How Alex Fielding and Privateer Space Are Taking on Space Debris - Ep. 196

0:10Hello, and welcome to the NVIDIA AI Podcast. I'm your host, Noah Kravitz. Take your protein pills and put your helmets on, everybody, because today we're talking about space. My guest is Alex Fielding, the co-founder and CEO of Privateer Space, a company that's on a mission to improve life on Earth by monitoring and cleaning up space debris. Privateer is creating the data infrastructure that will enable sustainable growth for the new space economy by allowing space operators to maneuver safely and effectively. Privateer is also part of NVIDIA's Inception program, which supports startups that are accelerating the adoption of AI across various industries.

0:48Alex is a serial entrepreneur who has worked with Apple co-founder Steve Wozniak on several ventures, including Wheels of Zeus, Ripcord Networks, and now Privateer Space. Alex has a background in engineering, robotics, machine learning, and AI, and is here to share his wealth of experience and insights on how AI can help us solve some of the biggest challenges facing humanity today. Alex, thank you so much for taking the time and welcome to the NVIDIA AI podcast. I am thrilled to be here. Thanks for making time for me. That's music to my ears. This is going to be fun. Let me turn it over to you.

1:23Why don't you start by telling us, you know, what inspired you to start Privateer Space and what the company is all about? Privateer is our, I say our, it's my third venture in space. I always thought I'd spend my entire career in the space industry, but life threw me a curveball and I ended up at Apple instead. So when I started my career, I was actually working on communication between satellites at different altitudes. So kind of Leo's, MEO's to Geo's, specifically to avoid atmospheric delay one, you know, sending signals home from various, various altitude satellites. It's the easiest way to kind of say it.

2:00But, you know, shortly after leaving Apple, I got really interested in building very, very small, low cost GPS kind of tracking tokens, things like Apple tax 20 years too early. Right, right. Was and I had a chance to play together at Apple in the 90s. And I said, hey, I'm working on this really interesting project. I think it would be great. Kind of, you know, help you keep track of your kids, your dogs, do fun stuff. We had a bunch of really interesting like gamification strategy back then for how people would use location as a part of their everyday lives. But this was before you could get GPS and a cell phone.

2:38This was, yeah, this was 2001. So we started the company and we started learning a hell of a lot about GPS. And one of the things that alarmed Woz and I was that, or was in me, was that half of the stuff we had in space in 2001 was trash. So half of it was dead satellites or we barely knew about debris in 2000. We were just tracking live satellites, dead satellites. Half the satellites on orbit were dead. And we used to joke with each other all the time over milkshakes that, you know, hey, someday we would be the first space, you know, kind of sanitation engineers and we'd ride up the back of the space trash truck and we'd toss these dead satellites into the compactor.

3:23And, you know, we were being somewhat facetious, but we knew that it would be an issue at some point. And fast forward 20 years later, my company Ripcord that we actually started at NASA Ames Research Center while I was at NASA, I was incubated in a weird way by the government. It's the closest I'll ever get through a tax refund. Starting Ripcord was an absolute blast. And we started in a concrete asbestos bunker, NASA Ames, with some other wonderful, crazy people like the founders of Cosmogia, which became Planet Labs. Made in Space was next door. One of the people on the Cosmogia team actually left to go start deorbit space in Italy, the magnificent Luca Rossettini.

4:05And our bosses at NASA at the time were Chris Kemp, our CTO, who's now the CEO of Astrospace, and Pete Orden, who's at the Breakthrough Prize, is just an absolutely brilliant character. And, you know, guess what? Just like the mafia, even though I left NASA to go work on vision-guided robots and deep learning, that company is on the way to go in public. It's got a new CEO now, and they're still doing great. I'm still on the board, and I think I'm still an employee. I probably would have fired me a long time ago. But this has given us a really interesting opportunity to go work on something else.

4:41And Woz called me out of the blue and said, hey, I'm really getting serious about looking at space debris as a topic. We should think about how we could go clean it up. And I was like, who's we? I got my hands full. This company's going public. No, man, jetpacks and, you know, rockets and backpacks. We'll throw them into the compactor. Yeah, exactly. And he was, you know, he was serious. And I said, well, I haven't given it, you know, much of a thought beyond the fact that it's an issue. So that's not, that's a half a thought. Right. And he said, well, just, you know, do me a favor. Go do some research.

5:17Is this possible? How would you clean up space? Who's working on it? Is there any way to get paid to do it so we don't end up building a nonprofit? Like, I mean, don't get me wrong. Steve is incredibly philanthropic. But I think he was realizing that this is not something you can undertake really lightly. You've got to really clean up and go. And I started doing research and what I found was somewhat terrifying, which was that we have far, far, far more debris in space now than we ever have. And every launch adds more debris. I mean, things even from the dispense of satellites, the housing, the packaging, the pieces that blow off an upper stage, those things float around.

5:58And if they are low enough, gravity and drag will bring them home and they'll mostly vaporize in the Earth's atmosphere. But other things, really small things like frozen propellant or paint chips or nuts and bolts and screws, we can't even see that stuff from the ground. So I went back to Steve. I said, well, there's a number of challenges. One, a lot of this stuff is very small and it's very fast moving relative to Earth. So, you know, you're talking about, you know, MVA or MV squared, but still plays a factor in space, right? So you now have a nut bolt or screw that's bigger than a bullet that's doing, you know, 7.7 kilometers a second.

6:37So it's roughly 10 times the speed of the bullet, but it's the same mass. So now you have an incredible, you know, impact potential, right? So we can't see them. We don't have a really good strategy for grabbing them. and they're they just go in the the known unknowns for risk you know that's right right so i said well we've got this problem the other problem we have is that believe it or not we don't agree on where things in lowered space are this is going to be fairly alarming for some of the listeners on this podcast but i was running with it wait what do you mean but yeah please continue well the reason for that is that there's a couple reasons so i won't go down the rabbit it holds too deeply.

7:21But one of the reasons is that when a satellite is dead or inactive, it doesn't have a transponder. So it's not talking home with us. We have to observe it. And then once we observe it, we don't have constant observation of our near-Earth space. So it's kind of like taking a radar dish that would look for an airplane and tilting it up at the sky at like plus or minus 40 degrees. And then you're waiting for objects to blow in and out of your field of view at 17 ,000 miles an hour. So this would be the equivalent of telling like a 747 pilot, there's a pigeon off your left wing at 300 feet, you know, turn right five degrees to avoid.

8:02And, but those are, those are the size comparisons, but they're three to 500 kilometers up for the most part. There's a, there's a high danger band between 300 and 600 kilometers where we got a lot of stuff. one of the challenges we have is that all those observational viewpoints you can't take into account the things you can't see or that you can't measure how do you you know if you can't measure it you can't monitor if you can't monitor you definitely can't predict it so we have these problems where we don't know the size the shape the material properties of the object so we don't know their actual astrodynamics we can't take an off-the-shelf astrodynamics model that would be somewhat predictive and say, let's guess where this thing's going to be in 20 minutes.

8:45Right. It's actually just the thing tumbling and or spinning at some interval that we can't measure. And now, guess what? When we hit it and we measure it, we think we know where it is. And then a few minutes later, wherever we predicted was going to go, we used models that assume that these things are bowling balls or spheres. And we don't have any satellites even that look like Sputnik anymore. So these things are like Coke machines with wings. So as a result, we are always wrong or almost always wrong when we try to predict where a body in space is going to be the next time we see it. And then finally, the impacts on it from radiation, from drag, from its actual angular momentum, those things change where it's going to be.

9:30Yeah. And we don't have a model to predict what forces from outside are being pushed on the object. And to kind of pause for a second, just to clarify the danger, obviously, debris floating in space, to me anyway, is just that's a bad thing, not a good thing, period. But what's the danger to life on Earth, to Earth itself, posed by all of these things floating around? Some of them, as you said, getting drawn in and vaporized before impact. Some of them not. Is it that they might all fall out of the sky? Is it that they're going to pollute the atmosphere? I'm sure there are things that I never would have thought of that you're wondering, should I say this on the air or not?

10:14No, it's top of mind for all of us because our hearts are in this place where we want to become a space-faring civilization. we want our future of humanity to be able to leave this planet to explore life and other opportunities elsewhere right but we are dangerously and perilously closed from cutting off this domain okay and the reason for that is and and it does impact life on earth we we do so much science from space today we have so many communications from space today there are these services that are in space, we don't do a good job sharing with, you know, kind of planet Earth where the services that we use are provided from.

10:55Like, you know, use your credit card at a gas station. Is it being processed by satellite? Is it your pure financial transaction getting done? Is it your starling dish at home? I mean, that one's kind of obvious. You have a dish on your house pointed at the sky. Okay. That's probably, you know, your satellite radio, your television, or something else or internet connectivity. But what people, I think in general, haven't realized yet is that everything that we have sent to space will ultimately, for the most part, most of it will come home in some way, shape or form. And we have some bumpers around how we predict and plan when space objects come home.

11:34What maybe wasn't totally obvious is that there's no obligation for things to burn up in the Earth's atmosphere. So, you know, friction on the atmosphere is totally dependent on the speed at which you reenter the Earth's atmosphere and the friction against it. It's kind of like, you know, hitting the ocean at 17 ,000 miles an hour. You're still going to burn up. So the thing that isn't always clear is that all of these satellites will come home. And today we don't even we that I'm using the royal space community. We do a good job, even though we know the morbidity and mortality rates. We know how much of the object will likely survive within some range of Monte Carlo values and simulations.

12:16We don't track, will those pieces that survive cross commercial air traffic? Right. Would they land on the lawn at your elementary school? Sky lab when I was a kid, I'm getting flashbacks to like, this thing's going to fall out of the sky and crush us. So we've got a picture painted. So this is where I ask, what are you going to do about it? What is privateer doing? How are you doing it? No, I think the general public only heard half of the privateer story. So when we started privateer, we said we have to remove the uncertainty. We have to begin to remove and solve for the known unknowns so that we can make space safer.

12:56So ultimately, we can make it more accessible. Yeah. And what does accessible mean? If you go to wayfinder.privateer.com, shameless plug, I'm available for free. You don't need that. It's cool. We were talking before we hit record. It's super cool. What you'll see there are a bunch of little dots, but what most people don't do is they don't zoom in and they don't click on the dots. So if you do, you'll get our near live real-time tracking of all those space objects. Soon you'll be seeing airplanes as well. You'll be seeing boats very soon. And if you keep zooming, you actually end up at Earth observing imaging.

13:32So you'll end up at basically Google Earth if you keep zooming in in Wayfinder. Right. And the reason for this is that while the focus immediately is stop people from crashing into each other in space, make sure that space remains safe and accessible by helping people maneuver around each other when at risk, the longer term play for privateer is to make sure that we make space data accessible without people needing to launch a ton of stuff into space when we already have things there that do the same thing. You know, we've got 50 something basically satellite based cameras looking at Ukraine right now.

14:06Right. They're not looking at totally different parts of Ukraine from their perspective. Why can't we share? My four-year-old daughter does better at sharing than the space community does. So Privateer is unlocking space accessibility by making all of the data that you see in Wayfinder, all of those images from our own data sets or from our partners available through APIs so that you can purchase the image. You can run image classifiers analysis on it. Maybe you just care about coastal erosion near your home. Maybe it's an ESG application. Maybe it's a global supply chain issue and your company has boats and, you know, tractor trailers and airplanes and things in their fleet that move things between locations.

14:49All of that stuff is available through Privateer, through APIs that you can consume through data exchange, and you can use these capabilities. So we're really going from space accessibility as the end goal. You have to solve safety to make it happen. There's a service on Wayfinder you can click on called Crow's Nest, which will show you the top 100 things that are closest to crashing into each other in space, ranked by 3D probability. and those predictions are good for 72 hours ahead in low earth orbit and seven days ahead in geostationary orbit and it's incredibly powerful because if you look at the 3d probability of those objects that they're going to collide and then you look at their distances and you look at their speed you can visualize the conjunction at the time that we think it's going to occur you can model and simulate it in the software and you can also look at the conjunction data message that has the long-term state vectors to be objects that tell the space operators how to avoid a piece of debris or how to avoid each other so they don't crash into each other.

15:51And we're not charging for that. I mean, I think companies charging each other to not crash in space or basically space extortionists, they shouldn't exist. We were joking that it's too bad this isn't a video podcast, given, you know, what we're going to talk about. And I'm looking at Wayfinder as you're talking and just encourage folks who are listening to go hop on wayfinder.privateer.com. It's really, it's fascinating to look at, even if like me, my knowledge of the domain is very minimal before starting to chat with you. So you described before some of the problems with, you know, how do you track an object that with models that were built for something much, much bigger and, you know, perhaps slower moving or just all of the, you now aware of all these different parameters that go into successfully tracking something precisely out in space that the models and take into account.

16:38So how are you solving? Are you building, obviously we're talking about AI, it's an AI podcast that's in the stack, but are you building hardware? Are you building new, you know, I'm gonna use the word cameras to be accessible, but whatever the equivalent is of an object that will look into space to see these things. What's the, I don't know, what are some of the technical things that you're working on or perhaps, you know, challenges that have come up that AI has been a part of getting past? AI and deep learning have been kind of pivotal technologies that we're using because there are so many variables.

17:16And those variables are, we're starting to use these technologies to improve space safety first. I mean, beyond that, it'll go to dynamic tasking and efficiency and all the great things you can solve for by solving safety. The safety things, there are things in Wayfinder and in the APIs that are available today that do things to predict and avoid collisions. So maneuver planning, telling each space operator, where should I maneuver to? That's the lowest risk position. And how do I get there? So the weird way it's like turn by turn navigation for satellites that they can either connect to their own pointing and propulsion systems or they can use ours.

17:54So we can connect directly to the space vehicles to maneuver them autonomously to avoid collisions. The other thing that deep learning. Just to stop you for a second, Alex, that's one of those things that my brain simultaneously like, oh, it's like a remote controlled car that makes total sense. I get it. And then the other part of my brain was like, it's a remote controlled satellite. I can't process this. In a weird way, it's like a remote control. It's like a drone. Yeah. Because they're in a 3D environment. And the challenge that they have, the drones don't generally have, is that there are many, many, many risks in their environment that are in that 3D box.

18:30and when you go to do a maneuver plan you're actually going to move your your remote controlled drone or satellite you then want to take into account the probability and risk of the maneuver you're doing so you don't want to maneuver into a space that's more risky than the one you're in that you're trying to avoid something in and you also have to take into consideration things where deep learning and ai are really helpful like operators don't want to burn fuel for no reason and fuel isn't in space. You have a very finite capacity for fuel. You want to extend your mission life. You want to make sure you have enough fuel to make it to the end of your mission life and accomplish the things you want to do in space.

19:07And you also don't want to, at any time, make a maneuver that is going to place your vehicle in higher risk with either another vehicle or debris. And what I think many people don't realize yet is that not all satellites even have propulsion. we have hundreds of satellites that have no propulsion they're they're placed and then they're naturally orbitally decaying and the original plan was over 25 years they would come from space home right actually we would use we would use you know all those wonderful things like like uh you know dragon radiation to help us get the thing uh including gravity at some point right so the first thing that we're doing is is using deep learning and ai to help predict and avoid collisions which that you can go demonstrate in the software live or you could subscribe and we do charge people to do their automated maneuvers for them definitely not extortion it's cheaper than probably having scientists go uh sit there and figure out where the lowest position maneuver is um and the other thing is is improving space weather forecasting because uh deep learning algorithms you know analyzing this data from space weather sensors to predict solar events solar flares, you know, coronal mass ejections, these types of things, they pose a threat to spacecraft and to astronauts on orbit.

20:26So we want to make sure that we take that into account. You know, certainly we talked a little bit about kind of autonomous navigation and enhancing autonomous maneuvers. And we do have a benefit that like Tesla doesn't have, right? When you're in space and you're maneuvering, we're not really worried about pedestrians. We're not worried about potholes. We're not worried about road closures, right? Only worried about the limited number of things in our environment that we actually can sense and detect. So call it, you know, roughly a hundred thousand things that are in play. They just happen to be moving very, very, very fast in a 3D, a 3D box.

20:59Though it's the equivalent of having, you know, air traffic control for space, maneuver planes where yes, you do have to worry about pigeons, right? It's an unusual thing. I don't know if that really answered the question on safety. Accessibility, almost all the data you see in Wayfinder, the Earth observing imaging, synthetic aperture radar, LIDAR, standard RGB high resolution photos of the Earth, all those things connect to image classifiers. So maybe you want to count the cars and parking lot. Maybe you want to look at the boats in the ocean and use those for new novel applications you couldn't do before.

21:38Maybe you want to combine signals intelligence so you can track people doing illegal commercial fishing or, you know, polluting our oceans or doing something that they shouldn't be doing on a construction site at night. There's a lot of applications there that are tied to climate. There's a lot of applications that are tied to making businesses more efficient. And the thing that makes Privateer so unique is that we help enhance the capability for people in space already to bring their technology to market dynamically so we can task those vehicles and use their capabilities. and soon we'll be doing uh you know even more with our own capability on orbit we've got a vehicle going up this november so we're all kind of on on high alert right now getting ready to ship stuff to the launch pad and ask to get it bundled up to go fly so and that will have nvidia hardware on board with some gpus doing some really interesting work on orbit and some things faster than sending the data down to earth and processing sending the results back up deep learning way out on the edge yeah just space yeah exactly i wanted to go back to something you mentioned before about uh the difficulty the difficulties of human beings sharing things with each other sometimes broadly and you're mentioning about sharing data in the space community you're talking about um you know you use the analogy of air traffic control for space and it made me think okay so you've got air traffic control towers at i don't know every airport and airfield or however many there are, you know, all over the world.

23:11Who decides who controls what parts of space? Or really, I think my question is, how do you and slash what's it like working with different governments and space agencies and regulatory buyers or whoever else it is who might have a say in, you know, well, why is this particular company air traffic controlling for all of space or whatever, however you sort of divvy up the regions. What's that whole process like, you know, both for you and maybe if you can speak to the broader space community about, you know, people who we've divided up ourselves into different nations and nation states on Earth trying to figure out how to manage space.

23:51It's really contentious. So I'm not going to give you a good answer because there is no good answer right now. We're dealing with the Wild West of space right now. and not every nation on earth that launches things into space signed the international space act or treaties sure um or the outer space act and treaties so what that means is that not every nation state that put stuff in space actually agreed that they would not weaponize space that they would not put nuclear things in space that they would not do hostile actions in space and we've seen time and time again that not only do nations do the same things they mimic the same bad behaviors they have on earth and space but they have the same hostile aggression towards each other in space which is it's heartbreaking yeah if you can't make space safe and accessible and if you can't get nation states to cooperate with each other on even coordinated maneuvers at a very basic level you're in deep trouble and space is the wild west and regulation in space is not as good as on the oceans and international water, even though we took 19, oh my God, 1500s maritime law is what we used for the space treaties.

25:07So you fly a flag behind your satellite effectively and you represent your nation state. Well, imagine that. Imagine a SpaceX satellite launched from the United States and a SpaceX satellite launched under Kazakhstan collide with each other. Is Kazakhstan at war with the United States over a hospital, you know, an accidental action? Right. These things haven't been solved for, nor has right of way. So it's crazy it sounds. I mean, if you're a sailboat on the ocean and there's a cruise liner heading your way, it's not the right of weight, right? The object that's not under power is the one that actually has the right of way, right?

25:43If you're under wind power, you win. If it's two sailboats and one calls starboard, you know who has the right of weight. doesn't happen in space. Yeah, we have those rules. So SpaceX is making many of the rules as they go because they're the largest operator in space. And I don't think that's a bad thing. I think they're really a disciplined, great team. They're doing a lot of the right things you would want to see a space operator that size doing. But they have to make these rules up as we go. And then we have to beg the other countries to follow. And what we're finding in space right now is there's a tremendous amount of friction between nations that have a tremendous amount of friction on ground.

26:21And they're doing hostile things in space that there are no space police. We just want the Space Force, but the Space Force is not the space, the international space police. Yeah. We even have situations in space to go down this, this nation state, you've, you've started the topic on nation state, nation state activities in space. And how do we set the rules? 40 plus percent of the objects in space right now are out of compliance with that international space act and treaties yeah what the hell are we going to do about it who's going to do it there are no there are no traffic cops in space so and we've done things like the artemis accords and the peace accords and a bunch of people have signed a bunch of paper but if no one says it's my job to enforce it they're all kind of signing it with their fingers crossed behind their backs and then it's going to be about the intent behind what they actually put on orbit and and they includes us right we're not anti-satellite the tests that we should have never done i i hope we'll never do it again in the future i don't think we will because we've led the charge to make sure that others don't but you know that doesn't stop others from doing it there isn't a penalty there isn't uh you know we're not going to embargo other nations uh based purely on their space activities.

27:37So not yet anyway. So these are all real issues that are surrounding space policy, governance, safety, and they're all bridging our path to accessibility. And the type of things that we're going to learn from outer space are going to be the types of things that heal our planet. Because so many things happening in space right now, I don't think the average person knows the The average person in terms of what we've, you know, put in the mainstream media knows how much of our climate research is happening from space. Right, right. And you need that more macro perspective. You need that holistic perspective on your sensing capability to look at Earth so that we can understand where the problems are and how to solve them.

28:21And we're, you know, we could be cutting off the gateway to those possibilities for a solution. My guest today is Alex Fielding. Alex is the co-founder and CEO of Privateer Space, a company, as we've been talking about, which is working on making space safer, more accessible, cleaning up debris. All the things we've been talking about, sort of tactical and also kind of philosophical, really, that, I mean, they have very tactical, practical applications. But a lot of kind of philosophical stuff that goes into literally exploring the unknown as Privateer is. Alex, while we have a couple of minutes left here, I want to shift gears.

29:00I'd be sort of remiss not to ask you. You've done a bunch of things already in your career, which shows no signs of slowing down. You've started companies, you've worked with astronauts and Steve Wozniak and all kinds of people. I guess sort of two questions. One, are there things about being an entrepreneur, doing the kind of work you've been doing, pushing boundaries and exploring new things that, I don't know, a story, an anecdote, kind of tips for people listening. A lot of the folks who are listening who are, you know, interested, obviously, with everything going on, certainly right now with generative AI, but more broadly, kind of the AI and intelligence explosion that we're experiencing.