From Snapshots to Systems: Why 3D Earth Awareness Is Becoming Critical Infrastructure | The Pair Program Ep97

Welcome to The Pair Program from Hatchpad, the podcast that gives you a front row seat to candid conversations with tech leaders from the startup world. I'm your host, Tim Winkler, the creator of Hatchpad. Join us each episode as we bring together two guests to dissect topics at the intersection of technology, startups, and career growth. All right, welcome back to The Para Program. I'm your host, Tim Winkler, joined as always by my co-host, Sean Leahy. Sean, so I was listening to my news podcast this morning and there was a headline about, you know, Blue Origin had a launch recently. I'm sure you guys are familiar, had a payload, a space mobile satellite, ended up in the wrong orbit. You know, just a kind of small, expensive oopsie there, but It got me thinking about how we spend a lot of time talking about precision in space, getting something exactly where it needs to be at the exact right time. And then we've all experienced kind of like the opposite down here on Earth. So my question for you is, have you ever had like an Uber or taxi or metro ride drop you off somewhere that was definitely not your intended destination? Yeah, but it's pilot error, I would say, with me being a pilot. So for our keen-eyed viewers, they'll notice my background's different. I'm in New York City right now. And when I first visited New York City in 2007, I was 17 years old and I was trying to get to Columbia University, which is on 116th Street. And I took the wrong line up, right? But there is a 116th Street stop on like the 4, 5, 6 line, if memory serves correctly. I could be wrong about that. But it's a mile to the east. It's through Spanish Harlem. And then when I got off there, I had to walk through and then go up a literal wall with all these staircases to get where I wanted. So it wasn't just missing by one subway stop or one long avenue. I was pretty much in the wrong borough when I got off. So pilot error there. I'm not sure if it was as expensive as the Blue Origin flub, but I felt pretty bad about it. Yeah, I feel like it's pretty easy to get lost in New York. I give you a, give you a pass on that. Yeah, I was thinking through it myself. I had a, uh, pretty sketchy late night metro ride coming home from DC where I fell asleep on the Orange Line and woke up at like the last stop in like New Carrollton, Maryland. And, uh, just got to say, it was not an ideal, not an ideal place to, to wake up and be around midnight. So Yeah, lesson learned. Never, never fall asleep on a metro again there. Cool. Well, uh, yeah, you know, this, this conversation at large, right, it, it, it brings into play, you know, a lot of these cases where those misses are— yeah, they're not just about, you know, precision. It's kind of also about a lack of, of context. Uh, so you're making these decisions based on incomplete or maybe outdated information at times, especially with like GPS. And I thought that was a good segue into today's episode. So today we're, you know, we're not talking about, you know, better pictures of the Earth. We're talking about this shift towards persistent 3D awareness of the physical world and what happens when that becomes something that we rely on as infrastructure. And so to tackle this, we've got a couple of guests joining us coming from, you know, different but complementary angles. One's building a next-gen Earth observation company from the ground up. Uh, the other spent decades evaluating these capabilities at the highest level of government. Uh, so first we've got Andrew Peterson, uh, co-founder and CEO of Array Labs, uh, building a system to generate real-time 3D maps of the world, uh, using clusters of satellites. Andrew, thanks for, for joining us. Yeah, thanks, Tim. Thanks, Sean. Great to be here. Good stuff. And then alongside Andrew, we've got, uh, Dave Gauthier. Is that— got that right? Former director of strategy and commercial at NGA, National Geospatial Intelligence Agency, and now advising companies like Array and helping them bridge the gap from innovation to adoption. So Dave, thank you for joining us. Welcome both of you to the Pair Program. Thank you. Thanks for having me. Good stuff. All right, we kick things off with a little icebreaker segment we call Pair Me Up. We go around the room, we rattle off two things that go well together. Sean, why don't you lead us off? Yeah. So mine's pretty classic. It's New York City and good, bad coffee. So wherever you go in New York, there's always a corner store or bodega and it'll have a full deli. They'll have bacon, egg, and cheese bagels. And they have these coffee machines that all look the same. They're just silver with the old school pull tap type thing. And the coffee's incredible. It costs $2. It beats the pants off anything else that's way more expensive. I'm sure there's lots of people who disagree with me on that. They're wrong. But the corner store bodega coffee in Manhattan and any of the Five Boroughs probably is unbeatable. And so I've been having really, really good mornings here again, get productive right off the bat. Yeah, that's awesome. It reminds me of Elf with the world's best cup of coffee, I think it was. That's right. World's best. World's best. People have confused me for Elf walking around Lormann. Just need that suit, man. Good stuff. All right, I'll jump in. My pairing this week is inspired by space and technology, kind of on theme with today's episode. I'm going to go with perspective and resolution. And it's kind of based on seeing some of those ultra high-res images from the Artemis mission. Which shot on an iPhone for a lot of those, which is wild in itself. But I think what stood out to me was, you know, how different your kind of like the under— your understanding of Earth becomes just kind of based on clarity and, and, uh, and angle. And there was one shot in particular that jumped out to me, which was like the Crescent Earth photo, that it just completely changed kind of how you see, see Earth. You know, it's the same planet but, uh, just a totally different perspective. So That's my pairing for today. I'm going to go with perspective and resolution. Let's pass it around to our guest, Andrew. Maybe a quick intro and your pairing. Yeah, sure. Great to be here, guys. And thanks, Tim, for the intro on the company, building ArrayLabs. I'm sure we'll talk a lot more about what that means and what we do. In terms of pairings, I also have a coffee. Connection with Sean. So we're already bonding here on the show, which is very good. But it is coffee and afternoons, you know, after lunch, you know, the next thing that happens is coffee. We're not quite sure where coffee is going to be, but it's usually some walk. We're having a conversation, trying to figure out how we're going to do something like, hey, you know what would be good? A coffee right now. And we just moved into a new office here in Redwood City, and there's like 6 coffee shops. Uh, right in a kind of a 4-block radius. And so, you know, whatever one is of those, we're going to get to, but, uh, but it's, it's pretty fun. So pick your poison. What, what's, what kind of coffee you going with? Uh, just straight black, a little iced Americano. What's, what's the coffee of choice to get that pick-me-up? Yeah. The only thing constant is change. Um, I'll not get the same thing twice. I'll, I'll try different things. What's the special? What's the thing? What, whatever sounds good. I've got a coffee in my hand right now. This one happens to be mocha, but I'm non-mocha exclusive. We can be anything. I dig it. Pretty happy for it. Keep those baristas guessing, right? That's it. That's great. Yeah, big fan of afternoon coffee or Red Bull or Celsius, whatever your poison is to get that second sprint, the second half of the day going. Well, again, thanks for joining us. And then Dave, yeah, quick intro and your pairing. Yeah, thanks for having, uh, me here too. Uh, like you said, uh, you said decades. I hadn't heard that before, but almost 30 years in national security and, and really doing remote sensing and data analytics of all kinds. Um, most of it from space, but, uh, like you said, I became Chief Strategy Officer at NGA about 10 years ago and really focused them on two things: AI and commercial space. I think those have panned out well for us, so I'm happy to talk about it. Yeah, and my pairing, I'm definitely with the vibes of Sean and Andrew today because for those of you who've watched Stranger Things, there's a scene very early on in the show where Hopper says, "Mornings are for coffee and contemplation." My kids even got me a Coffee and contemplation sticker because those things to me go really well together. If you open up, start your morning with the tyranny of the inbox, it's terrible. You need to start slow, think strategically, focus your intentions. So that's where I am. Shit. Shout out Stranger Things. Shout out Hopper. What a great character. He's pretty insistent about it, as I recall, the coffee and the contemplation. He needs it. He's very much for it. He does, man. Good stuff. Wow. 3 coffees I struck out. I should have coordinated beforehand, but— Yeah, Tim, you went very deeply intellectual there with yours. I was very, very impressed on that. I was just trying to impress Andrew. Just trying to impress Andrew. Not required. Not required. But— I'm sure the details on Array will equally impress. So I'm excited to hear more about it. Well, Let's wrap up, uh, yeah, that opening segment and jump into the heart of the conversation. So just a quick reset for our listeners on, on today's episode, uh, wanting to cover, you know, kind of three themes here. Uh, first, kind of starting with, you know, where traditional Earth observation breaks down. Uh, we'll move into what Array is building differently, uh, to tackle some of that, and then zoom out on what it means, you know, if this becomes infrastructure like a GPS or internet. And so We'll start with the old model. Andrew, let's start with you on this. We've lived in a world of 2D snapshots for decades. Again, from an engineering standpoint, what's the ceiling that we've kind of hit with that model? Yeah, so it's a great question. And it kind of comes down to the purpose, right? The person using the image, what are they using the image for? And this could be optical image, it could be a radar image, like some of the things that we're collecting, could be a 2D image, could be a 3D image, whatever that is. But, you know, these different, these different datas, they're not like, you know, they're not like flavors of, uh, of coffee or whatever we're talking about earlier, right? They're, they're— some of them are really good for certain things and really bad, um, at other things, right? And so if we think about where we've been for a long time, um, you know, with kind of the 2D imagery sort of world, it's really good for counting things. Like that's predominantly been a use case that people have done for a long time. You know, there's this many tanks in this column, there's this many, uh, jet fighters parked along this runway. And people, especially military, government, intelligence folks, um, they— these are really important questions to know, how many tanks are on the column and how many aircraft are on the runway. And so we're going to try to figure out that information as, as best as we can, you, you know, and so that's, that's kind of where this technology has lived. And if that's what you want to know, if you want to count something or maybe, uh, see if a building has been completed or not, or where this building might be, or if this building is still there, um, you know, after a conflict or something like that, you know, hey, great, we've got an amazing, you know, use case of, uh, of data there. And, and so we have a lot of 2D images of, you know, the tops of buildings and the tops of people's heads., but we don't actually live in a two-dimensional plane, uh, like Frogger. We live in a full 3D world. If you walk outside, most of what you see will be the side of things, which does not tend to show up terribly well in a two-dimensional image. Um, and so there's all of this extra context that you can have in the world if it isn't squashed down to a 2D, uh, photo and representation. And then that can give you more information, right? And, and I think the, the main thing here is instead of maybe counting, which is like one, one form of measurement. Are there other measurements that you can get from this scene if you can see it in 3D? Right. And so there's things like volume analysis, you know, how, how, how big is this pile of stuff and, you know, where does it need to go? Or line of sight analysis. Can these two things see each other? Or slope analysis and, and change or change detection and what's going on in these different places? Or, or what is this object really in 3D? Can I spin it around and can I look at it? And does that tell me what it is a little bit better? And if you are interested in those measurements, if those measurements show up in your business or those measurements show up in your operation that you're trying to conduct as a government customer, then 2D is insufficient because you can't measure all of those things. It is really difficult to get that sort of perspective and that sort of information. And so if you're building infrastructure or construction or underwriting insurance, or you're doing all of these things that require access to more physical world understanding or self-driving cars or Pokémon Go or whatever it is, right? These are all three-dimensional use cases where the full 3D measurement in one of these different ways ends up being like incredibly valuable. Um, and we've tried to find different ways to, to, to collect, uh, this data. Um, if we're lucky enough, maybe we'll fly a plane and we'll use a laser scanner, a LiDAR system to go map things out. Um, we've been doing that for, for a long time at, at great expense here in, in this country. We can talk a little bit about that, But if you want something that's up to date and cost effective and repeatable and can be accessed in any, any country or any place around the world, um, you know, airborne LiDAR doesn't quite, quite do that for you, or not nearly do that for you. And so that's where, yeah, there's, there's an opportunity to do something new here that we're, that we're bringing forward. And, uh, yeah, very excited to do it. Yeah, I was, I was doing research on this myself and, uh, I thought it was interesting, the, like, the crime scene versus live feed, uh, kind of comparison that they made— it's not like the data isn't wrong, but it's kind of disconnected from time. It's not real time. Dave, I want to catch your angle on this, spending your time at NGA. Maybe highlight some specific use cases, or what you can disclose, where you would say 2D imagery, those types of limitations can turn into failure during a mission. Yeah, that's a great question. I was going to take us back a little bit in time and talk about where 2D imagery got started from space. And really, it's a story of the difference between strategic intelligence, which was done from space, and tactical reconnaissance, which was done from aircraft, right, on the battlefield. And, and like Andrew said, there's different perspectives. With space-based imaging, you're looking straight down most of the time. You're also, you know, trying to get highest resolution. So you're very tiny areas, you're looking at one facility. Whereas with tactical reconnaissance, you're flying an aircraft over a battlefield, you're collecting, you know, oblique angles, you're seeing other perspectives, you're collecting a large area to try to get that awareness over a large area for the decision makers and the commanders on the ground. So What's interesting about sort of the breakdown is, as you said, is, is that 2D look from space doesn't give you the context you need to have that sort of enriched understanding of the environment. And the strategic intelligence was often this game of you, you get one snapshot every few days of a facility and you're piecing a puzzle together over years. To try to understand and, to quote a friend, steal secrets from the enemy, right? And the difference here is you almost need a continuously live picture of a wide area of the battlefield telling you what's moving and what's happening in a 3D experience to really have an advantage in employing troops and weapons and everything the right way on a battlefield. So yeah, I was again in, in some of like the, the pre-recording research was, you know, DARPA, a customer of, of Array Labs, uh, was pointing out a specific use case in this where, you know, in, uh, in 2D, you know, a tank, you know, under camouflage, um, and a flat decoy can look identical, but in 3D you can start to see, you know, height, you know, a little bit of volume and a little bit of that different terrain disturbance. Andrew, is this like, would you say, is this like a bottleneck more about like the— looks like the physics of the sensors themselves or how we kind of process the data? Yeah, it's definitely the sensors themselves. You know, there's only so many things you can do to collect imagery from space., or imagery from a camera or anything like that for that matter. And kind of comes down to some physics. Um, you know, if you collect a snapshot of a particular place, uh, particular, you know, from a particular direction, you know, you, you have your focal plane, you get a 2D image. Um, and so sometimes that tells you everything you need to know. Sometimes you really want to have something a little bit better than that. One thing that can help is maybe if you take two images from two different baselines, right? We've all seen this a little bit in, you know, you know, photo reconstruction and, and, and photogrammetry are kind of things that we do to be able to piece together a little bit more. And, and that's how things like Google Earth work. You know, you can fly through things and you look around things. It's because we've taken a lot of pictures from different angles, from different perspectives, and we put them all together. And, uh, That's great if nothing changes in the scene and you have time to be able to position your camera and then you go over here and you take another image and you go over here and you take another image and coordinate all of that. And that can be really good. A little bit harder to do that from your satellite, right? Your satellite's flying over. You can't stop and like, hang on, wait, I'm going to come over here and I'm going to take a picture. I'm going to go over there and I'm going to take a picture. Satellite's going, it's going to be overhead in a couple minutes. I hope you got what it was that you were looking for, which means you're pointed at the right thing. And maybe also there was no clouds because cameras and clouds, they're not amazing. I'm sure Dave is incredibly frustrated in times in his life where certain clouds are very uncooperative or not able to get what we were trying to go see. There's a member on our team that was managing Google's data collection with satellites, and he is also incredibly frustrated about, you know, those darn clouds getting in the way of all these things. And so our sensors, the kind of the approach that we're taking at Array is one, we're using radar instead of optical systems. Radar is able to provide a different sort of measurement. It's able to see right through clouds, just like, you know, the radio in your car still works when it's cloudy outside. I mean, you can, you can keep working with your, with these radar pulses that come down and bounce off and measure things., and the other thing that we do is we fly these, these radar satellites together as a cluster or as a swarm or as a, as a group. And as they fly over, they have different, uh, perspectives, right? They have different baselines, they have different positions as they're flying over with this, uh, with this flock. And from those long different baselines and different perspectives looking down into the scene, we can then, uh, have a pulse come down, reflect off the ground, bounce back up, get captured by all of these individual satellites simultaneously from different perspectives, and we can use that to triangulate the location of all of the different objects that were scattering that energy back in the scene. And that's how we can fix everything in 3D, then create these beautiful high-resolution 3D, uh, captures of these different locations. But we can do it, you know, at— in, in a split second, in a single pass, right as we fly over that scene, we can go collect this. And that's part of the magic of what we're doing that allows allows us to have this really, really high-resolution data system, this, this high-information sensor that hasn't existed in the past. So that's, I would say, one big, big advantage for, for this system. The other thing that I would say is that this, this system allows us to have a much larger power radar that we can operate than, than we would from, from a traditional system. In fact, this is something that, that Dave and I were talking about., you know, uh, recently. Um, but I don't know, Dave, if you don't mind, Tim, I'll ask Dave a question. Uh, you know, as you're thinking about kind of large satellites or different satellites that have existed in, in the past, you know, before, um, as you, as you've been working with them for, for, for, for a while, um, what feels different or similar about our approach, or what are some of the challenges with those traditional systems? Yeah, happy to talk about that. So the traditional systems, right, bigger was better, and all of the engineering and talent of the nation went into building bigger apertures, right? Bigger pieces of glass, bigger satellites. It— and at some point, you know, the engineering of that fails because you just can't launch something so huge. Maybe with Starship coming, we will be able to someday, but But, uh, so you're squeezing every technical ounce of performance out of the biggest thing you could fit in a rocket, right? And, and, um, but still you're, you're limited. And so it's actually funny that Andrew asked me this question because, um, I was getting my master's degree in aerospace and I was reading some DARPA papers for inspiration on a, on a thesis problem. And DARPA was playing with formation flying, right? And I was just like, formation flying is where it's at, because I can take— instead of one big aperture, I could take small apertures and put them at a distance from each other, and we could create a longer baseline and, and get some new, new imagery perspectives. So I was, you know, working on how to do that with optical systems. I would never even have dreamed of radar being capable of disaggregated apertures in small satellites, but That's kind of where the technology took it. You know, there's the microelectronics trend to get miniaturization better and better. Now we have the cost of launch way down, and you have garage manufacturing of satellite systems going on, and you can get a large number of satellites into a, you know, a formation flying array, and then get all the advantages of quality that you couldn't get from building bigger and bigger pieces of glass. So I like to say that the bigger your piece of glass for a mirror, right, to have an aperture, it's exponential costs to go bigger with that and you get like a linear gain in quality. But with a disaggregated aperture, you get exponential quality the more you add to it, but the cost is linear because you're just adding more small satellites. So it's, it's this huge advantage, I think, going forward. Quick shout out to our sponsor, Defense Unicorns. This one's for the problem solvers out there. They're hosting Warhacker, a first-of-its-kind hackathon built for the defense community. No buzzwords, no slide decks, just hands on keyboard solving real mission problems with real code. You'll be side by side with developers, engineers, and innovators from across government, industry, nonprofits, and academia, all hacking for the warfighter. It's happening June 16th through the 19th in San Diego. Got a real-world problem to solve or want to join a team that does? Learn more at defenseunicorns.com/hatchit. This is an interesting theme that I think Tim and I have seen through some of our previous guests too, of this shift to scale and like you just said, Dave, disaggregated but swarm-based, large number of assets performing together as opposed to a little bit more of a Cold War mentality around those exquisite systems and those single piece, massively complex, incredibly high cost, systems that used to be that was all how defense and intelligence thought. And now in the past decade or so, we've seen a move back to how do we commoditize things, but then, for lack of a better term, sew them together to get the performance back in aggregate that we would have with those exquisite systems. Andrew, I've got a question for you on data, because everything that you just described from a data perspective seems pretty intense. Imagery data already, both static images and video, is incredibly dense as a medium, way more times than text or machine data. And now you're adding that three-dimensionality to it, which is just going to explode its overall volume. And then you have to develop relationships between it, like altitude and different placement of objects. How does your team handle all the downstream storage processing and handling of that data? And then maybe as a follow-up there, how's AI changing that? I think most people think of AI still as a chat-based or text-based medium, but I think there's probably some pretty interesting applications to the imagery world. So take that wherever you want. No, that's a great question and very happy to talk about data, and, and maybe I'll also talk a little bit about the satellite and, and what we're building. So I have a, I have a model of our, of our satellite here. This is a, a 1/4 scale model. So the actual satellite's about 4 feet long. This is about 1 foot, 1 foot long. It's got a radar panel on, uh, on one side. This is what we use to transmit the radar pulses down to the ground. And then on the opposite side, it's got this deployable solar array. That we can use to, uh, to deploy this, this system. And this is one of our satellites. We fly in groups of a minimum of 4, uh, maximum that we're thinking about right now is, is 16. Um, and these satellites, you know, look at this, the scene from different angles and, and they transmit these pulses with this big panel here and it goes down and it comes back and We do something pretty interesting with our, with our radar that I don't think anybody else has done, uh, with theirs, which is we can use this big panel as both our, uh, our radar, right? To, to go send these very high pulses, uh, down to the ground. We can also use it as our data link. Uh, so we can actually send communication data through that same very, very high power, uh, antenna and get all of that data down to the ground and, and, uh, and, and use that to downlink data from the system. And that's incredible because it allows us to have just the massive pipe where we can zoom data down to the ground station and then that ground station can process that data. Yeah, we were doing some tests in our lab and I think we're going to be able to break a couple records soon for small satellite downlink. This should be pretty exciting. But that's sort of step one is you gotta get the data down to, to the ground. And, and maybe this reminds me a little bit of the time that I was at, you know, an earlier company, right? You know, I, I'm an aerospace engineer. I worked at a national, a couple national defense contractors. The first one was a company called General Atomics. And at GA, we make a radar imaging system that, that flies on board the aircraft that, that, that we would manufacture. We built about 4,000 of those radars. It was a low-cost, at the time, radar system. It was $1 million per radar. So we built about $4 billion worth of radar hardware with this program. That's how I learned how to do all the radar-type things. And I remember when GPUs sort of showed up around the radar sort of processing era, right? And we had always kind of had the, this radar. We, we could collect this data and, and, and radar imaging systems, believe it or not, have been around since the 1970s and, and some form or, or flavor. And some mathematician who figured out this, this technique of, of processing data in the 1970s came up with this amazing algorithm for how we were gonna process the data and it was gonna be perfect. And it's this algorithm called back projection. And the back projection algorithm produces perfectly formed and posed and focused radar data., and then they immediately said, this is ridiculous. Nobody will ever do back projection. This is not a thing that anybody will ever do. It's just too computationally expensive, right? It's just, there's no way humans, it's kind of like sequencing the genome or something, right? Like not in our lifetime, not in anybody's lifetime will this thing work where we can go process this data using this very complicated, but perfect algorithm. And so they made like 10 simplifying assumptions and every simplifying assumption, the image quality got worse and the compute Processing was a little bit better. And then every 5 years or something, computers would get a little bit faster and they'd remove one of these assumptions and we get a little bit better quality. And I remember at GA asking my boss, begging my boss for $4,000 to buy a graphics card. This was back when NVIDIA was doing kind of the GPGPU sort of thing. This was a little bit before AI, but there was this, you know, we were going to process data. Using these graphics cards. And my boss, you know, graphics card, no video output, just going to use it for processing data. And it's going to be great. And I did get the graphics card. We did use it to process the radar data and it just changed everything. In fact, we were able to use this algorithm before, this back projection algorithm that was this gold standard thing that nobody ever thought we were going to be able to do. And we started doing that at video frame rates, right? We could just crank through all of the processing required to do all of this data, you know, and analysis and share this in real time from our system. And that was the point in time. It was a big moment for me because it was like, okay, compute is not getting— compute is not getting slower, right? Computers are going to keep getting faster. There's this thing called Moore's Law. What's going to be possible tomorrow that's not possible today with the hardware that we have available today? And it's like, well, What's a harder problem than processing one radar's worth of data? Oh, it's multiple radars' worth of data. And we've been talking academically about multi-static radar systems, which is the term of art, you know, the proper term for what we're doing here with the formation flying radars. We've been talking about multi-static radar systems for like 50 years and nobody's built one. And the reason we haven't built one is because it's too expensive and too much compute and too difficult to process that data. And now it's like, oh, absolutely, like totally, totally a real thing. Totally possible. If it's not trivial, but it's definitely something that we can do. And then also the other thing you mentioned, Sean, is about once that data's down and once it is in 3D form, can you actually interrogate it? Can you process it? Can you get tools to go perform analytics on it? And I have to tell you, that whole industry has just exploded in productivity, right? The ability to take data and generate a meaningful inference from it, assuming it's high-quality data that tells you what want to know has never been easier and it's getting easier all the time. This algorithm that only a PhD and this is super small group over here could ever have written, it's like, oh, Claude can do that, right? Boom. It's got the article, it's got the report, it's got the paper and it's making suggestions, right? And so it's an amazing time to be alive and be able to have access to or to contemplate access to wider access to this data at higher resolutions, at just more detail than ever before, be able to get that down, be able to process it cost-effectively, and then be able to make inference off of that to help people understand whatever it is that they're trying to understand and what they're trying to know, whether that's for 5G base station placement or power line, you know, vegetation management, or again, self-driving cars or whatever it is people are going to do with this stuff. It's amazing to be able to have that data show up. And yeah, it's actually pretty reasonable for us to start computing that now, and kind of a fun place to be. Yeah. Very cool. I love hearing the pulling on the past experiences from like GA too, and applying that to this new venture. Very neat. If I could add to that for a second, it really is the perfect time for this. And I used to say, if you have a radar satellite, making a 2D image out of that data is about the dumbest thing you can do. To extract information because, you know, you're throwing away 999, you know, bits out of every thousand just so you can give something to a human so the eyeball can interpret it. So you're throwing away artifacts of anything moving and anything else that seems odd in the scene, but those are critical nuggets of information you want to retain. And we now have the computing power to just give the important bits to the, to the people who need them, as opposed to translating it into this image, 2D image, for then someone to use their eyeballs on to scan for differences. For sure. And Andrew, you kind of led into it on the, you know, some of the commercial maybe use cases. I want to get into that in a minute, but, um, I wanted to hone in on something specific to, uh, the resilience, uh, when we talk about this array or cluster of satellites. Dave, maybe you can lead this one up. A hypothetical scenario, you know, if Andrew loses one of these satellites in a cluster, you know, is the mission over or is it kind of like a self-healing kind of system? How does that work? Yeah, this is the beauty of the swarm technologies that are available now, and it's definitely self-healing and gracefully degrading. So you lose one satellite, you know, Everything can readjust to create, you know, a formation with just one less satellite. And then you can replenish, you know, you can add satellites back in. So the rapid replenishment or the continuous replenishment concept with scale manufacturing, there's a real advantage here to resilience as opposed to just having maybe one or two assets that are very expensive, and if you lose one of those, you, you know, you've lost half, maybe half of your capability. Very interesting. Let's, let's jump into the, the dual-use aspect of this. When we're thinking about— if we're zooming out, looking at this dual-use application, let's call it like a, you know, like a digital twin of Earth, if you will. So if we're treating, you know, this 3D Earth awareness as infrastructure, how does that kind of unlock who the customers are? Like, Andrew, maybe provide some examples of what, you know, maybe non-defense customers are kind of asking for this imagery. Yeah, and, uh, and, and there's— it's really nice for us in, in that segment, unlike maybe a couple other different, um, novel modalities for, for, uh, for, for different sensors that we see deployed in space, um,— who are like, maybe if we build this new sensor, you know, there's going to be this amazing market, right? We actually know quite a few people who are collecting 3D data, uh, today, right, from the airborne, uh, LiDAR, you know, use case, right? And, and what that looks like is you find a plane and you find a very expensive, uh, laser scanner that you bolt underneath the plane, then you— and, and then you go and you, you collect imagery, right, in, in the tri-county area. We're not entirely sure which tri-county area. This is kind of a crazy cottage industry, but somebody has this plane and somebody will bring this plane to you and then they will go and they will collect this imagery and they will— and then that will be helpful for you in whatever business operation you were, you were doing, right? Whether you're building something or you're planning to build something or you've built something and you're trying to keep an eye on it, right? There's this opportunity for 3D data to come in and make these things better. And we use this all the time. But the problem is, you know, Private planes are expensive, and these sensors are expensive, and it takes 6 months to get, you know, the sensor to the right place and get the data processed and get the data back to you if you rush it, right? If you don't, it's like years between the collection and the data, you know, coming back. It's, it's kind of crazy. Um, and so we know that there's value here. There's a lot of this data that continues, like, despite all of the friction and despite all the problems, that continues to get collected, uh, for these different missions and for these different use cases, um, but they're missing a key resolution, uh, component, right? And, and I think, Tim, you sort of started us on this journey of perspective and resolution and what it all means and bringing it together. But people think about spatial resolution a lot, right? That's how big is my pixel, right? And you want really, really, really high resolution. You want really tiny pixels so your object doesn't look pixelated. It looks clear and sharp and well resolved, and that's, that's great and that's amazing. But there's a missing component which Dave has really been pushing on quite a bit here, which is the temporal resolution component, right? I'm gonna look at this location, I'm gonna see this location as quickly as I can. Right now, for airborne LiDAR systems, that's like once every 5 years, and a lot can change once every 5 years. Or you could go recollect it if you wanted, at $50,000 per flight, but You know, people— and people do, right? But, but you're still not getting that any quicker than every 6 months or every year. Or maybe if you have a drone flight, maybe you can do that, you know, monthly or whatever it is. But, but you really gotta want that, and you gotta want to pay thousands and thousands and thousands of dollars to go get every one of those snapshots. And because that high cost, we don't do it as often as we would like to do it, and we don't get that temporal resolution, that, that cadence, right? You can't get this once a week. You can't get this multiple times a week or daily or intraday. That's crazy. That would be amazing. That would help so much to build this as this infrastructure piece. It's like having a GPS receiver, but you got to keep it stationary for a day for it to get an output. And you're like, "Okay, surveyors will like this and it's going to be really useful, but I'm not navigating through New York with this. There's no way." Right? And so how do you get that fast refresh, that fast update rate, and be able to do that in a— and like, the way you do that is you make it cost effective, right? So how do you get the cost of this thing? And I can tell you one secret, which is, you know, get out of the tri-county area, right? You have this very expensive asset, you're amortizing it over a small geographic footprint. Satellites are great, right? We can send this one satellite cluster up, it's going to see the whole world, right? And so anybody who needs that, we can go and we can collect and we can task many, many users all over the place. That really brings the cost down for us to be able to go collect this. And then the other part is just the way that we're able to do this with these long baselines and these radar systems ends up being way, way, way more cost effective than anything else that we would see from getting— from getting a plane over there, right? And this kind of makes sense, right? You know, if you, you want You want to take a really high-resolution photo of something, like send a guy with an iPhone, right? Like that's, but that's like, that's crazy. Right. You, and then you want to, so, so yeah, anyways, there's, there's a bunch of Pokémon Go players out in the world and collect that data. That's right. That's right. That's right. That's Andrew. It's not to me, it's not just a cost, you know, kind of equation too, but you're also adding fundamentally new or maybe at least enhanced value with that temporal aspect. Because now you have this images over time, multi-dimensional images over time. That's a whole other dimension for analysis, which— and now we get into classified stuff, but that introduction of that temporal continuity, or at least that ability to sequence things over time, really enriches the data. So it's not just cost, but it's a higher value data asset overall. Yeah. And it's more reliable too. You know, if you're like, imagine Google, you type in a search term, you hit go, and you need a credit card in a week to get your answer back. Again, like nobody's searching for anything, right? Like, can we get this and can we have it ready to go? And can we look back over time and can we look forward? You know, maybe, maybe we can predict, right? Certain things now. Now we have AI, now we have the ability to do this, you know, spatial inference and other things. What more can we do if we have this and it's always available and we know that we have this and then yeah, we have that look back sort of feature to understand change. Yeah. Yeah. So Dave, I want to pivot really quick to something that's been on my mind since we started the conversation, which is if I did my research correctly, you were the founding director of commercial operations at NGA and then also the chair of the first ever Intel Communities Commercial Space Council. It's an interesting niche because as you said, commercial 10 years ago in defense and Intel, those worlds didn't mix. You were a commercial firm or you were specializing in defense and Intel. There were a couple of exceptions to that rule. And now in dual use, which obviously Tim and I, a big part of the podcast is on finding the best dual use missions and companies out there, is, is huge. What was the first, uh, I guess, signal for you on the need to stand up that commercial-focused operations, uh, group within NGA? Oh wow, okay. Well, the signal wasn't really for me, it was for my boss, right, the NGA director Robert Cardillo. But I, I remember very specifically, it's one of those key moments in your life where you never forget. But we were, um, we were all of us in the green room at a GEOINT symposium before a speaking engagement, a panel talk. And somebody said, have you seen the Planet Labs Dove satellites pop out of the space station? Like, right? And we were like, no. And then they pulled up their laptop and showed video of these two size of a loaf of bread satellites getting, getting sprung out of the International Space Station. Doing a little tumbling and settling. And, you know, and it was like the aha moment was people in commercial industry outside of defense could build imaging satellites and deploy them and, and collect data. And it was— and they're this big, right? They weren't, you know, the satellites we were used to. It was a watershed moment, I think, in the geospatial industry. To realize if that's possible, you know, what else is coming. And pretty soon, you know, we were pretty— we were knee-deep in the conversations with Planet about the size of their constellation, what could the data do, what was the resolution, how can we get involved. And, um, you know, Planet was just the first domino in that, in that set of what we would call like new space Earth observation capabilities. Not far along came Capella, right? And Capella was the first small sat SAR in the US. Uh, there had always been commercial SAR systems, but they were overseas. There had always been, you know, there were a few commercial imaging systems like Airbus and, and others overseas as well. But all of a sudden we were seeing the rapid infusion of capital, the proliferation of small sats in both optical and SAR regimes.. And then a few years later, Hawkeye 360 kind of did the same thing with RF. And it was this wave of commercial is coming, right? And so, you know, the NGA director at the time said, we need a real program office here. We need a real focus on this because we want to be riding the wave, not crushed by it. That's awesome. That's such a good story about seeing that. Having an aha moment and then reacting to it, I think, in a way that, as you said earlier, has just panned out so well. I mean, between, you know, dual-use space and then probably AI and then maybe also UAVs, those are kind of the big three, I would say, super success stories across both commercial and defense intelligence. So thanks for sharing that, David. Absolutely. Yeah, I wanted to kind of transition into one more part of this conversation before we wrap up with the 5-second scramble. Andrew, I'll lead in with you on this because I think, you know, it's really interesting what you're doing in building something that hasn't really been done before. And always, yeah, want to pull as much insight from those types of founders as possible because I think just the most fascinating stories stem from that. But something that we've asked a lot of guests, you know, in, in maybe similar scenarios is like, how do you evaluate progress in something that maybe doesn't have traditional metrics tied to it, or like a playbook tied to it? Yeah, and this is a particularly hard thing to, uh, to answer in, uh, kind of the, the space domain, right? Because the lead times between creating a product and actually having that product show up in market are quite far away from, you know, what we see maybe in, in software or some of these other things where the build, test, ship, you know, iterate, you know, all these things. It's like, well, you know, there is a launch cadence there's a satellite and it's kind of complicated. We got to make sure all the right bits go on it and, and do it in the right way. Um, and so what do you, what do you, uh, how do you, how do you judge that, right? And a big part of that, of course, is the customer conversations. And you talk to folks and they're like, yeah, this would be amazing, this would be tremendous, this would be life-changing if we had this data, if we had this as this infrastructure layer, if we could reliably get data like this anywhere around the world as quickly as we could, that would just make everything that I am doing so much better. And I've heard that so many times, and that gives me and the team a lot of confidence that what we're building is going to be something that's useful for folks. And at the end of the day, that's all you can do, right? I think if you, you know, you can focus a little bit on business and you can focus about, you know, pricing or strategy, whatever it is, but at the end of the day, it's actually quite, quite simple, which is if you have people who are very happy with what you were going to provide for them, uh, the, the— all the business pricing things tend to like, you know, they, they, they come out okay, especially if we're able to do something that has this magic, you know, quality to, to it that, that we're— that I'm so excited by. And Dave, you know, is kind of mentioned, which is this like, it's, it's the highest quality 3D that we can find and we can do from, from space. It's very fast to go produce this. It's actually quite cost effective for us to go generate that as well. And so customers will always want things that are faster, and they always want things that are better, and they always want things that are cheaper. And if you can find those things and you say it, it's like, oh, all right, well, we'll, we'll do that then. Um, and then, you know, in terms of progress, there's incremental progress all along the way. You know, we built many radar systems at, you know, at subscale, tested them terrestrially. We're going to go do some drone testing here in a few minutes. And, you know, this is a really fun place for the company to be before we go and put this capability in space. But we're just driving forward as fast as we can. And but yeah, I think it's nice to know that what we're building has precedent. And has a need and clear need that we can understand from a customer that if we build it in this way, then we're going to have some pretty, you know, pleased folks. And so that's what we're— that's what we're doing. Cool. Yeah, well said. And I guess, Dave, you know, from an advisor perspective, is there like— yeah, what are you kind of watching? Is there a metric that you're watching closely that kind of tells you the business is working, is progressing? Why? I think one of the best things about this company and, and Array's approach is that they're not going after a completely new market that doesn't exist yet, right? There, there's an existing market for high-quality 3D data, and that's been served by air, airborne assets. And so if you, um, had the funding to fly an aircraft and you had the time to, you know, wait for the data, you were— you're definitely flying aircraft for those purposes. And it's very dual use. So, you know, I had the opportunity to talk to LA County about the aircraft that they fly and how much money they spend on doing these annual surveys and what they use the data for. And, you know, they're saving lives and property by calculating fire risk and vegetative growth and landslide risk and flood risk and all these other urban planning things that it will make things better for society. And they only get that data once a year. So the, the idea that you could replace that for a very rich area, LA County, getting it once a year and most counties getting it never, the idea that you could replace that with a globally available system that's constantly refreshing, maybe once a week, maybe faster, You know, this is just a fundamental change to humanity's ability to understand the environment and, and maybe be safer as we use that environment. Those are all civilian uses. And then you go to the military side and we've been using airplanes like JSTARS and AWACS and other things for moving targets. And, you know, this is an inherent capability to detect moving targets as well. So I, I see so many potential ways where this replaces something that was less capable with a more capable system that's available everywhere all the time, as opposed to on rare occasion I get to use it. So I think, um, from my perspective there, as the company demonstrates that this capability works on drones, on aircraft, space demonstrators, you're seeing progress. As long as this technology works, the market is there for it. Yeah, very cool. Um, well, I wish we had another hour because I've got a lot more questions I want to ask. Maybe we'll do a follow-up episode, uh, at some point. But just in closing, uh, Andrew, maybe just some quick hits on ArrayLabs. You know, tell us a little bit about, you know, how long you've been around for, company size, funding, some stuff like Yeah, funding. Yeah, first angel check into the company was from one of the co-founders of Google Earth, uh, a long time ago in the fall of '21. And I went through YC in '22, um, and, uh, have, have, uh, you know, been successful in raising for the company subsequently. Uh, now have 35 employees. We've raised $35 million. And, uh, yeah, it's, uh, it's been a wild ride as, uh, as we built and scaled things. We're kind of in an interesting place where, um, you know, the, the, the founders can no longer be in all the meetings, you know, doing, running around doing things. And we're starting to, to layer in that, that next layer of, uh, of people to help us go build all the things that we need to go build and keep an eye on all the things. But it's really gelling, and, uh, and that's really fun to be in a, in place where you have the resources to go fast and make things, and you have room to let people run. So yeah, it's been a really fun and rewarding journey personally. But really, everybody's super focused on that thing, which is, yeah, how do we go make other people's lives a little bit better, and what are we going to try to do to affect that? And so, yeah, it's been really fun to be part of this group of folks and try to get it all delivered. Very cool. When can we expect 4D? I'm just joking. There you go. You're just going to always want more. So if you go back to the time machine and all of our philosophical musings, you already know the answer. The fourth dimension is time, which is the resolution component we always forget about. So I think immediately after we launch this thing, we're going to have— we're not going to have not just 3D, but Fast Refresh 3D. We'll call it 4D, and we'll name it after you, Tim, which we discovered here. So thank you for that brilliant marketing insight. You got it, man. Yeah, happy to contribute. Yeah, let's wrap on the last segment here. So it's just going to be a rapid-fire Q&A, some business, some personal. Just spit out the, you know, the first thing that kind of comes to mind. Sean, you lead off with some questions with Dave, then I'll close with Andrew. All right, Dave, you ready? Let's go. Okay. You spent 30-plus years working with some of the highest technology projects in national security. What's one piece of technology from movies that's always laughably wrong whenever you see it on screen? Men in Black, the flashy thing. You can't erase someone's memory. Many security officers have tried to do that to me, but I still remember things. So good. Oh, that would definitely make contract offboarding a little bit easier if the ISO could just, you know, hit it. What was your dream job as a kid and how did you get from, from there to, to your career where you are now? Oh man, that's a deep one. So dream job was astronaut, and I think, you know, I grew up watching shuttle astronauts in the early days. A lot of kids saw Artemis II this past week or so. I think we've just created a new generation of wannabe astronauts, which is awesome. Because it pushed me into engineering, pushed me into the Air Force. I had a pilot slot, but my eyesight was too bad and that kind of derailed me. But I went off into the intelligence world and then realized, well, I can play in space anyway with satellites. So that's kind of how I got to be where I landed. Yeah. I think you're 100% right. There was a CNN interview with a kid on the day of the Artemis launch And he ended up dropping an F-bomb, which is a little much, but he was very happy that we were going back to the moon. He's from Florida, he gets a pass, I think. Yeah. What's one lesson from your undergrad electrical engineering days at RPI that you still rely on frequently? Wow. I would say that's persistence. So some of those classes were so unbelievably new to me, like the math and the concepts, especially like electrons and holes and how they both move around. So just the idea that if you keep working at it, you will eventually get the answer. And once you have that aha moment, you have that for the rest of your life. And it's great to know that just digging in and working hard can can push through. Awesome. Excellent. Applicable to all domains. What's the strangest or maybe hardest thing that you've had to explain to a room full of executives, generals, you know, the high-level decision makers? Well, the one fun thing is most high-level decision makers think that when they say something, people will do it. And I remember having to tell some agency directors like, you know, When you say that, it gets translated to the next round of people who intend for it to happen. And then when they direct it, it goes to the next group of people and there's this waterfall of apathy that finally, like, it just, the energy just dissipates into the government ecosystem and nobody does anything. Right. And they were shocked and appalled that that was my understanding of the system and that it couldn't be true. And I said, look, that's what happens if you don't have any champions who are willing to pick up the rock and carry it for you. You need to rely on your champions. And so there's a few great champions in the government world that do that all the time, over and over again. But absent that work, like, the system is designed for stability, and, and there's a reason for that— bureaucracies to be stable and competent. But they don't change quickly. Dave, every one of our defense tech listeners is pounding the table or their car dashboard in agreement with that right now. You're giving me flashbacks, but you're 100% right. We found our short clip teaser for the episode. I think. Yeah. All right. So if you were not in this industry at all, if you were completely out of intelligence, defense, imagery, aerospace, what industry would you want to be a part of? If I was completely out of this industry? Which might be hard because you said you wanted to be an astronaut when you were a kid. So this is pretty deeply attached to you. I'd have to go really, really out there and say, yeah, I think I would be a Well, I like to hike in the mountains, so I think I would be a hiking guide, something like that, right? Where elevation matters. Partial credit, partial credit, but accurate. Last one to close on here. Who's one person, living or dead, that you'd like to have a coffee or a beer with for a deep conversation? Wow. People always ask that and I've never had a good answer. Ready to go. So let me think, the one person to have, it has to be a coffee, right? Um, beverage of choice. Well, we talked about coffee earlier. I'm thinking it has to be a coffee. And, um, and let's go with Ben Franklin. All right. So yeah, there's a lot about Ben Franklin that is amazing. Uh, obviously a level of genius rare in this world, but also a level of irreverence. And so I think he'd just have some great stories to tell. Absolutely. Maybe the greatest and most famous amateur electrician of all time. Dave, thanks for doing the 5-second scramble. You crushed it. And we'll send it over to Tim Andrew. Yeah. One more though. We quickly forgot, Sean, the charity. Give him the charity question. Oh, I'm so sorry. Yeah. This is what happens when I'm traveling. Traditional final question here on 5 Second Scramble Day. What is one charity or philanthropic effort that's near and dear to you that you want our listeners to know about? Well, this is super easy because it's a nonprofit my wife started. It's called the Dyslexic Edge, and it's because we learned that children who are dyslexic They either struggle through school and end up dropping out, or they like have this moment where everything clicks and they do well in school and they're very STEM focused. And they, so half of NASA, half of MIT, they're dyslexic, also half of the prison population in the US. So you're, it's like a bimodal distribution. You either go this way or that way. So the idea was let's get the kids early learning how to read so they never fail out. And so I think that's a great thing to do. Very cool. Yeah, we'll give it a plug in, uh, in the show notes when the episode's released. Uh, thank you, Dave. Nailed it. Uh, Andrew, you ready? Let's do it. All right, let's do it. Uh, if ArrayLabs were a band, what genre of music would it play? Oh, uh, probably classic rock, I think. Maybe, maybe all rock, something like that. I think classic rock. Yeah, I like it. There you go. What kind of roles is Ray Labs hiring for over the next 3 to 6 months? Ooh, all of them, but a lot of electrical engineers, a lot of aerospace engineers, but we got a really fun group of problems to go solve. But yeah, people on the business side talking to customers, it's a whole range. All the jobs. What's one thing about the culture at Array that would surprise folks that are looking to join? Yeah, I think one of the things you get to do that's pretty fortunate when you start a company is you get to figure out who you're going to bring onto the team, right? And as a recovering engineer myself, I was thinking about what were the engineers that I really liked working with and who were super effective and made great teams.. And it was some combination of kind of what you'd expect, like, you know, smart people who work hard and, and, and, you know, really knowledgeable about their domains. But there was something else, which is kind of like the motivation for like why you're showing up and what you're trying to do with, with the technology. And I always found the people who really had a lot of empathy for the customer, and that could be an external customer, that could be an internal customer, somebody is going to use your stuff after you built it. That were just thinking about things in such a better way. And so, we try really hard to find people who are, you know, like self-aware and empathetic and also really great talented engineers, which is not like a strictly overlapping Venn diagram as these things tend to go. But there's— but the people who are there who really like doing stuff for other folks and who love solving problems, like we're— We're very happy to welcome folks here, but that's kind of who we are and pretty genuine about that. Yeah. Very cool. What's been one of the hardest lessons learned since starting Array? Oh, again, as a recovering engineer, you want to go solve all the problems, right? You're like, oh, I have some expertise at this and I'm going to go work on this thing. And there are many reasons why you must not do that when you're the CEO of a startup company. But one of them is you don't have time. The One is you got people who are like great at doing these things and you should let them run and go do the things that they're amazing at. And so I think that's something that all the tech founders sort of learn as they go. It's like, it's probably time to go do a different thing than this. You know, you kind of have 3 jobs and it's like, don't run out of money and make some sort of Progress, you know, goals and milestones and see that they're reached. And then the last one is hire well. As a CEO, these are like the big 3 things. And it turns out that the secret is the success of goal, right? So if you want to not run out of money, then you hit the goals and milestones. And if you want to hit the goals and milestones, it's all about, you know, who's at the company and giving them the responsibility and authority and to go make all the good things happen. So, yeah. Well said. Yeah. Letting go and empowering those folks. What was your very first job? Ooh, my very first job was a mechanic on a farm in Ventura County where I did maintenance on tractors and other vehicles. And that was a, it was a fun, very fun time. I couldn't do it anymore, but that was, it was good to get an exposure to. Mechanical engineering and all the things that make things work. Very cool. If you could put any sports car into space, what sports car are you going with? I would put my Subaru WRX in space that I had. I don't have it anymore. It was a while back, but it was a really fun car and it was great. It wasn't anything necessarily special. It was the blue with the 4-cylinder., a stick and, and it was, but it was, it was great. Indestructible. And, uh, indestructible. They kind of changed the design later on. I don't like the, the new design quite as much. So somebody should keep that in pristine condition out in space somewhere. We'll park it next to Elon's, uh, you know, Model S that's on its way to Mars. That's right. That's right. What's, uh, what's a person or who's somebody that, uh, in this kind of greater space tech ecosystem that you think would be a good addition to, uh, to our podcast? Oh, there's so many people. Um, I don't know if you've had Joe Morrison on, but he's worth, uh, he's worth a good time. I see Dave endorse that. Yes. All right, enough said. Joe Morrison, if you're listening, come join us. Another great— yeah. Uh, what's an activity that, you know, helps you completely unplug from, you know, the world of startups? Yeah, those are, those are the activities that I like the most. Anything that I'm sufficiently focused on to turn my brain off is a good time. Like most engineers or former engineers in the Bay Area, I'm a big rock climber. So that can be helpful. You're sort of problem solving your way up the wall. Maybe a little bit more rare, but, but still pretty popular. I'm a motorcycle rider. And so I like to go out to a track day whenever I can find the time to, to go do that. Yes. We're not, not a, not, hasn't been recent, but, but I, there's still a special place in my heart. I still ride on the street, which my co-founder really doesn't appreciate and reminds me many times that I need to stop doing. But so far, I was going to say both activities, you need to be very dialed in to, to avoid, you know, catastrophic errors. Yeah. Which is nice because it, it shuts off the, that's right. The other the other, the other, the other threats, right, that are always there on, on background. Yeah, that's right. Awesome. Uh, and then our closing question. Yeah. What's a charity or corporate philanthropy that's near and dear to you? Yeah, there's a, uh, a nonprofit that I like called Education Justice Academy. Um, this is a nonprofit that trains school board members and, uh, is actually a big problem that, you know, we have these things called school boards and they kind of oversee this whole, you know, thousands and thousands of kids. And the people that make up the school boards are like untrained volunteers that, that now are responsible for $50 million P&Ls. And this is kind of crazy. And so we should find some training. We should find some ways to, to make these people's lives a little bit better, a little bit easier, and make sure people show up that are, that are, you know, reasonable and can run and can get elected and can do good jobs there. So that's my, Yeah, it's a nonprofit that I think is great and they're trying real hard to make it easier for folks to have the right kind of training and knowledge and insight to be able to manage schools and school boards really well. Very cool. Yeah, we'll give it a plug. Dave, Andrew, thank you guys both for joining us on the podcast. It's been fun and we'll be rooting for ArrayLabs. Thank you. Thanks, guys. Appreciate it. Till next time.