What Can Intelligent Materials Do? - with Skylar Tibbits

[Music] yeah thank you so much it's a pleasure to uh be here talk with all of you i wish i could be there in person uh but it is what it is these days so i have a few main roles one of those roles is the editor-in-chief of the 3d printing additive manufacturing journal and this is an interesting perspective because it's a place where all things additive manufacturing the future of fabrication and the latest research come together so that's from you know art and design to material science computer science mechanical engineering etc and you'll see how our research is really connected with that academically i run a undergraduate program at mit in the department of architecture called the the art and design program so it's a design minor and design major for undergraduates and we get the best and brightest of mit undergrads uh that then also are interested in becoming polymath designers that can you know go off and create hybrids between computer science and biology and physics and math and engineering and and all things creative in terms of design but what i'll talk about most is our research at the lab the self-assembly lab um and where i would like to start to type to sort of dive into that is is ask a question you know what do we all think when someone says smart products you know we often hear that in any domain we hear smart shoes or smart cars or smart homes and that typically is associated with something like a nest thermostat for example when you hear smart homes or smart buildings that's typically a device centered approach to what we mean by smart similarly when we hear future of manufacturing or future of construction we directly go towards robots and i want to challenge that today i want to re-imagine what we mean when we say smart and what we mean when we imagine the future of making things that shifting away from the notion that robots and computers traditional forms of computing are smart that we need more batteries and motors and complexity and components and cost and often more failure in order to make something smart because we're typically thinking about this if something is quote unquote smart but it costs more and it fails more often it's not really that smart you might as well go with the dumb thing so we're always up against that how smart is the thing that we're talking about and i would like to shift our attention to materials simple everyday materials in our built environment that respond to each other that can assemble themselves self-organize respond to the environment sense actuate transform arguing that materials are true intelligence and that all of us as humans and all of the natural world around us is built out of materials whether those are cells or whether those are rocks or sand materials are the fundamental ingredients that build up capabilities behaviors agency and ultimately intelligence it's not computers and robots that we've become more recently accustomed to and so we're really shifting that definition and questioning it and that was really the impetus or the the recent book that you just heard about um it's really a guide to this new materials revolution there's lots of research and emerging developments happening in materials but more than that it's it's sort of changing the game and it's describing an ethos about a new relationship with materials in the environment in order to challenge or tackle some of the biggest issues that we face whether that's you know climate change or sustainability uh it's not about some holy grail new material invention it's not about some holy grail crazy new robotic technology it's about a relationship a symbiotic relationship with materials and our environment but how did i get here so i have a background in architecture and at the time i became very fascinated in software how computation and software became a new tool for design this is the first cad tool sutherland sketch pad developed at mit that then led to all sorts of sophisticated computational tools computation is basically everywhere in every discipline and it is no different in architecture and design that software and code specifically became a new language to design computation became a new way to imagine things a new way to draw a new way to analyze to simulate can look at every discipline from biology to economics to construction architecture and software and specifically computation has changed the game and there's a similar path in fabrication this was the first time a computer was connected to a milling machine in 1955 at mit and that led to digital fabrication so 3d printing cnc routers robot arms waterjets you name it so computers connected to machines in a way code became a new way to design and make so we can now imagine things we can computationally generate simulate things through code and we can send code to machines to automate the fabrication of parts and at this time i was building all sorts of generative installations doing just that i was writing code developing my own tools working with sophisticated software tools to create complex geometry sending that code to machines fabricating installations in different galleries around the world and you can see this in contemporary architecture you know the most uh well-known architects of frank gehry or zaha hadid or morphosis or many others use these tools these sophisticated software tools and digital fabrication tools to make what we all call mass customization and you can see that from 3d printing to laser cutting to waterjets where every part can be unique and you get free complexity supposedly and so we were doing this and we were making complex geometry solving those geometry problems computationally complex parts fabricated to make these complex installations and architectural structures but what we weren't talking about is the last mile problem which is an assembly problem so essentially by doing this you've created a big jigsaw puzzle you've created a logistics nightmare there's the blood sweat and tears that most people don't want to talk about you know how much energy people time in this case student labor went into assembling these very very complicated things that we designed so it became very clear to me that we have code for design we have code for fabrication we need code for construction what is the third part of this how do we reimagine the way that we assemble things not by part by part you know humans or robots assembling things from the top down but can we actually embed code into materials and the first way that i started doing this as a graduate student at mit i was in neil gershenfeld's lab the center for bits and atoms and we had a darpa grant so darpa is the defense agency it's sort of the the latest and greatest sort of almost james bond like technologies in a way and darpa is you know famously one of the uh initiate initiators of the internet for example and historically they've developed all sorts of wild far out sci-fi technologies so there was a grant called programmable matter and this was about a decade ago and i was working on this as a grad student and there was a number of other researchers and we were building quote unquote programmable matter but at that time about a year ago programmable matter or about a decade ago sorry programmable matter meant robots so robots at the smallest of scales and robots at the largest of scales and we were building them from millimeters to centimeters up to meter scale objects i was working on the largest of scales because of my background in architecture but i started to become critical of this you know at the time the dream was if you could just make everything out of robots you could program program everything like we do with robots and they could change shape they could change property they can go from one thing to another thing but that dream starts to fall apart especially when you're thinking about scale so look at it from an architecture perspective we don't want to build buildings out of robots because you don't want every brick to be a robot it'll be too expensive it'll fail too often it'll be too hard to assemble and it'll be really energy intensive so it's it's really a bad idea to build robots to build buildings out of robots and their scale in terms of physical scale and their skill in terms of numbers and neither one of those particularly works with robots so i was really fascinated in this and it became clear to me that the future of programmability the future of construction is not necessarily in robotics but it is in embedding information into our construction materials if we could have more agency more capabilities and more information in our physical materials then those materials could collaborate with us they could self-assemble just like biology and chemistry do all the time or they could sense and respond they could go from one shape to another they could transform they could have all of the smartness that we've all dreamed of but they're in simple materials so that is really what became the foundation of our research at the lab and now over the past decade has been many different projects and collaborations i put this slide here just to highlight that one thing that sets our lab apart is the scale that we work at we're really interested in the human scale world whether that's construction or manufacturing or product distribution assembly logistics we're interested in this messy macro scale human scale world there's lots of researchers in labs out there looking at very similar topics to what we're talking about with self-assembly self-organization programmable materials at very very small scales from synthetic biology for example or material science there are many researchers that are pushing the boundaries of what's possible at small scales one thing that sets us apart is the scale that we're working at we're particularly interested in these human scale processes but i think this is really important because if you ask most people they would say that we're pretty good at building things as humans you know we've figured out how to build rockets and fly to space or we can build cars or planes and fly around the world or you know we're pretty good at making things so why innovate here and i'm really fascinated by this question because if you look at every other scale if you look at the smallest of scales from nano scales or maybe even smaller you go all the way up and then you skip over the human scale for a second and then go to planetary scales everything is designed and built from the bottom up like there's no planetary 3d printers there's no one dreaming up the design of a planet and then fabricating it telling people or robots or machines to build this planet that i designed same thing goes for biology and chemistry there's no one that's saying okay here's the sledgehammer to put together your dna and make this human design functionality and construction emerge from the bottom up materials and their environment work together and all of those things come together to self-assemble structures but that doesn't really happen at the human scale at the human scale we predominantly use top-down design genius designers or code generating ideas and sending instructions to people or robots to build things so everything happens from the top down and i became very fascinated in that question why doesn't design and construction happen from the bottom up and that really is sort of the impetus of our research and what i'll try to talk about today so i've said it a couple of times one of our main categories of research is programmable materials and this is the idea that we take physical materials and we embed more capabilities more information in those physical materials so that they have the ability to sense and respond to act as sensors and actuators to physically transform based on information embedded in those materials so you could think about that as dna or i'll talk a little bit later about wood you can think about the grain and wood as a sort of natural code materials have properties those material properties can be linked with activation energies so wood swells with moisture or if you have temperature that usually activates metals or polymers and we can fabricate those material properties now with all sorts of new digital fabrication technologies in two-dimensional and three-dimensional complex structures with different materials in different configurations than may not have ever been naturally found before we can design and fabricate material composites that have the ability to transform they can sense some change in their environment whether that's temperature pressure wind waves sunlight moisture whatever it is and they can transform in very precise repeatable ways so we can design a simple material that then has information and agency to transform and that's what we mean by programmable materials and all of this started with our work called 4d printing so 4d printing came right after those reconfigurable robots i was talking about um and the back story here is you know we were starting to think that it's really exciting we can make these robots that transform but it's really quite frustrating how much we spent to make them and you know maybe they worked for 24 hours and there's lots of energy and lots of problems with this it doesn't scale so how do we do this in another way and it was right at that time that new technology was emerging called polyjet printing and so we started collaborating with stratasys on this project and essentially what we do is we multi-material print so with different materials we print in two dimensions and three-dimensional structures so depositing different materials in different geometric places exactly like that diagram i was just talking about with different materials they can be activated by different activation energies so what you're seeing here in the joint the clearish material is a hydrogel so we deposit just the right amount of hydrogel and the right geometries so that when it's activated by moisture it transforms from one shape to another shape and so then we started printing all sorts of things and dipping them under water so a line can transform into a cube or a line into the letters mit or a surface that folds into a truncated octahedron or a flat sheet like a record almost that folds into a hyperbolic paraboloid a saddle like shape using curved crease origami or a flat sheet that can expand and contract to create double curvature and this is a really big problem you can look at the automotive industry or you can look at the architecture space taking a flat sheet and creating double curvature so complex curvature without molding and forming is very very challenging in most industries and in this case we showed a flat sheet that could be easily printed with these two materials then activated with moisture and transform into a precise shape and it can do that repeatedly and so after this work you know the reason we called it 4d was because we wanted to add the element of time that there's this boom in 3d printing capabilities specifically in that case polyjet printing multi-material printing but they were typically printing static cold dead things you know just a another part that competes with traditional manufacturing in our case we wanted to print highly active things things that transform over time and so that's why we added the fourth we said 4d because we want to print things that transform and reconfigure over time and this really caught on a lot of researchers then started to doing research around this and we continued our work but many industries started to approach us and said this is great but we don't particularly care about 3d printing and we don't particularly care about plastics can we do the same thing with some other material can we do this with leather and wood and textiles and metals and fibers and you know many other materials out there and can we do it with other manufacturing processes and so that's why we then broaden this and we're much more interested in just generally how do we program materials what materials can we work with and what manufacturing processes and we've tried to demonstrate that this is a ubiquitous principle and we can use this across many different materials and many different fabrication techniques and we've collaborated with a lot of researchers and designers and software companies and applications companies over the years to try to develop this and so this has now become a major part of our our research so i've hinted at the legacy of wood so wood is obviously a natural material and has for a long time been used as an active material you can look at japanese joinery or ship building or even alcohol barrels and all of these examples they would the the master crafts people would understand the inherent knowledge of the material the grain the anisotropic property of wood so non-homogenous property to swell with moisture in order to make very precise strong even waterproof structures by collaborating and tapping into the properties of wood more contemporary example eames furniture taps into the the behavior of wood but forces the furniture in shape into shape so they use molds and they steam bend plywood to make the furniture which is really novel and beautiful and to this day is an exemplary piece of product design but doesn't actually tap into the agency of the material to transform itself and then a more contemporary example akimenges of the university of stuttgart and his researchers have shown amazing work where they tap into the wood grain of wood veneer changing the relative humidity in the environment and the wood opens and closes repeatedly based on that change in the environment but the challenge there is they're limited by the wood grain that you can find so what we started to do was to create our own wood we started to mix the wood and polymers together and extrude that and in the beginning we were making our own and then you can now order wood filament as a way to print with a material to print with and we print our own wood grain so we take the natural property of the wood to absorb and swell with moisture but we have designed freedom to design our own patterns so depending on the printed pattern of the wood grain you can get wood that behaves in different ways you can get curling and twisting wood for example depending on how you print it or you can get wood that folds to 90 degrees or you can get wood that transforms into furniture-like shapes and goes from a flat sheet into something like fish scales for example wood that you wouldn't naturally find in the forest converging and diverging grains but we have complete design freedom in order to make that and tap into the collaborative nature of the the wood in order to create that agency more recently we've collaborated with the product designer christoph guberon on these flat sheets and what's interesting about the flat sheet so here it's printed wood in the exact same way that you saw but we specifically made it flat because one of the challenges with traditional 3d printing is that the the taller you go the slower it is so the more layers the more support material the slower it is the less feasible it is for manufacturing so if we print flat almost like two and a half d printer then we can take advantage of the speed and customization and we can put it in something like a zip lock filled with moisture ship it anywhere in the world and on the other side it then transforms into these bowls and baskets so there's no manual assembly the shipping becomes the activation all the intelligence is embedded in the materials and it transforms into a 3d structure that would either be impossible or very time-consuming challenging to print as a three-dimensional object another material that we've explored is composites so carbon fiber or fiberglass or kevlar we collaborated with a company called karbatex and they make fully cured but flexible carbon fiber so it's really innovative in that it's highly strong it's lightweight but it's fully flexible so we can make active carbon fiber in this case by printing on it or you can bond or laminate to it basically we add different polymer polymer layers to the composite similar to the wood we create different grains and that grain based on the relationship to the printed polymer and the woven structure of the composite creates different behaviors when you change the temperature in the environment so this is carbon fiber that can fold this is carbon fiber that can twist and that's only the behavior changes only based on the pattern change that we printed it's the same exact material compositions so after that we started collaborating with airbus on a component for their engine so at the top of the engine there's this air inlet that brings in cool air but it causes drag so typically they would make an electromechanical actuator hydraulic or pneumatic actuator but that adds more weight it adds more cost and more potential to failure so you reduce efficiency by solving an efficiency problem which is not particularly good so we started developing this composite component that goes into this air inlet and it can open and close based on pressure differential so the faster or slower that the plane is flying creates a pressure differential inside and it can then go from one shape to another shape to open and close this air inlet controlling the airflow to the engine without any electromechanical hydraulic pneumatic components actuators etc it's just a material composite that has that embedded capability there's many projects over the years that we've explored around textiles so the first one was using a pre-stretch almost like a jack-in-the-box behavior if you stretch a textile like a lycra for example you stretch it around a plate you're embedding essentially potential potential energy into that you're you're basically winding it up like a spring and then you can add a material on top so you can print or you can bond or you can spray or stitch or embroider and you add a material that is essentially a constraint so when you release it in this case when you cut it out a circle jumps into a saddle shape a hyperbolic parabola so it goes from flat to 3d simply by the the stretch the force of the stretch and the pattern of the printed structure so that constraint pattern dictates the structure that you will get like pleating or tufting or rolling or curling then more recently again with kristoff the product designer we were asked to exhibit uh in the future of the shoe so in this case we were trying to challenge the manufacturing of shoes so typically if you look at your everyday shoe it's made of tens to hundreds of components that are manually assembled and there'll be many different materials and there's glues and stitching and forming and forcing these shoes together and they're very complex shapes and fairly complex assembly so we were showing that if you stretch the textile and you print this precise geometry when you cut it out it then jumps into the shape of the shoe with a single composite material you can get the full curvature of the shoe without any manual forming stitching gluing sewing etc then more recently we started looking at film so thin films that we could slit into these strips and when we recombine the strips you can make essentially macro scale textiles and these films are highly active with roughly 30 to 40 celsius or direct sunlight they can transform in seconds and they're totally reversible and we can make two-dimensional and three-dimensional structures in this case it's something called auxetics so typically when you have a material and you stretch it it'll shrink in the other dimension oxetic materials do the opposite when you stretch it it expands or when you compress it it compresses uniformly in all dimensions and that's a really fascinating phenomenon and we can create that with these materials so oxetic structures that transform based on temperature change or sunlight and this was the beginning of our work to construct our own textiles so right after this we got this a grant through an organization called afoa advanced functional fibers of america so obama had created these manufacturing institutes one on 3d printing one on composites another one on fibers and textiles and so we got a grant to do exactly that but creating our own textiles from scratch through different materials and fibers and so we started collaborating with a company called ministry of supply and we're trying to challenge this issue with clothing which is that you might go inside when it's super hot outside and it's air-conditioned inside or vice versa it might be super cold outside and it's really hot inside and really the only thing you can do is just add more layers but your clothing doesn't adapt to your body change and it doesn't adapt to the environmental change so what we did is we made our own fibers these are multi-material component fibers with different cross sections different ratios and colors between the material in the fiber the fiber is then twisted into yarns and then we see and see knit so we can actually change pixel by pixel in a three-dimensional structure the material and the fiber and the color every single pixel much like 3d printing but with knitting and this is how you get nike's flyknits or adidas knits or any of the knitting technologies is through this industrial knitting machine where you can change the fiber pixel by pixel basically stitch by stitch and so we do that in the same way just like the 4d printing we change the material properties we have our own fiber that has specific material properties and then we design the knit structure and in order to have different behaviors so we want garments that adapt and transform with your body so porosity change to make it more breathable for example or thickness change to make it insulation so hot or cold you could have a down jacket that then goes to a windbreaker for example depending on the environment around you gills and slits for ventilation so we want different functional mechanisms in our garments that then transform and we found ways to make this through industrial knitting and our own custom fibers mixed with off-the-shelf fibers to create these adaptive behaviors so all of that research is on materials that everyday materials you know fibers metals foams wood leathers plastics and we're trying to embed more capability more agency more information but all of that was really about shape change how a material that is in a simple everyday material can transform its shape based on its relationship to the environment another category of research that we study is on phase change so how things go from solid to liquid and liquid to solid may or may not be about shape change but we're really interested in this liquid reversible liquid transformation and up to this point you've heard me talk about 3d printing a number of times we then started to collaborate with the furniture company steel case and we were trying to challenge the issues with 3d printing so lots of people talk about 3d printing but there's this hurdle to get over it towards true manufacturing and scale and there's three main problems the speed it's too slow the scale it's too small and the material properties are really not up to par for standard material manufacturing industrial products so we were trying to adjust address that and if you think about furniture really 3d printing isn't connected to furniture today and and steelkas was saying 3d printing is great but we can't print a chair so that became the challenge for us how could we print large scale things very very fast with high quality industrial materials so we developed our own printing prop process called rapid liquid printing so essentially it's a large vat of gel the gel is 99.9 water uh it's safe you can wash it down the the drain but the gel basically eliminates the effect of gravity so you can print in three dimensions as if there's no gravity so you're it's like calligraphy you're drawing in three dimensions you can go faster or slower you can change the extrusion rate to change the thickness and then the material cures within the gel so the reason this is faster is because we can print without support materials we can print in three dimensions directly we can print very very large scale as fast as the machine can go and we can print with high quality materials like silicones to make inflatable structures or soft robotics for example we can print with off-the-shelf standard industry materials not special 3d printed materials this is a project we did with bmw looking at the future of the car and we were really interested in combining all of these functions from lumbar support to structure to crash protection all of the mechanisms that you might see in the interior of car combine that into one material that can morph to the shape of your body for comfort could change stiffness perhaps to to make it softer harder it could create massage features and lumbar support it could maybe even do crash protection like airbags so one material composite can morph its shape and function and this was only enabled by being able to print this fairly complex multi-chambered structure based on this this liquid printing process we've done many other projects around this technology from chandeliers and wall sconces and table lamps and toothbrush holders to shoes and bags and all sorts of things and then more recently we've spun this out as its own company around rapid liquid printing and they're developing this into a commercial system after looking at that printing process we realized one of the challenges is high temperature materials so so we could print you know chemically curing materials or low temperature so then we developed a similar process where we print with molten metal so high temperature metals literally liquid metal that we print within a powder bed and the powder bed is very much like the gel bath it supports the material so there's no support and this is even faster so we print because the metal cools in seconds we can print large scale structures on the order of meters very very fast in in seconds to low minutes it's then almost instantly solid so you can make these large scale metal structures like baskets and and furniture pieces and what's most interesting is that it's a hundred percent recyclable so everything you print can then be melted directly back printed again over and over and over again so it's almost like ctrl z you can undo print reprint and there's no waste in that process so we're really interested in this solid to liquid liquid to solid transformation another phenomenon that we study in this phase change category is a principle called granular jamming so granular jamming is like when you go to the beach and you jump on the sand and it's really hard but when you play with the sand it's really soft or if you buy coffee at the grocery store and it's a vacuum packed seal it'll be really rigid but when you open it you can pour it out like a liquid and that's because the particles get stuck they jam and they have nowhere to go and when they jam they act like a solid but when they have air space and they can move around they act like a liquid so this is a way to go from solid to liquid and liquid to solid without temperature change and it's infinitely reversible and almost instantaneous so there's a lot of other researchers that have explored this and we did a project with google around acoustic tiles and we did a project with steel case around reversible packaging basically a bladder filled with particles you pull a vacuum on it wrap it around a product like a table for example it then acts like styrofoam you can ship it but on the other side you can release the vacuum wrap it around a chair and now you have reusable styrofoam that's not just for one product it can be used over and over and over again but maybe more interesting and interesting thing than that we did a project with the group at eth zurich or matsuya kohler's lab where we were interested in using granular jamming for construction and there's a lot of advantages but right off the bat we saw two main challenges so one of those is that you can't have a bladder and you can't have a membrane so a bladder is susceptible to being punctured and the you know the membrane is is typically you have a vacuum pulled on it and so that's super energy intensive so you can't have that you can't be pulling a vacuum constantly so we need to get rid of those too so we developed a method of using simple rocks and strings so fibers and granular material can promote jamming in 2015 we did this project in chicago where we built this yellow bounding box and inside of that we laid rocks and string rocks and string over and over and over again and the robot is spinning out the string in very precise places so that when you remove the bounding box only the rocks with the string jam and all the rest of the rocks fall away so the string takes tension the rocks take compression and they get stuck that it acts as if it's like when you jump on the sand it's solid and then at the end of the exhibition we just wind up the string and the entire tower then disintegrates so you went from a pile of rocks and a spool of string to a load-bearing uh four meter tall column then to a pile of rocks in a spool of string so we can build architectural load-bearing structures that are totally reversible as if this was like reversible concrete we can build and disassemble build and disassemble so what's interesting about this is that concrete is great it goes from a solid to a liquid or goes from a liquid to a solid but it's not reversible there's a lot of waste and there's a lot of time spent waiting for that this is instantly solid totally reversible and more recent we we had a darpa grant and then an air force grant looking at using this for modular construction so we built a column like an architectural feature that then we rotated and turned into a beam or a small bridge for example that then you could walk across and the reason that we can do this is that we add compressive loads so basically we build the column fibers and rocks the same way we compress it it acts like a solid but then when we release that compression it goes back to a liquid and it disintegrates and so it's switchable you can then decide when it's a solid and you can decide when it's a liquid simply by controlling that compression so here we built a wall turned it into a slab and then disintegrated it and maybe most interesting we built a column that then we rotated again to a beam and we compressed it off axis and by compressing it off axis we showed that there's also this semi-solid state so it's solid you can load bearing you can walk across it but you can continue to morph it until eventually it disintegrates and collapses so there's a really interesting liquid semi-solid solid and in the middle it's actually sculptable and morphable so it's a really interesting phenomenon with just a dumb material like rocks and string okay so the last research category that we explore is around self-assembly as you would guess since our lab's name is called the self-assembly lab we're not only interested in materials that can transform their shape or materials that can go from solid to liquid but we're also interested in simple everyday materials that can assemble themselves go from one disordered state into an ordered state without humans or robots how can materials literally build themselves one of the first projects we did was a collaboration with arthur olsen a molecular biologist and we built these biomolecular structures the macro scale models of them this one's based on the poliovirus they're in these glass flasks you shake them hard and they break and you shake them a little bit softer and they come back together so they literally self-assemble you can be blindfolded you can be a little kid you don't have to know anything about this you can just pick it up and randomly shake it and it will always self-assemble if it's given just the right amount of energy not too much not too little just the right amount it'll then assemble and this was really fascinating to us to to our knowledge was one of the first demonstrations at a macro scale that self-assembly is not just a biological phenomenon or a small scale thing this can happen really at any scale and all of the blueprints are in the materials not in the human so the human doesn't need to know anything about it the parts are what are capable of assembling themselves but this was you know a dodecahedron structure and so there were some questions on could you only make repeatable things with parts that are the same so then we looked at a different environment we went underwater every part here is unique so everyone has a different geometry and the the nodes have different lock and key mechanisms and every part has only one place in the final structure so it has to find its neighbor and self-assemble to make the chair and the chair is interesting to us because it's a human arbitrary thing it's not that chairs just you know emerge from nature this is a human creation so it's top-down design but we can still use bottom-up assembly to create these arbitrary designs with very precise uh and arbitrary parts that have you know unique connections to them we then were interested in scalability in terms of numbers and scale in terms of size so we went to weather balloons so these are meter diameter weather balloons filled with helium and there's velcro nodes so each one of these tumbles around and we we throw a party we had people there pushing these up in the air we have fans creating this turbulence you need effectively brownian motion that they're moving around bumping into one another uh error correcting so you know weak local bonds break off stronger and stronger bonds build more successful structures and they start to connect and they build these lattice structures so over the course of the day or the weekend depending on how long you run it you'll get different structures so these are cubic lattices there's not one precise shape like in the previous examples but they can build many different shapes so they can build cubes and they can build these beams where they can build larger lattices so these are cubic lattices that could scale in any dimension it was built in this courtyard so it's constrained they can't just float all all around we can control the amount of energy and sort of guide you know if you need more units etc but we were really interested in this question about how do we build in the air space above a construction site typically you know we build on the ground and we assemble components with humans and machines but could we drop things from a plane or could we use the air space assemble things using wind for example flying around when the helium in this dies it then comes back to the ground and you're left with this much larger space frame structure beyond self-assembly we're also interested in self-replication so there's a really fascinating story about replication there's obviously biological replication more recently there's been computational replication through algorithms there's robotic replication where robot picks parts and makes another part and lionel penrose in the 50s showed wooden blocks that can self-replicate but there's not that many examples more recently so we did this project where we showed these simple plastic spheres that have metal balls inside that then can connect together uh assembling these circular structures they can encapsulate other ones they essentially make these circles they keep adding units getting bigger and bigger until they have instability points and then they divide so they grow and divide grow and divide continuously finding their own stability or equilibrium points and this is a really simple example of how dumb things like plastic spheres for example with just the right configurations inside can then assemble grow and divide much like cells so up to this point all of the research on the self-assembly side had been very academic you know this is what we considered our basic research you know more fundamental questions what's possible with self-assembly how do we scale it what if we have unique parts what if we do it under water what if we do it in the air what what if we study self-replication it was quite abstract and academic um most recently we were approached by a group in the maldives to take some of this self-assembly research and try to address the challenges that they're facing in terms of climate change and sea level rise and if you look at the solutions there's not that many one of them is to build barriers and walls so static man-made things that try to fight the forces of nature and that's almost never going to win and oftentimes it makes it much much worse there's many examples where man-made structures that were static and didn't adapt to the dynamic forces actually amplified the problem and probably most prominent is dredging so almost every coastal area or island nation around the world uses dredging so basically you suck up sand from the deep ocean and you pump it back onto the beach but dredging becomes an addiction you have to do it year after year it doesn't solve the problem it's super energy intensive pretty nasty and terrible for the marine environment and they do the same thing in the maldives and many other places so they can build an artificial island in about a month using dredging but the problem is that island doesn't want to be there and we became fascinated in the fact that at the same time that the island was built three other sand bars built themselves and these are you know large bodies of sand these these are small islands and locals will go there and they'll have a party they'll park their boats there their friends and family will come and they'll have an island for the day i mean these are large amounts of sand that build themselves and we know that islands and sandbars form themselves but the question is how does that happen when does it happen and couldn't we tap into it there's really fascinating researchers from around the world like paul kench for example that has been studying the morphology of islands and how they change over time showing that islands are highly dynamic that they're moving and changing and morphing all the time they're almost like living organisms very different from our man-made cities of you know new york and miami or malay and the in the maldives these are you know concrete structures that are not dynamic they can't adapt they can't get bigger they can't grow taller they can't get wider they're fixed in static in this super dynamic world of uh cyclones and tsunamis and hurricanes and waves and wind and rain but islands are highly dynamic and so i walked away from the first visit of the maldives thinking that actually i think the islands are the key to their success that they are a lot more sustainable and probably a better future for our built environment than our concrete man-made cities because these islands can change and they can get bigger and they can morph in the face of different forces so we started to study you know why do islands and sand bars form and there's many reasons this is very very complicated and by no means do we have all the answers but there's many different scenarios like wrap-around effects or channel effects and ramp effects seasonal changes and basically it's a relationship between the forces of the water so the waves and the current and tides in relationship to the geometry or the bathymetry underwater so how those two interact with one another is why you get accumulations of sand in one area versus another and so what we're proposing is to work with the forces of nature utilize the force of the ocean in order to promote the growth of islands and sand mars to build and accumulate sand naturally by tapping into the force of the ocean collaborate with nature rather than fight it essentially so in our lab we have big wave tanks and we study this so we pump waves we flow different currents of water and we place geometries underwater and we try to study how does changing the geometry influence the accumulation of sand what kind of patterns can emerge how do patterns emerge how do they self-organize where do they self-organize where do islands and larger accumulations and smaller accumulate accumulations or ripples occur and what we're proposing is to place underwater bladders much like a morphable artificial reef or a submarine meets an artificial reef so a geometry that you can sink and float you can deploy when and where you need it you can place it under water and then when storms come seasons tides change you can try to accumulate sand in strategic areas almost like gardening in collaboration with the ocean to guide the accumulation of sand so we've done two different field experiments over the years the first one we built ourselves and the second one we had fabricated in collaboration with a company called tencott the second one ended up being 20 meters long by four meters by two meters so much much larger these are made out of geo textiles that then get submerged under water so the most recent one we did in in 2019 so we go out there we have this large geometry we took the best of what we saw in the lab we then translate that to full scale and that in this case 20 meters by 4 meters by 2 meters we place it under water and then we study it so we're trying to do lab experiments where we we study it in tanks and then field experiments where we study it in the field the benefit to the lab experience is you can do hundreds of them there's also computation and simulation where you can do thousands or millions of them and then there's field there's nothing that compares with the field because it's the real location with the real forces but it's expensive and slow and the way that we monitor that is through satellite image drone video photos and physical measurements and so in this latest field experiment over the course of roughly four months we saw 300 cubic meters of new sand accumulate so this is quite exciting it's very very small scale it's super super early there's you know many many years of research to come but this is really promising to us that simply by placing a very specific geometry underwater by understanding the orientation of the forces they have two predominant seasons northeast and southwest monsoon seasons by understanding that orientation and the geometry we could guide the accumulation of sand based on the these geometries and so we're trying to create a system uh where you have these the capability to deploy these so if you have an island or if there's a coastal region and you know that you're going to have these storms or you're going to have these seasonal shifts you could deploy them in strategic areas and you can use the force of the ocean to then accumulate sand where you need it rather than destroy you can construct with nature so where is all of this research going i talked about in the beginning that we often hear the word smart products or future manufacturing and that typically means robots so we have a number of problems in front of us if we want to make smarter things it can't take more money more costs more complexity more failure to make something smart and the future of robotics or the future manufacturing as robotics is very much aligned with that that it takes lots and lots of energy and lots of complexity in order to build these things and then we're often building these robotic products if you take you know nike self-lacing shoes for example not that many people want to plug in their shoes at night so they we want the capability of smartness in that case i'm not sure we all need the self-lacing part but we often want the behavior but we don't necessarily want the baggage that comes with it so i'm proposing two things one of them is that there's a shift away from industrial robotics if we go back to that original grant that i was talking about the programmable matter grant about a decade ago if you look at all the researchers that were working on robots as programmable matter then and you look at what they're doing now almost none of them are working on traditional electromechanical robots almost all of them are working on materials whether that's synthetic biology or soft robots with pneumatic structures or 4d printing or programmable materials in the way that i've shown it materials are robots materials have agency they can have embedded information in programs they can sense they can react and materials are what all of us are built out of materials can have intelligence they can have all the capabilities of our living systems and our robotic systems in simple things so that is the one case that robots are shifting to materials and the other one is even if robots are going to be a near-term future in manufacturing construction we still should tap into this collaboration and agency with materials materials should be able to collaborate with robots materials are great at physics materials embody physical properties sense and respond to our physical environment can sense things that humans and robots often can't sense humans are great at creativity and collaboration adaptability robots are great at precision and repeatability so we really should create a symbiotic relationship where they all are collaborating but also tapping into the cues from the environment and collaborating to make a more sustainable environment so today we build smart robots and machines and tomorrow we build smarter materials and environments so i'll leave it there thanks so much [Applause] you

Channel: The Royal Institution

Views: 61,544

Keywords: Ri, Royal Institution, skylar tibbits

Id: -MsNjNeRFMY

Length: 52min 38sec (3158 seconds)

Published: Thu Aug 05 2021