Building a “Swarm” of 3D-Printing Robots
Matt McGowan: Hello and welcome to Short Talks From the Hill, a podcast of the University of Arkansas. Today, I’m talking to Austin Williams, co-founder and chief software architect at AMBOTS at the Arkansas Research and Technology Park. Welcome, Austin.
Austin Williams: Hey there, it’s great to be here.
MM: The company you work for primarily deals with 3D printing. Can you tell us what 3d printing is and how it works?
AW: Generally, when you think of printing, you think of ink on a flat piece of paper. With 3D printing, you’re laying down multiple materials on top of each other to make a 3D object. So you’re printing one layer at a time to build something.
MM: What are some of the benefits of 3D printing?
AW: So, 3D printing over traditional manufacturing has a lot of benefits, especially with the things that you can make with it. In traditional manufacturing, if you want to make a part you might use something like injection molding, or you’re going to have a mold and you’re going to inject some material into it to build something. But you can’t build a lot of things that way. With 3D printing, as its additive, building things from the ground up it can build much more complex things and there can be gaps and things and like helixes and all sorts of intricate designs inside that part. And you’re not really limited by the objects that you can print, like you are in traditional manufacturing.
MM: Can you tell us about some products that are made this way, so maybe a few things that we might be surprised to learn about?
AW: There’s a lot of really cool things that you might not think of that are 3D printed. One thing that might really jump out at you is NASA’s doing a 3D printing on Mars competition right now, where they’re 3D printing habitats for when astronauts go to Mars, they’ll have a place to live and work. So there are 3D printing things as complex as habitats like that. There’s also a lot of really neat things, like there’s a band that’s entirely 3D printed, the drum set’s 3D printed, the guitar, the bass guitar, entirely 3D printed, even a saxophone, where the entire saxophone is completely 3D printed and is fully functional.
MM: In a manufacturing context, what are some of the limitations of conventional 3D printing?
AW: So in my experience… I’ve worked a little bit in manufacturing, and I’ve seen how traditional tool factories are using 3D printing, and they’re mostly using it to prototype initial versions of their designs, and they’re using traditional manufacturing means to actually mass-produce those. And the reason they’re not using the 3D printers for everything is because it’s really great for prototyping new ideas, but when you get into mass production, it’s not quite up to the speed of traditional manufacturing. It takes a lot longer to output volume. With that said, where 3D printing is very beneficial is when you want to have one product, highly customizable, where you’re not out putting a bunch of volume, it doesn’t have to come out as quickly, but it can come out in a customized fashion so everybody can have that product personalized and tailored to them. So one thing keeping 3D printing from large-scale manufacturing is the speed of the 3D printing process and also the quality of the print jobs. You can get really high quality print jobs on a 3D printing system, but usually you have to trade off either the time that it takes to make the object or the quality of the object, and that’s the general trade-off.
MM: Tell us what AMBOTS has created and how it addresses some of the limitations of conventional 3D printing manufacturing.
AW: So with traditional 3D printing platforms, they usually try to set out to solve one problem or the other. They either want to solve the speed problem or they want to solve the quality problem. Because it’s hard to solve both in one system. So what we’ve done that we’ve distributed the system to multiple robots. So this robot can do high quality prints, this robot can do fast prints; you can print the inside of an object really quickly with less quality and you can use the fine detail on the outside of our product so that the inside can be printed quickly and the outside can have the quality that you need. So we’re providing that to make 3D printing more viable for manufacturing and we’re also doing autonomous assembly with gripping robots so that we can automate the assembly process as well. Which also makes some traditional manufacturing methodologies more attainable and less prone to error.
MM: Let’s talk a little bit about scale. With conventional 3D printing, one of the limitations is that it’s hard to scale up a printer. Is that right? How does that work? And how have you solved that issue?
AW: Yeah, so right now in 3D printing, you can buy a cheap, small box that can run you somewhere around $2,000, and like so for microwave size you might pay $2,000. If you want an oven-sized printer, that might cost you $10,000, and if you want a refrigerator size printer, that’s going to cost you $100,000 to a million dollars. And so you can’t have the best of both worlds. In any case, it’s going to be specialized, and if you have to do one thing that’s large, and most your stuff small, you’re shelling out a ton of money. There’s something you’re not going to use all of its capacity. So what we’ve done is made a modular 3D printing system. So at first you could buy a starter kit for our platform that fits in that starter range of a 3D printer, and you can print larger than the size of the 3D printers that exist in that price range today. But then if you want to move from say the microwave size to the oven size, you just buy a couple extra robots and a couple extra floor tiles for these robots to print in a larger area. And you can expand your platform so you can start here and then you can add more to get here to here to here. So what we’ve done is we’ve made 3D printing scalable, so you don’t have to buy individual machines to print different sizes, and you don’t have to shell out a ton of cash if you need a big printing system. And if you only need one large job and then the rest of your jobs are lots of small jobs, you can just split up the platform and have individual print jobs.
MM: Earlier you mentioned habitats on Mars and the NASA project. How would your autonomous robots or a swarm of robots, for instance, manufacture an airplane?
AW: So with our robots to manufacture something like an airplane, they would… they’d be able to work together to print that much faster than a traditional large format 3D printer would. So you might take, for like a human-sized airplane, you could take something like this and they make forklifts that work the same way that these things do, and so you just put this onto a forklift, and we actually found a way that we can we can hijack the controls of a forklift and control it with our software. So you just take this, you put it on a forklift, and you have a large-scale printhead. It could even be, you know, a metal printhead for printing the metals to the airplane and you could scale this up to that scale and cooperatively print an airplane.
MM: Can you tell us who stands to benefit from your research?
AW: So there’s a lot of people that can potentially benefit from this. For starters, other researchers that are researching swarm robotics, researching 3D printing, researching new manufacturing methods. Other people are entrepreneurs. When this technology gets to a commercial level, it really is going to be useful for entrepreneurs because there’s this big gap from the prototyping stage of… prototyping a product for your business to the mass production of a product. So there’s this big gap there that traditional manufacturing won’t touch because the volume of the products that these small businesses are needing to manufacture is too small for these big manufacturing companies to touch and so that there’s a huge untapped market there that our platform is really suited to solve for an entrepreneur to allow them to succeed and get to the point in their business where they can do scale manufacturing.
MM: What could be the long-term economic impact of AMBOTS and these autonomous robots?
AW: So there’s a lot these robots are going to be able to do, and what our long term vision for these is to have micro factories of these where every community is going to have these robotic factories and so the economic impact of that is… Instead of ordering a product and having it distributed and many different factories where there’s a huge supply chain to get that product to the end-user – that’s very costly – but with the platform like this where you’re only taking in raw materials into one factory and then fabricating those products in that one factory, local to the consumer, that makes the supply chain much shorter. You can get products faster, you can get customized products at a much lower price because the supply chain is so short. So it has a really good economic impact.
MM: How has AMBOTS’s affiliation with the University of Arkansas helped advance your goals?
AW: The university has really helped us to build this technology for the resources that we needed to be able to get this technology to a functional level without the university’s resources, equipment and things like that, it would have been very difficult for us to do this alone. So we’re really thankful for the resources we have here at the U of A.
MM: Austin, thank you for being with us here today, and we look forward to learning more about your work in the future. Thanks.
AW: Thank you.
MM: Music for Short Talks From the Hill was written and performed by Ben Harris, guitar instructor at the University of Arkansas. For more information and additional podcasts, go to kuaf.com or researchfrontiers.uark.edu, the home of research news at the University of Arkansas.