 Thank you very much. I just wanted to say first, thank you for coming to hear my talk and to let you know that there are a few firsts going on right now, so it's my first time being at CCC, so I'm excited to be here. So my first time giving a talk this late at night, so when I got my time slot, I asked myself the question, just how drunk will my audience be? So you can report back to me later about how drunk you are. And then the last thing that's a first that's a little more serious is this is my first time showing work in progress, so I'm a little bit nervous about that, so be gentle with me. So when we start talking about edible soft robotics and this exploration of candy as an engineered material, it was kicked off by just this thought I had while making cast silicone actuators that there's an intense similarity between cast silicone and gummy candies. And it's like one of those idle thoughts you have when you're fabricating something. But I was really captured by this like joking not joking place. I wanted it to be something that is totally ridiculous that I could also take very seriously. And so when I was trying to identify like should I do this project or not, since I've been a government contractor, when you propose a project, you have to like turn in these sheets about what your project is going to be. So I went ahead and made one about making candy soft robots. So this is my quad chart if I were going to try to get the government to finance my project. But for me, it was really about this place in the potential benefit zone that says I see food as the highest form of interactivity and that it that means it engages all five of your senses. And after you eat it, it becomes part of you. I couldn't think of anything more interactive than that. So that seemed like a good enough reason to go down this pathway. So kind of the next thing I do when I have a project is I ask myself, what's the history and the state of the art? I don't want to reinvent the wheel. I don't want to start from zero if I don't have to. So who is doing the work? What are the related fields if you want to make a candy robot? So soft robotics was obviously a starting point. Luckily, it's something I've spent a lot of time thinking about. So I had a good foundation there. And then the next one is molecular gastronomy. So molecular gastronomy is this idea of using food as an art and an experience, using the tools of science to get there. And the thing that was interesting to me about this field is that all cooking is technology. Just some technology is technology we're used to. So for example, a microwave. That was the cutting edge of technology not so long ago. And now it's just a typical tool. So as we look at this molecular gastronomy place, I'm interested to see which things that seem very far out there right now will become part of our regular cooking. If you look up edible robotics, the first thing that comes up is all health-related projects. So this, for example, is a robot that is made out of meat. It's made out of sausage casings. And its purpose is to, if you swallow a watch battery, it can do a lot of damage to your stomach. So this is supposed to go through your stomach and collect the watch battery and then flush it out of your system. Then the next one is kind of the other side of the same problem. So since you don't want to swallow a watch battery, but you might want to swallow a robot, they came up with a battery that's made from cuttlefish ink that you can swallow it and then it gets digested over time. So then to move from the health space to the art space, which is the other people who are making edible robots, there's this project called Living Food. It's inspired by the synthetic biology movement. So the idea was that they wanted to make living food that moved and breathed and like that would be part of the dining experience. Sadly, this was only a theoretical project, so there weren't any technological takeaways I could apply to my work, but it was interesting to consider. Then the other thing that happened is when I was talking to people about edible robots, people kept pointing me back at NYC Resistor, which is the hacker space that I'm part of. So I was like, oh, you know you fell in with the right group of hackers when your research leads you right back to your own door. So the top project is called Lady Godiva, and it's a chocolate that's a closed circuit made with edible silver foil. So when you eat it, the chocolate moans and you have kind of this experience. And then below is a jello piano. So the things about these projects is that they're dealing with sound, but I've always been much more interested in motion. So there wasn't a lot for me to take away from these. And then finally, there was this robotic big project, it was a large collaboration through an art space and a university and a private company. So they were making edible robots that moved. The Rice Krispy robot is Rice Krispy's covering a cable pole system. So I really like the motion and I like the form, but I don't really like that you have to eat around the cables. I was a lot more interested in the cucumber robot because the cucumber is strong enough that it can use its own material to be the cables. So it's like, this is where I want to be in the robotic space, but I didn't really want to eat a cucumber after it, been a robot. Like I wasn't really drawn to that. So that's why I was more interested in candy. So that's the state of the art as we go down this road. Then the next thing I like to do is consult with experts, which is a really interesting question when you're working on an emerging problem, an emerging field. Who are the experts? When I go down this path, I like to go as broadly as possible. Like I might get information I might not use, but the further I go afield into who I think experts are, the more surprising information and the better synthesis of ideas I will get across the problem space. So in this particular case, I consulted with Matthew Bargati, who is actually my collaborator at Super Releaser, and he is actually the person who convinced me to get into soft robotics at all. And he's also the person who told me that I had to make candy robots once I said it. Because I said, well, it doesn't really fit our soft robotics agenda. And then he said, part of our soft robotics agenda is to be awesome. So you have him to thank or blame about this project. And you can watch his talk from last year's CCC about soft robotics in general, if you want to learn more about the general field. The next person I consulted with is a friend of mine, Tim Rodriguez. He's most known for being a board game designer, but he also has a degree in food science. So one of the things that was really useful to talking to him about food science is that he took me down the pathway of looking at candy at the molecular level. Because it's actually all of these very, very simple systems in terms of the chemistry. And he had some great insights too about the relationship between why candy is sometimes not a very strong material and why vegetables are strong. So that's something that will come up again a little later. And then my friend Liz Hara, she is a professional puppet builder. I have worked with her at the Jim Henson Company and other places. And she is also an extreme candy enthusiast. She has eaten every kind of candy that is mass produced in America, which is quite a feat. And then one of the other things that's been really wonderful about talking to her is because she made edible puppets once. So she gave me this fabulous insight. I can tell you stabbing licorice into a hot dog does not make a durable puppet. This is the voice of experience. And then the other thing that's been so great about working with her is that she's very enthusiastic about the candy actuator space. So she's been texting me more information even while I've been here. So now that I've talked to the experts, I needed to bound my project. I needed to decide what the evaluation criteria was. Whether or not I was succeeding at making a soft robot. In this particular case, I decided you have to be able to eat the whole robot up to the computer control. As I said before, I'm not really a fan of the eating around cables. Also, it needs to be easy to reproduce. The primary reason that I'm doing this project other than to be awesome is to make a how-to to put into a book about soft robotics that I'm co-writing. And so what I decided easy to reproduce looked like was that it could use special specialty ingredients. You could have to go far afield to find the right things, but you shouldn't have to buy specialty equipment in order to do this project because I don't want people filling up their lives with all of these things. They're only going to use once or twice to make a candy robot. The other was that it had to be not too fussy to make. When I talk about fussy to make, I don't know if any of you have ever made a meringue, but it's incredibly sensitive to the humidity and you have to adjust everything very carefully. So I wanted the robot to be something that it wouldn't be temperature or weather dependent. I wanted it to elicit an emotional response that was strong. And for me, when you deal with candy, you either get two directions. You get this joyful space, this fun and happy space, or you get something that's really gross, like snot or those jelly beans that taste like vomit. So those were the two most exciting places for me in the edible robotics place. And then I wanted there to be a baseline-fit flavor standard. So I didn't want it just to be technically edible. So a lot of people I talked to about this were like, well, you can eat fabric or you can eat paper. And I was like, I think that's really not in the spirit of the project. Like I think that there has to be a baseline deliciousness that makes this an appealing thing. And then it needs to be durable enough to make the day before. I don't like the pressure of having to perform making something to the last minute right before I serve it to someone, especially for something as complicated as a robot. So I wanted you to be able to make it the day before and then serve it. And then also, there was no need for long-term stability, though, because once you get into the place where you're making a food actuator, long-term storage just sounds like dusty, dirty, handled by a lot of people, which then went away from kind of the delightful edible component I was looking for. So those are the specs that I'm going to measure success against. So now, since I got into technology, kind of through being a fabricator and through my experience working with materials and like direct relationship with the material world, it makes the most sense to me to spend time with the materials and to touch the materials and to experience what they do. And aside from it being my comfort zone, the other thing I wanted to bring to the process was that I have seen many, many people build projects out of inappropriate materials. And you can see that they're spending a lot of time over and over again trying to make something do an object that it cannot achieve. So because I don't know anything about candy as a engineering material before I started, I needed to spend time with the materials to know what appropriate use cases were. So first thing I did was start acquiring some materials. So Calloustians, the place to buy hydrocolloids in New York City. I'm going to say hydrocolloid a lot. That's what makes a gummy candy, a gummy candy. So just so we have a definition to start with, a hydrocolloid is a stable suspension of hydrophilic molecules. So most of them are sugars, but some of them are proteins. So for example, gelatin's are proteins and they're a hydrocolloid. So moving on, I acquired a bunch of hydrocolloids. Now what I didn't realize when I started down this pathway, gelling of hydrocolloids is a foundational problem of molecular gastronomy. I thought that I was going into a space where everything was going to be mostly solved. But what I found instead is in this book, this is 22 different options for gelling agents. So there's a lot of ways to achieve this outcome. And I'm actually going to read directly a quote from the book because it made me laugh. Unfortunately, relatively little theory exists that can help cooks know which gelling agents or coagulants will have synergistic effects. We can recommend several good combinations based on trial and error in our kitchen and others, but more information would be helpful. There are many possibilities out there awaiting discovery. Now the reason why this quote is funny to me is because I work in soft robotics. And essentially, if you swap out kitchen for lab and like one other noun, it's exactly what you would read in an academic paper. So I started on my hydrocolloid path by doing pectin. Now pectin was a good case because it actually illustrated what doesn't work. It's a hydrocolloid that is made of monomers that are very dense and very small. And they only have a single point where they join to the other ones. And because there is no motion within the monomer and then there's only the single point of connection, this makes a gel that's incredibly brittle and breaks apart under any kind of snapping force. So it's totally inappropriate for soft robots, but it is very good for learning about hydrocolloids. Number two iteration or number two material exploration was I burned the gelatin. This was horrible, it smelled awful. And it was also really interesting because it reminded me that I was working with a protein. I think that we don't really think about how the materials within candy are things that are kind of surprising if you pick it apart a little more. So I made my third version, my third batch was too much gelatin. And this was yet another reminder that I was working with an animal product because the outcome looked and smelled like and taste like skin, which is not a desirable outcome in candy. So now we get to some slightly more successful gels. So this is Jell-O-Brand gelatin. It was a good compromise between flavor and ease of handling because it already had sugar, it already had some measurements done for you. But this is an example of an elastic hydrocolloid. So what happens is collagen is a tri helix form. So it's inherently very, very stretchy as you can imagine the space between the three components of the tri helix stretch out. Well, when you make gelatin, it breaks into these single strands. So they're all like a free floating spiral. As you treat the material, the gelatin, as you heat it up to a certain point, you actually start making these bonds again and you start getting tri helix sometimes, but not all of the time. So you end up with this combination where the tri helix parts are very stretchy still. All of the free floating ends that didn't make bonds end up in this tangled mess that a chemist told me is a lot like spaghetti. So when spaghetti is really hot and moist, it slides out and you can pull some strands away. But as soon as it gets cold, that tangled mess is stuck together and becomes a very, very solid thing. And so that relationship between the tangled mess and the tri helix is what gives gummies. They're like very, very signature texture. So this is an experiment that I made out of gelin. Gelin is a different kind of hydrocolloid. It's a plant-based. It actually is a bacteria that grows on the bottom of lily pads in nature, which is pretty interesting. Now the great thing about gelin is that it makes this extremely stretchy material, but at the same time it's very tender and soft. And when they were talking about synergies between hydrocolloids, it's often combined with carrageenan to give it a stiffness. But the thing that's interesting about mixing hydrocolloids is that you don't get an average between the two. When you start mixing them, you get entirely new behaviors that aren't reflected by the individual. So that is a future experiment. So I went back to the Jell-O brand gelatin recipe and then I was starting to be like, okay, this is working a little bit. So I'm going to go ahead and cast it in a soft robotic mold. So this is dusted with cornstarch. Traditional gummies are actually made in cornstarch molds. They press metal shapes into cornstarch and then pour the gel directly into it. This is not food safe, so please don't do this in the future. So when I do future robots, I'm probably going to make silicone molds, but this mold was just one that I 3D printed. So then my seventh experiment making gelatin base, the making hydrocolloid ballast gels, was I used what they called the Harry Bowel recipe. I don't know if that's how you pronounce it here. But the thing that was really interesting about this one was that they sub out regular white sugar for other sugars. You use glycol and sorbetol because it changes the moisture amounts in the mixture and you can eliminate some of the water when you switch the sugar source. So unsurprisingly, this one was the most delicious one that I made. And I think part of that was that it has added acid. So you can use citric acid or you can use, I can't remember the name of the other acid, but a lot of different food acids can be used, but it really improves the flavor profile of your candy. Oops. So in the future, I have some plans for testing the assessment of these gels and I didn't even get to it because the gel problem was so much deeper than I anticipated. So I'm planning to test the durometer, the elasticity, the flexibility, and the durability of all my future iterations of the candy making. So in the store-bought material side, I went to Economy Candy. I bought a lot of different candy. I spent way too much money on candy. Also, I learned don't eat your mistakes if you start working in candy robotics. I ate three iterations and then I didn't feel so well. So I'm not a purist when it comes to soft robotics. I also love hybrid robots. So I did some tests where I drilled into chalky candies to see whether or not they could be spacer components, which successful, they were totally clean and had durable holes and I could drill in a variety of sizes. Inspired by the Octobot, which is a soft robot that runs a liquid across, I think it's a platinum plate and that creates gas. I was like, oh, maybe I can make an explosive gas actuator by using pop rocks, but you can tell that it does not have any explosive action at all. Nothing happened, just a very slow fizzle. So if I want to make a chemical reaction robot, I'll probably go to Mentos and Diet Coke because the internet has really well documented that that works well. Another thing I tried was buying pre-purchase candy and then heating them. So one of the things that's interesting about this set of hydrocolloids that I'm exploring is that they are not single state, like one time reactions. They actually can revert, so you can heat them up and cool them down and it will cause the same chemical reaction set over and over again. So actually there's lots of do-overs if you're working in gummies. And so these ones were microwaved and I thought that would be an interesting option if people didn't want to make candy robots from scratch. And then fruit leather. I was inspired by the fact that there's a lot of soft robots that are made out of airtight fabric that doesn't flex but just holds air. So this is me using a compressor to pump air into some fruit leather. And it also made a pretty good seal just with its own stickiness, which was a surprise. So I said something before about vegetable matter. A lot of the things I used for pulling force that were made of candy didn't hold up to the amount of force I needed to cause the actuation. So I tried dried mangoes, which are excellent. You can, they fit the taste profile. They go well with candy. But you can tie it in knots. You can bind things off. You can use them as cable controls. Excellent material for this space. And then also there's these licorice vines that are hollow so you can pull things through them. They're also airtight so you can use them as tubing. So if you wanna make a pneumatic system. So I'm gonna move on now to, my next plan in this process was that I was going to do an iteration of a wide field. So usually what I do is I take all of my promising materials and I'll do a single iteration of each of the promising materials. But the thing is that in this case, I ran out of time. The gel problem was so much bigger than I ever thought it would be that I only got through one iteration so I wanted to let you know that right now. So in addition to iterating on all the promising materials, I wanted to do one iteration of each control method that is popular in soft robotics. So that's cable controlled, pneumatic, hydraulic and chemical reaction. The only one I got through is cable controlled. But this is my first iteration of a cable controlled robot made of candy. So ta-da, it kind of works. And then when I do my first iterations, I have to go back and assess what worked and didn't work. And one of the things that I found about a cable controlled candy robot was I wanted a lubricant. So I tried oil, oil was terrible. It was a solvent. It worked a lot like water. Everything started breaking down and getting stickier. And then I was like, oops, of course, sugar dissolves in moisture, should have thought of that. But sometimes when you switch from one domain to the next, the things you know don't transfer with it. I tried corn starch because I was like dry lubricant seems the way to go. It got a little gummy as it rolled. So it was good short-term lubricant but you couldn't have high cycles. With confectioner sugar, it was the most promising. So confectioner sugar is a mixture of very, very fine sugar with corn starch. The second iteration would be a narrowed field based on my first set of iterations, TBD. And the same thing is true to getting the final soft robot. So repeat iterations till your specs are met. No excess and iterate. Yes, you're done, TBD. So the thing though about this whole conversation is I really wanted to take the risk of showing work in progress and showing unfinished work. And the reason why I wanted to show unfinished work is because it incentivized me to show you my failures. I think that innovators spend a lot of time talking about how we're willing to fail. But what happens is you either get all the way to the end or you give up on the project and there's not any airtime for what your failures were. And the other thing I was interested in, why I wanted to give this talk is because I'm talking about candy robots today but my deeper interest is in systems and processes. And particularly how a robust process can give you a framework in which you can succeed on something that's never been done before and also how being in a good process means when you get stuck, there's a way to correct for that. There's a way to get unstuck. For me in this process, when I get stuck, I can always go back to an expert or find a new expert and it means that I never hit a bottleneck where I stop innovating. So hopefully people will think about processes for working on emerging technologies that are from their toolkit. I'd love to hear about that or use this one to work on your own problems. Some final thoughts. I think design thinking and other processes yield effective pathways through the unknown. I think people in general should work on emerging fields wherever you find them. It's where you find the most exciting problems and the most interesting opportunities. And then in general, I want people to reach out if you wanna share or collaborate about edible robotics. I could use more help in chemistry. Control systems for soft actuators is a juicy problem or you tell me, maybe you're an expert in something related to this that I haven't even thought of yet. And also please reach out if you wanna talk process. So this is me and where you can find me on the internet. And if you wanna talk to me about the other things that she introduced me before, I like to talk about soft robotics, spacesuits, costumes, and puppets, so. Thank you. Thank you.