The Space Economy Has a Labor Problem | Ethan Barajas

Speaker A: The next major bottleneck to space being that labor force being the ability to actually do things on orbit. We can send up to three and a half tons of cargo to the space station, uh, and that takes a crew of four astronauts like seven days to unpack and seven days to unpack. Labor is going to be a massive part of the space industry and that the labor largely will not be human because we need the humans to focus on what they're very good at, which, which is the science side of things.

Speaker B: This is Ethan Baragas, the co founder and CEO of Icarus Robotics, a company building robots that can work in orbit and take routine tasks off astronauts hands. We talked about why cheaper rockets do not solve the labor problem, how robots can learn to operate autonomously in zero gravity, and what has to happen before space becomes a real economy.

Speaker C: Welcome Ethan. Thank you for joining me today.

Speaker A: Oh, thank you so much for having me. I'm really excited for the conversation.

Speaker C: Yeah, me too. Great. Um, so everyone, ah, in the space industry talking about rockets and launching rockets and things like that, um, why do you think that labor in space is actually the next bottleneck?

Speaker A: Yeah, so the more and more launch capacity that we get volumetrically and mass wise to actually send things to orbit, the more we can do on orbit, we can send up larger infrastructure, we can send up more mass. And the more and more companies we see now working on the other side of the equation, which is reentry. You have companies like Varda Inversion, um, and you have SpaceX with their starfall program. And everyone's announcing the reentry side of things. And this offers a brand new paradigm to what we've imagined space as before. Previously it's you send it, it burns up in the atmosphere after it loses power or instrumentation dies. And now that's entirely changed if we can bring things back. Um, and it much more mirrors the logistics enterprise that we see on Earth where you can ship things, you can receive things, you can send them places, you can bring them back. And when we think about the Earth and when we think about Earth as a company, we see half of the Earth's GDP as an entire whole comes from labor, comes from people doing things, people fixing things, maintenance things, repairing things. It's hands on manipulation. And when we look at the space industry, there's been 700 astronauts in all of history. You know, there's 47, around 47 active astronauts right now. And if we look at some of the most famous photos that we all know and think about, about servicing and labor in space the one that pops to my mind and the one that's in the Detroit airport is, you know, you have an astronaut out on a tether on space shuttle fixing the Hubble telescope with a wrench. And you start to think about how feasible is this in reality for the rest of all of our large infrastructure problems? You know, the ISS is coming down at the turn of the decade, we're sending new commercial stations in its place. And it's not just one, there's multiple companies talking about it and multiple countries talking about this as well. India with ISRO and their station, uh, you have, you know, China and the Chinese station Tiangong right now. And so there's only going to be more larger infrastructure. We're talking about putting data centers up there at first small compute nodes and then large gigawatt size data centers in the future. And if you think about all of these things, they all need maintenance, they all need labor. And so this is why we see the next major bottleneck to space being that labor force being the ability to actually do things on orbit. And so, um, when we look at where the space industry is going and we take the Earth as that model, we can predict that more than half of the space industry's GDP eventually will come from labor because it's just that hard to put humans up there. Um, that's exactly what we're building. Um, Icarus was made to take the work off of astronauts in the meantime while they are there. And then in the future be able to do work in places like geo, where the radiation is so high, astronauts will experience a life's worth of radiation in an hour in GEO because of its placement in the Van Allen Belt. And so we can do work in those environments that humans literally cannot.

Speaker B: I realized something kind of ridiculous. I spent a huge part of my week in conversations, podcast interviews, sponsor calls, guest research. But after a few days, half the important details were buried somewhere in my notes or completely gone. That's when I started using MeetGig. It joins my calls, automatically records everything, describes it, and gives me structured summaries with action items and key decisions. But what really got my attention is that Meat Geek now has AI voice agents that can run, um, conversations for things like recruiting screens, support calls or discovery conversations. It honestly is an AI meeting intelligence platform. It's free to start, no credit card required. Go to MeetGeek AI and check it out. Link is in description.

Speaker C: Yeah, yeah, I had, um, I guess I prepare some questions for you for later about the future and uh, you know, going to other Places um, outside the orbit and stuff like that. So that's pretty exciting. And um, from what I read, the astronauts cost is huge. It's about 130,000 per hour, which is insane. And um, um, like why is it so expensive?

Speaker A: Yeah, so a lot of people, they point to, you know, oh my God, it must be the up mass of food and water and things like that. And actually the reality is human space flight is expensive because of safety. Um, we care so much about astronauts lives. Number one, it's very expensive to send up food and water and life support and all these things that keep astronauts alive. But people forget about the costs incurred by the hundreds of people, if not thousands of people on the ground working day to day to quarterback them through their operations and what they're actually doing, uh, on orbit. And so it comes from that mission control and the people there, alongside of all the other consumables spent with human spaceflight. And whenever something is involved with human spaceflight instead of a satellite and something goes to the space station because of all the crew and safety acceptance testing that you have to do for it to be safe, to work around astronauts for it to be safe so it doesn't explode or break or break something on the life support systems that's on the station. Everything becomes far more expensive because of all that safety testing you have to do. And so, you know, over time, launch costs dropping, actually don't drop the cost of astronaut labor. Uh, it's everything else around it that is making that price so high. And so a lot of people say well, you know what happens on starship economics and what happens when this becomes very, very commonplace. It just, the safety part doesn't go away. And so for us when we're building our robot, it's actually far more expensive to test it to make sure it hits NASA's human space flight specifications to work with astronaut crew. The H fit testing that we have to do, the battery pack that we have to qualify for a crew rated environment versus a space rated environment. Two very hard things, but one is just that much harder.

Speaker C: I see, okay. So it's all, all the combined things to keep people back and forth, logistics things, food safety, all of that stuff that makes sense. Okay, and can you give us a little bit of uh, a few examples maybe like what exactly astronauts doing today, uh, in international Space Station, uh, or else out there in the orbit.

Speaker A: Yeah, I mean I'll give you a great example. And it's one of the largest tasks that we're you know, on orbit to do in Q1 of 27. So we're handing over our robot at the end of the year to NASA and to the teams over at Voyager and ISS National Lab to get qualified and integrated onto SpaceX's rocket to take us up in Q1 of 27. And once we get there, one of the largest tasks that we'll be taking off the astronauts hands is, you know, cargo bag manipulation. Every 60 to 90 days, depending on resupply cadence, we can send up to three and a half tons of cargo to the space station. Uh, and that takes a crew of four astronauts like seven days to unpack and seven days to repack. And at this point they haven't even really unopened those cargo bags and taken out what uh, they need for science experiments and things like that. They've taken away the immediate things that are marked for immediate stowage. But when it comes to things like a science experiment, anecdotally we've heard from the astronauts that we've worked with that if there's a two hour experimental method, the first hour and 30 minutes will be spent on going to those cargo bags in different places around the vehicle to get things needed to carry that experiment out. And so that's exactly what our robots will be handling in the beginning. And then over time we've actually specced other different tasks like the maintenance tasks, things like seal inspections, taking photos. That just takes time. Things like changing out filters for water and for air, things that just take time because our astronauts are on there for a very specific purpose. There's some of the most highly trained technicians in the world, if not the most highly trained technicians. They spend years and years in training and we spend millions of dollars training them to operate in this environment. So we want them to do the research in the manufacturing that only astronauts can do. The number one cancer therapeutic right now on Earth was researched in space. And we changed the way it was manufactured based on that research. Things like the next generation of semiconductors, US Semiconductor, they have their furnace on orbit on the iss. The next generation of fiber optics. You know, a company called Flawless Photonics just pulled 11km of Zblan on orb. You have far less energy needed to push far more data because of the crystalline structure of that fiber optic cable. You have companies pushing the boundaries of superalloys and different metals that we can't make here on Earth. And so that's the stuff we want the astronauts to focus on. And so if we can offload all of those menial tasks and at the same time build up A corpus of data, of manipulating objects of unknown mass, objects of shifting of center mass. This allows us to take something massive off the astronauts shoulders, but build up this intelligence layer for autonomy that scales out beyond just the International Space Station, but to all of those other maintenance layer tasks that I talked about on the outside of the station. Whether that's maintenance, the station itself, maintenance things like data centers doing servicing of other satellites, whether that's refueling, whether that's changing out electronics boards or different science payloads, everything, everything comes down to the dynamic coupling problem of picking up another object and manipulating it dexterously while in orbit.

Speaker C: Yeah, I for some reason thought that astronauts are scientists and that's what they do over there, but apparently they spend a lot of time just doing like, um, you know, maintenance and uh, moving bags around and fixing stuff. That, that's kind of crazy. And yeah, I didn't realize even that that's what's happening over there. For me it was like a, you know, like a military tank crew where everybody has their specialty. And uh, that's what I thought. But you know, apparently it's not interesting.

Speaker A: Everybody, everybody forgets about the housekeeping, you know. Yeah, the last thing you want is, you know, your PhD to be not being doing the science that they're trained to do, but you know, literally taking the trash out in some cases.

Speaker C: Yeah, exactly. Yeah, that's way too expensive. Okay, so let's talk about your company. Um, and can you tell us exactly what you're building, how the robot looks, operates and things like that?

Speaker A: Yeah. So the best way to conceptualize what we're building visually, like in your head, if you can imagine it, is think about a drone with two bimanual manipulators and a perching arm. Or you can think about something similar to an underwater rov. And the reason why it's designed like this is so it can be free flying between visiting vehicles, like say for example, a Cygnus vehicle or a cargo Dragon vehicle that brings up cargo to the International Space Station as well as the International Space Station or whatever that that infrastructure vehicle is, whether it be an orbital carrier or warehouse or whatever might be. And so the robot that we're developing first is for those use cases that I talked about a second ago, the cargo bag manipulation, manipulating basic science, plugging in, plugging out, and doing those very, very basic tasks that eat up lots of astronauts time. And we're lucky enough to be working with the folks over at Voyager that are building Star Lab, the commercial station, to actually deploy our robot to the International Space station, starting in Q1 of 27 to test some of these ops for the first time. And in the, in between, we're doing a bunch of fractional testing. Uh, in our office we have something called an air bearing facility that allows us to replicate three degrees of freedom. So you get X, you get Y, and you get yaw, and we can test partial microgravity conditions that way. And we get to use industrial gyroscopes that allow us to test roll, pitch and yaw of the robot. And then those all culminate in a test that we're doing in about three months from now, which is a parabolic flight. Uh, a plane goes up to about 5,000 meters and it comes down to about 3,500 meters, and it falls at the same rate of gravity. And that allows us to experience free fall the same way that astronauts are constantly experiencing free fall on the International Space Station. So for the first time ever, we'll get to experience all six degrees of freedom in free fall in microgravity, uh, on our robot. And, uh, it's going to be very, very exciting. Six months. Very, very nervous. Six months. Um, but the thing that I can say is that we're already excited about the next deployments after the International Space Station and what it looks like, taking all of those learnings that we'll get from each one of these discrete tests, and then that deployment from Q1 27 to Q1 28.

Speaker C: I see. Okay. Uh, yeah, I wanted to ask you about how you test this on the, um, on Earth, because, uh, there is no gravity over there. And that creates its own challenges. Like, people can, uh, somehow fly around, hold things. Your robot will have to do kind of the same or have little, like, engines to manipulate. That's going to be interesting to see, uh, how that's going to work out. Okay, so can you give us couple of examples, uh, of like, cargo manipulation? I know that's what. That's one of the, like, tests that you're going to carry out on the station. Are there any other tests that you're trying to test, like your technology, the mechanical manipulations, uh, and things like that?

Speaker A: Yeah. So right now we're in this really awesome period where we get to work with both the folks right now on the International Space Station that operate it, the ops planners, mission control, the astronauts themselves. But then also we get to work with the folks building the next generation of space stations for NASA, um, and the commercial sector. And the things that we get to look at are all of these legacy systems The International Space Station has been orbiting, there's been a person orbiting our heads for the last 25 years, continuously. Um, and we've taken legacy aviation hardware from the 1970s and that's majority of what the space station's built out of. And everyone thinks it's this far future advanced thing, but it's been around for a long time and we've just been upgrading it and keeping it alive. Um, but these commercial space stations, obviously, number one, the thing that you look at first is flight heritage. What's been to space from nuts, bolts, washers, all the way to full systems that haven't hurt anybody, that have worked properly, haven't failed. And you go there first because you don't have to spend all that money requalifying things, safety wise. You can take it, it's worked in the past. But for the things that haven't quite worked the way that we want to or we haven't optimized, these commercial stations are optimizing. And so we're actually bringing up a NIST board with us which is, you can imagine a board with a bunch of the next generation of switches, toggles, ah, cargo bag closeouts and hatch closeouts, things like the next generation of filters. Um, and so we can actually make these, and co these with these commercial stations to what does it look like for a robot to actually manipulate this in the future? How can we actually automate this bit better? Or can we actually have this entire system removed for something that is a fully autonomous, um, so astronauts don't have to worry about in the future. And so that's a great example of something that we're excited about that isn't super core to our business right now, but in the future will be, and we can actually start to actually modify the infrastructure itself to be Robotics first, instead of just human first, or at least human and Robotics First. And I think another really exciting thing, just because it's kind of, you know, where I got my start in the industry, my very first internship ever, I was at NASA growing plants autonomously in a nano lab. And so to be able to do you uh, know autonomous plant growth chambers when you're young, that sticks with you for a while. And it was right around the same time that I was doing that that the KTRUDA research came out on that cancer therapeutic that still to this day is the number one cancer therapeutic. It made $29 billion in revenue just last year. Um, so seeing that was very, very exciting to me. And for us to have the opportunity to work with some of these biological experiments, um, that are doing protein crystallization, that are run entirely autonomously with microfluidics and just being able to plug in power and data and show that in the future, if we had an unmanned platform or even a break in crew time where we had an uncrewed period on the vehicle robotically, we could continue this sort of science that the ISS has been putting forward that you can only do in microgravity. Um, and so once again, it's not as core to us as that cargo bag movement because it takes up less time. But for me it's really, really, really quite exciting because I see this is where the application layer of space going to is creating products that you can only create in space and then bringing them down for the good of humanity.

Speaker C: Yeah, yeah, it reminds me a story of uh, uh, the Mars mission with Curiosity and Opportunity, right? Like those were really like labs on the wheels and they could do a lot of scientific, uh, analysis and figure out like, what is this land, or what is this rock, what is this? And stuff like that. So that, and that was like 20, 25 years ago. Uh, I can't even imagine what NASA, uh, and companies like yours building today, um, and to advance all the science and things like that. So that's also exciting. Uh, it's really nice to see that you testing stuff like that also to make it possible. Um, and um, my next question is about artificial intelligence. So your robots, um, use that technology, right? Can you tell us a little bit how exactly you're using it? For what? And uh, is it like for autonomy or for analysis or for anything else?

Speaker A: Yeah, so this is how we build out autonomy in our system. Uh, our main belief is that autonomy isn't something you just code into a robot. It's something that's earned, not given. And so you take the cutting edge of robotics terrestrially right now, and you look at what it is, and it comes from robot learning and things like VLAS and foundation models. And basically, if you want to think about your favorite humanoid robot that you have terrestrially here on Earth, whether that's figure, whether it's optimus, whether it's the stuff physical intelligence is doing, whatever it might be, uh, at the end of the day, they have a bunch of humans teleoperating these robots collecting expert examples of tasks. And while we're collecting these expert examples of these tasks, we're capturing video data of the robot actually carrying that task out, as well as the telemetry of the joint angles, the states, the velocities of the robot doing that task. And we can take that. And the same way that an LLM predicts the next best token or the next best word, for example, when you go ask it a question, what we're doing with VLAs and robot learning is we're taking those examples and we're actually predicting the next best frame in the video. So if I know to pick up the cup and put the cup away in the cupboard, I know terrestrially on Earth, the physics is if I drop the cup, it will fall. The interesting thing is in space we have this zero G environment where the cup floats. Uh, if I poke it, it'll float off in the distance and I can't exactly regrab it. Um, and so this is a big bottleneck when we try to use the cutting edge of where robotics are terrestrially in space, we don't have that data set en masse. And right now photorealistic sims like Isaac Sim, there's still a massive sim to real gap, no matter how much synthetic data we have. And so you have to augment this with actually true learned gathered data in distribution data in its environment. Um, and so at first our robots are actually teleoperated. Um, so we have a human operator in the Loop 24, seven that's allowing us to collect these expert examples. So a great example of a task is moving a cargo bag from a visiting vehicle to node A or one or whatever it might be and we move the bag from A to B. And if we carry uh, and collect enough of these tasks of this operator moving a cargo bag from one position to the next, or plugging in an experiment and plugging out an experiment, or doing something like changing a filter and then putting the new filter in. What we can do is we can then take those 80, 150 examples and train the autonomy. Because now we have the physics base, the in distribution data collected alongside our simulated data and our photorealistic SIM that we have. And this is how we reach autonomy for these high touch tasks. And you move from an operator that has to sit there 24,7 operating this robot to do every single discrete task to an operator that can control maybe a fleet of these robots and can actually execute those high level primitives and say, hey, I need this robot to move this experiment from node A to node B. I need this robot to move this cargo bag, uh, with X tool in it from the visiting vehicle into Columbia or whatever it might be. Um, and so that's how we're training autonomy. So we've built this pipeline, terrestrially with our sims and with all of our fractional tests. Um, and we're building or um, going to use that same pipeline with our in distribution data that we collect while we're on orbit and deploy out autonomy through this phased stepwise program.

Speaker C: Mhm. Yeah. Like AI in physical world is extremely hard. Um, I've talked with few people in the past, uh, who's building drones and autonomous vehicles and um, like augmented reality, like devices that you wear on yourself, wearables and every time it's conversation that uh, it's really hard, you have, it's challenging to collect the data. You have to combine real data with simulated uh. And uh, I think yeah, once we get through this bottlenecks it will become easier. Right now like there's probably no essentially captured videos of how this physics works in zero G. Ah, to train AI. And that's a very interesting problem to have. I'm sorry for you, you have to be the first one to solve this.

Speaker A: Uh, well, very luckily we have some of the best engineers on the team working on the problem. And I think that like as far as the people that we've brought onto the team and the engineers that we work with, we bring on people from you know, both classical aerospace backgrounds that have worked at Northrop and Blue and like NASA on the OSAM mission, all these things. But then also very uniquely we bring on people that have worked on surgical robotics to work in the operator interface and people that have worked in heavy machinery because those machines, they pick up massive amounts of earth and this shifts the center of mass of that machine and there's a massive controls problem there. And so we go select people from these different industries that can actually work on this one really hard problem on orbit. And I uh, think that's one of the most exciting things about working at Ickers and working on this problem is that it is a very, very tough problem, but that attracts the best people to work on the problem because they want to solve it. Um, and I'm lucky enough to work with those guys and learn from those guys every single day.

Speaker C: Yeah, for sure. I bet. So I also wanted to clarify that your company actually um, sells is it robot, is it services. Ah, is it specific tests that you're trying to do? What is it exactly?

Speaker A: Yeah. So right now to drive adoption, we're selling the service. Um, the last thing we want is to sell hardware and then the hardware breaks. There's adoption issues and friction where one of our customers has to train their own operator and have to go through the training process. Of these autonomous behaviors and those high level primitives. So we're taking all of that uh, off the shoulders of our customers at this point and what we're doing is saying, hey, you guys have a cargo vehicle coming in 20 days. Normally it would cost you this amount of time and this amount of capital uh, to move that cargo and unpack that cargo. Um, and we just work off percent capture model of what that looks like depending um, on each of the task spaces. And in the future as the technology matures and as we can drive adoption and people get more familiar with the systems, uh, then we'll look into on a case by case basis of what it looks like to sell the hardware and actually sell the robot itself. Um, but for right now we want to own it to drive adoption as much as possible. Make sure there's an easy transition between, you know, human labor and robot labor on orbit.

Speaker C: Mhm. Yeah. Yeah, okay, that makes sense. I also wanted to touch upon since like this is like non existent market yet, or market is just forming for this type services and, and tools and robots and stuff like that. So um, how did you even figure it out? How did you came up with this idea, uh, to break into this market and how do you know it's even real and there's going to be a lot of need for that?

Speaker A: Yeah, I think the biggest thing that we did as we were starting this company, uh, is for the first six to eight months we didn't build anything. We didn't focus on the technology. We focused on talking to stakeholders and people with the problem. So we talked to people at NASA that were ops planners, we talked to the mission managers, we talked to the astronauts themselves, we talked to everyone upstream and downstream, from the scientists that wanted to put their payloads up there, to the people building the commercial stations to understand their financials and understand the revenue model of that station. And from those conversations we were actually able to get enough commercial backing and backing from people like NASA who gave us a cargo bag from the International Space Station to test our robots on, um, to be able to see if we can unpack them and unzip them and what robotic problems we'd run into. Like the folks over at Voyager are working with us as our mission managers. Voyager previously bought a company called Nanoracks. Nanoracks has done over a thousand implementations of experiments from researchers and other companies like ourselves to the International Space Station. And they're the only company currently with volumetric space, the Bishop airlock on the International Space Station. And so working with them they say, hey, we can help you get the robotic payload up there, we can work with the astronauts with you, we can help make sure that this experiment goes well, that this tech demo goes well, and you can serve the customers and people that you need to on orbit. And so we've been able to work with all the right people and get all the right backing to push the idea and the company forward to the place that it's at right now. And then over time, as we've de risked on the commercial side of things, that this was even a problem to spend the time technically on. The amazing thing has been that the tailwinds of people looking at space as an investable piece of infrastructure instead of just research and development and exploration, um, after things like the SpaceX IPO and after things like the US Space Force being created, um, and people realizing that this is a core part of infrastructure. And we've gotten inbound from not only folks within the US government, but within other agencies as well as that are excited about the ability to manipulate things on orbit. And so it went from us very much researching is this even a thing? To getting backing to getting buy in from people within industry to then as we started to push out our first demonstrations of robots working, of doing certain things like a coast to coast demo from New York to San Francisco, things like moving around cargo bags, um, we started to get inbound from folks about the actual problem itself. And we realized it was much larger than just moving around cargo bags on the International Space Station or a commercial station. But it's this problem that's going to plague the industry as it moves forward and as more mass goes to orbit. And so it became very, very easy for us to make the bet that labor is going to be a massive part of the space industry and that the labor largely will not be human because we need the humans to focus on what they're very good at, which is the science side of things, which is understanding what happens to human biology during space travel while the robots take over the very menial stuff and extract the value out of the space industry in that very, very unique environment on orbit because it's cyclic. The more applications that we can find and produce en masse on orbit, the more money that's been put back into the ecosystem, back into the industry and that drives the space economy. And so we'll be very exciting. And you know, it happened recently where people said we need manipulation on orbit and they came to us, uh, on a very big proposal to be their partner. But we'll Be very excited when more and more people start to look at their entire logistics enterprise of whatever mission they want to do in space. And they say, hey, I need an Icarus robot for X, Y and Z in this part of this entire conops of what I'm trying to do from

Speaker C: A to B. Yeah, interesting. So yeah, this is really going to be like, uh, analogy, um, with Earth industries, maybe like, um, you know, plumbing company or construction company or something like that. Uh, when you send the whole crew basically with your cargo to do or like moving company. Right. Uh, your truck goes with your furniture and there's a couple of people who carry and place items in your house, whatever you want. Uh, yeah, that's interesting. My simply simplified mind only works with analogies like that, so.

Speaker A: No, no, I agree with you. That's the way that we think about it too. One of the very first ways that like, we were able to make sense of it and it seems like it sticks with most people is if you think about a construction site, uh, to have someone, you know, operating the, the backhoe or the dump truck or whatever it might be, it's so valuable to have this machine doing something that can arguably do it better than humans. You know, it can move so much more Earth than a human can. Um, and that's the way that we think about it. So with the more robots you have on orbit, the more that you can do, and the more that you can do, the more humans you can actually have on orbit and the more things that you can have in the space economy.

Speaker C: Yeah. And then from a product to service standpoint, uh, how do you build or manage your roadmap? Ah, like where the company will move what you need to develop next. If this is like a test stage, MVP stage and market is still forming, how do you do that?

Speaker A: Yeah, so at first with this deployment, we have our very clear set of what we need to do, uh, in the robot that we have. And you can kind of consider this as the NVP of de risking those tasks actually being able to be done by a robot. And we're working with our partners to get that done. And so we're very, very focused on the fact that in about half a year's time from now, we're handing over this robot to get sent to the International Space Station. Um, and then from the beginning of 2027 to the beginning of 2028, that's execution of those tasks. And we really, really do have to nail, uh, that execution. Uh, but the exciting thing is we know that if we do get it right. We're already working on the productized version of this robot based on the learnings that we have from that deployment and from some of the other conversations we've had and the things that we've been added to on the commercial side of stuff. We already know that we're working on the EVA version of this robot, um, and we're working towards that deployment outside of the pressurized volume of the iss. And so we let kind of the market dictate where the technology needs to go. At the end of the day, our core competency is solving the dynamic coupling problem of fine grain manipulation on orbit and that physics problem and being able to m manipulate not just cargo bags, but things like electronics and science payloads. And so what that allows us to do is say, well, we have this amazing capability and this amazing platform in the IVA environment that we can move around inside a vehicle. And the next desire from our customers and people and partners outside of that IVA environment ecosystem is the same exact capabilities of manipulation and dexterous manipulation and solving that dynamic coupling problem, but just with radiation hardened and temperature hardened hardware. And so that's immediately what the next tech dev timeline goes to. And so the way that we thought about it as a company was how from whiteboard and funding to launch, do I get this robot, arguably one of the most complex robots ever deployed, and definitely the robot with the most compute ever deployed to space. How do we get this on orbit as fast as possible? And when you talk about the IVA environment, well, you have far less hardware requirements where you don't have the radiation that you do outside, you don't have the temperature swings that you have outside, the solar radiation pressure. All these things that make space extremely hard outside of the physics, uh, are a little bit abstracted away. Um, you still have to go through a bunch of different hoops because it's now with crew instead of being outside the space station. But it allows us to bring the hardware up faster and solve the problem that we actually care about, which is that robotics problem. And once we have that robotics problem solved, then it becomes an engineering execution problem of building those hardened components for manipulation outside of the station. And there's been robotic manipulators outside of the space station in the vacuum for longer than I or most people have actually been alive. Um, and so there's a playbook on how to do this. And so we're always searching for the greenfield problem and the thing that others haven't solved before. And we're doing that in an easier environment where when things break an astronaut can come and fix it. Um, because we know that after that it's an execution problem to move to the EVA environment.

Speaker C: Yeah. Outside I guess of the station. It's also a lot of things, uh, even like think about satellites and they break and who takes care of them. Right. Uh, so that's also a big problem potentially on the horizon to deal with. Yep. Yeah. Okay, interesting. What um, kind of trade offs you had to make in terms of what functionality to build versus not to build for this phase.

Speaker A: Yeah. So the biggest thing and like as an engineer you always want to push the envelope as much as possible. Make, make the most elegant solution, um, to the, um, and I think the biggest trade off is, and this is quite funny because it kind of gets a bad rap in the media cycles and when people are talking about it, um, teleoperation, uh, for us, our customers, the people we work with, they don't care if the science experiment is plugged in by a unicorn, a robot or an astronaut. They just need the science experiment plugged in the manufacturing module, they just need it plugged in. They need the sample from that to go study. They don't care if the cargo is unpacked, ah, by an astronaut, a robot or whatever it might be. They just need the task completed. Um, and so for us, when we realized that the thing that we were most excited about was immediately rolling out autonomy and this massive, massive challenge of what that looks like. Um, and we're still very excited about it. It's still very core to our company and still one of the core problems that we're solving. But for this first deployment, in order to get it into our partners hands as fast as possible and get the capability there tested in Heritage on orbit as fast as possible, that's one of the trade offs that we had to make where we say, hey, we're actually going to focus on being really good at teleoperation, even though this sounds a little bit backwards because the more time that we spend there on making sure we have an amazing teleoperation stack that can work at the latency of S band radio, that can still give our operators really great feedback that will play tenfold into the future when we're building out that autonomy and we're collecting that data and building that pipeline for those high level principle, uh, primitive tasks like moving around the cargo bags from A to B or plugging in those experiments or changing those filters. And so that was a very big trade off in like kind of the design timeline of we Would have loved to immediately have worked on that problem, um, but now it just gets pushed to a little bit after the deployment. So when we look at the productized version of the robots, after this MVP goes up, then we can have that as a core capability inherent to the robot itself. Um, but it was more important for us to get the robot on orbit and get Heritage as fast as possible.

Speaker C: Yeah, yeah, that makes sense. Yeah, that's a smart move. You want to be testing in real environment as soon as possible and then build on top of it. That's what the MVP is all about right there. Just to get out there and start. Put your hands on it. Yeah, exactly. Yeah. Okay. And then, um, out, uh, of all those maintenance tasks and, uh, fixes cargo movements and things like that, do you think anything humans will still be doing potentially in the future? Or the plants are like automating completely everything and only like scientists go there and do the science?

Speaker A: No, for sure. I mean, like I said a little bit earlier ago, like, our core part of our thesis is the more robots that you have on orbit, the more humans you can have up there. Um, because they can do the things that truly really matter. And they don't have to focus on the housekeeping or changing the filters or making sure the life support is working. Um, they can focus on truly doing the science. Um, and I think there's so many things right now. We don't have automated wet labs terrestrially on Earth. Like, we haven't removed the scientist out of the loop here. We don't have automated factories. 100% automated factories. They don't exist. You still have humans in the loop. Um, and to say that we're going to do that in space immediately, uh, is an impossibility. And so we'll take away the menial tasks, the things that we really don't need extremely skilled labor to do. So the people that do spend, you know, five, 10 years of their life in training terrestrially before they even get sent to space, and then they get to this environment that slowly kills them. Over time, it degrades their heart tissue, their bones deteriorate the same way that if you had osteoporosis they would, um, your ocular pressure changes. Is a very bad environment for, for an astronaut. Um, and that's why, like, you see, the most time anyone's ever spent on orbit is like, I, uh, believe it's Peggy at this point. And then she's had about 500 something days. Um, and the reason is because it's a very bad environment for Humans. And so if we can have, in that short amount of time, in all of those years they've trained to operate in that environment, then focus on the things that truly matter and that truly like keytruda, will affect hundreds of millions of people, then that's the most impactful thing for them to do. And that's what we want to unlock by taking away those tasks. So it's most definitely not, you know, remove all the astronauts and it will be fully automated. There'll definitely be fully automated use cases for sure. But the human will never be removed. Out of the loop.

Speaker C: Yeah, yeah, yeah. I think it will enable more people to go up there and just expand and do more stuff simultaneously in parallel. Different sorts of things. Especially as you mentioned, um, commercial, uh, stations will be there, data centers will be there, like travel to the moon or whatever. All these sci fi things that we dream of today will be available sometime in the future. And they all need humans over there as well, I'm pretty sure. Okay, cool. All right, so a couple of like, um, sci fi futuristic questions for you. If we like look at the uh, orbit and maybe moon in the next like 20 years, how do you see these operations going in space? Robots, humans, how they operate, who does what? How does it look in your mind?

Speaker A: Yeah, like when I close my eyes and I think about this, this is the exact point that I was just talking about a second ago, where the things that we can automate with robots, like let's say, you know, these microfluidic experiments for protein crystallization for therapeutics, or things like the manufacturing of semiconductors, um, or you know, things like creating fiber optic cable to be used on Earth or the moon or whatever it might be. Those are handled in, you know, pretty autonomous factories that are just working on specific use case and purpose built for that specific use case. And that's amazing. And I see the other side of the equation where the humans are pushing the boundaries of science where we haven't found. They're looking for the new killer application that might not be protein crystallization, that might not be fiber optics. Um, and then also I Hope within that 20 year time frame, I hope eventually I can buy a ticket to get up there to the International Space Station and go visit for a week or something like that. Um, and you know, right now NASA's driving extremely hard on the moon base. That's our goal, to put American astronauts and humans back on the moon. Um, and not just to visit, but to stay. And so I think if you give us that 20 year timeline that could be a very real possibility where we're seeing humans living and working not only on, you know, commercial space stations and habitats in low Earth orbit, but also on the surface of the moon. And I think that's just a very fun future that I want to be a part of and I want to create.

Speaker C: Yeah, yeah, it's great. And also, um, it's fun to dream about things like that. Like that. Right. And see what actually will become true over there. Yeah, interesting. Okay. And also with, like, with the recent news about SpaceX, it's also really motivating to be in the space and building in this space. Uh, so, yeah, it's definitely been like

Speaker A: a massive driver for the entire industry.

Speaker C: Yeah, absolutely. I think now only a lazy person didn't hear the news and, uh, doesn't know what this is about and, and how huge it is. Uh, overnight became overnight. Right. Uh, anybody, I don't know if anybody would imagine that that's what would be happening. So, yeah, anyways. All right, Ethan, thank you very much for information. It's been great, uh, learning about what you're building over there. And at the end I usually have my innovation Q and A for everybody. So I have three questions for you. First question is, can you define innovation in a few words?

Speaker A: Yeah, I think M. To keep it very concise, innovation at its core is turning some new novel idea, uh, into a true actionable solution, uh, that you can implement.

Speaker C: Okay, great. Second question is, uh, which innovation in the human history you think changed the world the most and why?

Speaker A: I think, uh, if you look at, throughout all of human history, it has to be, uh, leveraging electricity. Um, whether that was for, you know, we had, you know, in the 1920s we had electric cars and then we went gas and now we have them back. You look at the infrastructure that you and I are talking to each other through right now, um, being able to harness that, that core part of our every single day lives at the state that we were, I think, is foundational to everything from now moving on into the future. M. When we talk about what we're doing and the reason why data centers are even looking to go to space, it's because we can use solar power at a much higher rate than you can here on Earth. And you can generate more electricity for things like compute, and you can have that power. And so I think if you talk about any innovation, it's the ability to use electricity for electronics across every single use case.

Speaker C: Yeah. Okay, great. And the last question is, uh, which technology or tool or software that we use today, we will be laughing at, ah, ten years from now.

Speaker A: Um, I think, uh, like in ten years from now, probably one of the largest things that we'll laugh at, uh, is the idea that we ever, ever, ever used a single use rocket. Uh, you know, you think about like, Werner von Braun and you think about like the Saturn rockets and how amazing they were. And like, even right now we look at them, ah, man, this is amazing. And you look at some of the small, medium launch, like vehicles that we use right now, and the more and more that space becomes a part of our daily lives, not everyone realizes that, like, you know, use GP gps, that's from a satellite, you go swipe your credit card, the PNT timing signal that payment's not going through without a satellite. You know, memory foam that was made by NASA, Invisalign braces that came from NASA technology. There's so many things that we look at now that we have no idea that the space industry's done for us. Um, and then just looking at like, single use infrastructure and being like, can we actually believe that we threw away the entire rocket after we were done with it? Or can you believe we just burnt up the entire satellite after three years? Uh, I think we'll look at that and say that was laughable.

Speaker C: Yeah, yeah. It's kind of already getting there. Yes. All right. Okay. Ethan, thank you very much for your time. I hope, Uh, I will. Well, I will keep in mind your company and check out, uh, what you're doing. It's very interesting to me with the space industry and where things are going. So that's my, like, uh, personal, um, uh, interest there. So thank you for your time.

Speaker A: Thank you so much. It's been absolutely amazing. I really appreciate it.

Speaker C: Yeah, absolutely. Take care. Bye. Bye.

Speaker B: I hope this conversation gave you fresh ideas and inspiration. If so, please share it with your friends and colleagues. Don't forget to follow, subscribe and leave

Speaker C: a quick like comment.

Speaker B: It helps more people discover this episode. See you next time.