AWS in Orbit: Empowering exploration on the Moon, Mars, and more.

[ Music ]

>> Maria Varmazis: I'm Maria Varmazis, host of T-Minus Space Daily: Empowering exploration on the Moon, Mars, and more with Lunar Outpost. And today we are bringing you the next installment of the AWS in Orbit podcast series from the 40th Space Symposium. In this episode, I'm speaking with representatives from Lunar Outpost and AWS Aerospace and Satellite, and we're going to be speaking about building systems for ambitious space objectives and how AWS can enable and support that. Gentlemen, welcome. I'm so glad to see you both.

>> Salem El Nimri: Thank you.

>> AJ Gemer: Thanks for having us, Maria.

>> Maria Varmazis: Yeah. AJ, let's start with an intro, please.

>> AJ Gemer: Absolutely. So my name is AJ Gemer. I'm the cofounder and chief technology officer of Lunar Outpost. We're headquartered in Arvada, Colorado, but we have over 150 employees across three continents, with offices in Luxembourg City, Luxembourg and Melbourne, Australia as well.

>> Salem El Nimri: Nice to meet you, AJ. My name is Salem Nimri, and I am with AWS Aerospace and Satellite team, and I'm the resident chief technology officer for that team.

>> Maria Varmazis: I'm so jazzed about this chat because it's a CTO-to-CTO chat. And as I said, I'm just really glad to be a fly on the wall for this [laughter]. AJ, why don't we start with you telling me a bit about Lunar Outpost and the incredible things that you all are working on?

>> AJ Gemer: Absolutely. So at Lunar Outpost, you know, we are the leaders in cislunar robotics and mobility services for the Moon and for other planetary bodies. So when we started Lunar Outpost, we knew that folks who wanted to operate in cislunar space were going to need mobility as a service on the lunar surface. And so we set out to develop our line of planetary robotics and lunar rovers. And really pleased to say that we flew a rover only a few weeks ago on the Intuitive Machine's IM-2.

>> Maria Varmazis: Congratulations! Huge, huge!

>> AJ Gemer:

>> Salem El Nimri: This is amazing.

>> AJ Gemer: It was. It was a fantastic experience.

>> Salem El Nimri: It's a great success and it's a huge achievement coming from a commercial company, so this is a great.

>> AJ Gemer: And so, you know, the pace that we've kept up during that time has been incredibly rapid. We now have four more missions going to the Moon with another map rover later this year, on IM-3, and culminating in the lunar terrain vehicle, providing mobility for two NASA astronauts and a wide range of NASA, International Space Agency and commercial payloads and cargo.

>> Salem El Nimri: What I like about Lunar Outpost is their approach to all of this mission in terms of building the rovers. From the design, build, test, deployments, operations, it's definitely unique. And they are leveraging the latest technology, including the ones that are provided by the cloud, AWS, in terms of achieving this. So congratulations, that's big success, and for pushing the whole industry to innovate a little bit faster.

>> AJ Gemer: And that's exactly it. We really want to see as much happen in the cislunar space as we can with the time that we have. And so to do that, as you say, we need to accelerate that pace, we need to be sure we're leveraging the latest and greatest technologies to the best of their capabilities. And all of those things add up to, you know, enormous value to our customers. And that commercial customer support in turn helps us to do more missions more rapidly and do more cool things.

>> Salem El Nimri: Yeah. And with your success, they reached TRL 9, which is test readiness level 9, and this is hard to reach. And this means that they are qualified as a company and their products to operate and launch into space and expand even more. So maybe you want to cover some of the things that you've done in terms of your latest mission and the successes you've achieved. We'd love to hear about that.

>> Maria Varmazis: Yes, please, I would love to hear that, yeah.

>> AJ Gemer: Absolutely. So this map that I mentioned that just launched a few weeks ago, we called Lunar Voyage One. It was our first map rover, about one meter cubed in size and 20 or so kilograms, and it allowed us to validate a lot of our in-house developed technologies -- things like are onboard computers, our sensing systems, our software, and especially our Stargate command and control ground software system.

>> Salem El Nimri: That's amazing. And what I like about what Lunar Outpost did is that their availability, they exceeded the availability times. How many times did you have in your availability with the Stargate system?

>> AJ Gemer: Oh, yeah, Stargate had something like 99.9998% uptime.

>> Salem El Nimri: That is impressive.

>> Maria Varmazis: How many decimal points is a reality there?

>> Salem El Nimri: Exactly. That is impressive, that is impressive.

>> AJ Gemer: Well, and what's great about that is like that exceeds the uptime reliability requirement even for class A and crewed mission. So this is a very safe system that we've now proven out and achieved tier L9 on the lunar surface, and we can apply that to the lunar terrain vehicle when we actually have astronauts on board who can't afford any downtime.

>> Salem El Nimri: And we're happy that they built the Stargate system. And a lot of their designs and simulations on AWS, leveraging AWS services that we have, from AC2 to EKS to DynamoDB and more. So it's really exciting. And I like what you said about the communications and your partnerships with Nokia. Because communications to the Moon are challenging.

>> Maria Varmazis: And it's also a huge need, yeah.

>> Salem El Nimri: Yes. If you want to look at it from the way it is done right now, you are dependent on the infrastructure that is on Earth on the ground with the Deep Space Network, especially for tracking and navigation. So that brings a lot of challenges. And these guys are working in an area where it is really difficult. It's not like there is a GPS system around the Moon, that doesn't exist. And you can't bring a compass and put it on the rover, because it's not going to work. Because here's the fun fact about the Moon: There is no magnetic field around it.

>> Maria Varmazis: Right, right.

>> Salem El Nimri: So nothing is going to work, so you are dependent on this infrastructure. And you guys are basically pioneers when it comes to being on the lunar surface.

>> AJ Gemer: I think here on Earth, we often take things like GPS or high-bandwidth communications a bit for granted. It's easy enough to take a 4-K video and send it to your friend. But to do the same thing on the Moon is much more challenging. And so these are the types of technology challenges that we attacked right away at Lunar Outpost, worked on our solutions, as you said, tested them very thoroughly, learned from that testing. And as we saw on our last mission, you know, the technologies are ready to go and ready to unlock the Moon for all kinds of commercial and scientific uses.

>> Salem El Nimri: Yes. Can you speak to us about the science experiments with LV-1?

>> AJ Gemer: Absolutely. So on LV-1, we had what was called the "resource camera" that was provided by MIT and NASA Ames. And that was a multispectral imager that had the ability to take pictures of the lunar regolith or rocks on the surface, maybe if we were lucky even some ice in a permanently shadowed crater and characterize that and return science data. Now, as we know, the IM-2 mission tipped over, the lander tipped over on top of our MAPP rover. So we were not able to actually drive around on the lunar surface. However, the MAPP rover did survive that hard landing and operated for almost three hours in a permanently shadowed crater.

>> Salem El Nimri: That's impressive.

>> Maria Varmazis: That's impressive, yeah.

>> AJ Gemer: Yeah, it's one of the harshest environments in the universe, and now we know that map can handle it.

>> Maria Varmazis: Dang.

>> Salem El Nimri: It's a leap. You always learn from each success, you're pushing the envelope. I'm excited about all the artificial intelligence integrations you guys have in your plans. So if you can give us a hint or a sneak peek on what's coming down the pipeline?

>> AJ Gemer: Well, on the topic of navigation, that is an excellent application for AI and ML on the Moon, for a number of reasons. First of all, although you've probably seen some pretty pictures of the Moon, and they look like very high-resolution, in fact when you zoom in, the very best resolution is usually only about two meters per pixel. So you can imagine for a rover that's a half meter long or even a meter long, that's not sufficient resolution to plan your path out entirely in advance. It gives you a good starting point. And we have some digital elevation maps so that we know what slopes to expect and things of that nature. But really, navigating on the Moon is all about taking in new information, processing it quickly, and updating your plans in near real-time on the edge, on the rover. And so that is an application that AI is really well-suited to identifying new obstacles which did not previously appear in prior maps, and then charting a new course, a new safe course, around them. And it's not just obstacles like rocks or craters, it's other things like areas of shadow or of sunlight. The MAPP rover is solar, and so passing into a shadow for too long will cause our power to run down. And so taking in these really multivariate problems and effectively doing a multivariable optimization to find what is the safest path, taking into account, you know, lighting conditions, thermal conditions, RF communications, as well as the terrain itself, is an excellent example of how we can employ AI.

>> Salem El Nimri: And I like what you said, "permanent shadow." Balancing that, that power cycle, charging the batteries, and moving forward, that's really impressive. And I like what you guys are doing, I guess. You're taking pictures and you're building these 3-D terrains and then you navigate the obstacles and figure out.

>> Maria Varmazis: Is that Stargate; is that what that is?

>> AJ Gemer: Yeah, Stargate has those functions for building out essentially a world map as the rover drives. And actually, something else that we can do there is use a digital twin of the MAPP rover in simulation to test out a variety of different paths that we could drive along before we actually command the rover to do it.

>> Maria Varmazis: Wow, that's cool. I was going to ask about how y'all use AWS to enable any of this. Can you tell me a bit about that?

>> AJ Gemer: Absolutely. So it's been absolutely great working with AWS, I'll say.

>> Salem El Nimri: It's mutual.

>> AJ Gemer: You know, certainly in addition to the spectacular stability and uptime that we enjoyed on our mission, I would say, you know, having that robust and reliable backend has allowed our engineers to focus on innovation and solving those, you know, very specific problems that are specific to operating a rover on the lunar surface. We know we can trust AWS to support our systems to work flawlessly on a mission, as we saw. And that allows us to give a thousand percent focus to the mission at hand. So we had an excellent mission with Lunar Voyage One, and we're looking forward to our next MAPP mission, Lunar Voyage Two, at the end of 2025.

>> Salem El Nimri: And I love the way they built their Stargate system. If you think about it, I think it should be a model, because you can replicate it and simulate what is coming in the future, operate for now, or collect the data and run it back and see what happened. So it's like almost you can do a pre, post, and current status all at the same time leveraging the resources that you have.

>> Maria Varmazis: That is pretty remarkable, yeah.

>> Salem El Nimri: Yeah, so that's really good.

>> AJ Gemer: Very true. And we had a very interesting experience with that during mission operations for Lunar Voyage One. When you're in the mission operations center -- and I, you know, was privileged to serve as one of our four flight directors with a whole team of operators -- the data is coming in so fast. And, you know, we've set up Stargate to make it human readable and, you know, easy to interpret, make it operational so you can make decisions based on that. But still, after-the-fact, as you said, you're always going to want to go back and look at, you know, inspect that data very closely and see what sort of trends and things you can learn from it. And I expect the data analysis even from Lunar Voyage One to go on for months or even over a year. And so, you know, the AWS support of the Stargate system makes that possible and makes that easy and intuitive to do. Every time someone has a new question about something we saw on the mission, we can go right in, pull that data, you know, visualize it in many different ways, and draw our conclusions from it.

>> Salem El Nimri: Yeah. And I love what you guys are doing, I mean, for humanity. The Moon is not a destination, it's a stop, it's a first stop. And from everything that Lunar Outpost is doing, all of us, we are learning so that we can prepare to the next milestone and the ones above it. And I really look forward to seeing how we can help Lunar Outpost with our AI systems and services that are built on AWS. I mean, I would love to pick your brain on what you see the future is coming for AI and machine learning for lunar exploration.

>> AJ Gemer: Oh, absolutely. So in addition to the sort of on-mission functional applications I mentioned with obstacle identification and charting new courses, you know, we have applications for AI in our technology development and things like predicting component reliability and things like that, which are increasingly relevant for longer and longer missions. So for example, the lunar terrain vehicle is designed with a lifespan of 10 years on the lunar surface. And so there's a lot of predictive analysis and modeling that goes into building a vehicle that can withstand that harsh environment for that long. And so these are good examples of how we can take and test data as we develop our technologies and our systems for LTV and ensure that, you know, we have the reliability and the lifespans that we need on the Moon.

>> Salem El Nimri: I love it. And when you look at the future of what's coming down right now with agentic AI and all the agents, I can imagine one day we will have an agent that serves as a geologist, an agent that serves as a geochemist. Of course, you learn everything from what we have here on Earth and deploy those systems to the rover so that they can navigate and do things on their own. And you'll find the rover that talks to build an agentic mesh within it to say, you know what, I'm going to go on my own, I can chart my path. And then the agentic system for the geologist that is loaded on the rover says, I see a feature over there that I think we should go there. It talks to the agent that is for navigation and says, you know what, this is the safest path based on the cameras and the 3-D map that I developed. So I love that we are heading in that, and we are leveraging also the learning that we had on Earth to bring it more close by. Because this is needed when we go also to Mars.

>> Maria Varmazis: Yeah.

>> Salem El Nimri: I mean, with the Moon, it's like, what, two to three seconds delay? When we go to Mars, it's going to be seven minutes to 40 minutes, depending on the orbit.

>> AJ Gemer: So bringing Mars into the conversation, it's been fantastic to have all of the learnings that we have from the Mars rovers. But there are also some very important differences when doing science on the Moon. So on the Mars rovers, we have these rovers with very long lifespans. They have a relatively similar to Earth day and night cycle, and fairly benign temperature extremes. And what that means is you have the luxury of time to pick out your science targets. So the Mars rovers have a large science team. They're constantly poring over new data, new images, things like that, and flagging where they want to go next. And sometimes those operations just to get to the next science target could take many months. On the Moon, because you have the 14 Earth day day-and-night cycle, you need to have much more rapid operational decision-making. And that certainly includes science investigations. So something that we can do with AI on the Moon is help extend humans' perception, their sense of the Moon and of the environment they're operating in. So, you know, a human can see in certain wavelengths, they can see out to a certain distant.

>> Salem El Nimri: I love it, yes.

>> AJ Gemer: But things like very challenging lighting extremes, the harsh shadows on the lunar surface, seeing things far off in the distance or at the microscopic scale and in other wavelengths essentially allows human scientists here on Earth to understand that environment much, much more quickly than ever before. And that's how we can maximize our science decision-making as we traverse and explore on the lunar surface.

>> Salem El Nimri: That's exciting. And one thing, before we came here, you told me about your next mission that's going to be around the equator of the Moon. It's completely different than the South Pole.

>> Maria Varmazis: Yeah, yeah.

>> Salem El Nimri: So I would love it if you could highlight on some of the things that you're doing there. And hopefully like with AWS, we'll try and support you guys to make it happen.

>> AJ Gemer: This next mission, which is called Lunar Vertex, and is funded through the NASA PRISM program -- it's actually the very first PRISM mission, PRISM 1A -- will go to this site called Reiner Gamma. And that's at about seven degrees north latitude. So as you said, very close to the equator. A very hot environment compared to the Lunar South Pole where Lunar Voyage One went. And we will drive around at this Reiner Gamma site and explore the local magnetic field there and hopefully determine where it came from.

>> Maria Varmazis: This has been so vaccinating. I have to ask, when we look towards the future and zoom way out at the incredible challenges that you all are taking on at Lunar Outpost, I mean, when you think about establishing human presence on the Moon and beyond, like what is driving the vision that you all are going towards? And, you know, what does scale and speed mean to all that?

>> AJ Gemer: I've heard the saying, "space is for everyone," right? And I find that to be particularly true. It brings people together across all kinds of borders and boundaries. I think humans have this in-built desire to understand the universe that we live in. I know I certainly feel that. And so with the time that I have here, I want to understand and learn and see what exactly is out there. But more than that, I want to set the stage for the future as well. And that's where the speed and the scalability comes in. There are technologies, services, and capabilities that once they are established will become, you know, key enablers, and we'll see just an explosion of economic activity to fund the exploration, as well as scientific activity. Every time we do one of these space missions, we learn something new and unexpected. And immediately the people who work on it want to go back to the drawing board and design the next mission that investigates that new piece of information that much more thoroughly. So the faster that we can do that, the more cost-effectively we can do that, that's what Lunar Outpost really seeks to enable. So we're looking beyond the Moon already to Mars. Our lunar terrain vehicle, the Eagle, will be equally capable and operable on Mars as it is on the Moon. And so, you know, that provides an excellent mobility platform for rovers and explorations of the future to Mars and even beyond.

>> Salem El Nimri: Thank you guys for all the great work that you're doing. I have one question. I'm interested in your opinion about like what is the next rovers you're going to build? Some people say they're going to go for drones. I know there is no drones that you can operate on the lunar surface, so I call them "hoppers." Are you planning to develop some of these so that you can cover bigger distances, go inside craters, and do more exploration? I'm interested to hear what you guys have in your inbox.

>> AJ Gemer: Absolutely. And in fact, if you zoom out and you look at the big picture, where I see us going next is swarm robotics and cooperative robotics, right? Especially robots that are heterogeneous, that are different from each other, and each have unique capabilities. So an example I often like to give is, in the past, say the Mars rovers, you have one extremely high-value robot operated by a large team of humans. But we would like to change that paradigm to where one human can operate many different types and a large number of robots. And that's how you maximize the exploration, maximize the workforce capability that a human can provide on the Moon. And like you say, this addresses some really cool lunar challenges and explorations. One of my favorite is lunar lava tubes. So we know for a fact there are lunar lava tubes with skylights, you know, that we can kind of peer down into. But, you know, mankind has never explored inside of those lava tubes. And that's an example where swarm robotics would give us capabilities to, you know, see what's down there.

>> Maria Varmazis: Get in there, yeah.

>> AJ Gemer: It could be anything in there.

>> Salem El Nimri: And I love this. That's a great idea, swarm robotics. It will enable you to deploy more robotics, and you don't need to wait for many years to build the perfect robot that has all the instruments and payloads on it. So this way you can just deploy them. And if you connect them through an agentic mesh, they can talk to each other and say, you know what, I discovered something here, and that rover can come over and basically explore and help us and you just move.

>> AJ Gemer: Bring in a different robot that has different sensors or something that is well-suited to whatever that first one identified.

>> Salem El Nimri: And this way you can extend it; you can explore further distances. Because the first robot will say, I finished here, I'm going to go somewhere else, and the second one will come in. Well, this is amazing. This is great.

>> AJ Gemer: And swarm robotics also helps solve a number of the challenges that we're facing today on the Moon, things like positioning, navigation, and timing. The more nodes, the more communication relays that you have, the more timing apparatus that you have to, you know, accurately identify the location of any one of those swarm robots, you know, it helps refine that model that we have of the lunar surface. So simply by virtue of having a swarm of robots on the Moon, we will understand and learn and discover new things almost regardless of what the sensing capabilities of any individual robot is.

>> Salem El Nimri: They can all talk to each other.

>> Maria Varmazis: Amazing. Gentlemen, this has been so fascinating. I want to give you both an opportunity to provide any concluding thoughts before we wrap up. Salem, why don't I start with you and then, AJ, you can finish us out.

>> Salem El Nimri: Well, it's been a pleasure working with the Lunar Outpost. They have a team of geniuses. And to be able to sit here and say, with AWS, within AWS, we are developing services that are reliable and can service and help explorers such as AJ and others, it's just a privilege. So thank you very much, AJ, for that.

>> AJ Gemer: Thank you. And likewise, back at you. I mean, having AWS support on this, you know, this is what the new cislunar economy looks like, is having these capabilities, being able to work together rapidly, each of us solving a piece of the puzzle, and supporting each other. And, you know, the overall outcome is just an absolute, you know, sea change in what we can do out in space. And that's what we're on the precipice right now. So it's great to be a part of.

>> Maria Varmazis: Thank you both so much. I appreciate your time today.

[ Music ]

That's it for this episode of AWS in Orbit by N2K Space. We'd love to know what you think of this podcast. You can email us at space@n2k.com, or submit the survey in the Show Notes. Your feedback ensures that we deliver the information that keeps you a step ahead in the rapidly changing space industry. This episode was produced by Laura Barber for AWS Aerospace and Satellite and by N2K producer Liz Stokes and senior producer Alice Carruth. Mixing by Elliot Peltzman and Tré Hester, with original music and sound design by Elliot Peltzman. Our executive producer is Jennifer Eiben. Our publisher is Peter Kilpe. And I've been your host, Maria Varmazis. Thank you for joining us.

[ Music ]