TLDR

In this episode of BidCast, Jason Kam interviews Mike Horton, the co-founder of GEODNET, discussing the innovative technology behind their low-cost base stations, the competitive landscape in the geospatial market, and the implications of tokenomics on network growth. They explore the impact of GEODNET on agriculture and drone industries, marketing strategies, and future projections for revenue growth, particularly in the robotics sector. The conversation highlights the importance of decentralization in technology and the challenges posed by international competition, particularly from Chinese firms.

Chapters / Timeline

• 00:00 Opening

• 00:44 Introduction to Geodnet and Its Technology

• 03:45 Understanding the Cost Differences in Base Stations

• 06:43 Market Competition and Hardware Evolution

• 09:34 Tokenomics and Network Coverage

• 11:19 Marketing Strategies and Community Engagement

• 13:14 Empowering Small Businesses with Geodnet

• 16:40 Manufacturing Challenges and Decentralization

• 18:46 Revenue Growth and Customer Base

• 23:48 Future Projections and Backlogs

• 27:45 Understanding Customer Dynamics and Tokenomics

• 31:40 Exploring New Use Cases and Revenue Opportunities

• 35:23 The Role of Robotics in Future Growth

• 40:35 The Impact of Robotics on Market Dynamics

• 45:16 Innovations in Indoor Positioning and Satellite Technology

• 50:44 Competing in the Global Robotics Market

• 57:33 Financial Sustainability and Future Outlook

Transcript

<< Jason Kam (00:44) >>:

Yeah. All right, guys. Welcome to another episode of BidCast. I'm your host, Jason Kam, aka, @MapleLeafCap Today is May 28th, 2025, 9 p.m. Hong Kong time. BidCast is being live streamed to BidClub members. Questions are from the members and my own. Today I'm speaking with Mike, the founder of GEODNET Mike, welcome.

<< Mike Horton (01:04) >>:

Hey Jason, good to be here.

<< Jason Kam (01:06) >>:

Yeah, it's really early there in the West Coast. I really appreciate it. Let's dive right into it. I guess you know, this is the question that came up to my mind when I first started doing DD on a name. What is really the difference between a $25,000 base station that your competitors install versus the $700 one that your users install? Like, is there any difference at all?

<< Mike Horton (01:33) >>:

From a performance point of view and technical capability, there's no difference and in fact in many cases our station will be better because it's tracking more satellites and more signals. From  why the cost is lower,  one is we're not trying to make money off of base stations and that certainly is a core part of the business of some of the competitive  networks. And then the second piece is we're using the most modern integrated circuit technology.

So as you know, integrated circuit technology sort of follows these different process nodes and the silicon that's inside of our base station is on like an eight nanometer node. And so it's therefore it's lower cost and lower power. If you compare that to sort of a traditional  base station, that may not even be an integrated circuit at all. It may be an FPGA with a bunch of discrete RF components that drive the bomb cost up well beyond the price of a GEODNET mining station.

So essentially we're just using much more modern technology than those base stations. And that modern technology is born kind of out of the opportunity for GEODNET, which is that this modern technology is now in things like the latest Apple phone and the Android phones. These higher  capability GNSS chips have made their way into mobile devices, into cars, and the networks that support them just haven't caught up.

because the network you have to refresh physical things on the ground. It's actually much harder than, you know, replacing the watch on your wrist to go install a base station in a remote location and get it back connected to internet. And so that's where the magic of Deepin has come in order to not only refresh and create a better network with all the modern signals, but also a much bigger network now standing at about three times the size of the nearest competitor and do that in a short period of time. That's really the magic of Web3 and tokenomics and

crypto rails and all this stuff.

<< Jason Kam (03:30) >>:

And it costs almost nothing to run. The electricity cost is nil.

<< Mike Horton (03:32) >>:

Yeah,

it's a couple watts, two, three watts of power.

<< Jason Kam (03:36) >>:

Why are they pricing it like it was back in the 90s or 2000s versus kind of dropping the cost dramatically and upgrade to tech to kind of be more purposeful to you?

<< Mike Horton (03:45) >>:

Yeah, I mean, think that,

well, when you have a huge company to feed and you have an existing business and reputation, these things, you can't change these things overnight or you,  you know, you essentially go bankrupt or you destroy your business, right? mean, those businesses have been built over 20 years. have incredible reputations that they, that, you know, price is an important part of how people think about things. And a lot of the customers of those businesses are

You know governments and cities and people they have contracts and they've done You know most favored nation pricing agreements and all kinds of you know contractual and business reasons You just can't go in and mess with that not to mention just the fact that if if you drop your revenue line By a factor of 10, you know or more in some cases Your business it just doesn't work the same way anymore. So those are reasons. I think a good analogy would be

<< Jason Kam (04:40) >>:

Hmm.

<< Mike Horton (04:43) >>:

Sort of the, think about where Nokia was when the iPhone first dropped. That's kind of where we're at. You know, I'm sure Nokia at the time the iPhone drops said, hmm, interesting. This could be a problem, but let's, you know, wait and see. yeah, we wait and saw and things changed.

<< Jason Kam (05:02) >>:

And are there any sort of other

competitors that you kind of see in the market because they do go to a lot of trade shows from China that sort of build similar tech with similar price points even cheaper?

<< Mike Horton (05:13) >>:

On the hardware side, think the price will continue to come down. So a base stations is a network effect thing. you really what's powerful about GEODNET is just getting this network established and running and so on. But on the device side, I mean, prices are continuing to come down. New devices are coming out. mean, consumer robotic lawnmower, you know, for seven, $800 with the house is when you look at that and compared to our base station, our base station looks expensive because it has

In a sense, even more equipment in it to do the job. So I think the hardware costs are coming down. That's enabling a lot of  new types of physical AI devices. That's what I always say is like now the sensors are really ready. mean, before, why hadn't all these robotics and drones kind of been able to go mass market? A lot of it had to do with the sensing technology, be it GNSS, the chips like what's in our base station weren't there.

The LiDAR sensors, the cameras weren't there to really enable it. The understanding that the compute engines that to process all this data weren't there. And now, now those things are. And so these things can have their moment. And it really is very analogous back to the sort of the iPhone analogy back before the iPhone. You know, people tried to build those types of devices over and over again, and they couldn't quite get it to work because the compute memory screens.

all the elements, batteries, weren't quite there to pull off something that can meet the sort of overall zeitigest of the consumer. You would come out with a device like a Compaq IPAQ. I had those, they were based on Intel. Intel at that time had a good arm silicon, crazy, they sold that off. But they had it and they had this device, the IPAQ, it was pretty close, but it was a little heavy.

was a little clunky and it had this sort of junky version of Microsoft Windows on there. It wasn't a very great experience. You had to use a stylus. So, you know, just that small change to getting the form factor right and getting the component cost down, making it thinner, that sort of enabled it. I think that's where we're at. You know, the sensors are there. A lot of these devices are still a little bit clunky. You look at the videos of the humanoids, you see them, they're still...

pretty clunky, but we're getting really, really close to make that transition to very mass market devices.

<< Jason Kam (07:42) >>:

So with

the happening that you're doing with the geotoken emission, would seem like the hardware cost will also keep going down. the ROI, assuming price remaining the same, it can actually still make a lot of sense. Is that the right way to think about it?

<< Mike Horton (07:57) >>:

That's a way to think about it if there were hexes to fill. But our tokenomics are set up on really an incentive around a one high quality device per hex.  a lot of regions that are of interest where most of the revenue generating potential for the network is, the vast majority of hexes are in fact already filled with a high performing miner.

that has earned its NFT, which means it kind of has locked the hacks and has obtained the governance rights associated with the NFT. So another way to think about it is really the GEODNET network is almost complete. Our tokenomics don't really encourage people to deploy additional devices in places that already have  complete coverage.

<< Jason Kam (08:37) >>:

 And for those hexes...

And to the extent, if we were to go into a hex and there's a lot of miners deployed,  do these devices get saturated  based on significant usage and therefore requiring more miners to be deployed into those areas? is there like, there's no limit? Okay.

<< Mike Horton (09:05) >>:

No, no, no, no, no,

you're a geographic coverage problem.  so we're trying to cover geography. We're trying to cover geography that will, you know, be a benefit to our customers and that will let them generate revenue and therefore us generate revenue. And you only need one miner per hex to satisfy. could put all,  1.4 billion people from India in one hex. If you could squeeze them all in my ones that one station could, could cover.

every mobile device. So there is no saturation. It's very different than  wireless networking.

<< Jason Kam (09:34) >>:

wow, okay.

Yeah, brap entrant. Interesting.

That's interesting. And then to the point of distribution,  of the 17k devices that you sold, did you spend any marketing on those, selling those devices? Or like social marketing, or is it kind of happening organically? Okay.

<< Mike Horton (09:55) >>:

Most of the markets are sure. I mean,

my goodness. Yeah, I would say so. You know, we're not, you know, running TV ads or things like that. Not in a traditional Web 2 sense, but and certainly our investors feel like we don't do enough Web 3 crypto native, blah, blah, blah. I hear that every day. But from my point of view, from a Web 2 guys point of view, we've put in a tremendous amount of effort to build up

<< Jason Kam (10:16) >>:

Yeah.

<< Mike Horton (10:25) >>:

you know, community channels, be it discord, telegram, be it attending,  web 3 events being doing podcasts, what have you to help people understand the value proposition of GEODNET, why it's an interesting project to mine, why it's an interesting project to follow. Definitely. Absolutely.

<< Jason Kam (10:45) >>:

It's mostly those DePin investors who bought the device and set them up.

<< Mike Horton (10:51) >>:

Yeah, absolutely. There's probably a good 80 % are individual  deep end enthusiasts who've purchased devices and set one or two up. We did a poll on Twitter recently, or on X I should say recently of how many devices you have. And I think like two was the most common answer. So maybe you get one and you get one for your friends. So it's a very diverse network. There are a few people who've deployed more.

but they're in the minority. And then there's also some Web 2 entities that use GEODNET, including a number of the dealers of Ag products now  and  drone products. So in Ag and in drone, we've had a number of dealers now, more than a handful, who've also deployed quite a lot of devices in support of their business operations. And that's pretty neat because you have a person who's running a drone services business or selling equipment to farmers for precision agriculture.

and they are able to make money selling our service. They're able to make money selling their equipment and then they're able to make money actually deploying stations and mining. So it's been really great for them because it's helped them grow their business quickly.  And they've been able to leverage the full power of the GEODNET value proposition. So those guys are super loyal enthusiasts and have kind of been brought into the web three. Like a lot of them we've gone through. see

On our YouTube channel, we post videos on how to set up a wallet, how to swap, how to use quick swap, type of stuff, how to use raydium. We do that for the specific reason to support those guys. And,  you know, they love GEODNET. They just love it because it gives them now a footing to compete with the quote big guys. know, so if they traditionally, you know, someone, a smaller  service provider  of equipment and Ag might be competing with the John Deere dealer.

And they wouldn't necessarily have access to the John Deere network. They wouldn't have access to those resources. Now with GEODNET, they do. And it's been super effective competitively. They've been able to really grow like crazy. And at the same time, these big companies, they are very extractive. That's  a sort of standard thing, I think, in American business culture is that as these companies make it into the Fortune 100 or whatever, they just become more more extractive. So there's a lot of frustration.

in these fields, but geospatial stuff, like you have to have infrastructure to play the game. And that's very, very difficult for a small business to build infrastructure and GEODNET and the web 3 tokenomics around it really solve that in an incredibly elegant way. And I just think that that  is going to continue to pull pretty hardcore web 2 people into the space. And we expand plan to expand how we've done, how we had this success with guys and drones.

And in agriculture, we're going to expand it into other areas, into a more broad robotics play to just bring in all kinds of robot developers who have challenges of  the development cost, of servicing, of traceability of their equipment, how it's working in the field. And Web3 and the GEODNET platform can help solve all of that. So we're going to have more and more platforms like that come up.

<< Jason Kam (14:03) >>:

Yeah.

I definitely want to get to robotics in a second, just on this topic, the last one I want to ask is just roughly ballpark. If we put a GEOD Token aside, how much cash dollars have you spent to actually make people aware of the 17k device that you have deployed? How much cash dollars did you spend to get those out the door? And then maybe for the next 17k, do you expect to spend more or less marketing dollars?

<< Mike Horton (14:30) >>:

boy, I don't know if I have a good number off the top of my head. I guess I wouldn't think, yeah, maybe a million or two. Yeah. One to 2 million.  a lot of sweat actually. Yeah. A lot of the team, ⁓ you know, responding to discord posts, I would say has been the, the, you know, the thing. And, and, and it's not a huge team. So probably not in dollar terms, probably not.

<< Jason Kam (14:35) >>:

Yeah, so is it like a couple million? Is it like a couple hundred K?

Okay, and it's a lot of sweat equity, guess, then. You just like...

<< Mike Horton (14:59) >>:

a huge number and sweat equity terms and emotional effort and 24 by 7 customer support on Christmas Day. Yeah, a lot.

<< Jason Kam (15:07) >>:

got it

Yeah.

And those devices that you sold, the hardware, you sell them at breakeven, gross margin-wise?

<< Mike Horton (15:18) >>:

Yeah, so we don't sell the, GEODNET Network doesn't sell the devices. The GEODNET Network has partnered with a company, HiFix, who developed, who I'm involved in, in transparency, but who developed a device. They then turned around and gave it to a manufacturer who's been building it and then supplying it to a number of distributors. there about 10, 11 distributors of the device. And it's a certified device. So every device has  a crypto chip in it with this  private key that creates a...

cryptographic signature of the data so we can tell that the devices are authentic and those you know gets sold through a distribution channel and that helps a lot because countries have different languages, you  know border policies, different certification, RF certifications and we've gone through that now in all different kinds of places whether it's Anatel in Brazil or this and that you know it's just on and on and on ⁓ and I think that that type of decentralization

is very critical and we're going to do more of that. We're going to enable more of that in robotics as well. think this type of geospatial and physical AI infrastructure, people are going to see more and more that's super important for every  country if it wants to be able to maintain sort of its real sovereignty and peace, needs to have a capability to build, develop, maintain, service physical AI machines.

<< Jason Kam (16:40) >>:

And only the device manufactured by that manufacturer can farm $GEOD And you will not be turning that on to any other devices that's sold by anybody else.

<< Mike Horton (16:49) >>:

No, no plans to do that. think that one potential exception, well, a couple of exceptions. I we've been starting to have these devices manufactured in multiple geographic locations. So we added and initially everything was built in China last year due to trade disputes between China and India became quite problematic. So we started in India and this year we've been working to get it going in the United States. It's taken a little longer than I had hoped. It's not.

not trivial today to build things in the United States, but we're working on it. And I think the lights at the end of the tunnel, then we'll start to be able to crank out production in the United States at a reasonable cost. But it's been a pretty massive effort actually to figure out how to even make a relatively simple device like the, like the miner  and get that built effectively. So I think the US has got a ways to go to catch back up manufacturing.

<< Jason Kam (17:44) >>:

But if you get a huge order tomorrow, you don't expect this trade war to put an issue on lead time and everything else.

<< Mike Horton (17:50) >>:

Well, I mean, it just depends on what's tweeted tomorrow. I mean, no, I wouldn't go that far. work, we work through it. We have capacity in India. We're building it in the US but I think to say that, to say something broad like that is I think not, not reasonable. The trade war is very unpredictable and that's why we're making the investment. it was just looking at it from an ROI point of view, I think we would

we wouldn't put all the effort into diversifying the thing. But I think this trade war, we don't know where it's going and  you don't know where trade war ends and other things start. We don't know. So it's, I think, really important that supply chains and decentralization occur. Again, especially on things like robotics and physical AI, decentralization is, I think, paramount to national sovereignty and security.

<< Jason Kam (18:44) >>:

Moving on to the revenue side, guess last I checked us. Yeah. Yeah.

<< Mike Horton (18:47) >>:

⁓ can't make a quick analogy on that, Jason, just so the audience really understands what I mean, because

maybe it's not being clear enough. You know, in the Ukraine today, that decentralization is down to the household level. So people are making what Ukraine is doing to support the four million drones per year that are being used in that conflict. They are having individual homes and a deepened style build drones.

So kits of parts go out to homes and people assemble the drones in their homes so they don't have a centralized place of production to be targeted.

<< Jason Kam (19:24) >>:

You can do that too. I mean, it's actually not that hard to build a drone if you have all the components in place. Yeah, yeah, yeah. You can just drop a bomb beneath it. Yeah. And it's kind of scary.

<< Mike Horton (19:27) >>:

It's really easy, isn't it? It's super easy, yes. Indeed. Yeah, it's super easy.

That's the way I feel you. And that's why it's really important it's decentralized because that's only the tip of the iceberg. What's going on there is a small microcosm of what could come. It's serious technology.

<< Jason Kam (19:44) >>:

Yeah.

Yeah.

This is a good segue actually, for the drones. guess you probably work with a drone company today, but of the $3.5 million to $4 million of ARR, can you break that down for us? How many logos, how many customers, what's their churn like, and everything in  here?

<< Mike Horton (20:09) >>:

Yeah, so we have about 10

big customers now that constitute a lot of that. And then we have about 30 plus more smaller customers that are probably in the 20 % range of revenue. So about 80 % is divided among 10 bigger customers. And those are all people who have pretty established businesses and high precision GNSS products and services. The smaller customers, I think,

are all folks who are kind of coming in, a lot of it through the drones and robotics side that are trying to build up new businesses around these types of devices and services. And so I'm pretty optimistic on a lot of that. First of all, we have,  I think, really big impact on both those types of customers. on the bigger customers, I think we've been really enabling for them to launch new capabilities and new services.

for them and they really depend on GEODNET and the smaller customers I think were kind of creating almost helping them create their business together. So I think that, you know, if you're trying to do, you know, a startup to sell robotic lawnmowers, mean, GEODNET's a game changer for you because you can reduce your bomb, you can make it much easier to deploy the product, you can make it much closer to that iPhone experience you're not telling. You know, in that space today, if you go...

On Amazon, you look at this things, you're either burying a boundary wire, which in the United States for our yards here is just not a practical thing to really do, or you're trying to up a base station. And in a lot of places, setting up a base station isn't really super practical. that leads to a really high return rate and a bunch of problems. And that's been a big limiter for the market. And we solve that problem completely.

<< Jason Kam (21:54) >>:

Yeah, that's

⁓ Would you say historically the Grove have been signing up new logos or was it because existing customers just paying you more dollars?

<< Mike Horton (22:14) >>:

I'd say, you know, it takes a good year from when we sign up a new customer to when we get significant revenue. So most of the growth you see at any given time is from customers we signed up ⁓ with a while ago. And for bigger customers, there's another period on top of that for them to just get comfortable before they even sign up. we have a lot of customers. Yeah, it's been years.

<< Jason Kam (22:28) >>:

Yeah.

Wow. So it's like a two year sell cycle for large customers.

<< Mike Horton (22:41) >>:

For large customers can be a good two years. And even then you're still at the early phase of the revenue growth. it's web 2 takes time. It takes time. There's safety considerations. There's  geographic, there's regulatory, there's just integration. There's them training their own sales cycle. There's changing the way products operate to take advantage of having a network available as opposed to telling people to go out and set up a base station.

Yeah, it takes time, it takes education. But I think the good thing is we have a lot of things that are being using GEODNET right now. In fact, we went through a long list of them yesterday that are still in sort of trial phases right now, deep in trial phases, have been on trial for a long time. So those are all things that can come off and start to monetize. And we're  kind of going to see some of that over the second half of the year and into the beginning of next year.

<< Jason Kam (23:23) >>:

Yes.

<< Mike Horton (23:39) >>:

that are already, we know they're using it all the time. We see that the performance is good. So those give us confidence in revenue growth.

<< Jason Kam (23:48) >>:

Yeah, so that's an important question because the 4 million annualized ARR is as of now, but it doesn't take into account of signed contracts and potential backlogs. Roughly speaking, how big is that backlog versus your formula ARR?

<< Mike Horton (24:00) >>:

I mean, I think what we feel comfortable saying is that we do see a good path to to 10 million in ARR in the near term.  think beyond that.

<< Jason Kam (24:09) >>:

Neutromous and

like, like your end.

<< Mike Horton (24:12) >>:

 you know, it, we don't control the rate at which some of these things happen, right? We're, we're, we build a platform, people build on top of that. They roll things out. So we, I can't, I don't, don't certainly don't want to publicly be stating an exact timeframe for that We, push as hard as we can. We do the things that we,  can to support people, but we're a taker a lot of times on kind of how quickly these things actually end up, you know, turning into revenue.

because we're not out there. The foundation itself doesn't directly sell subscriptions. It relies on a network of OEMs and resellers to sell subscriptions and it takes the time it takes ⁓ for these guys.

<< Jason Kam (24:55) >>:

And almost more than doubling of that, that's new customers, mostly new customers.

<< Mike Horton (25:01) >>:

Well, certainly the near term stuff that we see is all things that we already are working on and that have already tested the network. things I'm not even, I don't even really sort of think about,  at least in the near term, you know, so they're saying what's going to happen over the next 12 months.  As I sort of mentioned, even the smaller customers, you, I mean, they may turn on, but the revenue won't be that big initially.

<< Jason Kam (25:12) >>:

Yeah.

<< Mike Horton (25:28) >>:

So everything that we sort of see in the next 12 months, we are working on today. We have people are using the network today. That's not what, so we're not, if we're talking about, we just did the drone show last week. There's some new things that came up. really interesting. That is stuff that will start to impact, you know, later in 2026. It's certainly not going to impact ⁓ the rest of this year.

<< Jason Kam (25:35) >>:

Mm.

Yeah. I  was going to say something.

And then I guess the 80 % of it goes to the burn, the 20 % goes to the foundation. That 20 % is used to spend on salaries and marketing and all these things that you need to do to keep the lights on.

<< Mike Horton (26:11) >>:

Yeah, and AWS. that's something that, know, our AWS is a major expense for us because all of this data is streamed real time. And we also store it all. So this is something in the future we could look to decentralize to enable full decentralization. And that's something that might make sense. But yeah, so when you're connected to GEODNET, you need to have the data goes from

<< Jason Kam (26:13) >>:

Okay, yeah

Yeah.

<< Mike Horton (26:39) >>:

from some station somewhere through the cloud. In the cloud, gets processed, quality checks get done. ⁓ It may get moved to different datums in a sense. there's some processing that goes on. And then it goes from the AWS down to the device. ⁓ And that time from when the data is first observed by the satellites to when it arrives to the drone, let's say,

<< Jason Kam (27:01) >>:

you

<< Mike Horton (27:08) >>:

could be less than one second, ideally. It can go up to like five, but ideally it is less than one second and our key customers really want to see it less than one second  all the time. So that's actually pretty demanding latency requirement. And so we end up running in caster servers, the technical term of art is called a caster server. So we run caster servers in every single AWS availability zone. So that second layer of processing is a decentralized network in and of itself.

to relay data from stations to users.

<< Jason Kam (27:44) >>:

Has there been any turn of your customer base historically? Nice. 10 million is a really magical number because if I do my math correctly, after a NIC's happening, your annualized inflation of a token at the current prices come out to just about 10 million dollars. So you will be at a point where  the ARR dollars that you generate to the burn roughly equals to the emission that you pay to the miners.

<< Mike Horton (27:47) >>:

None so far.

Mm-hmm.

<< Jason Kam (28:14) >>:

It's a complex equation, then beyond it, the beauty to this, in my opinion, is that the ARR is completely  orthogonal to how you inflate your tokens. It's completely in your hands on how you drive the ARR. And there are, I think, a lot of venues how you can grow that in terms of signing up new customers for existing use cases and increasing the take rate meaningfully in providing more services.

And then there are some crazy new cases about robotics, about drones, about everything else that we kind of haven't talked about. I guess  I want to go into that. the first question in that area is, am I right about my math, right, of what I just laid out?

<< Mike Horton (28:54) >>:

Yeah, you are. I don't even

really look at it as inflation exactly because the way I look at it is we started off with a billion tokens and that is kind of the max supply. yeah, I mean, you could look at it as inflation is, but there's not really that the token itself is deflationary. mean, we're already down below a billion. it's not like the total. I always look at things in terms of thinking about

the business in terms of FDV, in terms of all tokens being liquid, because that's where we're going to be at some end state. So in that respect, I think it's more appropriate to look at sort of are the rewards being distributed in a way to create an end state that we want. And the end state that we want, to a larger degree, we've achieved. We have the world's largest network. have coverage covering a huge percentage of the population of the world. And we've used...

less than half the mining tokens to get that initial network bill. And so, and we're on our third half at the end of June, as you said, so the amount of emissions is going to come way down in order to support that network. We're not necessarily, I mean, there a few places we want to add more coverage and we're working on that with kind of specific programs. We have a super hex program to get coverage. So we're working on, there's some locations in Japan we desperately need to fill.

We could definitely have been working on Brazil and we've gone really, it's really gone well in Brazil recently. So we now have a very nice network in Brazil. I'd hazard to say, I don't know if we're the best yet, but I think we're probably there.  We've done that in India now. We're definitely the biggest and best network in India. So there's not too many places left that we're really  gung-ho to immediately create a dense network on.

at filling in the US is a very big country, but, not a lot. So you look at that and you say, look, we've, we've created the end state that we want. And when we created that initial token allocation, you think of it kind of like a, how much do I need for these different tasks? Well, the way I look at it is we've, we've built what we need with less than half the tokens that we sort of planned. And of course there's a sustaining part of it for sure, but

A lot of these things, initial point of airdrops or whatever you want to incentive mechanisms is that bootstrapping function. The bootstrapping is done. That's the key thing to sort of highlight. The bootstrapping of GEODNET is done and we're working towards now a state where you're really looking at the network as this utility to build additional functionality. And that ties back into our robotics roadmap.

got this global precision layer, we've got this base thing that can give any device an absolute accuracy to a centimeter. Now let's help people build things on top of that.

<< Jason Kam (31:50) >>:

Yeah,  in particular to that revenue point then, I laid out kind of three, I know there are like different small areas I want to go into later, but like on a high level, new customers on existing use cases, existing customers taking up the take rate on new services you provide them, and then extremely new use cases that is not even in your numbers yet or in your kind of previous use case. Which of those three excite you the most? can you quantify it for me?

<< Mike Horton (32:12) >>:

All right.

Well, I'm a technology guy, so I'm always the most

excited about the new use cases doing the newest things.  I think from a blocking and tackling point of view, the existing market that's been there for a long time for high precision geospatial GEODNET's doing super well in that space. And I think it's, there's no end in sight to where we can go with that. We have some very exciting things coming.

 that we'll be showing at the ESRI user conference. So ESRI is a big kind of mapping organization that runs community of people in what GIS is called, Geospatial Information Services, a very professional organization around a lot of like public utilities, water and wastewater, energy,  you that you need high precision mapping services. So we've gotten better and better tied into that community. And I think there's a lot upside there.  And then

This Ag thing, mean, lot of good things coming on Ag We have a wonderful  relationship with USDA that has really been excited because 1.8 million farmers in the US don't do precision Ag or something like this. mean, the exact number,  we did a webinar recently with the USDA and a couple of our customers and we went through some of the numbers, but because the technology has been so expensive, it's really...

All this automation and farming has really benefited the big farmers. can think of people sort of like in Iowa, South Dakota, those states where you have really large farms, but smaller farm operations haven't been able to afford it. And GEODNET changes all that. And that has a benefit to those farmers, especially in bad years, because it reduces their input costs and labor costs. But they have to be able to get over that upfront cost of getting the equipment and services and paying for it. So I think there's a lot of potential in farming. I think there's a lot of potential in construction and survey.

And those are the kind of classical use cases that the technology does. So no end in sight to what we can do on the very traditional business case for RTK.

<< Jason Kam (34:29) >>:

Yeah,

I suppose Mike when you build a budget for this year and next year and as you look at the sort of ARR and order backlog trajectory, I mean, it certainly is true that you and the team sort of lay out different venues of growth. And I guess from the looks of it, when I listen to you, the path from 4 million to 10 million, it seems like it's a hodgepodge of different efforts. Like there's no single biggest driver that would take you there. It's kind of everything else working together. Is that correct? Okay.

<< Mike Horton (34:54) >>:

Yeah, that's right. At this

point in time, is no customer that's super dominant. That could change. One of these things could accelerate  faster than others. But at the moment, no, and I don't foresee it.  Just because, like I said, the web 2 things take time. And so you're kind of working through growth curves with each one of these customers and adoption.  So yeah, I would agree with that statement.

<< Jason Kam (35:23) >>:

Yeah, I mean, I've heard you publicly talked about  the sort of, you know, the GPS device in the car and I heard you talk about wing bit. heard you talk about the Android position app. Are those big drivers for you going from four to ten or are they outside of

<< Mike Horton (35:37) >>:

No, they're outside that. Yeah, those are all outside. Those are all things. I mean, those, we definitely have our share of things that were,  I don't know, we're trying out that never have been done before because going back to the first point that you made in this conversation about, how is this expense and the change in expense and the cost of this type of equipment.  So that's very beneficial. The reduction in cost of equipment is very

<< Jason Kam (35:40) >>:

Okay, can you quantify? Yeah.

<< Mike Horton (36:07) >>:

beneficial to the accessibility like getting the small farmer to be able to compete with the big farmer and have the same tools and techniques, right? That's kind of what that does in that market. But at the same time, having access to globally referenceable centimeter accurate position, which is what we provide, globally referenceable, absolutely accurate centimeter position, that's a crazy capability. That's like being able to, you know, as a human, if you want to put it in human terms, it's like being able to remember.

you know, where everything is in your life down to the center of your accuracy and be able to translate that to your wife or to your friend and then them having that same knowledge instantly. Like I can share that knowledge with anyone instantly by having this reference map. So it's a crazy kind of capability that you can do different things and we've toyed around a lot with different augmented reality ideas.

So we have a demo, showed it in that talk on the Solana mobile where we show a piece of augmented reality content. And the case of the demo was like a GEODNET coin. And we put that coin in front of the door of one of our offices in Asia. So that coin is sitting there in front of the door of the office. And you look at it with a phone that has GEODNET and the coin shows up in front of the door, just like it's supposed to. You turn GEODNET off and you look at that coin with

GPS and the coin is like where's the coin it's actually it was off maybe 20 meters to the left on the wrong heading as well so I think that what do you do with that well I don't think anyone really knows yet to be honest with you but we think it's pretty cool capability I mean it should be very useful whether it's for meeting your uber or for finding you know more specific locations I mean the or geocaching ⁓ but maybe it's a gaming thing or maybe

I don't really know, but it's something that you cannot do today that we enable and we're like, well, let's try to stimulate some activity around that. I think tying into the Solana Seeker phone is going to be a really kind cool way to test it because you're going to have a lot of people, people who are buying the Seeker, I think are a self-selecting group that are like really  edging into new stuff and want to try new things. So we'll see where that goes. But  yeah, there's a lot of that.

Those things are not really in our revenue, short-term revenue horizon. We don't know how big they'll be. We don't know when they'll...

<< Jason Kam (38:35) >>:

Are they more

experiments or are they more like actual real business efforts? It's like this is cool.

<< Mike Horton (38:40) >>:

Some are

experiments. I mean, I think augmented reality and reality is an experiment and we've kind of interviewed a lot of different people in that marketplace. And we do think the experiment is probably the right category because a lot of things haven't really taken off yet in that space. I think when you look at drones and robots, I would say that's less of an experiment and it's more about, you know, it's just still pretty early. Like humanoids aren't really, I mean, they're monetizing for a lot of really cool demos, but

Actually, like in our business,  talk about, like, I like the dogs because dogs are generating revenue. Like we have customers like Drone Deploy and their name says drones, but in reality they're doing sort of mapping services using a combination of photogrammetry  and, you know, back-end cloud processing to extract data out of imagery. They do a lot of, they have customers who collect data all the time on the robotic dogs. So the dogs are...

a little bit more actual revenue-generating machines today than humanoids, but all that, I think that's a category that I wouldn't categorize as experimental. It's more a question of timing as to when that explodes. I am totally, 100 % convinced that we are all gonna be interacting with a lot of robots on a daily basis within five years.

<< Jason Kam (40:01) >>:

Yeah,

I do want to touch on that because that feels more real than not because with billions of dollars Silicon Valley is pouring into it.  Let's say standing by your end or maybe next year you can't commit to it. But I'm going to say for you, let's say you're standing there and you're looking at 10 million ARR, all of which coming from new customers, existing customers, drones, your existing use cases. And then you look at robotics. How soon and how big would that be for you?

And are you already talking to the largest robotic companies in the world, either in China or US, about this effort?

<< Mike Horton (40:34) >>:

We're definitely talking to some really good names and I think, I mean, if this stuff goes consumer and that's my bet, so I'm focused a lot on consumer applications, I think it can be ginormous. i mean, Yuan runs around talking about it'll be the biggest product of all time. That probably is not an exaggeration that this market will be the biggest market of all time. ⁓ I just think that

It replaces physical work. mean, that's the thing. mean,  you know, depending on what stats you believe, like 70 % of the GDP is physical work. And, and, and also the ROI on these things is fantastic. There's a, mean, there's a real ROI on robotics. It's not, I don't know. There was, it was there an ROI on getting your iPhone. I guess there was, cause you could send messages faster and you could, you know, you entertain yourself more cheaply or something. I don't know, but.

At end of the day, the function was still the function you had before and there wasn't an ROI. There is an ROI with robotics. There's a real like monetary ROI when these things work where you end up, you know, they pay back quickly. And then when they work with each other and share data with each other, you get a network effect between them. And so I think that that will have another benefit to it that you also don't really see with.

sort of traditional smart devices. So yeah, I think it could be very, very, very big. And a lot of the stuff that has been sci-fi or has been sort of quote military technology or whatever, as it comes into consumer hands is it's gonna revolutionize the way we live.

<< Jason Kam (42:15) >>:

Do you feel like those potential customers feel that your technology is vetted and of matured enough to be part of that spec in day one in commercial production?

<< Mike Horton (42:26) >>:

Yeah, I think all these things are pushing on Obtanium, so they're always looking for new edges. Where we are working on our roadmap most strongly to intersect those devices in a more concrete way is to crack some of the indoor problem and some of the GPS denied problems. So if you think about where GEODNET works really well today, it's outdoors. And when you go indoors or when you go into an environment that's quasi indoors, like...

you know, super deep urban canyon downtown, the signals get all messed up and are too weak to really effectively use. So we're tackling that in two different directions. The first direction is incorporating vision and LiDAR sensing into our stack. We made an investment into a protocol called Rover, which we couldn't be more excited about, which is doing HD mapping. So they have a deep end device that combines a LiDAR and a camera with GEODNET and allows you to build.

HD mapping data for simulation training and actually just having a reference map for autonomous vehicles to use. And that technology stack we're working to bring into our positioning stack. So turning it around the other way, I have HD data available. I've either trained on it or I have a reference map. How do I do localization with the combination of RTK and LiDAR and camera so that I can get now high accuracy anywhere. So that's one major direction to solve sort of help.

build absolute positioning capability indoors. And the other, which is equally exciting, is that our ground stations starting in June will start to be updated for tracking low-earth orbit satellites. so low-earth orbit satellites are much closer to the Earth than the traditional positioning satellites. So traditional positioning satellites include the US GPS, the Russian GLONASS, European Galileo, and the Chinese BEDO satellites. And those are 12,000 miles from Earth. Low-earth orbit satellites are like

 you know, three to four hundred miles from Earth, so about, you know, thirty to forty times closer. The signal that then hits the ground is about a hundred times stronger. So those signals, so the same frequency band, same physics, but now a hundred times stronger means they can penetrate into indoors. And so you start to see that a lot of the problems that GNSS faces indoors may be solved with these low-earth orbit satellites.

We're really, really excited. excited. have a contract with one of the foremost companies launching these things that is really, you know, well set up, I think to yes. Well, this out of the launching the low Earth orbit. So you need specialized, you know, the people think, oh, Starlink Starlink is a communication satellites different than a positioning satellite positioning satellite works a little bit differently. Although there've been research efforts to look at how to reuse the.

<< Jason Kam (44:57) >>:

Satellites. Satellites or robotics? Okay. Okay. Okay.

<< Mike Horton (45:16) >>:

the communication satellites for positioning, but fundamentally normally when you do positioning, like the way GPS works, it's kind of a different type. But there are groups,  one here in the Bay Area that we have a contract with that is going along and they're already using GEODNET data today, is launching 300  low-earth orbit satellites. And those will, I think, really help bring the absolute global position layer ⁓ to be easily utilized with GEODNET in the indoor environment.

<< Jason Kam (45:46) >>:

How much of your current  mind space or R&D dollars are going towards these new robotic efforts versus  signing up more clients for existing ad efforts, for example, or uping the tick rate of existing customers?

<< Mike Horton (45:59) >>:

Well, in terms of R&D dollars, almost 100%, in terms of  what is going on in terms of  BD, it's much more of a mix. So  I think in BD, still have a, we're trying to hire some more BD people as we speak, so we have some.

<< Jason Kam (46:20) >>:

Sorry,

but for you personally, your mind is like mostly robotics. Okay, wow. And robotics now, sorry to cut you off, robotics now is how much of your ARR?

<< Mike Horton (46:23) >>:

My mind? Yeah, mostly robotics and drones. Yeah. I mean, think that

It's not maybe I would have to give you. mean, there's also some things that are hard to decide whether you would call it. What do you call  auto steering in cars and in precision act? Right. Like a das functionality. Is that robotics or is that autonomy or not? I mean, it's a kind of a blurry line. So I would say anywhere from if you want to take a very strict definition, 10 percent.  And if you took a looser definition, maybe 30 percent, something like that.

<< Jason Kam (47:01) >>:

Yeah,  it's early, I think that's the optionality that would step Function ARR and it just makes this narrative or this target, the network that you built extremely attractive  to potential investors, I feel like.

<< Mike Horton (47:03) >>:

That's early. So early.

And maybe

if you include drones, it's even higher. I I think of drones as aerial robotics. think most people also would agree with that characterization. So, and drones is quite a space for us.

<< Jason Kam (47:32) >>:

Yeah, from what you from all of what you know about today in terms of the progress of your potential customers all across different sectors and your target short-term target of 10 million ARR. Let's if we push forward maybe a couple more years,  what would be a ARR number that would make you happy? Do you think?

<< Mike Horton (47:51) >>:

I mean a billion. mean, yeah, no, I'm not joking. mean, I don't think I'll be particularly happy until the vast majority of robots in the world are built, you know, have, have, or we're working with the GEODNET platform. Yeah.

<< Jason Kam (47:53) >>:

That would be very nice.

And the path from 10 million to a billion, most of it will be robotics.

<< Mike Horton (48:14) >>:

Absolutely.

<< Jason Kam (48:16) >>:

Okay, so you're betting the firm on it effectively. This will. What is the point, if you can give us a timeline on what is the point where most people will go, holy shit, this is entering commercial production, you think?

<< Mike Horton (48:18) >>:

Yeah, absolutely.

⁓ that's so hard to say. You know, the easiest way to think about it is  probably that iPhone type of moment where,  or maybe it's actually the Android moment where people realize, hey, now the second guy's come in and this is obvious that everyone's doing that. For me personally, I started to really see it when I started to better understand what...

our customers were doing with drones and the level of automation and capability that drones have, I started to really see,  in many ways, the revolution around drones is much more  significant than maybe people realize. And so that gave me the clue. And then the second thing that gave me sort of the reality check that this was really very, very inevitable  is the progress of self-driving cars.

So self-driving cars have gone through the valley of death and come out onto the other side now. And there was a period, I think, when many in the valley here were starting to doubt. I mean I started working on self-driving car stuff, God, back in like 2015, 20, I mean, that was a while ago. That's a decade ago, right? And then there was a hype cycle around that, of, you know, like all these hype cycles where everyone thought it was gonna be 2021, you know, everyone would have a self-driving car. And it took longer, it just took longer.

You see us, on the other side now. mean, in San Francisco and in LA, they are all over the place. Waymo is doing a remarkable job of, I think, building trust in their product. So that's a hard, hard problem. I don't know that people understand how hard a problem that is to get to where that is and all the compute and learning and regulatory things that go together to pull that off.

It's a big thing, technology has been able to overcome that. And I think it's going to start getting much easier. So for me, are little, but those are early signs.  But I think we're right there on the cusp of this stuff, just really mass market.

<< Jason Kam (50:40) >>:

Mmm.

So maybe this year or next year, your existing contracts will be really buying into this from all the things that you've talked about gets you to 10, 20, 30, 40. And then when robotic hits, this is when it goes S curve.

<< Mike Horton (50:53) >>:

I think

definitely the robotics is the thing that will take it exponentially, particularly consumer robotics. I think that's the most powerful thing.

<< Jason Kam (51:03) >>:

This ties to another question. know we're running low on time. I plan to keep a very strict schedule for you. I've heard, because I'm Chinese, I've heard a lot about the advance of robotics in China. know, Unitree, Da Jiang all of these guys are well-funded and quite sharp. It would be shocking to me that there's no competitor like yours coming from this side of the world. Have you come across any? Do you feel like you are competing with them? Do you deal with any Chinese customers? Like I said, yeah, yeah.

<< Mike Horton (51:31) >>:

We,

yeah, so certainly we have a strong knowledge of the China market and definitely today and small robots,  they're all made in China, right? And Da Jiang so for folks on the call, Da Jiang is DJI and DJI is a superpower company. I don't think people fully appreciate or understand in the United States.

<< Jason Kam (51:33) >>:

Haha.

<< Mike Horton (52:01) >>:

just how incredibly competent and powerful and technologically complete Da Jiang is.  They make their own system on a chip, which is what we see as created the competitive advantage for their products. That's why we are pushing so aggressively to develop specific system on a chip that can reduce the cost power weight of these robotic solutions. Localization is a

big computational driver, all the math that's done to do the positioning is really intense. So they've done that. They made that investment there, know, many for drone. mean, if you look at that thing, sub 250 gram completely automatable by API  has just an amazing set of capabilities. So yeah, I think in terms of the  the current market situation,

Yeah, the US has got a hole to fill. we think that deep end and the deep end movement can help fill that hole quickly on its own. If that doesn't happen, that China will continue to sustain a remarkable advantage and be in a position of  very strong power. So it is imperative that deep end succeeds. It's imperative that these things start to move faster.

The US doesn't do centralized planning, right? We know that. Everything, you go to the DMV, know centralized solutions don't work in the US. I don't know they work very well in Western Europe either. What works really well is great competitive market dynamics and sort of decentralized solutions, if you will. And so I think that super important, super, super important, but yeah, today in the robotics field, China's way ahead, no doubt.

<< Jason Kam (53:49) >>:

So another way to read this, very helpful by the way, another way to read that is that you're not in that market. They have their own RTK geonet like solution that they're using.

<< Mike Horton (53:58) >>:

Well, we are in that market in the sense that those types of equipment can connect to our network when they're exported to Europe and into the US. So DJI today is a very popular piece of equipment in the US. of lots of the mean we publish videos on how to connect GEODNET to DJI drones and those those videos are really popular because lots of people use those drones.

Now there is a potential ban of those starting at the end of the year which creates another market opportunity. That ban is on the books and law that FCC will no longer approve new DJI drones starting at the end of the year and that will have a major impact on the drone market in the US. When you look at Europe, mean Europe, the Chinese cars are becoming extremely popular there. So those cars all need to have

ADAS functionality, that means they need to have RTK, that means they need to have potentially been trained with HDMAP data products. So those are definitely opportunities that are lawful, that are super easy to support because they're growing.

<< Jason Kam (55:03) >>:

that you don't plan to operate in mainland.

<< Mike Horton (55:05) >>:

We can't operate in mainland. They have geospatial restrictions. So you can't set up a geonet station in mainland China. That's why you don't see any on the map

<< Jason Kam (55:13) >>:

Yeah, makes sense. I know we're really low on time. Just last question for me. I mean, you're talking about potentially competing against these Chinese players. These guys have potentially billions of dollars of R&D budget. You have a budget of, I don't know, for multiple fund raises, $10 million, maybe more. I don't know. Do you plan to do more fund raises? And what are the terms like historically? Yeah, yeah,

<< Mike Horton (55:32) >>:

I don't know why we're competing against the Chinese players. We're not competing against them. that I

think the US is competing against them. We would like to stimulate an ecosystem so that drones and robots can be built in the United States competitively. If you try to build a DJI Mini 4 drone in the United States today, there is no way to do that. It cannot be done. There's not an ecosystem of parts.

There is not a system on a chip that can do that. there is none and that's just reality. There's no way to build that functionality in the US today. We want to fix that problem and we want to be part of that solution, but we're not going to go out and build DJI Mini 4s. That's not our goal. It's our goal is to leverage our network to  build a system on a chip architecture that can enable people to make those here.

so that the US has a sovereign capability to build small drones and millions of units like DJI does. That we're very, very interested in. But DJI will remain a friend and a customer of ours because they sell drones all over the world. They need RTK and we have an RTK network. And I think in terms of building out geospatial infrastructure, the Asian manufacturers, they don't want to be building out that infrastructure. That just sort of from a commercial...

I'm a commercial company making farming equipment in mainland China. I don't want to be building an RTK network overseas. That just amplifies all the fear and paranoia people have around that. They want to be able to just connect. And we have that connectivity so that type of equipment can connect and be used. And that helps them sell equipment. It helps the farmer  be able to... The guy in Turkey who's got a farm and wants to use an FJ...

know, tractor steering kit that lets them get connected and run. So that's a win-win. It's not competitive at all.

<< Jason Kam (57:25) >>:

Yeah, and just last question for our audience. What's your monthly burn currently and do you fund it through internal VC raised capital and ARR?

<< Mike Horton (57:32) >>:

Yeah, currently, exactly. So about 200 to 300 K a month.

<< Jason Kam (57:38) >>:

Got it. And you don't have to sell tokens basically to find yourself because of the ARR plus of it.

<< Mike Horton (57:41) >>:

Not, no,

I mean, you know, I think we're definitely want to ramp that revenue so that 20 % can, you know, cover all the foundation needs. The foundation is pretty, pretty thin, right? But I go back also to that AWS comment, that is a big expense for us. So it's something that, that, you know, there are a couple of options for something like that. One is ⁓ to get the 20 % to be big enough to cover that. Another would be to, to, you know, leverage the decentralization to take care of the...

further decentralizing the network. ⁓ But we do have data storage and data processing costs are a big expense for the GEODNET Foundation. Running a 17,000 station network that delivers real-time data  all over the globe, it's pretty intense and the latencies have to be low and we use a lot of AWS resources to run the network.

<< Jason Kam (58:34) >>:

Very helpful call. Thank you so much, Mike, for the time today.

<< Mike Horton (58:38) >>:

You bet,

Jason. Glad to do it.

<< Jason Kam (58:41) >>:

Yeah.