Subscribe
Share
Share
Embed
DRONE ON explores how drones are reshaping the world. Hosted by Bryce Bladon, the podcast documents the tech, economics and people piloting the world's largest standardized drone imagery network.
Why the Uber for Drone Data is Upgrading the World Map
===
Bryce: [00:00:00] Welcome to Drone on the only podcast in the air and on the airwaves. I am your host, Bryce Bladon, and on this show we explore the tech, economics and people building the world's largest standardized drone imagery network. Each episode we explore how drones are reshaping industries. Creating new economic opportunities and literally changing how we see the world.
On today's episode, we examine a company revolutionizing the way we capture and access imagery data. With the Uber for drone data, we explore why the company's CEO believes current imagery solutions are broken, how his platform connects drone operators with businesses needing high resolution spatial data, and what advancements in drone and emerging technologies like blockchain might play in scaling this vision globally.
Today's guest is co-founder and CEO of Spexi Geospatial. Bill Lakeland. Bill, why don't you tell the audience a bit about yourself and how your background led to this topic?
Bill: Okay, great. And hey Bruce. Good to see you this morning. Yeah, so Bill Lakeland. I'm the CEO at Spexi. I've been in area mapping for quite a long time now about [00:01:00] my will here at university.
It was my passion one or two University with area photography and mapping to help optimize the planet. Environmental impacts. There's just a bunch of really good sort of environmental reasons why I got into this in the first place. It was very compelling and I got into it , and. That went down the rabbit hole as far as it goes.
And so the story evolves sort of flying airplanes and large format n cameras, lots of parts in the world, all over North America. A lot of governments mostly buying the data. Uh, and there's always the question over the years of doing this that how do we get it at a higher resolution and how do we get it more frequently?
And it comes down to costs and scale. Airplanes can be everywhere at anytime. They're very expensive to operate. The camera's very expensive. The processing is. Takes days, sometimes months, uh, to, to get through the system. And so there's just a real bottleneck scale by using existing technologies. And so I've always been trying to figure how to, how to level this up.
Drones came online back in 2007, honestly, when I started looking at them. But really they didn't may start making an impact till about 2018. Regulations started to [00:02:00] change. It was easier to fly drones. More at scale. The technologies became more miniaturized. We're not bothering the general public. They're very accessible, meaning every corner of the earth, we can access drones that are capable of collecting imagery at a better resolution than a million dollar mapping camera.
And so it became time to actually start building a system to put these things to work. And the pilots that are passionate about flying drones and just want a reason to go out and collect imagery or just. With their drones in the air, are we giving them really cool visa to do that? Give them purpose and also some form of payment?
And so we started building a software application to do that, this level up, this game to to provide higher resolution, more frequently collected data. We started building this in 2018. Actually, it was still little bit early, but as time progressed, some technologies advanced for us to truly scale as drones became less than 250 grams, which was a really important threshold for us.
They don't bother people, the altitude they're flying at and they are distributed all across the globe at, uh, numbers that make sense for us to actually fly cities very quickly. A blockchain technology really came online in a way [00:03:00] that, that severe powers, decentralized networks, the supply side of the system, it unblocks scale for us.
In terms of an ization piece, as well as an authentication of the data. So being able to prove the immutability of the data being collected using decentralized pilots, other decentralized hardware. Xi doesn't own the hardware. It's not a closed system that we can, uh, control. So blockchain has actually unlocked a scale to be able to authenticate that data.
We call that proof of capture. So there's a bunch of technology pieces that came together, and I'm gonna say that really happened in 2022 to really turn this into a reality. AI being another big part of that as well.
Bryce: Got it. Well, what you're building right now is the world's largest drone imagery network, but I'm really curious about what it is.
Today and how it functions and, and then maybe you can paint me a picture of what it would look like when it's fully realized.
Bill: Yeah. Today we're still in the early days, so right now with Spexi has launched a pilot network that will be decentralized in the near future in the agreement with Layer Drill and Foundation.
And so the Pilot Network will be shifting to that location where Spexi will be the first adopter of the data coming [00:04:00] out with our girl Foundation. So the pilot network though, is currently active in North America. We're in over 200 cities, uh, flying very frequently at a very high resolution. And when I say that it's like greater than better than three centimeters in resolution, the status quo typically for rare plays is around 10 centimeters.
Satellites of s you get is really around 30. That's very expensive and hard to get and it takes time to get that type of data. So we're talking so much more detail, um, because it's a decentralized network. The cost of acquisition is significantly lower. These pilots are so distributed across the globe that we could access really high resolution data at low cost because pilots are already in the communities and towns that we need the data acquired in, right?
So we're actually activating local citizens to fly. So we have to send an airplane a thousand miles anymore to, to fly a remote city or town. There's already pilots there, right? So these networks are spun up. They're already activated, they're flying, they've been collecting data for over a year now. We we're 4 million acres in data collection and yeah, hundred thousands of flights already in the can and flying actually right [00:05:00] now.
And it's pretty exciting to see this grow and scale at the pace that is growing up.
Bryce: Maybe you can walk me through what would be the difference between using a platform like Spexi to try and map something versus trying to use like a satellite. You, the one thing you did hit on was that the satellite has, uh, you know, resolutions of up to 30 centimeters and on Spexi you could do three or possibly even more.
What else is there that differentiates the two approaches?
Bill: The
Bryce: big one is
Bill: like tasking the system to be able to collect quickly, so it's hard to task a satellite to collect. Quickly, and it's very expensive. So the cost difference is 50 times, essentially the, the cost of fresh year. That's one aspect, not just the resolution.
Right. And so with that higher resolution and that that ability to collect faster, there becomes more real time use cases that were just never unlocked before. Mm-hmm. Use cases in, in relation to AI and large geospatial models that. Can actually use current state imagery at a resolution that's high enough to point out things for business systems.
So show me the missing roof [00:06:00] shingles on a roof. Show me the state of the sidewalks. All these sort of like enterprise analytics that every company needs and uses to function day to day, whether they realize it or not. People move around the world, robots, cars, all these things move around the world. If they can have a higher resolution digital world in their brain to make decisions before they're actually in motion.
It creates more efficient motion, and that comes with a huge amount of value. So we're able to collect data at a high enough resolution and high enough frequency to inform those digital models that enterprises rely on to make day-to-day operational decisions.
Bryce: So it sounds like it comes down to three things, at least in comparison to planes and satellites, and that's speed of data, quality of data.
And unless I'm misunderstanding something, cost of data, is there any other major differences? 'cause it seems like the trifecta.
Bill: Yeah, I mean that really is, and that's always been the trifecta I've been shooting for all these years. Airplanes can fly very high resolution. They can get three centimeter and below.
That's doable. Very [00:07:00] expensive and difficult to do at scale. Airplanes aren't everywhere. They burn a ton of fossil fuels, so it means they produce a lot of carbon to do this and the turnaround for those data sets just takes a lot longer than from drones. So it's those sort of performance vectors are make it unscalable compared to what it is that we're doing.
Bryce: Well said. You, you had mentioned some use cases coming online over the past few years, but where are you seeing the most traction in terms of industries or or use cases for this data?
Bill: A lot of like property insights at the core. And I, I tie this back to, you know, Uber's always got this last mile delivery problem and it's actually more like a last hundred meter delivery problem.
Whereas if Google's good at getting you generally to that location within a hundred meters. Then once you're in that last hundred meters, where is the door to that building? What is the roof that I'm looking for? Where is that hydrometer that I need to find? Where is that level of granularity? Right? And that's the type of information we can extract.
And so where, when we're talking about that last hundred meter that comes to city [00:08:00] operations, whether you're in parks or whether you're city planning, whether it you're a parking officer, all these sorts of things need this high fidelity information that translates to insurance companies trying to underwrite.
It translates to. 9 1 1 systems trying to get the right ladder to the right buildings. They know where the third story is, or if there's blockages and alleys to be able to access. And I'm getting close. Show me an image of the door so I know which door to knock on these little, it's a detail that are so important in that last set of meters that apply to solely different vertical markets, right?
And so it. The problem is more of a generic problem, but there's early adopters that are typically governments, fire and rescue 9 0 1 systems insurance, a lot of operational companies, logistics companies like Ubers and Door Dashes, and all these companies that do have that last mile delivery issue in their face.
Right? And we're working with partners to scale through these segments. And really solve this last a hundred meter problem
Bryce: that to that last mile problem. Thank you for making it both a metric and otherwise there. But, [00:09:00] uh, that last 100 meters, do you think that is the problem, like the drone imagery market is gonna solve in the next three to five years?
Or is there anything else on the horizon? Because I see that as kind of the culmination of. Of what a lot of those startups are promising, and I can see that taking up a lot of value, but I wonder if, if that's the extent of it.
Bill: Yeah. Honestly, the imagery is only as good as the machines that are extracting information from it.
So really creating a digital planet, a large geospatial model like an LLM, but not, but a geospatially charged model, creating something and updating this very quickly and frequently is, in my opinion, it will solve the this logistic issue that is so prominent with so many different organizations. But it's also gonna help streamline just robots moving in a more meaningful way through our world.
And so optimizing operations for so many different levels of enterprise government and just humans, I want to create a digital bunny to, to show me where the nearest Starbucks is in the new city. And I'm putting all my glasses in an [00:10:00] augmenting reality. And this bunny needs to be trained on a digital world so that it knows set of walk through the real world and not walk you off a sidewalk, not walk you into a park bench or a tree.
It's actually. It's taking you through the city to your Starbucks or whatever, wherever it is you're trying to go to. And so those things need to be trained very accurately and the power of what that is pretty hard to put a pin on right now, but it is an absolutely ginormous market that is, is the one we're truly after.
Bryce: Absolutely. All my mind immediately goes to is how Nvidia currently the world's most valuable company, in addition to all this AI LLM stuff, has been quietly building the concept of an omniverse, which among other things is. It's very hard to describe as something I haven't read up on in a while. But my understanding is basically a metaphysical space for large scale industrial conceptualization and manufacturing, among other things, and the main challenge that Nvidia has been running into with that concept, data quality.
How do you ensure that [00:11:00] things at the, say three centimeter range are exactly where they should be, especially for all the, all the great possibilities that AI is supposed to unlock? It does seem like. Data quality is perhaps one of the, one of the big things that people point to a lot of the time. And, and in my own personal experience dealing with LLMs, bad data can, can be the death nail of something.
Bill: Yeah. If being at the grown level to start building those foundational elements to a large model is actually very important because there's so much data out there and it's all, it's not registered. Refer together. And so there's different accuracy components, there's different resolution components and things don't fit.
And so this is why it's been so important for us to like hyper standardize what is they're building. Every drone is flying the, exactly the same way. Every part of the world is collecting data. So what we're doing here in North America will be the same. That's in US really in South America. If we want to add complexity, IE fly lower or fly higher resolution, which we C certainly can do.
That is applied broadly across the spectrum so that it's. The same type of data, the [00:12:00] same accuracy. And so systems can actually now reliably build on top of this, uh, in a consistent layer, in a consistent way. If we're just crowdsourcing data, meaning you go fly your drones, however you fly them, we'll take your data and turn into a product, it becomes impossible.
To actually put all these pieces together and make a reliable product for other systems to build on top of. So the standardization, the protocol that that we'll be living inside of Layer Drone, the set of rules and the technology to collect this at scale in a very standardized and meaningful way is so important for the, how valuable this data can be in the future.
And it being a foundational element to what I've described.
Bryce: Absolutely. It does seem like the standardization would be. Crux of it here. If you can allow me to get a little, I dunno, weird in my metaphor about this. How I've always understood what Specia and lone is doing. And boy, this metaphor is gonna get mixed fast.
The concept of web one to web two to Web3 kind of comes down to the idea of the ability to go from Web3 being you can now own things on the internet. Web two being you can copy and run [00:13:00] things on the internet web thing, web one being the ability to interact with the internet. I miss. Representing it, and this metaphor is only gonna get messier from here, but on the world map side, around the turn of the 20th century, most of the world map had been mapped.
But what was being solved with what I like to call world map 1.0 was standardization in a way. There were entire sections of ocean that were still extremely difficult to navigate Around the 20th century, the concept of nautical miles and things of that effect, latitude versus longitude. One was very navigatable.
One was not in in big, open stretches of water. Mapping 2.0 in my mind has always been the Google Maps ification of, of, of the world map as we know it. The ability for an individual to interact with the world map at any scale, at any point and, and get meaningful visual information outta that. What I hear with what I'm gonna be calling World Map 3.0.
The standardization and the overall, I dunno what to call it, other than ability to utilize that data, is, is kind of how I'm conceptualizing it in my head. And to be clear, these are just [00:14:00] mixed metaphors I am using to try and understand where things are going and how they got here and, and the patterns from history that tell us about it.
Bill: That's a super interesting way to describe the pathway and I like how you've done that actually. That makes a lot of sense. And I would actually more say this is a three dimensional like. Google Earth created the 2D version and there's a two plus D where there's some horizontal, there's some oblique looking, there's some 3D adage to certain components.
True. 3D is what it is A we're created, right? Yeah. We, we are the 3D version of this, and in fact, I would say four D and in fact, I don't wanna jump on a four D abstract thing here, but four D really means there's a temporal aspect to this. No, you're pretty rough. And so then if we're collecting everything every three months, or we're collecting everything on demand, then it's very temporarily charged.
So it's, it's the 3D and it's constantly up updating and refreshing that can be cycled and tuned into two different timeframes, right? And so that, that is what generates a four D amounting application. I'd say that's where we're at.
Bryce: I love that. I was trying not to add too many [00:15:00] dimensions to my metaphor, and that's where it fell apart, because you're absolutely right.
If, if you're a customer accessing the network explorer right now, uh, or the actual network data that is. One of the key overlays is to check out what did this, what did this imagery look like three months ago if, if it was collected. And I think if you go to spec.com, you could even see some of those comparison photos.
But I digress. We've only got a couple minutes left and I gotta take you to the funnest segment of the show. But first I did just want to ask one last question for you, and that is, what would you say has been specie's biggest? I don't wanna say success. Let's say lesson so far, you guys are operating in a very unique space, and you have been one of the few companies in this space for a long time, long being relative, but drones have only been around for a few decades.
In my mind. You've been at this for a while. You used to fly planes. Now you fly drones. Now you have an entire company that flies drones, and now that company has an entire decentralized network that it builds. On top of that is hundreds, if not thousands of drone [00:16:00] pilots. That's the biggest lesson you've learned.
Bill: The biggest thing for me is timing and adoption and technology. I think I've always been trying to push the envelope as as quick as I can. Like I said, in 2018, when this platform first launched, we were too early and we were too early because drones weren't small enough yet AI wasn't capable enough to be able to extract information fast.
For the value that this needs to generate certain systems. Blockchain technology wasn't there to be able to decentralize this, which adds a piece of scale. And so it's like pacing this at the bleeding edge of where, where it can be versus where it can be adopted in the world is. One of the biggest learnings I'm always keeping a very close eye on, and I think we've awarded, um, to bridge that very closely.
The data set we're collecting truly, it's always makes you a little bit sad in a way because I love imagery, I love photographer, it hurt and I got into this as a photographer and I love science. This is a science of photography. And then how do you scale that, right? That's where I came from. And environmental impacts, all these sorts of [00:17:00] things.
Really what this is becoming is the highest resolution, highest all the three image layer of the planet that no human actually even looks at. This is all, this is all for computer and robot like interactions, right? Extract information in real time to make like minute by minute decisions, operational decisions.
So I mean, to be able to do that means we can't just generate this thing and assume it's gonna go. We need to work with AI companies and partners that are solving other problems so that this data can be leveraged in the way it's intended to be leveraged. Right? And those are all moving at different paces and trying to pace all those pieces simultaneously so that this can explode at the biggest level that it can is to me, is one of my, one of my bigger learnings.
I'm costly learning this. Right? And I find it very fascinating. I find it. Really cool to work with some partners that are doing really, really interesting stuff, partnering with ground-based data sets that were this 3D from the air. But ground is also very important. Marrying technologies to, to [00:18:00] maximize value in what this is.
It's very exciting. So I would say that's probably one of the bigger learnings. And also just the ability for Web3 to unlock scale. I wasn't expecting this, if you go back to my earlier business plan, but my first business plan that I wrote in 2017. I was talking about service providers and data collectors, right?
And so is like crowdsourced 2.0 kind of a space. I never expected blockchain to be the enabler for scale in this, in the way that it has, which is, it gives me goosebumps every time I talk about it because it was like I'm missing piece that I was always looking for that I didn't realize was actually evolving in real time at the right time.
So from a Tanya perspective. That just plays right into the story.
Bryce: Absolutely. I'm gonna summarize some of what we said here, but to put it simply, it sounds like Spexi company has been around since 2018, arguably conceptually a little early, but the product, the platform, solves a very real problem. Current imagery solutions are expensive, they're slow, or they're not at the resolution that modern or emerging applications require.[00:19:00]
What is perhaps most interesting to me might be, 'cause it's partially related to my job, is where Spexi goes from here and how you scale that Now that the opportunity in the moment does seem right now that the Web3 technologies, the drone technologies and, and even the regulations arguably have never been more agreeable.
Exactly for folks interested in becoming a pilot, to using the Spexi app and flying automated flights, the Spexi.com to learn more about Laird and, uh, how the network is scaling in a decentralized fashion. Laird, are there any other resources or places you'd suggest to anyone interested learning about this stuff or maybe even a, a potential customer for the, for the network?
Bill: Yeah, customers Spexi.com, they'll have all our contacts in there, all our social channels, everything will be tied to all these, so it's very easy to get ahold of us. We're very friendly bunch, so we'll make it very available. So those two channels, Bryce, are definitely the best locations to go to to find the method of communication that's referred by whoevers there.
So, but yeah, Spexi.com definitely for for customers. And layer drone.org [00:20:00] for pilots for sure.
Bryce: All right, Bill, you're a very important man leading a company with dozens of people. Do you have two minutes to play the Bryce is Wrong?
Bill: I'd love to.
Bryce: Alright. I'm gonna tell you three statements about myself. Two of them are true. One's a lie. You have to figure out.
Bill: I feel like there's a red herring in here somehow.
Bryce: So, oh, I love. I have a career of raising red herrings, my friend. In 2017, Bryce's Media work helped the International Fencing Federation leverage augmented reality and 360 video to promote Tokyo 2020.
In 2018, Bryce's blockchain work was leveraged as a fundraising tool by WikiLeaks. In early 2020 when COVID threatened to delay the Tokyo 2020 Olympics, Bryce worked with the IOC to create the mascot Olympics where mascots from prefectures from around Japan competed against one another in a COVID safe version of many Olympic events.
Bill: Two Olympic items on here. Did I?
Bryce: That is correct. One with [00:21:00] the FIE, the International Fencing Federation. It's French. If you're wondering why that acronym doesn't add up,
Bill: I was wondering about that. I'm like, I'm not familiar.
Bryce: The other with the IOC, which would be the of the Olympics.
Bill: I'm gonna go with the first one,
Bryce: unfortunately Bill, you can find that work on the New York Times new media page. It was the story of the day in late 2017. We did a feature with, oh gosh, I'm forgetting the. Fencer's name, but a, a Brooklyn fencer Peter something. This is embarrassing, who basically did inner city fencing Academy for a lot of people who wouldn't otherwise get the opportunity.
It was a great little video. It was the third one. The IOC does not like you doing anything that might remove the dignity of the games, so.
Bill: Okay. Very cool. Well, I should, I was close. I was close, but
Bryce: it's all right. No winnings for you, but we'll get you next time. Thank you so much for your time today, Bill.
I really appreciate it.
Bill: Yeah, likewise, Bryce. Good to see you.
Bryce: Good to see you.
Thanks for listening to Drone On. [00:22:00] Subscribe wherever you get your podcasts. Get a new episode every week and leave us a five star review on your podcast app of choice. You can learn more about our sponsors at Spexi.com. That's Spexi.com and LayerDrone.org. Find out how you can contribute to the world's largest drone imagery network too.
Thanks again for listening.