[Automated Transcript]

Gary: Our ultimate goal is to be like the Carnegie's of the next century and be the steel mills of the next industrial economy in space. That's in a grand vision scale. So that power converter that we use for our boundary can be used on a lunar surface for power distribution. And other applications that can run the machinery. 

Deep: Hello, everybody. I'm Deep Dhillon, your host, and today on your AI Injection, we're exploring the future of space manufacturing with two guests, Gary Calnan, CEO and co founder of Cislunar Industries, and Walter Schroeder, CIO and also co founder of Cislunar Industries.

Cislunar is pioneering space based metal recycling, transforming orbital debris into valuable resources for sustainable space development. Thanks so much for coming on, guys. 

Gary: Yeah, thanks for having us. This is good to be here. 

Deep: Let's get started by, breaking down a little bit about what Cislunar Industries is all about.

Like, what's the problem you guys are trying to solve? Why do we need to recycle metal in space? 

Gary: Yeah. And I mean, I think, this is a great.

intro to give you the back story on the company because it helps to set the stage for why we think it's necessary in the first place. So around, 2017, I discovered the international space university and I always wanted to get into the space industry. 

And around that time, also, we were starting to see, that entrepreneurial space is really starting to become a thing. SpaceX was starting to launch rockets and I started to research. What's going to be needed for humanity to expand beyond the frontier where we are in space to push that envelope of human expansion into space.

and I became aware of the need for resources. If we want to live beyond Earth, on the moon, in a space station, on Mars, or wherever. Fundamentally, we need to be able to use the resources that are available where we are. We can't bring everything with us and resupply from earth especially the further out we get So there's this whole concept called nc2 resource utilization using the resources where they are And I learned about that.

as I was getting into the industry and realized that finding a way to use those resources is going to be fundamental So when I went to the space university, which is where I met walter and where the company was born we literally just sat down in front of a whiteboard and said, okay Resources are important to have an industrial economy.

We need the whole supply chain from extraction to production. So what does that look like? we started to look at, companies extracting resources talking about, mining asteroids and the moon. there were companies talking about manufacturing things, but in the middle of that process, we didn't see anybody who was going to tackle the problem of taking those mined resources, and then converting them into something that the manufacturing companies could use.

Basically, if you, for the analogy on the earth, this is like a steel mills that take iron and turn it into beams, that seemed like a really, natural place to go where that was definitely going to be a very important piece of the puzzle right in the center of the value chain.

as I was there, I started to learn more about space debris the nature of it, the quantity of it and how it was growing. I really didn't know much about it before going to the space university, but we saw that as an opportunity to, use space debris as the first thing to process.

So before we even have mined materials from the moon or the asteroids, we can mine the space debris field and use those resources to learn how to make things in space and manufacture with and start to build things. at the same time, we're helping to solve the space debris problem, which right now is, you know, it's a growing problem.

It's bigger than probably most people realize, as we have more and more satellites going into orbit, this becomes increasingly challenging. the old stuff that's up there is just a big cleanup problem, right?

This is a way to take that economics of having to spend a lot of money to clean something up and flip it on its head, where that now becomes valuable. So why do we need metal processing in space? one reason is to create this, robust economy that could support human expansion into space and derive those resources to support people out in space, without having to haul it all up there.

And the other is so that we can find a way to deal with space debris in a economically sustainable fashion, where people have an incentive to go out and get that space debris. Bring it to us. We can buy it from them. If there's a place to sell it downstream we become just like a recycling facility.

Deep: So I want to dig into that but I want to capture the context a little bit more. So you said a number of things that I want to understand better. 2024 that's causing. more, folks to get into space and put more things into orbit.

And why do we even want to get off the planet? cause last I checked, none of it seems particularly hospitable and it doesn't seem like a fantastic place to hang out. at least until you get a hundred light years out, which seems pretty far. how much of this is just techno futurism and optimism and what's driving it all have the space agencies made some kind of fundamental shift post SpaceX where they're realizing we really want to build this ecosystem of private stuff.

I realize that's a lot of stuff, but tell me why are we even having this conversation? what are the forces that cause. Cis Lunar and all the other companies in your potential ecosystem to suddenly arise.

Gary: there's a lot of big questions in there and some of them are philosophical. I mean, so you have the context in the background of why it's happening now. Part of it is just the launch costs are going down and that's enabling, more, activity at a lower price.

There's a lot of commercial value even before humans get off space. Just to have satellites in orbit doing things for Earth. Whether this is analysing the climate, or tracking ship on the water or taking pictures. There's all kinds of things that business and government need space data for. And so there's a proliferation of satellites, as it becomes less and less expensive to get those satellites up there.

And they're becoming more and more capable as well. And, just more quantities of it. I mean, SpaceX has their, internet constellation star, Starling, and that's thousands of satellites already. And I think they're aiming for something like 40, 000 satellites at its full capacity. And that's just one of them.

So one constellation, there's a lot of stuff there that makes, increased activity up there, a thing, which makes the problem of space debris more relevant. so that's one thing. Your bigger question was, why should we do it at all? Or maybe why should we do it with people?

It's a question that comes up a fair amount you don't inherently have the desire to go to space yourself it's kind of hard to explain it But I always felt like human beings have this innate Drive, to explore what's over the next hill I want to know what's there, and go there myself, to feel it and experience it.

So there's, that element of human desire to go out and experience the universe. That I think is part of the driving factor of it. 

 There's that piece and then there is also a resources piece. As we have the ability to launch more things into space.

And create the capabilities to manufacture things in orbit, there are certain kinds of structures that you, you can build, you know, capabilities on that could be benefits to people on earth. which say, for example, space based solar power is one of the big projects people talk about.

there are different opinions about how valuable it is, but you need a very large structure in space, collecting solar radiation, then beam it back to earth. this is a, carbon free way to get power back to earth where the sun is always shining. and so there's a real potential climate benefit to having some of these capabilities.

in order to build those things. We're not going to launch all of it up from rockets and assemble it in orbit. The best way to do it is to build those big structures in orbit with factories that build those things in space. If you can get your materials from space, that also further reduces the cost of delivering, the materials and then building out those structures.

Deep: So back to like, how does this whole funding world work? have the space agencies like the European, the American, the Indian, the Russians, are we all on the same page that we're trying to build an economy out there as opposed to having nation states, do the work?

Yeah. 

Gary: I think there are different philosophies depending on which nation you're talking about. I don't know if you're going to get the same philosophy out of the Russians and the Chinese, but in the rest of the world, for the most part, there is an interest in leveraging, the commercial sector to, create capabilities.

 The space agencies, they also create capabilities too. So they're sort of seeding some of this and trying to create customers in the initial pieces of a market so that once this gets going, there will be commercial companies that can sell to each other, even space to space, not just space to ground, where most of the commerce is at the moment.

there's funding to build up this ecosystem. It's another economic development area for every country. Right. And you also have prestige and Scientific benefits definitely come back to earth, just inventions that happened for space that are useful on earth. that happens throughout the space age.

everybody wants their piece of that puzzle, but there's also, a space race to make sure we're also on the moon at the same time that our rivals are. So it's not like we're all doing it together, but there's a bit of a competition. it might even go out to Mars eventually.

Deep: Let me sort of describe what I'm imagining happening and tell me how far off I am. somehow Cislunar, is going to get some kind of platform in orbit. now you've got this thing that has to catch stuff that's spinning around the earth at, 17, 000 plus miles per hour, which I suppose the platform itself is also spinning at that same, orbital velocity, and then you have to kind of creep up on, OTs that are floating around and grab 'em, maybe like a net or a magnet or something.

and then you grab onto them. there's some robotics that has to navigate you there, grab the thing, handle the thing, and then it sounds like you're gonna try to melt the metal down and build a forge or something into ingots.

So then you get that problem. And then once you get these ingots. I guess you just hang on to them until somebody comes and picks them up or something. walk me through the, scenario. it seems pretty futuristic. 

Gary: you have some of the elements, right there.

So just lay it out, at its basic state. One of the projects we did recently was called a metal propellant ecosystem with the space force. that was a good backdrop for it. the vision as we've always had, it was. our focus as a company is on creating the metal processing piece, at least in this part of the business.

we have partners and there's a number of companies out there. Astroscale is one, KMI is another one. Starfish Space. And some of them are going after space debris, astroscale in KMI in particular are really focused on being able to capture and remove space debris.

so they're the guys that would go out and grab the object with a spacecraft. once they acquire that object, we would go after not little pieces, but think of things like large rocket stages that are still in orbit. For old satellites that are still in orbit that are dead now.

Yeah, it would attach that thing, they would figure out how to detumble it. There's a lot of stuff that has to happen there. it's not a trivial problem, but they're working on it. They're investing in it and they're making progress. They've done demonstrations on orbit even. then they would bring that back to the platform.

we would match orbits with the platform, just like a SpaceX capsule does when it reaches the international space station, same concept. they would have that object with them. when they get to the platform, a robot arm on the platform would grab that object from the spacecraft and attach it to, a truss Now we don't really anticipate necessarily owning that platform. We might rent it from a company that's built an industrial park in space. we would just be one of the industries on that platform. once that thing is there, we would bring that spacecraft into a garage, or hanger, if you think of it that way.

Another kind of volume of space that keeps it enclosed and inside that volume of space is where we would cut up the object, remove the metal that we can remove. Figure out what's a waste that we can't use, figure out parts that maybe can be reused like a junkyard. and the stuff that's just metal that we want to recycle.

We then feed that into our foundry, which is our core expertise. We melt that down and we turn it into feedstock for manufacturing. So that could be like wire for 3D printing, ingots that are then further processed into other shapes like beams or tubes or sheet metal, the things you would go to a mill on earth for, you know, metal mill, we would want to make in space.

And then, we might have a market for, we make a truss that we sell to somebody and they assemble it for whatever they want. the other key piece we're making is metal propellant. this is what Space Force was interested in, which is a really interesting angle there's a company called Neumann Space and others out there that, make electric propulsion systems that use solid metal as the propellant the electricity turns that metal into a plasma. The plasma creates thrust. what's cool about that is we can harvest a fraction of the metal from each piece of debris that we collect.

Turn it into propellant, the spacecraft that goes out and gets the debris. And then that fuels the next mission. We only need like 10 percent of that object turned into fuel, to go out and get the next one. That means 90 percent of it is left over to, make more propellant for other missions.

Like what the space force might want to use it for, for maneuver of their satellites. And that's kind of how, it goes downstream from there. So that same company that gets the debris would have the trucks that then take, that material to the customer. on the same platform where we feed another manufacturing company raw materials and then they make something, 

Deep: So I just want to understand what are the inputs to your box and what are the outputs? Are you generating power or you're assuming you've got power? 

Gary: We assume we've got power. 

Deep: What's your, and then you assume that this truss, this big piece of space debris is nearby.

Do you have to figure out how to grab it, cut it up and pull it in? Or are you assuming somebody else has done that? 

Gary: The other company would bring the debris to the platform. That truss would be part of the platform. then we would move it through our process. 

Deep: so it's so input is.

Big chunk of stuff. Yep. Output are the ingots. Is that 

Gary: it's wire and rods, tubes, sheet metal, materials for manufacturing fuel. 

Deep: Okay, and then are you also taking your output and dealing with the storage of that output somewhere? 

Gary: Yep, 

Deep: and then at that point someone else comes and picks it up and does something. Got it, okay. 

Gary: This could be also ported to the lunar surface. 

Deep: If you take all of the space debris that's floating around and you assume that by some miracle or some amazing technology, you can recycle 100 percent of it.

what's the cost of, actually taking that comparable amount of metal up from Earth and getting it out there? 

Gary: a couple of years ago when we were estimating the population of space debris, I think it was at 10 million kilograms of intact space objects that were up there.

And that's grown because there's new satellites, the launch costs, Right now, and this will go down over time is something you think of like it's 10, 000 a kilogram. 

So 10, 000 times 10 million. I don't know 

Deep: 10 billion? 

Gary: Yeah, something like that for the existing stuff.

we're at a population of 10, 000 satellites in orbit We're expecting that to grow to upwards of 100, 000 by the end of this decade. So, and those satellites have a lifespan that's five years, a lot of them do anyway, some longer, but most are shorter.

the plan is to deorbit those in the atmosphere, but we could just as easily keep taking from that flow of materials. 

Deep: I see 

Gary: a flow that's getting bigger and bigger. And we would take satellites as they reach end of life and harvest those materials from those satellites instead of deorbiting them into the atmosphere.

Yeah, that's 

Deep: an interesting point because for the past 50 plus years we've been decommissioning through sending them back and burning them up in the atmosphere. Now, on Earth, you have this whole problem of, picking through a bunch of gook until you get to the materials you actually want.

What's going on out there with respect to the space debris? imagine you're not getting perfect little chunks of aluminum. it's probably mixed in with all kinds of stuff. how are you guys thinking about that? 

Walter: So there's a process of robotic separation and Gary mentioned earlier the garage where this would take place. another benefit of having a garage as well, that you don't pollute further. There's a certain pollution acceptable within the material that you melt down.

For example, depending on the coating, sometimes you need to take it off. Sometimes you can leave it on when you melt it. when we go further we would also think of taking other parts that are not metal based to use them for recycling and generating new objects or new materials.

For example, one of our big projects with DARPA was on building a metal. processing economy on the lunar surface, one thing you can find as a resource on the lunar surface is, iron, if you mix it with carbon, you get steel, which has much better properties.

So, some of those components that you would think of maybe pollute, anything that needs to be recycled with circuit boards or other parts of a satellite have carbon in them. So by using this and mixing it with iron that you find naturally on the lunar surface, You could create steel 

Gary: and there's Almost no carbon on the moon. 

Deep: Well, we got lots here. We got a lot. We don't want it out of our ecosystem, but we definitely don't want it in the atmosphere. 

Gary: And I think to clarify when we've always conceived the disassembly of a spacecraft or, an upper stage, for example, we imagined that we would have a robot arm that's like, okay, remove the copper wire, right now our assumption is that that can't be fully automated.

We would have a human on the ground That's okay. That's that's the copper wire do the copper wire removal routine, there would be some automation then the human would get back in after that's been done but there are definitely areas where I expect that automation will increase over time.

And this is sort of where that AI, yeah, 

Deep: let's talk about that. Cause that seems like the 1 thing of everything you described. It seems really straightforward to me. But before we get there, how are you guys building stuff or are you in concept land and how do you build, are you physically building or are you in the simulation world where you're trying out your systems and all your engineering working in a simulation space?

Gary: we're fundamentally builders. So we, build and test and iterate. That's, that's how we learn. That's how we move quickly. Our team are a bunch of makers. All of our engineers are makers. That's their nature. 

Deep: Like physical stuff. Yeah, absolutely.

Like we try, 

Gary: we test and try it. And there's some conceptual stuff up front but we don't model everything first. Right. 

Walter: The room where I sit on the right hand side is a room where we store things that have already flown in a parabolic flight and have been tested.

on the left hand side is what we're building. So we're building currently hardware for the ISS, and to go on satellites. this is a maker space here. 

Deep: Okay. That's, that's kind of interesting to me kind of in and of itself that, that you're not building in simulation land. So that tells me that You probably have something really immediate that you're building.

So maybe let's talk about that if you can, like what's your immediate goal? What exactly are you building and for whom, 

Gary: I think our very first project that really got the company off the ground was NASA we got a phase one SBIR research, contract, six month contract.

we had a little bit of matching funds from an investor. instead of giving them a report at the end, we built a prototype foundry that would work on the ground to show them the concept. we took simulated space debris, which was like a strip of aluminum. And we did a live demonstration where we melted aluminum, and made little ingots that look, like this and this is sort of representative of what one of those fuel rods we talked about earlier could look like. NASA wanted us to do that because they wanted us to study the possibility of recycling large objects from outside the space station.

not trash inside the space station, but actually like upper stages and stuff. they could use that in their moon to Mars architecture, which envisions reusing materials both on the lunar surface and on Mars and beyond eventually. 

Deep: So let's talk about this foundry that you built. maybe describe it for us. what does it look like? How big is it? what's standing next to it? Like a bunch of soda cans, did you, really simulate one of these truss situations, out in space?

we were simulating on the, 

Gary: on the first one, we were simulating what it would be like to cut the skin of an upper stage off because we figured, like, rocket stages have a lot of aluminum. Most of them do either aluminum or steel. and the skin is almost all metal. So we took a strip of thin aluminum.

this is representative of. The kind of alloy you would use to make a rocket body. that was the first demo that we did in that first phase one, the phase two, we took that concept, to a parabolic flight, which is where you fly up and down and have a little bit of, simulated microgravity for 25 seconds at a stretch.

and we took this machine on one of those flights. the next version of it, was about the size of a small refrigerator, mid size up to your chest kind of refrigerator. that was to make a rod in zero G.

So we, did a couple of tests with that and it uses electromagnetic induction to heat up the material. it uses electricity to, put a magnetic field inside of the metal that heats it up and melts it down then we cast it with a continuous casting process, which in theory would allow us to build things that are as long as we need them to be, because it continuously has these rollers and you move the material through this is a process that's done on earth for things like an 

Deep: arc welder, 

Gary: no, this is just taking molten metal and moving it through, this process.

Deep: Okay. 

Gary: And in microgravity, you don't have the usual benefit of gravity pulling it down we have various ideas on, different stages of how we can do that, in microgravity. in this demonstration, we had a plunger that just pushed the metal through. 

Deep: Okay. 

Gary: in the future, as we get a little bit further down the road, we're also testing a concept where We have multiple coils that are, this was also tested on the parabolic flight, multiple coils that create electromagnetic field, and you can control the intensity of these different magnetic coils to change the position where something should be.

So you can actually, without touching it, you can move metal things through a space, even molten metal things, with just electromagnetic fields. And this could be applied to other things too that are kind of interesting. Attaching to an object that's nearby. And, you know, sort of like mini tractor beams almost, but, but for, for processing, we were trying to replace the gravity, you know, with electromagnetic fields, basically.

Deep: once the molten metal makes it through, then you have something that chops it up into little ingots or something. 

Gary: you end up with like a rod. That's this, like this one that I was showing you is there some small inch diameter kind of rod.

Or it could be different dimensions, right? Whatever we have, whatever we set up the casting wheel to do. but then we would further process that into wire or, or other shapes. we can extrude it into other shapes. We have a NASA project, we're working on to make a system that can take molten aluminum and directly extrude a wire inside vacuum.

Because they want to be able to do that on a lunar surface. So they're taking what we worked on already on these other projects. Saying, Hey, can you do that to make a wire extruder for the lunar surface? And then they put us with some other, companies and departments at NASA. they're basically taking lunar rocks.

Lunar regolith is what they call it. They're rich in aluminum. They're processing those aluminum oxides, which is how the chemical properties are when it's in the lunar regolith format, they're extracting out the oxygen because we can use the oxygen.

That makes it not an oxide anymore, but now you've got silicon and aluminum. they further process it to extract the aluminum. they give us the aluminum in molten form. We turn it into wire, we hand that off to the team that does wire additive manufacturing, and they're going to take that wire and print something with it.

they want to show this end to end experiment, make it all work together and take that to the moon. right now we're just making the wire extruder for them for that purpose. 

Deep: oh, this is fascinating.

So this is really, a whole robotic exploration economy they're trying to build you need to harness whatever's there, combined with whatever you can get there as cheap as possible to build, period. Now, what is it that they exactly want to build? it seems like propellant's a big deal because you land on The moon, if you want to go to the next spot, you need fuel.

what's that ranked order list of stuff they want to build look like? 

Gary: They want to get oxygen and hydrogen out of the water that's on the moon, oxygen out of the regolith that's on the moon that could be used as chemical rocket propellant, like when you see a rocket launching from earth they also want to extract oxygen for humans to build a base for a permanent human presence on the moon.

so do, China and Russia too, by the way. So there's a rivalry happening here a bit of a race to get there. it's for structures as well. So that's where we come in to make structural materials, whether that's to hold the stuff that they're harvesting, like tanks kind of thing, or to build, the actual structures for bases.

and these could be structures to put machinery underneath. You know, this could be places where people live. There's all kinds of infrastructure you would need for a base. instead of hauling it all up from Earth, we'll haul the first ones up, because, they need to be perfect. But once we get it there, let's start building stuff with what we have there.

towers to get electricity. there's all kinds of things that you can use the local resources for that are pretty basic. 

Deep: Got it. So, you basically get grants or a contract from NASA or a space agency. NASA, Space 

Gary: Force, DARPA, ISS National Lab, 

Deep: and then you're basically in progressively more, detailed and elaborate prototypes, but your ultimate goal is to wind up with one of your systems on, the ISS or on the moon or something.

Gary: ultimate goal is to be like the Carnegie's of the next century and be the steel mills of the next industrial economy in space. That's in a grand vision scale. that's where we want to fit into the puzzle. we also have this other thing that's come out of this whole thing, which is power because we power our thing with electricity.

Right. So that power converter that we use for our boundary can be used on a lunar surface for power distribution. And other applications that can run the machinery. it can be used right now to where even we have commercial contracts for this part of it, which is not around the recycling, but it's funding the company, to supply satellites with these power converters that are a high power level needed for our foundry, but it had an application, in the rest of the market.

So this is a great way for us to get there. If you're trying to think about how do we fund this? We have a commercial spinoff that has evolved out of this and has also expanded the scope of the company From just metal to metal and power. So if we have metal and power in that ecosystem We're at a pretty critical piece of the value chain We're not generating the power, but we're taking that power and changing it into the form that it needs to be to do a job.

Deep: Okay. So we're a few minutes into the podcast. Usually we bring up the AI question sooner, but I really needed to get the whole context. Cause it's quite different than normal. So what are the problems? you're facing how are you using machine learning and AI in your systems today?

I imagine there's a bunch of robotic stuff, but in your foundry or manufacturing systems, like, what are some of the problems that you're seeing that? You know, we're machine learning or AI can help 

Gary: Walter. Do you 

Walter: want to 

Gary: jump in on 

Walter: Yeah, sure. I can start.

an AI is as good as the data you train it with. many of the problems we are facing. There is not so much data available yet. Right? , but will be eventually. so where there is little data is the environment of space. this is micro gravity. This is radiation.

different temperatures. so this is a field where the AI could be trained much better than a human who is used to a very different environment and often takes a judgment based on previous experience. The environment, where we train the system already is, this first test that we do with, manipulating an object in three dimensional space and microgravity.

we have a system of different coils arranged around the metal object. we are trying, to make this object move and control its position in microgravity. That's what, Gary earlier said, kind of defying gravity 

Deep: the coils? Like, what does that mean? I'm picturing a metal ingot with a Yeah, so you have a coil that's 

Walter: the size of my fist, right?

And it's just spin. So now you have four, five of those arranged around an object, right? and so you can change the magnetic field. Which has an on the object within this circle or sphere of coils. with the change of the frequency of the magnetic field, you can now control the position of the object.

You can also heat it, but it also allows to, have an impact on the force that, that is, On the object in which direction it would move. 

Deep: So wait, is there physical contact with the object or are we talking about magnetic forces? 

Walter: Yeah, contactless.

So it needs to be an object that is out of metal and can be magnetized basically. 

Gary: the metal doesn't have to be magnetic. Okay. Because you're inducing a magnetic field in the target object. So aluminum works fine even though it's not magnetic. It becomes magnetic in an opposite way. As you run the coils, 

Deep: and then there's differences between the different metal properties and composition that cause you to have to manipulate, to change the forces that you're applying in rotation, 

Gary: the forces, the way it, the force are created in the metal changes depending on the temperature of the metal.

so that also is always changing. And it's a constant feedback loop. Which certainly could be optimized with AI, I'm sure. 

Deep: Interesting. so you're envisioning some machine learning to like, learn that process across a variety of different materials and compositions that you're going to pick up in space.

Walter: Yep. 

Deep: And so how are you getting your training data for that? are you using our understanding of physics and properties and compositions to generate synthetic training data? or are you physically sticking lots of different objects in the thing and spinning around and just gathering data based on your trial and errors or something else?

Walter: Be early in the process there. Right. So we have done it in parabolic flights, but the microgravity isn't as good there. So, it's more of looking ahead of where fine control would go which, yeah, when we have it on the ISS, we would collect as much data as possible and that would help us to, improve.

The model that we are using to control the position 

Gary: there is a set of data that has been generated around the material properties of, pretty much any conductive material out there on the ISS so far, but only on very small samples, like maximum eight millimeters in diameter spheres.

So it's limited data from a perspective of manufacturing larger things. Which is why we're taking it the next step. We also have consultants, this company called G space that, is focused on, the data around space and microgravity and how to analyze that.

So they're helping us with some of this stuff and they have data sets around that, it's an area that will definitely grow as we do more. 

Deep: it seems right for a physics based modeling approach, Like in simulation land, This is an area in machine learning that's growing quite a bit.

There are scenarios Where you can model much of the physics and you'll probably still need to fine tune your model with real data, but you need to get the model pretty far before you get to that point. Right. And so I could imagine you, modeling, maybe doing a finite element style, compositional.

Structure in simulation land where you mix in different metal bits. And then based on that, you can model the physics of the, forces that you're applying and their manipulations. And then based on that, you could generate a whole ton of synthetic data, bootstrap your model, and then now try it out on actual stuff.

Gary: Yeah. And actually, to that point, for that coil system, we work extensively with the Colorado school of minds as one of our partners on that project, they have Andrew Petruska there, is a, he has a, a group that he runs that is specializes in this magnetic. Manipulation with multiple coils topic, 

Deep: for other 

Gary: applications as well.

those guys definitely did a lot of modeling of how it should work, to get to the point where we could even try to control it in the physical space. So in that case, they did a lot of modeling upfront more than we did for just how do we cast something like that? We just tried a lot of different things.

we did some analysis too, but That was more trial and error. 

Deep: So you have this metal object and you can spin around and reorientate. You can move 

Gary: it up and down and make shapes with it. So what's 

Deep: your immediate goal there is your immediate goal to like.

carve it up, cause ultimately you're just trying to get it into the forge. 

Gary: The idea originally was that, when you are in a foundry and hot metal contacts the melting chamber, it wears out rather quickly depending on what materials you're using.

on earth, these things are serviced all the time and you can't do that when you're space. So if we can figure out a way to use contactless technology to both melt and move the metal into a process that's less likely to degrade, that just makes the lifespan Of the foundry longer and more reliable.

Deep: float it and then hit it with your energy beam and then melt it right there, 

Gary: you melt it and control what they do in the space that they have a system of spaces right now that has two coils like this. 

Deep: Yeah. 

Gary: And they superimpose a high frequency and a lower frequency, the high frequency heats the low frequency controls it?

I first it puts it into position, it raises the little ball up and holds it there with the fields, and then another frequency hits it and it melts it. then they do different things with frequencies to wobble it and that sort of thing.

you only need a small force to control position in microgravity, because nothing's pulling it, besides those fields. You need more power to melt it but you can do both of those things. So we don't need to have contact heat melting. you just use electromagnetic energy and that heats up the material.

Deep: And that same electromagnetic energy you're going to use for cutting too, like when you have to cut off the original raw. 

Gary: there's a couple of different possibilities. You could use that. Actually, there are ways to use electromagnetic energy, you know, melt a certain area, but it's more likely to be an electron beam welder that's turned up hot enough to cut instead of weld could be laser, but E beam is a good, technology for this. E beam needs a vacuum, that's why we don't use it much on Earth, and also creates x ray radiation. but who cares if you're in space, there's plenty of radiation anyway. and it's, it's robotics, 

Deep: So going back to the AI question, there's this challenge of manipulating objects in a contact list. Microgravity environment. Are there other problems? that you're anticipating? 

Gary: Yeah, the other big area that could be interesting is characterizing the materials before they're acquired by the company we hire to get the object.

we think we know what an old upper stage is made of. based on plans and what they said it was originally, but maybe it's changed or there's still some fuel on there or maybe, I don't know. There's, there's a lot of different questions. So you would want to have a way to read with remote sensors as much as possible.

Verify first, what is it that could be a visual comparison. And then what is it composed of? that might require different kinds of, Remote sensing technology. I think AI is a great application there to sift through that data and learn how to, tease out maybe from things that we didn't think could indicate a material, but actually do.

Deep: be using, regular imagery, multispectral, hyperspectral imagery. 

Gary: And then you get a little closer to it where you're still not touching it, where you might use, non destructive characterization of materials.

But ideally, if 

Deep: you don't have to get close to it, that's a huge, 

Gary: exactly. Cause then you could target which one you're going to go after before you make the orbital maneuver to get there. Right. So there's probably like a distance thing first. And then when you get close, you really characterize like, okay, where do we want to grab onto it?

You know, has this, the thickness of this part gotten dangerous to be the part we latch onto, whatever. you don't want it to break up after you hold on to it. So there might be some analysis there. I think there's going to be a lot of pieces for that that are relevant.

And then even after we get it into our process, there's going to be some materials analysis, sensing that I think we would want to do on the front end before we bring it in and also to verify what we've created on the backend. Both of those, I think are ripe for. Machine learning or AI kind of application.

Deep: that's a big project in and of itself. Like somebody building remote sensing of all the objects in space, from a vantage, which would let you understand material properties so that you even have the data to do. 

Gary: There are a number of companies working on remote sensing of where the objects are and how big they are, not necessarily the material composition yet.

so that's interesting, like 

Deep: even just tapping into the data where it is, what it is. Then you have your inventories, you can look it up, what the original properties were, now, you combine that with LIDAR or something so you can reconstruct its current, spatial geometries.

from that, you might be able to go back to the original model and start to understand what chipped off and what's there. 

Gary: exactly what you're saying, for that piece of it. But really understand the material characteristics would be valuable.

Now, certainly with the new satellites that just reach end of life, we'll know what they are because the company still exists So, you know, you can just ask them or we would have a contract up front for it or something. So that would be. A different scenario, right?

Deep: now is this a problem that you guys would be tackling this material characterizer thing? Or is that something maybe it's kind of in your scope But it's kind of in whoever's grabbing this stuff. 

Gary: we need non destructive analysis of materials.

for any in space manufacturing technology. we have to verify what we're making in space remotely is actually what we think it is before somebody will be willing to use it. there are other companies working on this kind of stuff. 

Deep: And what about like machine learning or AI, like within your foundry itself?

So like from the point at which you, grab the raw materials. Is there something in the actual foundry itself? Like maybe the control mechanisms or, something else where you see some AI needs? 

Walter: Everything that moves. Or is movable, consumes energy and by optimizing movement. yeah, you can save energy. And power. So here you have always a way of using AI. I know that from a friend of ours, who is programming the Canada arm and she spends a lot of time of finding the perfect movement.

For the arm to use the least amount of energy, but also, um, looking at forces that would be applied to a work piece so that wear and tear is minimized. there's thousands of ways of moving a certain thing, especially if you're going into robotics and looking at this disassembly of satellites. so there's a room for optimize with AI.

Gary: And there's also, what sort of relates to what Walter was talking about earlier, but you know, we're going to go up. We're going to model what we think is going to happen when we melt these ingots on the space station and re solidify them into a shape. but it's likely that our model is going to be wrong.

And also as we go and change and get different materials that are slightly different than the last one, every time we melt something down and we want to form it into some shape, we want the crystalline structure to be the way we want it to be and everything, it's There's a lot of dynamics in there on how, how fast you heat it, to what temperature it gets to, how it moves through the process, how fast it cools, all that stuff.

there's, there's so many dynamics that I think it's another area that's ripe for optimization 

Deep: awesome. Well, this has been a super fascinating conversation. One thing I'd like to end with is. If we jump out five or ten years into the future and everything you guys are executing on, As far as what you're dreaming and scheming it all works out What does the world look like from your vantage 

Gary: Yeah, this is this is the dream that keeps us going all the time.

I think I mean, it's I see having permanent presence on the lunar surface, the beginnings of an actual economy up there where you have not just materials being made to build stuff and maybe ship back to Earth, potentially, or to other, things we're building on orbit, but, but actually business to business transactions happening there for, to support the humans that are on the space.

We see, multiple commercial space stations emerging potentially by the end of the decade. And if this works out, like we would like to see that happening. like in 10 years, definitely there are plans for multiples of these. I want to see also large structures being manufactured on orbit for various applications.

People are starting to test out in space, you know, solar power in space, large antenna arrays to, that can look into deep space or can, Assess larger areas of, you know, the environment where we're managing traffic. mean, I think the activity is going to go way up in the optimistic plan. Talk about a hundred thousand satellites on orbit.

By the end of the decade, if that's happening, we need maneuver for all those satellites and sort of a vibrant industrial economy that's starting to emerge within 10 years. It will still be the beginning, but and, 

Deep: and with respect to CisLunar, your forges are on the moon, on the space station in asteroids.

Gary: Yep. 

Deep: Yeah, 

Gary: maybe asteroids by then. I don't know. But for sure, I would expect us to be on the lunar surface with an initial test facility or initial capability there. I would expect us to be in space stations to process their metal waste and also on a platform to start processing satellites at end of life.

Probably that before debris, but you know, and eventually space debris. I also see our power systems being all over the place too. So that's the other piece of the business that we're working on. 

Deep: Exciting stuff. Wow. This is cool. I do have one thing that I'm kind of wondering about.

So you're, you're on the space station, you have metal debris, you want it processed. I understand the value of recycling it. But what are the, what are the forces kind of encouraging people to just shove stuff back down into the atmosphere? I imagine there's like a tension between the longer arm play of trying to recycle stuff versus the shorter arm need because there's just so much junk out there to start shoving stuff down into the atmosphere so it burns up, like.

Gary: I mean the most important, the most, the strongest thing is that there is no capability to recycle right now. And we have to keep it clean. So the first plan is, you know, right now the plan, the only plan available to them is to de orbit it. Or if it's in a higher orbit, too high of an orbit, you just put it in a graveyard orbit, which is not exactly perfectly safe, but that you put it in a, in another storage orbit, basically where it's just dead satellites.

And how are 

Deep: they doing that? They have, like, is that the same, like, how do they get it? How do they decommission? They save a little bit 

Gary: of propellant and then they move it into that orbit. And then the thing is, like, they just, 

Deep: like, push it up there or push it down. Yeah, whenever you 

Gary: accelerate a spacecraft and you're in orbit, it goes up higher.

Deep: Uh huh. 

Gary: If you slow down, it goes lower. So that's, that's how orbits work. It's kind of counterintuitive. But you just get it to the right orbit and you leave it there. 

Deep: But it's a physical push. 

Gary: Yeah, it's a physical push. You need to save propellant. And sometimes they're broken. And so now they're going to need like someone else to come push them, which is a new business area that the other companies are working on.

Deep: So does that, to what extent does anything exist right now where, where these little spacecrafts are flying around and moving debris in and out of. 

Gary: There are, um, there've been a few missions. I don't know how many, just a small handful that have shown to be able to move a spacecraft into new orbits with another spacecraft, I think north of Grumman.

Might be the only one who's actually done this so far. Astroscale is close to being able to do this. Starfish is another one that's close there. So there's a couple of companies out there that are, you know, getting close to do this. The Chinese might have done a mission that can do this. I think they did one as well, where they moved their satellite up into, but it's only a few so far.

Deep: Got it. Awesome. This is, I don't know what you guys are up to.

It's just kind of wild. I feel, I feel like 

Gary: it seems like complete science fiction, but it's actually. Yeah, closer to reality. 

Deep: There's just a lot going on in this sector that I don't think 15 years ago, there was that much going on.

Yeah, 

Gary: I mean, now to 15 years ago is a huge change. 

Deep: Yeah, so, I mean, it's kind of wild to imagine some of this, and you can imagine a million cool things coming out on Earth, like the fact that if we're really recycling in zero orbit, then chances are we can probably send stuff into old landfills and like other places.

Right, yeah. Right, or into the oceans would be a great place to get going too. Yep. Awesome.. Thanks so much, gentlemen. This was, this was just so interesting. 

Gary: Yeah. Thanks for having us.