Morning, everyone. Our next presenting company is Duos Technologies Group, trading under the ticker D-U-O-T on NASDAQ. Through its wholly owned subsidiaries, Duos Technologies, Inc., Duos Edge AI, Inc., and Duos Energy Corporation, the company designs, develops, deploys, and operates intelligent technology solutions for machine vision and AI applications, including real-time analysis of fast-moving vehicles, edge data centers, and power consulting. Presenting today are CFO Adrian Goldfarb and Chief Technology Officer Jeff Necciai.
Good morning, everyone. My name is Adrian Goldfarb, as you mentioned, Chief Financial Officer. Actually, this is my second tour of duty as the Chief Financial Officer. I thought I had retired after taking the company up to Nasdaq in 2020. I had planned to retire in 2022. The company had other ideas, so I've actually come back for a second tour of duty. It's interesting. The last time I was at Gateway was about five years ago. That was pre-Nasdaq. We went up to Nasdaq, and now I'm back again. You're not gonna hear much from me this morning. I just wanna cover a couple of things. Our story has changed quite considerably.
There are a number of people I bumped into this morning that are familiar with us, and we were always kind of known as the railroad tech company. So the story so far is that, you know, from the standpoint of the big picture, we've been involved in the technology side for many, many years. The company is largely a software and technology company that has been applied into a very specific railroad application. Since I came back as CFO in May of this year, I've been working with the management team. About 90% of the people at Duos are new in the last few years, including the new tech team and Jeff, who I'm gonna introduce in a moment.
And we are bringing the company forward in a number of different areas, including the two new divisions that were described previously. So we have the Duos Tech business. Jeff leads that business. He's gonna talk in some detail about that. We're also moving, we announced this morning, into the edge data center business, and we'll be talking a little bit less about the power business, but these are all burgeoning industries, and we feel that the company has a tremendous growth potential. So, I'm going to stop here, and turn the presentation over to Jeff Netchai, who is not only the company's Chief Technology Officer, but actually has been tasked by management to now lead the Duos Technologies business. So thank you, Jeff.
Thank you. Thank you, Adrian. I appreciate it. Good morning, everyone. Jeff Netchai, Chief Technology Officer at Duos. You know, I met the folks at Duos about four years ago, sort of for the first time. It was kind of after a long career in the intermodal gate automation industry. And, you know, I was struck immediately by the sheer potential of this company in Jacksonville. You know, I'm not a railcar mechanic. I'm not a railroader by trade, but even I knew that, you know, railcar inspection really hasn't changed that much in the last hundred years or so. But I have this company in Jacksonville, suddenly I see what they're doing.
I see what they're doing with machine vision and precision optics to be able to let railcar inspectors and operators inspect these railcars at track speed. I mean, this is technology that's usually confined, you know, typically confined to either an assembly floor kind of inspecting what I like to say, inspecting labels on Gerber baby food jars. They're taking the same technology, and they perfected it so that they can, you know, perform inspections on rapidly moving, really larger-than-life railcars, and it was just absolutely amazing to me, so to be honest with you, I was really struck by the potential of this technology, and whether our new CEO, Chuck, whether he hired me or I just refused to leave, I can't really remember.
But I'll tell you what, I'm really glad I stayed, because what we're seeing today, as Adrian hinted to, is we're finally at the culmination of all the work that we put in over the past four years. So in those past four years, we've really transformed the technical solution. We redefined our business model, and we kind of reimagined the way we work, in general. Through heavy investment and really heads-down discipline, we've added features and commercial viability to the already pretty cool, technologically cool, railcar inspection portal. We've basically made it more sellable, made it more stable, and certainly scaled it so that it can actually be used by lots more railroads and other associative types of companies. You know, and also during that time, we've looked at this technology.
We've discovered some things about ourselves. We do a lot more than just railcar technology, right? Just looking at railcars. We take this same technology, and specifically what I'm talking about is the acquisition of high-resolution images at track speed. Okay, what track speed is anywhere from 60, 70 miles an hour, all the way up to 125 miles an hour. It's really a feat in and of itself. It requires a lot of illumination. We have purpose-built illumination. I don't wanna get into too much detail here. I promised Adrian that I wouldn't. It's hard to talk about this stuff. It's hard not to get excited about the technology sometimes. But we know that we can take that same technology and apply it to other industries. For example, like truck inspection portals, right?
If I can do it on the railroad, I can do it for a truck that's moving 15 miles an hour through a portal over the road. That's something that I have experience with. I've spent 20 years with my last company, doing just that. As I said before, we automated gate systems, and that's part of what we did, and the advancements that we made in the artificial intelligence that we apply to those images are just nothing short of astounding. We've completely revamped the way we do that, so it's made it, like I said, a lot more sellable to a lot more people in much less time. You know, Adrian mentioned something. It was the technical solutions. I think I've spoken to some of you here.
By the way, this is new to me. I've never heard of something called, you know, financial speed dating. It's pretty incredible, pretty interesting. But some of the conversations we had earlier, I could tell that, you know, folks are familiar with our business model in the Railcar Inspection Portal. We're the railroad company. We're the... You know, I've heard a few things said. I really don't like to look at us as a railroad company because the technology just happens to be one application for the railroad. This technology that we use, that we develop, that we perfect, can be and will be applied to a lot more industries than just railroad. So, what we like to say is it's intelligent technologies, and even more specifically, it's really, intelligence in moving things.
Intelligence in motion is what we like to say. You know, in that time, and I will change slides, I promise. I'm gonna move forward. But we also worked hard to align our business strategy with our strengths and core competencies, and that ensures that our efforts, we want to ensure that we're focused on efforts where we can truly excel. To that end, we developed two new lines of business: edge data centers and infrastructure and power, and we'll be talking about those in just a moment or two. Before we get into those lines of business, though, I really want to talk about our cornerstone product. We talked about it a little bit during the speed dating we had this morning.
But just some of you may have heard of the Railcar Inspection Portal that Duos. I know it's been about four years, I guess, since we talked to the group. But we still make the Railcar Inspection Portal, and just keep in mind, as I'm explaining this to you, the technology is applicable to so many other things. But when we speak of railroads, most of us generally don't imagine them as being heavy adopters of high tech. But that's really not the case. According to the Association of American Railroads, each year, freight rail companies invest an average of $23 billion on technology to bolster and modernize their privately owned nationwide networks. Now, they do this to increase safety and increase throughput and to make the railroad more efficient.
That's where our railcar inspection portal actually comes into play. What these railroads do is they collect and analyze data from tracks, wheels, and locomotives to help keep freight and rail to help move goods in the safest way possible and also to optimize their network and to adhere to regulations. Duos Tech's railcar inspection portal is an AI-powered visualization and inspection system that helps railroads improve and optimize their operations as well. Utilizing a combination of optical and other sensor-based technologies, the RIP scans railcars at track speed, and that track speed, again, is up to 125 miles an hour. It acquires high-resolution images, resulting in a 360-degree all-around view of the top, the sides, and even the undercarriage of those moving railcars.
It utilizes advanced artificial intelligence machine learning models to analyze each railcar, looking for anomalies and defects. It reports those defects to operators and inspectors within 60 seconds of scanning a railcar, and it provides a comprehensive, easy-to-use, remote visualization and inspection capability to railcar inspectors. This is a lot of value that our railroad companies get from the railcar inspection portal. The railcar inspection portal and AI remote visual inspection capabilities also help in cross-border operations, so not only mechanical failures, not only defects all the time, but sometimes we're looking for illegal riders. Every day, our AI models detect several illegal train riders heading north out of Zacatecas and another site in Mexico.
And our systems even produce reports specifically designed for helping Customs and Border Protection agents easily identify offenders on those trains. So we're talking about AI. Obviously, a lot of folks are talking about AI today. Ours is, you know, it's, there's two types of AI. There's AI performed on structured data and unstructured data. We're dealing with images, which are classified as unstructured data, and what we're looking for is, anomalies and defects, like I've said before. We've developed over fifty models that look for these defects, and over 80% of these are focused on identifying those conditions, which are most likely to contribute to a derailment. Consequently, these are the same inspection points that are mandated by regulatory agencies such as the FRA and Transport Canada.
Such inspection points covers wheels, axles, brakes, springs, and various components located on the underside of the railcar. Now, you have to understand, most of us that work at Duos, we're not railcar mechanics, and we're not railroaders, so we're a technology company at the end of the day. We know engineering, and we know software and AI, so how do we effectively create artificial intelligence models? What we've done, one of the big changes that we made in the last four years, is we've hired on certified mechanical railcar inspectors. We use these folks, and let me tell you, they really know their stuff. I mean, they're telling me about all these railcar mechanics and their, you know, the mechanical underpinnings of the railcar, and I'm constantly amazed and blown away by these guys.
But we use them as subject matter experts to help build our artificial intelligence models. They work with our labeling team, they work with our data science team, and they make sure that, not only do we have a model that's effective and up, by the way, 95% accurate, they just do a fantastic job in being able to deliver on those promises. Let's see. So... Sorry, I got to rely on my notes. So this, at the end of the day, artificial intelligence models, it's all about, for what we use in the Railcar Inspection Portal, it's all about safety. And that's the one of the primary reasons that our railroad clients use our system, is to increase safety, and obviously, they're interested in increasing performance as well.
The railcar inspection subscription service, which is a new service. I guess it's about going on a year old right now. What it does is it leverages the information that's coming from those railcar inspection portals that we have on the rail networks. It leverages that information and makes that safety data available to other companies. It could be other railroads, it could be railroad owners, railroad lessors, shippers, but it makes that same data available. That's without the high cost of hardware investment, which has already been done. So it is truly a software as a service. So, we feel that that's gonna help a lot. Obviously, it's not as expensive as putting in a lot of hardware for a railcar inspection portal, but it gives you all the advantage.
If you're a railcar owner, you could see the condition of the railcar. You can actually make a plan to maintain it. Now, there are indications that the railroads are standardizing or will eventually standardize the safety data across the industry. Duos is kind of well-positioned to be there when this standardization occurs. We would like to actually lead some of that standardization, and we're well positioned to do that. I believe last year, we scanned over 8 million railcars, just last year, and that number for this year is expected to be around 11 million. No one else in the industry has as many scans, as many railcar scans, as many pictures of railcars as we have, and certainly not as comprehensive.
So we're talking to the right folks in the industry, not only railroads, but some of the industry players. We definitely want to be at the forefront of any of that industry standardization, and we think that's gonna happen. So I mentioned in the beginning, and I know that Adrian mentioned the two new business units that we have. One is Duos Edge AI, and this is about edge data centers. One thing I want to make clear is that it's not new to us, right? We've been building and deploying and operating and maintaining edge data centers for a long time. In fact, we were early practitioners of edge computing and designing and deploying our railcar inspection portal technology in the mid-2000s.
Since the RIP relies on immediate storage and processing of enormous amounts of high-resolution image data in real time, backhauling that information over a wide area network or to the cloud is simply not practical. So we designed and developed localized track-side edge data centers containing all of the compute power required. To this day, our railcar inspection portals process data at the edge, utilizing our innovative edge data centers that we've actually perfected over time. So, we get into this edge data center business, which I've mentioned a couple of times in the speed dating stuff, this morning. It's kinda like a hotel for servers, and I know you're all familiar with that model. But we do have some expertise in that area, and there is some crossover to the rest of our business.
For example, our business runs a 24/7 network operations service center, and all of these edge data centers, they must be monitored. The environmental controls must be monitored, and that fits right into our 24/7 service model in the NOC center that we have. There's a gentleman by the name of Doug Recker that we brought on to head up that business unit. Doug has a fantastic record in building edge data centers and data centers in general. I think he built and sold two companies so far, and we're really looking forward to Doug to be able to do fantastic work with the edge data center business, and we're looking forward to that. The edge data centers, what are the key drivers for edge data centers?
There's first and foremost the proliferation of Internet of Things and artificial intelligence across industries, and you can see it on the news almost every day now. Folks are using artificial intelligence for lots of different things. It's being used in lots of different industries. So why does that have anything to do with edge data centers? What companies are realizing more and more is that it's too expensive, and there's too much latency to take that acquired data and analyze it somewhere in the cloud. It's just expensive from a data standpoint to ship all that data to the cloud. So a lot of that storage, processing, and acquisition happens at the edge, and that's what folks are pushing for. That's why edge data centers are so prevalent. You hear it bundled with things like, "Well, we're doing AI.
We have a lot of GPUs. They're all in edge data centers. Retailers and businesses of almost all kinds are pushing it to the edge. The other thing that we're using edge data centers for. They're key for delivering broadband and internet infrastructure to unserved and underserved communities. Duos recently sealed a deal to deliver EDCs to a Texas regional school district for the purpose of bringing high-speed internet to some of those rural areas. Don't wanna completely run out of time. I wanna leave some time for Q&A. We were gonna talk about... Is there anything you want to say about the power?
Just keep rolling.
The second division that we have is-- Oh, I'm sorry. It's the power, Duos Energy. I'm sorry.
So the Duos Energy Corporation, again, much in line with the other two divisions that we were talking about, there's some key tie-ins with the rest of the Duos, Duos Tech company. Number one is about 33% of our employees right now actually come from the power business, so they're dealing with power plants. I think they've installed power plants pretty much all over the world. So 33% of our staff includes those people, so there is a lot of experience in that area. And we were talking about AI, and we were talking about GPUs. These are extremely power-hungry kinds of things, and so, allowing, you know, getting off-grid power to some of those EDCs is a perfect alignment to those two divisions as well.
I think I'm gonna stop there for power, and I'm gonna pull Adrian back up. I'm sure you have questions that have something to do with financials, and you don't want me answering that.
Thanks, Jeff.
You're welcome.
I'll pass this mic around for anyone who has questions.
Just before you do that, let me just switch forward a couple of things. So, there's a presentation that's gonna be released. Fe, is it today? Yeah, this presentation is gonna be today. Obviously, we've only really scratched the surface here, so please feel free to reach out to Fe, organize some time with us. This is the team now that's been assembled. Obviously, CEO Chuck Ferry, myself, who you know, Jeff, who's gonna be running the Duos Tech business. Doug Recker on Duos Edge AI. So, that's the team. So a very, very experienced management team. Second of all, there is some numbers in there. Our numbers have, quite frankly, been nothing to write home about for a few years.
We're unfortunately, like a lot of small micro caps, struggling to get to break even using cash. That is on the cusp now of changing. So I'm not gonna go through these numbers in detail. What I'm gonna say is that we do have a very, very clean cap table. There's about fully diluted share count is a little over 14 million shares, but about half of that is owned through two series of preferred stock, which are just other ways of holding the common equity. So once again, we can talk more about that. Either come visit us or set up a... So with that, thank you, and if there's any questions, we're happy to take them.
So focusing on the EDC business for a second, can you give us, maybe a bit more granular look into the unit economics of that business? And then also, secondarily, the sales cycle. Obviously, with the legacy rail business that you've been operating, it's a very specific customer set, even though the technology's broadly applicable. How do you figure out which customers you wanna go after outside of maybe just the education market?
Okay. So, the Duos Tech business obviously is complicated. It's a lot of technology. The Duos Edge AI business, quite a relief from my standpoint, is actually a very simple business. It's broadly like the real estate business. So we will basically manufacture these Duos edge data centers that are a little bit bigger than normally goes on a railcar inspection portal. We deploy them into the ground. That is about $1 million to get that done. The unit economics are those will be expected to produce around $400,000 a year in revenue. The contracts on those are very long, typically five-10 years, and the margin on that is about 71%. So as you asked about also the customers, the customer set on the Edge AI or...?
Mm.
Yeah. So the customer set is typically the tier three and tier four markets. So, there's expectation that there will be tens of thousands of these edge data centers that we deploy over the next few years. Our goal is a little bit more modest than that. It's about 150, and as we mentioned, it's in the tier three and tier four markets, so that's typically things like, initially, schools, hospitals, out, particularly out in rural regions. The announcement we made this morning talked about the fact that that was in a school district. I think it's called Region 16, which is based in the Amarillo area. However, there are other larger companies which, for example, Exxon, who may take portions of those edge data centers.
So the business model is made up from that, but it's a business model that goes cash flow positive very, very quickly. Other questions? Great, so we'll be around the rest of the afternoon. If you do wanna schedule some free time with us, please do, but thank you very much for attending.