next 10 years?Josie Hughes: That is a tricky one.Yiğit Mengüç: That is a tricky one. If I look at it retrospectively, and I was trying to remember, Surya, did you go to the soft robotics meeting in Switzerland?Surya Nurzaman: The one in Switzerland in 2013?Yiğit Mengüç: Yes.Surya Nurzaman: Yes, I organized that.Yiğit Mengüç: I think it is interesting because at the time it felt like everybody was there, and how many participants did we have? Less than 100? Eighty maybe.Surya Nurzaman: Maximum 100, I think.Yiğit Mengüç: If I remember correctly, about 5 years ago, soft robotics was the second most common key word at ICRA after AI, or machine learning, or something. That timing is 2013 to 2018, so 5 years. Even then, I started having conversations with colleagues in the community, with other professors, saying, “Are we over soft robotics? Is it overused?” That is where I feel like it is now too. I think that we are not satisfied with saying, “We're doing robotics,” and soft robotics is now too big of a term. Josie, you are bringing up this idea of learning models, and large learning models, and LLMs, and these as examples from traditional robots and where we need to interface. I think that soft robotics will have a similar branching out in 10 years.It has an opportunity to do this branching, and so the soft robotics committee might end up being the trunk. I think that it will try to split off into more subdisciplines, but this subverts the goal of being more interdisciplinary. I don't know what this prediction means, but that's my feeling, that's the direction things tend to go, is communities get big enough and then you subdivide into niches, but I don't know if that's the right way to approach it if we're having this philosophy.Surya Nurzaman: Thank you Yigit, Josie?Josie Hughes: I think if we were here in 10, 15 years discussing the same things, then we have done something wrong. I think the conversation needs to shift and I think it is right that it shifts. I think our question in the next couple of years is to work out what is the shift that is going to propel us to make this shift to empower the next 10, 20 years? Both in very practical terms, in terms of technology, and grant writing, and resources, but also more philosophically. As we understand the paradigm, we are working better, and as we have other developments in terms of LLMs, in terms of fabrication, in terms of hardware, I think it is important that we try and work out what is this next shift and what is the next development. Is it interdisciplinarity? Is it an application? Which I probably do not think it is.What is the shift that can change, or advance, or augment what we are doing now to define the next period of soft robotics. What is the soft robotic revolution that is going to really change the paradigm of the way we are thinking? Is it another mindset? Is it another piece of technology? Is it bringing in a different discipline? I think it is important that it changes and it evolves, I think it is very hard to say what that evolution looks like and where it should be.Surya Nurzaman: All right. Thank you so much, Josie. Do you have any message to the people who want to start exploring this field?Yiğit Mengüç: I will pick up on a thread that Josie said some time ago, which is this idea of the creativity in the work. One thing I have seen from that is that if you are going to be creative in the work, I think the work of creativity requires the work of growing into yourself, being confident enough to see that all of us, all humans, all of us come to life with our unique perspective and that includes soft robotics. Having the confidence to be like that, your personal interests, you are very specific like, “I think this is cool and I'm not really sure why it's important,” is the place to start.Research is important, but you have to start with that first little spark of, “This is really cool,” and starting with that personal excitement, without explanation, without rationalization, without the science side yet is the thing that will feed, I think what Josie is saying, this idea of feeding the goal of evolution of the field, the goal of changing, and growing, and being able to discover the applications for sustainability and societal good. That starts with being able to recognize and protect that first little excitement, and that is where new students usually have a lot of excitement and I think they get overwhelmed with this need to show that it is important. There is a tension of funding it and explaining it.Josie Hughes: I think that is a really nice perspective. I think that coming into it, being as inclusive as possible. I think diversity is important, and I think it is great that we can make a contribution to robotics in general through this by having it as a more creative space where we celebrate interdisciplinary approaches and different ways of thinking. If we are all thinking the same way, we cannot approach these difficult topics, so I think discussions are really important, I think the community is specifically important in soft robotics because we cannot tackle these problems alone, and we need these discussions to change and also challenge our ways of thinking.I think the more discussion we can have, the more we can bring the community together and the more we can support a future of interdisciplinary diverse researchers who have the opportunity to exploit and accelerate creativity and thinking in terms of the science, in terms of applications; I think we have a really exciting future ahead of us. It is exciting to see where soft robotics is going to go in the next 5 to 10 and future years.Surya Nurzaman: I think we have had a very interesting discussion. Thank you so much again for participating in this discussion, I really appreciate it.Cite this article as: Nurzaman S, Hughes J and Mengüç Y (2023) A roundtable discussion with Josie Hughes and Yiğit Mengüç. Robotics Reports 1:1, 50–56, DOI: 10.1089/rorep.2023.29004.snu.FiguresReferencesRelatedDetails Volume 1Issue 1Oct 2023 Information© Surya Nurzaman et al. 2023; Published by Mary Ann Liebert, Inc.To cite this article:Moderator: Surya Nurzaman, Participants: Josie Hughes, and Yig˘it Mengüç.A Roundtable Discussion with Josie Hughes and Yiğit Mengüç.Robotics Reports.Oct 2023.50-56.http://doi.org/10.1089/rorep.2023.29004.snucreative commons licensePublished in Volume: 1 Issue 1: October 19, 2023Open accessThis Open Access article is distributed under the terms of the Creative Commons License [CC-BY] ( http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.PDF download