EPISODE 136: SOFT ROBOTICS

Transcript

Andrés Almeida (Host): Welcome to Small Steps, Giant Leaps, your NASA APPEL Knowledge Services podcast where we dive into the lessons learned and real-life experiences of NASA’s technical workforce. I’m your host, Andrés Almeida.

Think of an octopus flattening its body so it can squeeze itself between coral. What if a robot could do that? A novel realm of engineering called soft robotics is being studied at NASA’s Langley Research Center. Engineers are working to understand the limitations and capabilities of soft robotics that could one day support space exploration, including missions to the Moon’s surface. 

Our guest this episode is Jim Neilan, an engineer and project manager at Langley. He’s going to tell us about the latest in soft robotics and what we’re learning so far.

Hey, Jim, thank you for joining us today. 

Jim Neilan: Well, thank you. I appreciate the opportunity to talk today. 

Host: Can you give me some background on soft robotics?

Neilan: Yeah, so soft robotics is quite the field. So soft robotics themselves are robotic systems that are made out of materials like fabrics and plastics and elastomers that allow the body or the shape of the robot to really easily conform and comply around harder objects in the environment. 

So there’s a lot of interesting technologies that are coming out of that and being developed, not necessarily for soft robotics bodies themselves, but being developed in other research areas that are also very applicable to the field of soft robotics. One of them is liquid metal circuitry. So they use gallium alloys that are very pliable, conductive metals, that can be used in soft robotics and allows the software robotic system to have circuitry that’s flexible and bendable, and stretchable at times, and also, they can also heal themselves a little better if there’s damage and there can be healing involved there so that they can correct issues that happened or damage that happens during operation which is, which is pretty incredible. 

Another aspect is that’s coming along in soft robotics is thermal thermally pliable plastics. So these are thermoplastics, that when heated, they can be very pliable. And then when cooled, they harden and stiffen. So that allows, that gives some strength to soft robotics in certain in certain areas, and in certain use cases that we’re looking at. So it’s very interesting. It’s both, you know, on the surface and subsurface, and it’s in the electronics, and it’s in the actuation, where we use either water or air or other liquids to be able to move these robots around. 

Host: So, it sounds like that’s kind of inspired by nature. Can you think of any examples where that are inspirations for soft robotics? 

Neilan: Yes, in nature is a big inspiration, inspiration point for soft robotics. There’s lessons that we learned from nature. So you know, we spent, you know, billions and billion years here, evolution solving some very, very hard problems in the natural environment. And right now we look, as software roboticists, we look at snakes and birds and fish, and cephalopods, and sharks and insects and even people to give us inspiration on how to develop soft robotics, due to the fact that you know, nature has really solved some really difficult problems over the years, with the soft systems, us included. 

Host: What would you say are some of the biggest challenges in soft robotics development. 

Neilan: One of them is control. So, a soft robot system contains a theoretical infinite numbers of degrees of freedom, infinite degrees of freedom, which means that the body, any part of it can move in any direction really, at any time. And that’s a bonus. But it’s also a minus, it’s a pro and a con because the ability to move in any direction allows us software, but kind of a versatility and the ability to be robust and operate in unknown environments in dangerous environments. In environments where there’s, you know, unknown shaped objects, we have to either manipulate or move around.  

But we also have to know with certainty how the robot is going to move in a lot of these instances. Especially in space operations, we really would like to know how our systems operate so we can predict and determine how they’re going to function in the space environment or on the lunar surface. Soft robotics, because they have these infinite degrees of freedom, that makes the control and the modeling we develop for the control very, very difficult. So we can’t easily predict how the robot is going to respond in every situation or in situations that we’re familiar with. And so, that’s very challenging and requires a lot of, a lot of mathematics to be able to model and then predict within certain certainty how the robot is going to act in the environment that it’s operating in. 

We also have, you know, different sensors that can help us that are with the robots, you know, remotely and to be able to look at look at the ways so there are, there are ways to solve the problems of modeling in in control of these soft robots. But they all really involve quite a bit of rigor, and also a level of uncertainty, that’s not going to go away. So there’s, this is what we’d like to refer to as a risk tolerance. It’s that, well, we might not know the exact answer. But, you know, maybe that’s good enough. You know it’s good enough to not have the exact answer. And to let these systems operate.  

Host: So, can you give us some examples of critical needs in space where soft robotics can be beneficial? I know you talked about the Moon, the Moon’s surface, but how would that look? 

Neilan: Sure, well, and I include this might be a little apocryphal in the in the in the field, but I include inflatables in this as well. So an inflatable system, in my mind is a soft robot too. So inflatable systems are really great, because, you know, we have habitats that we can inflate and easily store and then they can can expand and you have this this this volume of livable area, which is really great for transporting materials to surface and mood. For instance, there’s a lot of there’s a lot of work and benefit in compliant manipulation. So I mentioned, you know, having the soft bodies kind of move in various different directions. So what that one big benefit of that is having a manipulator or a robot gripper be able to do is to be able to form around a strange object we didn’t originally plan for. In traditional robotics and space robotics, we know exactly what we’re going to move around. And we know exactly how to open up the gripper and close around a very known piece of geometry, a connection point. 

As we move further out and stay longer in the space environment, we really have to have robotic systems that are versatile and able to operate in unknown uncertain environments, and soft robotics, their compliance, their material compliance gives us that versatility, and ability to deal with these unknown environments. So if we find a rock, for instance, on the lunar surface that is shaped very oddly, and we want to sample that rock and or move it, a soft robot, a soft gripper can form around that particular piece of geometry that we didn’t originally plan for, and grip it and move it or take it as a sample these types of things.  

And there’s other examples of that there’s, there’s also areas where, like vine robots, if used, if used in the space environment, can actually explore caverns or explore rubble piles, by being able to move in unplanned directions in a responsive manner, so it’s interacting with the environment. And as soon as it hits like a rock or something, it’ll start expanding in a different direction. If you’ve ever looked at or have seen the vine robot research that folks like, you know, Stanford University are looking at, they’re developing these vine robots for, like, search –and-rescue, types of type of work. So there’s large rubber piles or buildings of collapse, and these vine robots will snake through under the rubble and be able to find passages and possibly find, you know, trapped victims, you know, that are in the rubble.  

So that type of a type of mobility, in what we call like a morphological response, or a embodied intelligence response is how we define the community. That type of you know, capability is very, very valuable in unknown environments for search and rescue, obviously, here on Earth, but for space exploration, it can it can go into these weird areas that we don’t necessarily have a specific system designed specifically for, that these robots are able to actually go and explore and, and look at various structures on the Moon, rolling and bouncing.  

There’s a lot of work that’s being done in you know, how do we cheaply and effectively move robots around on the surface without having, you know, large power sources and wheels that can get stuck, like you know, we see the, the rovers on Mars, their wheels will get stuck a lot, right? And there’s, there’s soft robotics research that look at rolling and bouncing and even the snake-like sidewinder type of motion, or tank tread type of motion with these soft robots that are able to kind of distribute their weight over the surface of the Sands or the regolith a lot more efficiently and then that getting caught In the wheel is digging in and getting stuck in different compact, compactness of the various regolith material, that that solves that problem or can solve that problem with these different ways of mobility. 

And we look at biology that’ll solve these. Like sidewinder snakes, for instance, they’re very efficient moving across the desert, you know, over dunes and what have you. And so that same type of type of biologically-inspired work is going on in soft robotics, for mobility. Other areas where soft robots really kind of shine in really, we think, as researchers think that can really add benefit to space exploration is, you know, temporary structures, temporary habitats, these robots can kind of move around the surface and then link up and rigidize.  

And then because because they’re, they’re either run by a fluid or, or, you know, like an air or a liquid or other types of actuation, we can actually pipe you know, water through these soft robots, once they form the habitat, and then all of a sudden, we have a really nice radiation barrier, because water is a really good radiation barrier. So these types of things we’re thinking about, and the different types of use cases, that soft robotics, you know, offer the space space exploration world. 

An upcoming one is assistive robotics. So there’s, there’s work in robotics in assistive robotic spacesuits, and using soft robots for the next-generation spacesuits that do not only being able to pass liquids around in the body for cooling, heating and cooling of the spacesuit, but also in helping astronauts with moving objects around and walking around in service. Moving around in the spacesuit is hard. It’s labor intense, even in zero-g environments, there’s mass you’ve got to move around.  

And on. If you look back in history, like Apollo 14, for instance, when they’re on the surface of the Moon, Edgar Mitchell and and Alan Shepard were walking around the surface and they were had a wheelbarrow and they were walking towards a specific crater that they were looking at, they got exhausted very quick, very, very quick. And so having assistive robot being able to help an astronaut move on the lunar surface of the Martian surface, is really helpful specifically on the lunar surface, because it’s such a weird, low-level gravity environment that it can be very trying to, to be able to efficiently walk. And you see, when astronauts do walk on the Moon, they’re bouncing around, you know. But when they’re pushing equipment or carrying equipment, they can’t really bounce and gait changes, can be very exhausting. 

And when you get when you’re in an enclosed spacesuit environment, and you start if you start hyperventilating, and sweating and all these things, bad stuff can happen. So you have to be very, very careful. Astronauts really have to have to have a lot of mental training and physical training to be able to operate effectively on these EVAs. So it’s pretty incredible.  

Host: So, it’s a nascent field, and you’re studying this at Langley, but does that mean you’re also working with Johnson Space Center in human spaceflight? 

Neilan: We’re not working directly with them. But we’re very familiar with their work. And obviously, I’ve talked a number of times, and in regards to different projects, there’s so I had mentioned like habitats and inflatables, and different types of, like soft systems. And there’s also what we call a heart a hybrid soft robot, where we have some hard components and soft components. And that’s, you kind of get to the better of both worlds there. 

And at times, and, and we’ve talked with Johnson in regards to even the connection tubes, from spacecraft to spacecraft or spacecraft to rover, and discuss different different types of robotics, both in a hybrid sense to, to be able to make that really an efficient and interesting way of transporting people without having to go outside in the space environment to get from a habitat habitat to a rover, and what have you. 

Host: In your opinion, what makes a good robotics engineer, thinking about all these things that you just mentioned? 

Neilan: So, you know, robotics is a multidisciplinary area of study. It really is. It’s, it takes computer science, it takes mechanical engineers, it takes electronics engineers, it takes biologists, right, and when we look at how do we kind of mimic the biological world and some of our robotics so we talked with, you know, biologists, marine biologists really is a big one. And especially in looking at how just studying the octopus right and the octopus is incredibly versatile. All intelligent creature, and being able to kind of look at how they use what’s called hydrostatics in their arms to be able to have this really interesting high dexterity type of tentacled arm be able to manipulate, you know, things on the ocean floor. Or taking that and we’re looking at how do we build a technical arm to be that versatile, and, and yet still be very precise, to be able to do robot construction tasks on the surface of the Moon or in orbit.  

So we have these types of robot arms, or sometimes they’re called continuum arms, continuum robotics, where we have these systems that are both kind of a hybrid soft and hard type of robot arm that are biologically inspired.  

And so it takes these different fields and having a very broad, you don’t necessarily have to have a real deep understanding until you get into a certain area that you really want to learn, but a very broad kind of, you know, review, from all these different fields. It also takes, or I should say it also, It’s also very beneficial to have a lot of different skill sets, you know, folks on your team is a big one, because you can’t be an expert in every single one of them. So if you have mechanical engineers, and you have biologists, and you have electrical engineering, you have computer scientists and you have, you know, people that are interested in developing these new systems in these soft systems, and then working together to take very interesting solutions in their particular fields and bringing them together — those are the types of the types of skills that really make a great robotics engineer is to at least understand that it is no set single field that will be able to really get you to the answers that you need. 

Host: Are there any demonstrations planned for the public to kind of get excited about? 

Neilan: There are are so many interesting bits of work out there. And there’s other research groups, like, you know, at Ames and Ian Johnson and Glenn and the other centers in JPL that have been doing and are still continuing research in these fields.  

Host: Including inflatables. 

Neilan: Yeah, inflatables at Marshall is a big one, right? With the inflatable systems and Langley and more. So I’ve been working hand-in-hand with inflatables. And deployables is a big one at Langley. 

So the big demonstrations are kind of like taking, taking a lot of the separate types of technologies for soft robotics and having them kind of demonstrated in a smaller scale on, for deployable boons, for instance, or inflatable habitats, or for JPL. They’re, I mean, they’ve developed the gecko foot. So the gecko, this foot, they have all these microscopic little hooks on geckos and geckos can cling to smooth surfaces very, very well. And so JPL and a number of other researchers kind of got together and they and they made a robotic gecko foot that is able to do, I think it’s like a three-by-three square that can lift 10 to 15 pounds or something like that. It’s really incredible. And so they have these different soft gripping mechanisms that really do some incredible things. And those are the types of demonstrations that you see that are upcoming.

Host: Can you explain a little bit about the lab where you are? 

Neilan: Sure. So what I do is, for the past 10 years, I’ve been working on, you know, autonomous systems and robotics for for Langley. And the most recent project that we are working on is a project called PASS or Precision Assembled Space Structure. And this is an in-space assembly demonstration project working on the robotics and the autonomous systems to be able to build the next-generation space telescopes out in orbit at the LaGrange Points. Because it’s very hard to have astronauts out there for very long periods of time that those particular locations are hard to get to and hard to get back from. We know that our robotic systems are the ones that have to go out and build that.  

Now what we’ve done is we’ve made a robotic gripper, a soft gripper for the PASS project that allows So the end effectors of the robot arms to be able to handle very, you know, different-shaped objects. So they have one gripper that can deal with multiple-shaped objects. That’s one benefit from the softness of this robot. And also be able to do to apply pressure and heat to be able to bond the construction elements. So there is an adhesive that’s that has been developed, and that we’re testing in the lab that we can that we can basically glue construction components struts, for instance, together in the space environment, using this specific gripper. And we can also disassemble this bond.  

So it’s a glue, it’s a material, it’s an adhesive that you apply temperature and pressure and it cures. But if you apply temperature and pressure again, you can weaken it, and then separate the components and then go reuse them somewhere else. 

So the soft gripper is what is the gripper that’s providing this capability. And it’s kind of opening up kind of the research field in regards to reusable space components. So once we build a large structure in space, and then when a structure is end-of-life, what do you do? Well, when it’s in low-Earth orbit, it’s typically burnt up. 

But when it’s not in low-Earth orbit, it’s kind of put into a parking orbit with what satellites go to when they’re done, which is arguably still not the best solution, right? Because you just have space junk flying around. And there’s tons of that, or you can reuse. So that kind of opens up that field, well, okay, well, if we have this platform, we can reuse it. If we have to reconfigure it, take it apart and maybe go somewhere else, then that capability is there. Now, there’s a lot more to that than just doing that, that there’s a there’s a cost tradeoff and all that. But that’s the idea. And that’s where the research, the lower technology readiness level research, where we are, kind of we take these ideas and go, “Well, what if?”  You know, “What are these different capabilities? What could we do with this? And what does this open up? And that’s an exciting part of the research because it’s the what-ifs that really come from, like left field, for instance, that are like, “Oh, wait a minute, we can do that.” And then it kind of spins off into something else, which is really exciting. 

Host: And that’s really exciting. I’m actually can’t wait to see what comes out of this. All this research. 

Neilan: Yeah. Same here. 

Host: I bet. Before we let you go, what was your giant leap? 

Neilan: Well, so my giant leap was a series of hops, that I think added up into one big leap. 

Host: Small steps! 

Neilan: Small steps at first. And that really started when I was five years old. And I realized I wanted to work for NASA. Of course, a five-year-old, you want to be an astronaut, right? You don’t wanna do anything else, you know? 

So with that in mind, I, you know, was always involved in the sciences, math and science fields and all that and, and also, you know, really enjoyed philosophy and, and kind of spun those into a plan of action to get into this, you know, the space industry, specifically working for NASA.  

What happened though, is, I’ve always been a tinkerer for years and years as a kid and then in college, and then I decided that robotics was a way for me to have hands-on work in exploration. My giant leap was when I realized that I wanted to use these systems in exploration. I used to be a physics major. So I was a physics major in my undergraduate and then computer science in my graduate. And then I was like, “Well, how do I meld these two?” And that’s that was really like, you know, what I want to explore. And robotics really allowed me to go places that I couldn’t go and I started out doing marine robotics. And I’m still pretty interested in marine robotics because it’s a lot easier to do that than lunar robotics for instance. The payoff’s very immediate, you know? You can build something, go into water, and see what you get. So when doing that I realized that space robotics and autonomous systems was my place. 

Host: We can’t wait to see where this goes. Thank you so much, Jim, for being with us today.

Neilan: Thank you. I really appreciate the opportunity to chat about this work and chat with you. 

Host: That’s it for this episode of Small Steps, Giant Leaps. For more on this topic and on Jim, visit our resource page at appel.nasa.gov. And don’t forget to check out our other podcasts like Houston, We Have a Podcast and Curious Universe. Thanks for listening.