The power of daily Earth imaging

Click here for Notes and Transcript

Time Markers

00:00 – Episode introduction00:34 – Terran Orbital00:58 – Welcome01:26 – The Planet Vision03:28 – Looking to the left06:35 – The Kind of Imagery Matters09:12 – The Nature of Partnerships11:42 – Digital Twins15:32 – Analytic Capabilities16:29 – Questions About Privacy20:00 – Data Bottlenecks23:56 – Space Takes – AI & GEOINT38:59 – Space Takes – Starship43:42 – Space Takes – Jared Isaacman

Transcript – Ashley Johnson Conversation

David Ariosto – Ashley Johnson, it is great to have you on the pod. I want to kind of just like jump right into this and talk about planet. It’s such a such an interesting company. This is an earth imaging company, and sort of the mission, if you correct me, if I’m wrong, is to sort of image all of Earth’s land mass every day.

Ashley Johnson – Let’s make change visible, accessible and actionable.

David Ariosto – Exactly. Well, let’s get into that. So tell me a little bit more about the vision here, like short term, long term, and why what you do at Planet is important. Why those working at Planet, I come to work every day.

Ashley Johnson – I’m so happy to talk about that, and really excited to be here. David, thank you for having me.

Yeah. So as I said, planet’s mission and vision is to image the Earth every day and to make the change going on on the planet visible, accessible and actionable. And those are really important words. When I think about the teams we have working at Planet, we have an entire Space Systems team which is innovating all the time on everything from the satellite bus and radio to bring the data back to Earth, on the payloads, to come up with new and interesting data that we can produce as to what’s going on on the planet. And that’s all about making the change visible. And then we have an entire team of software engineers and data analysts that are figuring out how we make that data accessible. So that’s everyone from our data pipelines team that are doing the rectification and making the data, what we call analytics ready, all the way through to the teams that are building the platform, which is our tools and APIs that make the data accessible to our customers, which are governments, they are researchers, universities, commercial entities, and making sure that we’re serving up the best, most validated possible that we can.

And then it’s our ecosystem of partners and data analysts around the world that are making that information actionable, so understanding what it is that our customers are trying to solve, whether that’s, you know, disaster response to a situation, situation that’s happened, or understanding the changes that might lead to a future disaster. We work with an ecosystem of partners that are taking the data and making it actionable to serve those use cases and to really create a broader understanding of the changes that are happening on the planet and what we can do about them.

David Ariosto – Well, that’s I think I want to start there, because there’s a separator here with planet that, I think, you know, maybe for the layman consumer might not necessarily see this, because, you know, you could, you could look at Planet and say, Okay, this is an earth imaging company. They’re doing things that other companies are doing. But you mentioned this concept of sort of looking to the left, which, which I really like the way that sort of characterized. So So first, like tell me what that is, and then we can kind of get into its implications.

Ashley Johnson – Yes, well, let me give credit where credit is due. This is a phrase that Robert Cardillo, who is our global defense and intelligence strategy leader. He was the former head of the NGA, the National Geospatial-Intelligence Agency for the United States, and he what he really appreciated about planet for a long time was that we weren’t only collecting data of specific points in time and and trying to understand those locations that we already know we needed to be watching, but where we are gathering the data about everywhere on Earth. What that enables is when something happens someplace that wasn’t a previous point of interest. You don’t just get the information of what happened after, but you actually get the information of what happens before. So that’s what he calls moving to the left on the timeline.

David Ariosto – That’s just that this seems monumentally significant in terms of what you can do. I mean, essentially, you have the capacity, almost like look back in time, to sort of piece together. You know, little bits of information that tell you why something happened the way it did.

Ashley Johnson – That’s 100% right? And that is what planet makes. Planet unique in a way that, you know, if I, if I put my CFO hat on for a second and think about how we talk to analysts of, you know, competitive moats, what’s going to make it hard for somebody to catch up with what planet is doing, even as technology is innovating at a rapid pace all the time, it’s that we have that historical archive that dates back over seven years. We have over 3000 data points for every point on Earth, so that when something happens, you can say, well, what were the things that led up to it when there’s a military situation? Well, were there things happening around that area that might have been early indications that something was changing and something was about to happen in circumstances like Lahaina, where you have these fire of a magnitude and catastrophic impact. Well, were there early warning indicators in terms of the nature of the crop biomass in the region, and the amount of dryness in the soil reaching new levels, or soil temperatures that are reaching historic levels, and those compounding effects that lead to something like what happened? Well, that’s not the only place where those circumstances exist. So now governments in other places that are concerned about, well, could this happen to us? Can use those indicators, or what we call planetary variables, to say, Well, how do the variables in my geography stack up, and how much at risk am I and what should I be doing about it? So that ability to look left and have that Archive of Data is something very unique to planet and something incredibly valuable.

David Ariosto – Also, also, when you talk about that, I think it’s really interesting in terms of the discussion that we had about the nature of what we’re looking at, because just that, like the kind of imagery matters, and that’s changing pretty dramatically. So it’s like in the past, what we’ve seen is imagery that matters to the human eye. And, you know, most important in that context is sort of that vector is, is resolution. That’s different. Now, at least that’s, it seems to be changing with the advent of machine learning, AI, even quantum powered technologies that are starting to fold into these things in which the human eye is not that’s, that’s only that only factor that matters. It’s almost like what a machine can interpret. So, as one of you could, as one of you could, could point to me, then, in terms of that evolution, how that’s changing, and what you know, what that eye in the sky now looks at is just different.

Ashley Johnson – It’s actually, it’s even, what is the nature of the data that we’re trying to capture, if you’re if you’re trying to capture a picture for a human to look at. As you said, the most important factor is, you know, how strong is the resolution?

And for commercial entity, there’s, there’s a limit on actually the imagery that you can capture. So 30 centimeters per pixel is what the regulators will allow us to capture. And beyond that, you get into government classified satellite imaging. But even moving beyond the regulations, the information that you can get that is machine learning ready, rather than eyeball learning ready, is much more interesting. For example, we have a new satellite we launched in August of last year that captures 400 spectral bands in the image. And this is a hyper spectral sensor. It’s It was created by NASA JPL planet worked in partnership with a group called carbon mapper to integrate it into our satellite bus and launch it. And now what it’s doing, it’s collecting data over areas, that is that enables us to tech detect methane leaks.

So this is not something that’s visible with the human eye, but we can capture this information and feed that information back to governments or back to energy producers concentrated animal feeding operations to make them aware of the circumstances and what that would mean for the surrounding communities and what they can start to do about it. But that is not something that any resolution can capture for the human eye. That’s something that a machine can detect through the spectral bands that we capture, and then we can do processing on top of it to make it visible to the human eye. So that’s one vector that makes this really interesting is you can see the things that a human couldn’t see. But the other is…

David Ariosto – But before you go into the next point, I wanted to get one question in terms of what you just said. You mentioned JPL Jet Propulsion Laboratory in Pasadena. This is part of NASA. I think the nature of sort of government private partnerships is a really interesting one to suss out here, because in the sense of this sort of commercial explosion within this space economy. There are all sort of these, like government scientists that there are, you know, tried and true, relied upon individuals. And so that symbiosis between commercial and government. I think sometimes when people think of things, they think, you know, governments in one place and commercials in the other, and one does one thing well, and the other one does another thing well and and never the twain shall meet. Um, oversimplified, admittedly, but, but just like the nature of those partnerships, I wonder if you could speak to that and how planet has kind of looked to that as in part of its model.

Ashley Johnson – Yeah, I’d say obviously public private partnerships have been, you know, the source of a lot of innovation, especially those things that are going to have much longer time frames commercial entities, especially those of us who are in the public markets, you know, our investors tend to not have the level of patience that it takes for, you know, some of the timelines that we’re we’re talking about, and something like a hyper spectral sensor may not be something where it’s immediately apparent what that end market is going to be.

The market is right now. Yeah, right. But the value to science and the value to research, and ultimately, the value to governments is more obvious, and so the government’s ability to fund those research programs. But then be open to working with commercial entities at that point where a commercial entity is going to move much more nimbly, think about how you can do it much more cost effectively, and really be thinking about what is the value to the economy, to creating a market that ultimately will fund this going forward and and across all industries, there should be that natural handoff point from the government funded research to where the commercial part, the commercial world is going to take, take it to where it can really be most impactful.

David Ariosto – It almost seems like it just, it finds its way to the market at some point, even if you don’t quite understand, I mean, even just like that, the nature of the nature of CAT scans, frankly, in terms of the old, the old Apollo program, and you know, that was just sort of looking for a belief, were for gaps in terms of the spun aluminum. And it ultimately led to advances mental technology. So it’s like, it’s almost, you never know where it’s just like the science for science sake is going to kind of lead into commercial sector. But I think when terms terms of planet, one of the things when I hear about you talking like looking left and looking at, you know, a whole scope of new kinds of imagery that not only humans but machines are looking at. I think of digital twin technologies, and I think of like the broader construct of, like, how you operationalize this from a strategic standpoint, whether you’re a commercial company or whether you’re a policymaker, like getting, getting the level of granular insight, not only in terms of what’s happening now, but like, what brought us to this point, and then using that to extrapolate and devise strategies, to think, to combat everything from deforestation to supply chain questions. It just, it just strikes me as we’re like, at the vanguard of this brave new world in terms of understanding who we are, and planet is like, right? Like you’re at the tip of the spear.

Ashley Johnson – Absolutely, and I referenced earlier, the fact that we have a platform. It has tools and APIs in it that enable us to, you know, what we like to say is democratize access to this data, and whether the funding source comes from government or, you know, commercial solution providers. Ultimately, the goal is to have an ecosystem of, you know, people that are taking this data and understanding, what can you do with this? So I’ll give you an example of one of our one of our partners down in South America. So in Brazil…

David Ariosto – I was just gonna ask you about this. You beat me to it, but yeah, please go ahead.

Ashley Johnson – So they work very closely with the Brazilian Federal Police, and so what they’ve done is they’ve combined the analytic capabilities that we built on top of our data and we expose as an API, which is road and building detection. They use that to detect deforestation or potential for deforestation in the Amazon. So if you think about it, deforestation happens after you’ve moved big machinery into an area, and that means somebody has to build a road to bring that machinery in.

Well, we can detect that road with our daily scan of the Amazon, and with those road detection, marry that with what that what they had on their side, which was access to the permitting database of the Brazilian federal police. And so that enables us to bring the signal to signal to noise ratio way up, because anything that’s a permitted deforestation initiative immediately gets silenced, but anything that is potentially something that’s not permitted gets flagged, and then you can either task a high resolution satellite, you can fly a drone, but you understand the nature of that activity. And so if you’re talking about the Amazon and deforestation, that can be anything from illegal mining, drug and narcotic narcotics and weapons trafficking to simply a cattle rancher trying to, you know, cut down trees to create more grazing land for cattle. You want to respond to that as the police differently, depending on which one of those things it is.

So they get the tip from the daily scan and the analytics that’s been enhanced by our partner in the region, and then they cue a satellite or some other source of. Information to give them the greater depth of information that they need, and then they can respond. That’s enabled them. It enabled them to capture over 3 billion US dollars worth of contraband and significantly decrease the illegal deforestation that’s going on in the Amazon. So that nature of we have that broad scan data, we make it digestible through our analytics. We integrate that with an ecosystem partner that ultimately sent enables the end customer to do things that are valuable for constituents, for their governments, and, quite frankly, for the entire world.

David Ariosto – And this has led to interventions, Thousands of interventions.

Ashley Johnson – Three thousand, over three thousand interventions, absolutely, they’re a great partner and a great showcase for the types of things you can do with this amount of data, which to some can can feel very overwhelming, but with these analytic capabilities. And you know, frankly, we should talk about this with the advent of AI, the speed to which you can get to that type of solution is accelerating at a pace that’s unprecedented.

David Ariosto – All right, so now I kind of want to, like, step back a little bit and delve dove into, kind of, like, the practical implications, how this is actionable and like, how this, you know, it can be used as sort of a force for good. The question I have then is, like, when we kind of step back for a second, we get a sense of, like, how much of this sort of we talked about this term, this emerging techno sphere, essentially like this, these circling satellites and, you know, Earth imaging systems that provide us such a much more profound understanding of who we are as a species. There are inevitably questions about privacy that come into this, this, this, this equation. And you know, the nature of the nature of not being able to hide is a good thing when it comes to things like deforestation, but you could see how this could be used in different ways, in some more sort of those Orwellian constructs, whether it’s on the commercial side or the government side. So I guess my question is like, how do you how do you safeguard that? How do you kind of protect it against those inevitable and very legitimate concerns?

Ashley Johnson – Yeah, so, so, first of all, let me dispel a few concerns that people have when they hear about this. At, you know, we’re our Pelican fleet. When we have it fully optimized, we’ll be able to image it up to 30 centimeters per pixel of resolution that enables us to read, you know, the numbers on a tail of an airplane, and to see windshields of cars, but we’re not tracking any particular, you know, individual per se at that level of resolution. So we’re not identifying who a human is through that level of resolution. So that’s one element here. The second element is, we’re capturing revisit every day, but we’re not like a video camera, you know? We’re not just this giant security camera sitting around the earth and then track somebody.

So I just want to just spell that for anyone who’s really not familiar with what the capabilities are of commercial satellite imaging. But then you get to a really important point where there is a lot of data in here that can have broad implications individuals, when you pair that with other data sources, as well as that activities of countries and commercial entities. And for those that are trying to do good, you want to make sure you’re safeguarding and protecting that information. And we spend a lot of money and time and resources on securing our satellites all the way down our pipeline through to our data delivery interfaces. So, you know, fully encrypted end to end, with a lot of security around it to make sure that our data is only getting into the hands of the people that we intended.

David Ariosto – That’s a lot, oh, sorry…

Ashley Johnson – There’s a lot of it. I was just gonna say the flip side of it is, yeah, for some people, they’re really not happy with this, but that’s because what they’re doing they’ve gotten away with, and they they’re, you know, they’ve liked that. So whether that’s, you know, having methane leaks made made public, and the amount of methane coming into the air, not everyone’s excited about the fact that that’s going to be fully visible. You know, we work with Humboldt County and other counties in California for monitoring how much marijuana is being grown and is it within the permitted amounts. And that’s also something that people would love to fudge the numbers. That’s not possible anymore when you’re when you’re being monitored from space, and we can do calculations of that crop biomass to know where you know more production is happening that is than is what is permitted. Same thing on building construction activity. So, so yes, there are people that aren’t going to be happy with the data that is more generally accessible, but from a security and privacy perspective, we are very careful with this information. We have an ethics committee that is thinking about what are the potential downstream implications of the work that we’re doing with different entities, and how might this go sideways from our corporate and personal efforts? Six so that we don’t think about it after the fact, but we’re actually thinking about it beforehand and putting the right safeguards in place.

David Ariosto – So in the last bit of time that we have here, I kind of want to get back to that what I sort of piped in a little bit earlier in terms of just the sheer amount of it that you’re looking at. Because you know, when you talk about this level of imaging, and you talk about the inclusion of AI in a much more sort of profound way. And you talk about, you know, the various sort of vectors that are not being under consideration, that are not, you know, purely related to resolution. It’s a lot to process, and there’s latencies and you know, and how you deal with that, and the nature of how the industry is changing. I mean, I recently saw Eric Schmidt of relativity spaces saying that, you know, he focused on least, at least in the tweet, had focused on on some of this because, because of data processing concerns on Earth and, you know, energy constraints. But I think it’s a real question in terms of, you know, this is such a growth market, and there’s such a need for so much of this, this imagery at at speeds that are just, you know, there’s a bottleneck potentially there. So how do you, how do you address that?

Ashley Johnson – Yeah, so this is something we think about all the time, as we’re downloading 40 terabits of data every day, terabytes. But there, there are a couple of ways that we’re looking at it. So, so one is just simply optimizing the data that we have in such a way that we’re serving up the most relevant data. We’re moving into, you know, lower cost, lower energy consumption, storage area that that data which is not going to be of use to the you know, whether that’s because of cloud cover or other reasons, but we are also moving upstream from that to the satellite.

So in the Pelican that we launched in January, and this Pelican is our next generation high resolution imaging satellites, we included an NVIDIA chip, so Jensen chip, and with that, what we can do is start to optimize what data we communicate back down to earth, so and then how quickly we communicate it back so, for example, if you, if you thought about, you know, just a classic, genericized military or even economic use case where you want to understand the amount of activity going on in a particular area, how many planes on A runway, how many cars in a parking lot, that type of thing, you might capture an image that is 80% cloudy and you can’t get any of that data from it. You want to know that right away, so that you queue the next satellite to retake the image. But you can have an 80% cloudy image, but the 20% that is not cloudy has all of that information, so you don’t need to cue the next satellite, and you can use the AI in space at the edge to communicate the data back that says the answer is six there were six planes on that runway, or the count is 300 cars in that parking lot.

David Ariosto – It says or changes the role of the analyst back back home, so to speak.

Ashley Johnson – Exactly, you get that information back in real time, especially as we we’re developing satellite to satellite communication capabilities. You take the latency from what is now hours down to minutes, and ultimately down to seconds, and reduce it from you have to produce an image that gets downlinked rectified, and then an analyst looks at it to what’s the data that I needed to know, and what do I want to do next? So from an analyst world, their job becomes much better. They’re thinking, they’re thinking about what comes to the right so I know what happened. I know the data today. Now I need to think about what does it mean and what comes next. And that is a much more powerful job for an analyst than just staring at imagery and submitting counts into a database.

David Ariosto – I think that is a good place to leave it. Ashley Johnson of Planet, thank you so much for joining us. This is just like it’s such an insightful and pleasant conversation. I think, I think what you’re all doing over there is terribly exciting.

Ashley Johnson – Well, thank you so much. I really enjoyed it, and it was really great to meet you David.