$SEE.L：欧洲刚刚让DMS双寡头成为必选项。为什么它只交易在自由现金流的11倍？| Hugo Navarro

Today we've got a really interesting one. We've got Hugo Navarro from Undercovered and Undervalued on. This is his second time pitching; the first time, the company he pitched got acquired. I don't know if it was 2 days, 2 weeks, or 2 months later, but it was fast.

This is a different one: Seeing Machines. The ticker is SEE, and it trades in London. Full disclosure: This is not investing advice, and this is a foreign stock. There are disclaimers all up and down, including at the end of this podcast. Hugo is super passionate about this idea and has written it up multiple times. You can find a link to the write-ups in the show notes.

1. Can the regulation slip or get watered down?

He pinged me multiple times saying, “Let's do an episode.” Finally, I thought, “You are so passionate about this. I've got to have you on to discuss it.” It's a complex story hitting a regulatory inflection, but I think it's really interesting. Hugo thinks there's huge upside here. So I'm going to let him explain all of that to you.

Pretty well. I hope this one goes like the first time. We covered NCR Atleos, and it was bought in about 2 months, so I hope it happens the same way.

That is the dream. The last time I saw you, you were coming off a whirlwind. You had just been in Vegas, and you flew—

Overnight to New York. I'm looking good right now.

Before you dive into what they are, a disclaimer: Nothing on this podcast is investing advice. That's always true, but it may be particularly true today because we're going to be talking about an international security, which obviously has increased risk, taxes, and all that sort of stuff. We're not financial advisers. This is not investment advice. There is a full disclaimer at the end of this podcast and in the show notes.

The company we want to talk about today is Seeing Machines. The ticker is SEE, and it trades in London. I will just say, before you dive into what they are, that I think you're really passionate about this idea. I saw you back in April or May, and you said, “I want to talk about Seeing Machines.” Then you emailed me in June with its PR announcement, saying, “This is starting to work. Acceleration. I want to talk about it.” In July, you said, “Come on, man. Let's do it.” I thought, “He is so excited about this. I've got to have him on to talk about it.” So I'll toss it over to you. What is Seeing Machines, and why are you so excited about it?

2. What Seeing Machines does, and why DMS is harder than it looks

First, a bit of context on how I got into Seeing Machines. I started looking at this company randomly. It just appeared—I don't think it was even one of my screeners—and it came across my desk. I started doing some research into it, and I thought the ramp-up was being completely mispriced by the market. That has worked well; it's been a 50% return since I entered the company.

I think we have a very good asymmetric return for the second leg of the thesis. The quick pitch on Seeing Machines is that it is the leader of a 2-player duopoly in what's called DMS technology. Basically, DMS technology is software that checks your face while you are driving, or while you are doing another type of activity, to ensure that you are looking at the road and avoid any type of accident.

Despite this seeming easy to replicate, the technology is very hard to develop. There are only 2 players—

3. The math: fixed opex, Europe now, Japan and the US later

You cut out for 1 second. Your sound cut out. If you just want to say what you said, start about 10 seconds ago.

These 2 players, which are the only ones that have the best technology, have both taken 20 years and hundreds of millions of dollars to develop it. These companies have been listed for decades, and they've been losing money for decades. Seeing Machines has spent, I think, half a billion dollars on research and development over the last 2 decades, so it's very expensive to develop this technology.

What happens now is that, in Europe, it is mandatory to have this technology in every car. This means that these 2 companies, and especially Seeing Machines, will start to produce free cash flow. My estimation is that, for fiscal year 2027—their fiscal year ends in June, so when I refer to fiscal year 2027, I mean June 2026 to June 2027—they will make around $20 million to $40 million in free cash flow. We'll explain the variation later.

The market cap is $330 million, so at the midpoint it trades at around 11 times free cash flow. I believe that can grow at a high-double-digit rate. There is a very high operating-expense base, around $55 million per year, that is largely fixed. Extra revenue doesn't mean extra costs.

This means that, with the European regulation now mandated, they have all the programs in place and should make free cash flow. When the next leg comes—Japan and America, probably around 2030—every extra dollar of revenue will go straight to free cash flow. Instead of $70 of extra revenue from Europe producing only $20 of free cash flow, $70 of extra revenue from Japan and America would produce $70 of free cash flow. That's massive.

In between, there is subscription optionality from the fleet segment. That's basically this technology applied to fleets, like truck fleets, Amazon, Caterpillar, and stuff like that. It also adds optionality because it's very recurring. We'll dig into that later.

Basically, this is a thesis where there's huge operational leverage, a duopoly with what I believe is a very good competitive advantage, and hard-to-replicate technology. There are some risks and some reasons why this is cheap, but we will dig into those during the podcast.

At a high level, you've got this oligopoly—basically a duopoly—trading at, let's call it, 10 times your estimate of forward free cash flow. Not only that, but the free cash flow kind of explodes because it's all operating leverage. Europe alone has regulatory requirements kicking in, and Japan—I think it kicks in in Japan in 2029. Is that right? Maybe—

2029 or 2030.

4. The seatbelt manufacturer analogy

Okay, perfect. I read—and I shouldn't let people know this—that Hugo has done tons of write-ups on this company on his Substack. I will include a link to one of them, whichever one Hugo thinks we should direct you to.

You had one line that I loved. I might be slightly misquoting it, but you had a line like, “It sounds crazy. You're buying something before 16 million cars are mandated to get it.” You might say I'm crazy. But what if I told you to buy a seat-belt manufacturer in the 1960s or 1970s? You'd say, “A seat-belt manufacturer? No cars have seat belts in them.” Then, a year later, 80% of the cars are rolling off the line with seat belts.

I thought that was an awesome framing of it because you are right: I look at it and say, “All cars will have this.” But sometimes a regulatory requirement happens, and suddenly they've all got it. I'll pause there because I thought that analogy was so great, and I want to let you talk about it if you want to.

That's why I think this is so interesting. First of all, this technology, outside of being a great thesis in my opinion, saves a lot of lives. One of the main costs for insurance companies is the insurance angle, which I think is very interesting in the long term. This reduces the risk of a catastrophe by around 90% when you are on the road, especially for truck drivers, and it reduces the insurance cost of cars.

For example, in the U.S., Tesla's self-driving cars—I think Lemonade, one of these fintech or insurance-tech companies, offered a huge reduction in premiums to drivers who had that self-driving feature. I think this will happen in the future because it will become clear that this has huge cost savings for insurers. It will also save a lot of lives.

In Japan, for example, they are pushing this regulation for 2029 or 2030 mainly because there's been a huge increase in accidents involving people looking at their phones or being distracted on the road. That's increasing the number of deaths we're seeing on the road, and there are lots of associations pushing for the regulation.

5. My pushback: what stops a new entrant or an in-house build?

Let me find my first pushback. You say this is regulatory-driven in Europe, and all of them are coming online—literally as we speak. I worry that you've got this company that's been in a duopoly with another company that's kind of winning it.

When it’s a niche market, they dominate, but I worry that when you expand it to 16 million vehicles and it’s regulatory-driven, all of a sudden, it’s not like it required these guys, who have spent cumulatively $200 million in capex over the past 10 years. That’s a lot of capex, but I would almost guarantee the capex you spent 5 to 10 years ago is wasted.

Could I come up with a competing product for $40 million or $50 million using state-of-the-art technology that comes in when, all of a sudden, 16 million vehicles need it? Or do one of the big manufacturers look at this and say, “I could outsource this when it was a niche thing on high-end or custom vehicles, but now that it’s required, I’m just going to build it in-house”?

6. Naturalistic data, Mitsubishi Electric, and the accuracy gap

You mentioned Amazon as a customer. That’s not on the car side, right? That’s more on the fleet side. But I look at that and say, you get all the data happening inside a car, lots of machine learning and AI. Why would Amazon outsource this to someone instead of just building the product in-house? So I threw a lot of competitive responses at you. I’d love to hear how you think about that framing.

That’s a great question, and we need to frame it and understand why the thesis is compelling. First of all, Seeing Machines and Smart Eye are already in those 16 million vehicles. We are already in those contracts, and these vehicles usually last 3 to 5 years, so on this first leg there’s a low risk of replacement.

But, as you said, there’s a real risk of another player trying to take share over the long term. I am expecting a third or fourth player to appear, because this usually happens in the OEM sector. Some reasons why this is hard: many players have already tried to get this solution right.

For example, Mitsubishi Electric is a client of Seeing Machines. Why? Because they couldn’t develop their own solution. For context, I’ve researched a lot with industry insiders and had many conversations with management regarding the technical side of the technology. They got great results in, we could say, lab tests, but when they took the solution into a naturalistic environment, it worked really badly.

They trained all of this with synthetic data. Basically, you generate data and train it on that, but it doesn’t work well in a real-life scenario. Seeing Machines initially developed this solution for the mining and trucking sector; it only later became relevant for the automotive sector as a whole.

They have billions of hours of footage of truck drivers and mining employees using this technology. That’s extremely important: they have tons of naturalistic data, and that’s why they have developed the solution. If you compare it with Smart Eye, Seeing Machines has much better accuracy. That’s key in my opinion.

Well, I certainly hear you, but if I was just thinking off the top of my head, you said they’ve got great data inside. Every Uber I hop into has someone recording inside. That’s more for safety, but there are 5 companies providing that that do have the data.

I would think about Tesla and a lot of cars coming in. That example—Tesla has its own DMS solution.

There was a huge recent problem with Tesla because some people put a video out on the internet showing that they were fooling Tesla’s self-driving system with a plastic head. That’s the current level of Tesla’s accuracy regarding DMS.

7. The elephant in the room: a $55m convertible due in October

Okay. Let me go to a slightly different risk. We can come back to the business, but I think the elephant in the room, when I looked at this, is that they have a $55 million convertible that is due in October. You and I are talking on August 25, 2026.

They said—I believe it was in their earnings deck, which came out on August 11—“We’re in the late stages of renegotiating this thing.” The convertible loan is with a customer, and they’re fully supportive. I hear all that, but I’ve done markets a long time, and for a company to let a $54 million convertible loan get within 2 months of expiration is lunacy.

I’m going to say there’s no company that would do that unless they absolutely could not roll it. They’re putting on a brave face. You could say there was this huge inflection that they’re in the middle of that really juices these results for them, so they can do it.

But I look at that balance sheet and say, “I hear the explosive nature. I hear all this.” I look at the balance sheet and say, “This is the balance sheet of a company that’s distressed, or there’s something I’m missing.” I’d love to talk about the convertible.

I hear you on that. I think a lot of investors are waiting for the convertible to get refinanced before investing in this. First, a little bit of context: they started the refinancing process around April or May because the timing was really bad for this convertible.

Basically, they needed to report KPIs in order to show potential lenders that this ESL thing is really happening. Even this surprise was not reflected in the share price: they were going to receive these royalties, ASP was going to hold up, and Q3 and Q4 have outperformed. I’ve been talking with management since April on that note.

The signing has slipped multiple times, mainly because due diligence has taken longer than expected. But, as they’ve said publicly—and I published an interview recently—they are in an exclusive period, meaning they are now with a final lender and are in the final due diligence process, including final documents and all that stuff.

They are also in contact with Magna in case there’s a bit of a delay or something like that, so they can work it out and get an extension. I think this will get solved before October. It’s a risk, to be honest, but I don’t think there’s a high probability of this being resolved with dilution.

8. Footnote 21 and the accelerated royalty payment

No, I definitely hear you. I just look at that and say, “Man.” It’s strange timing. It’s weird. I was flipping through their semiannual statement for December 31, and at the end of it, it says “debt financing facility.” Then there’s another item that says, “We entered an amendment with a major customer project that accelerated a royalty payment, but in exchange, we need to basically give the customer money back.”

I look at all this and say, “Look, it’s a—”

So—

This is footnote 21 of the semiannual report. I try not to, like—

The royalty acceleration. You mean that one? They received $15 million.

Yeah, $14.1 million. They’re not giving money back to the customer.

Let me explain that, because context is really important there. Basically, there are minimum volume guarantees under many of these contracts with OEMs. This contract was—you had one vehicle, and suddenly the production program for that vehicle went below the threshold that triggered the contract, meaning they had to be paid those minimum royalties.

This isn’t related to the loan. It’s because one customer went below the required volumes and basically canceled a program. They had a car and stopped producing it, or they thought they were going to produce 100 and produced 20. That triggered the contract, and they received the payments in an accelerated way.

Okay. Look, I totally believe you. It’s just the way the footnote reads. They don’t do a lot of calls, right? Normally, I read the earnings calls. I think they only do an earnings call.

The way the footnote reads, it says, “We get an accelerated payment. It improves near-term liquidity but gives rise to future payment obligations to the customers.” It seems like that was a liquidity move, but that totally makes sense.

Yeah, that’s the thing: it was poorly explained. I raised that with management, and the explanation was basically that there was a project that was canceled, and they had the legal right to those minimum payments, so they got them accelerated.

Let me go to this. This is a regulatory-driven thesis, right, as we’ve talked about?

Yeah.

Vehicles are accelerating because Europe said in July 2026, “This is when the cars have to start having this system rolled out.” I do wonder about this. Regulatory-driven theses can be really interesting, but this company knows that regulatory theses can be delayed or modified.

It feels like this is happening, right? It happened in the past. You’re seeing the KPIs in flux. But I worry if the companies come out and say, “This is too onerous. We’re having too much trouble with this.”

Could there be changes to the regulation where, all of a sudden, this company has this big convertible that’s counting on the acceleration, and then the acceleration stalls out, or—

I mean—oh, go ahead.

In Europe, I don’t think there’s that risk. In OEMs, this is a multiyear period. When you design a vehicle and equip it with a certain type of software or camera, until you stop producing that vehicle, it will have that software. Therefore, that’s 3 to 5 years where nothing is going to change.

9. Robotics: $20 of silicon versus $20,000 chips

After that period, especially in Japan and the U.S., we could see OEMs pushing back against this regulation, but this risk is the second leg of the thesis, not the first one, we could say. But it’s true what you say. We’re seeing some complaints from customers, but I see that more as bullish because it will require higher-quality, higher-accuracy systems. That’s what Seeing Machines offers, rather than just trying to take this away. It’s very difficult to take this away now.

You mentioned higher-quality systems. Let me ask a separate question. One of the growth areas for these guys—the core of the thesis, if I could put it that way—is regulatory-driven car growth, right?

One of the growth areas that you mentioned is robotics. They’ve got a robotics play that they’re working on. You said that, in robotics, their edge is the same as what they have in cars: very high performance at very low cost, right?

I believe this is on the robotics side, not on the car side. You can tell me if I’m wrong, but you say, “Hey, Seeing Machines’ solution runs on a $20 piece of silicon at the edge. It’s on the robot instead of having to run on $20,000 NVIDIA chips that are kind of in a data center.” That sounds awesome: low latency, low power, way cheaper, and all that sort of stuff.

I guess my question is, why does Seeing Machines have a unique edge in robotics? Robotics is a hot sector. There’s all sorts of money pouring into it. If you’re saying, “Hey, this one company that’s kind of adjacent in the car field has the only way to do a $20 robotics chip,” I’d say that seems a little suspicious to me.

Robotics is still really early-stage. I assign practically zero value to it. It’s just some optionality. Let me frame a little bit how this robotics segment appears.

Almost 2 years ago, Mitsubishi took a 20% stake in the company. Mitsubishi Electric—not the part that does cars, but the part that does robots and other things for factories—bought a stake and started to develop a plan along with the company in the fleet segment and in adjacent markets. Some of the adjacent markets they wanted to work on were smart factories, robotics, and humanoids.

They recently did a pilot. It’s basically Mitsubishi paying the company to develop new solutions, and if that ends up working, Seeing Machines will get paid a royalty. Mitsubishi found the technology very interesting because it can deliver decent performance at very low cost, but it’s still not proven. It’s not yet commercial, we could say.

They’ve built some prototypes, and Mitsubishi likes them, but this is really early-stage. I assign practically zero value to it, although it’s exciting over the long term, especially because Mitsubishi spent £4 million buying a stake in this company. I think the price was similar to what it is today.

10. Smart Eye versus Seeing Machines: software only or full system

Let me go to a different question. You mentioned Smart Eye earlier, and look, I do a half-day of prep for these podcasts, so I could be completely wrong. But based on my loose Googling, quick reviews, and everything, I believe Smart Eye sells a really cheap system, right?

They basically say, “Hey, carmakers, here’s the software. You go figure out your car, your infrared system, and all this sort of stuff. You install that yourself; we do the software.” I think Seeing Machines says, “Hey guys, we’re going to charge double to triple what Smart Eye charges.”

It’s around 70% more.

70% more. Great. But we give you the whole package, right? It’s not just the software. Here’s our camera, here’s our infrared—everything all together, all working together, almost.

Not exactly. Let me give a bit of explanation. Smart Eye has a pure-software approach. The market really likes that. Seeing Machines has a systems approach, meaning they have a team that does software, a team that does optics, and a team that also does what’s inside the camera.

The reason Seeing Machines can charge more is because they reduce the cost of the overall system. If Smart Eye offers you software for $4 but your camera is $25, and Seeing Machines offers you software for $8 but the camera suddenly costs $20 because the software can work better when it’s developed for the camera, the overall cost is the same or slightly lower.

Seeing Machines develops a systems approach, meaning it can take cost out of the hardware and take an extra margin out of that side. That’s also why it performs better in terms of accuracy: it builds the camera for its software.

It’s, we could say, Android versus Apple. Apple builds its hardware for its own system; Android just develops the software, and everybody that builds a phone plugs it in. A systems approach works because it has been developed to work on it.

Okay, no, that’s perfect. It is funny you mentioned Apple, because anytime you talk about system integration, the first thing that pops into your mind is Apple.

I know you’ve spoken to people in the automotive industry. When you’ve talked to people outside of Seeing Machines’ management team, because I think management is the one that relays that full story, have people in the industry vouched for it? Do they say, “Hey, we prefer the Seeing Machines model,” or, “Yes, you actually do save money even though you’re paying more for the hardware”? Have you gotten that verification?

I have not been able to confirm that because people who work in the automotive industry work either with Valeo, Magna, or Tier 1s. Most of them don’t know if they are working with Seeing Machines. It just goes through a Tier 1. They’re like a Tier 3, we could say.

They don’t really know what’s going into their car. They just know that Valeo makes it work, they comply with regulation, and it’s okay.

11. Why no tier one ever bought them

Gotcha. Let’s go to Tier 1. This is an interesting component. One of the things that jumps out to me is Magna, I believe, is the one that has the shareholder loans, as you said. They do business with them and put them through.

Why didn’t Magna buy them? Why doesn’t this belong as part of a Tier 1?

Yeah, that’s hard to tell. Neither Smart Eye nor Seeing Machines are part of a Tier 1. Probably it’s just because these companies have been burning money for so long, and it didn’t make sense to have them in the group. What are you going to pay them right now in royalties compared to what you had to spend over the last 2 decades to develop this solution? It wouldn’t have made sense to buy them back.

It made some sense to have an exclusivity agreement, like Magna had, or some stake, like M&G is doing, or some partnership, like Valeo had. Valeo sold its research and development team, basically, to Seeing Machines, and they made the partnership that way.

There are some related transactions, but nobody really ended up buying the company, maybe because there was huge uncertainty about whether this regulation was ever going to end up coming.

12. Fleet: Guardian 3 and trials that keep not converting

Gotcha. What else should listeners be thinking about? Again, you’ve published 6 articles on them in the past year. I know you’ve got the management calls, but I can only get up to speed so much in a day. What else should I be thinking about, or should listeners be thinking about, when it comes to Seeing Machines?

Something that’s key for this full year is the fleet side of the company. Automotive right now is low risk, we could say. The ramp is already here. We will see some more royalties coming in, and we will probably see growth stabilize around 2.5 million vehicles per quarter.

Not all of that is coming from Europe because these cars are platforms. If you develop them in Europe but sell them in the U.S. or Japan, they will come with the technology just because you develop on platforms. That’s some unexpected volume that’s flowing through.

Let’s start with the fleet side. The fleet side is basically the Guardian Gen 3 solution, which is a camera that costs around $500 plus a recurring monitoring fee per year. That’s something you put in your truck, and it checks if you’re asleep or distracted. It has a very good return for the fleet.

The trucking company, because I’m sure the trucking company gets discounts on insurance, their drivers are safer, and all that type of stuff, right?

Yeah, and especially, it protects against tail-end scenarios. This is for heavy trucks, especially if they carry very expensive stuff. If you have a big accident with one of these trucks, you can have very big liability.

They’re doing many trials, but the problem is those trials aren’t converting. They had Guardian Gen 2, which worked well, and they released Guardian Gen 3 about a year and a half ago, but the uptake has been slow.

The reason for this is that we’re in a big recession in the trucking industry, practically worldwide. There are higher insurance costs, higher diesel prices, and higher costs for everything, basically.

13. My last pushback: at some point it is them, not you

So, discretionary costs—we'll say something that you don't really need to run the business—and that's capex. You're going to delay it as much as possible. Well, let me push back on that, because the whole push for internal cameras, right, would be: A, it's going to save you money on your insurance; B, they say all that.

Not always.

But if you're—I guess, if I'm going and pitching a product and saying, “Hey, you install it and it's got all of these benefits,” and one of the benefits is that it's going to save you money in some way, shape, or form. And I know some of these, and we'll come to competition in a second, because my other question would be: I don't know about the consumer-car side, but I know on the trucking side there are a lot of systems like this. So why should this one even get purchased?

I guess my push would be: okay, it's a recession, but in a recession, if you say, “Hey, I can save you money with this,” that's the first thing people are going to sign up for, right?

Yeah, but it saves you on tail events, we could say—the liability, for example. If you have one of these devices and your truck driver is asleep and the truck crashes or something like that, you can show that it's not your fault as a company. So if something really bad happens—

Does that—that is an interesting question. I haven't thought of that. So, if you're a trucking company and you have a truck driver who falls asleep, the truck driver is liable, not you?

I think that's the case. That's because, I mean, if the truck driver—

If you can prove that the truck driver did something wrong.

Yeah, was negligent, like it happens in aviation. They have the same product for aviation with Collins Aerospace, and that's also going slowly.

You know, one of the main things, apart from the tracking, one of the problems with the trucking industry is, as you said, there are many solutions that offer kind of the same telematics solution for fleets. The thing is that those solutions have very bad DMS, and Seeing Machines has very good DMS. So, for those that really care about their drivers being distracted, they will buy Seeing Machines, but that's a small part. That's the problem they're facing.

I guess my push would be: you said they're not converting customers, right? And I think in your report you mentioned Amazon and Caterpillar might be 2 customers that they're on the 1-yard line with and haven't—

Caterpillar is already a large customer. I think there's more pilots with them.

Because I just, you know, when I hear a company saying, “Oh, we're in a recession. Nobody wants to—a trucking recession, whatever type of recession—nobody's buying our product,” I kind of look at it and, more often than not, it's not been, “Hey, the—”; it's you. It's not them, right? It's like, “Oh—”

It's just 1 of the problems they are facing. 1 of the reasons is that the upfront hardware fee doesn't seem to be working that well, so they might try a more recurring solution.

Basically, the CEO and management team have been completely focused over the last couple of years on the automotive side. Now they say, “We have this business. We are going to try to solve this business,” because this was already working in the past. It's just that we launched a new product, that product got some delays, we launched it at a very bad moment for the industry, and we have some problems converting larger customers.

Their main problem is with larger corporate customers, very big ones. That's what's happening. So they are trying to do new things. They have some very large pipeline deals. If any of those end up converting, the outcome for this full year will be very different.

14. The balance sheet: receivables up 120

No, that makes sense. It's just that, as I've gotten more jaded, management excuses have fallen a little flat and gotten less and less interesting to me.

Let me go back to the balance sheet. I don't believe they published a balance sheet for their June quarter. I think they just said cash.

It's just a trading update.

Say again.

It's just the trading update.

But it looks like working capital really ballooned in H1, right? And that makes sense. Revenue was up 45% in the first half of the year, right? But at the same time, accounts receivable goes from $11 million to $25.3 million. So revenue is up 45%, and accounts receivable is up 120%.

I'd love to hear your thoughts, but again, it comes back to the balance-sheet issue I talked about, right?

They reported cash. So, yeah, working capital has been a problem in H1. But that's—

Well, let me finish, and then—I mean, oh yeah, and I see it in the fifth bullet of that deck. It makes sense that working capital is up because revenue was up 45% in the first half of the year.

I see a company that's growing quickly, burning money, has this near-term bill, and receivables are building up. I say, “Oh, you know, if I was a forensic accountant with an Altman Z-score, I'd say, ‘Uh-oh, things are getting pretty crazy over here.’”

Okay. First of all, working capital is increasing because, once the quarter ends, they get the reports of how many cars they have and the royalties. They do the due diligence to confirm everything is okay, and then I think it's 60 to 90 days from the end of the quarter until they get paid.

They get paid by car manufacturers, so there's a very, very low risk of default. They set up a receivables-financing solution recently, but they have not used it so far. I assume they saw no need for it at the moment, which should also speak positively to the progress they are seeing in refinancing with their new lenders.

As they grow, we will still see high receivables, especially because they report at the end of the quarter and get paid 60 to 90 days after the quarter. But the risk is low, and they have that solution to finance the receivables if they need immediate cash.

I understand that the balance sheet looks ugly right now, especially in a company that is just on the brink of free-cash-flow generation and a growth inflection.

15. How much operating leverage is left in Europe alone

Let me go to the last question. I mean, I think the big thesis for you here is the operating leverage, right? And I'd like to put robotics and the Guardian systems to the side for a second, because I think people could probably see I'm a little more skeptical. I think they're cherries on top, right?

I think the real thing is the automotive production piece, and you can see this—I'm looking at the chart right now. Production goes from 488,000 in Q4 2025 to 2.1 million in Q4 2026. And again, you mentioned that it's a summer Q4, but it is exploding upward.

At the same time, revenue is going up 45% in 2026, growing even faster, and automotive revenue is up 135%. Adjusted EBITDA losses in 2025/26, right? Yeah—

Not much, but it's basically adjusted EBITDA break-even. It's still a loss.

I guess my question to you is: how much more can they do, operating-leverage-wise, on just Europe? If I said, “Hey, I don't know if Japan's coming on in 2029 or 2030. I don't know if the U.S. is ever coming,” and I was just betting on the European piece, how much more do cars grow, and when does this flip to your $40 million in cash flow number on just Europe?

True. Basically, the difference from this full year to the next one is that the next full year will probably be 10 million cars, compared to the much lower figure of around 5 to 6 million for full-year 2026.

Last quarter it was 2.1 million vehicles, and that was March to June, before the regulation kicked in. Many OEMs were already prepared, but some were not, so we should see a bit of a bump in the July quarter, the July-to-September quarter. Probably it will be around that 9-to-10-million range.

If fleet doesn't perform well—for me, fleet is an important part of the business—I don't think the story is just automotive. On automotive, I think it's probably $20 million in free cash flow for full-year 2027 if fleet doesn't perform well. There will also be small increases as L2 autonomous vehicles drive demand for this, and those types of cars are increasing mainly in the U.S. Outside of the regulation, that will drive demand up.

16. Does full autonomy kill the DMS story?

So, on the fleets—what? Okay, you mentioned L2 autonomy driving demand, which makes total sense. I think about a Tesla and the classic story. You mentioned the guy with the plastic doll head.

They ping you if you're not paying attention. You're supposed to be—

But as we go into a more autonomous world, does that actually kill Seeing Machines? If I thought about L5, where the car is completely self-driving, there's no need for me to pay attention. So Seeing Machines doesn't matter at all in that world.

Is this something where it's really hot right now, but if you were really believing in the autonomous story, this is actually a huge negative for the company?

I mean, on DMS, I think this is a good time for that because even if we get L5 in 5 years, regulation will be slow to keep up with it, especially in Europe. We will probably see requirements for people in the driver's seat for a long time, so that should keep DMS demand.

Seeing Machines is already thinking about that. Management is thinking about that risk because, if we don't need anybody and completely trust autonomous vehicles, there's nothing in the news regarding an autonomous vehicle that did something weird and, say, your grandma is scared and doesn't want to be in a vehicle where there's not a driver at the wheel.

So if that happens and there's no need for DMS, they are developing a new solution that's basically 3D vision on the car. It's still early, but we're still early for the need for that. What they're really good at is vision systems.

The solution here would be to replace the expensive sensors that you have throughout the car. For example, for your seat belt, there's a sensor there that costs a couple of bucks, and in each of your seats, it's a couple of bucks. If you can replace that with cameras in the car that can detect whether your seat belt is on or not, rather than depending on a sensor, that would be their long-term optionality or long-term alternative.

I don't know if I'd want to be seeing that. I don't want to be in a world where the driver monitoring system that we built on data is gone, but we install cameras that detect whether your seat belt is on. That doesn't sound like a great world.

One other thing: obviously, the majority of this, as you were just talking about, is the European car-driven story.

17. Chinese OEMs selling into Europe

You know, I'm domestic. I know you're in Spain, right? Am I remembering that correctly?

Yep.

Yeah. The thing I keep hearing about domestically—and I've thought about this with the risk for U.S. auto manufacturers—is that Chinese cars can't be sold in the U.S., right? They're just banned—tariffs or whatever it is.

In Europe, I keep hearing that Chinese electric vehicles are taking it over by storm.

That doesn't mean the ADAS is happening. China has to follow local European laws. But I do ask: who's doing the Chinese ADAS when they're selling cars in Europe? Because if they've all got independent players, à la Tesla, you see where I'm going.

So first of all, China is following ADAS too. DMS is also mandatory in China. It's a huge market there.

But no, is Seeing Machines doing it, or do they have internal players?

Seeing Machines? There are local players in China. Most of the Chinese OEMs have their own, but there have been lots of complaints regarding the poor performance of DMS in China and in Chinese cars in Europe.

I've been told that Seeing Machines is looking to maybe use its software on those Chinese cars that are going into Europe, but the problem is that Chinese OEMs don't pay a lot. So maybe they just will not pay up. There is a risk that if Chinese cars completely take share away from European OEMs, the total addressable market in Europe will be reduced a bit.

Okay, cool. I threw some hard questions at you.

18. Licensing the fleet software to telematics players

One thing I would like to touch on is the fleet side. I think this is really important, and I think it has a lot of future potential. For me, the thesis is that the DMS ramp-up in Europe is priced in. That's the base case. I think that's the margin of safety: if that doesn't derail, I should not lose big money.

I think fleet has huge optionality, not only on their classic business model, we could say, of selling the hardware, but also on licensing it. They are going to license their solution to hardware developers, and they are already working on a deal to start doing that.

That solves a lot of their problems because, basically, with competition, you stop competing with telematics players and instead integrate into them. That means a telematics player with really bad DMS solutions can have good DMS solutions that work better with clients and improve safety. Clients will demand that, and Seeing Machines gets a royalty for it. The market is smaller, but it's also higher margin.

I think that should be some optionality that we see over the coming years. They are already close to a deal with a Taiwanese or Japanese player that's going to develop its own hardware solution, but its software didn't work as well. So they wanted to put Seeing Machines' software on it as a white label. We could see a pure royalty model, like on the automotive side.

Have they landed one of those deals, or are they just in talks right now?

Not yet. They have not yet landed them. When I first talked with them, I raised the same problem you mentioned to me: okay, you have this great solution, but first of all, you have a very small surface. You are not a big player, so you cannot compete on the other offerings that these telematics players are offering.

You have a great solution, so why not license it? Why not partner it? They've understood that this was a very interesting opportunity, and they have started pursuing it. So far, they've gotten interest.

From my understanding, the deal they're pursuing came from inbound interest. It was not them pursuing other deals, but they are now pursuing other players to do this type of licensing deal on the fleet side.

19. CEO incentives and the overpromising track record

Last question. Your write-up mentions that the CEO has a first tranche of performance units that vest, I think, at the end of this month. They require the stock to be at a certain level, and they're almost in the money. Ignore the first tranche, because at this point those will play out, but I think he's got more tranches.

Yeah, he has more tranches, and at very high prices. The poor guy needs this to work—and work really well—to make money on those. I think he has some at 20p or something like that, which, over about a 4-year time frame, if they execute on fleet and automotive, they could achieve that.

Basically, with 50 million in free cash flow and a 20× multiple, you get to a price 3× higher than the current one. So I think if they execute, they can achieve those targets, and if they execute on those, he deserves to be paid really well.

I've spoken with him quite a lot. He's really hardworking. He takes calls on Saturdays and Sundays. They're always traveling, talking with investors, working on licensing deals—everything. They are very hardworking people.

But some things I want to mention around management: if you talk with people who have been invested in the stock for the last 10 years, they have a very bad opinion of it, mainly because management usually overpromises on timing. That's something I've come to understand better as I've spoken with them.

They tend to overpromise on timing because it's hard to know how long it will be from when you get a deal that's practically closed to when you receive the paperwork. It's very variable with these OEMs. My experience is that they tend to be right about what they will achieve, but they tend to be a bit late—or sometimes a while late—on timing.

Let me push back slightly on 2 things. First, on the management incentives, I should have gone there earlier because it is so rare to see stock-price incentives in a European company. As soon as I saw that, I was like, “Oh, that is really interesting,” because it's so rare.

No one can ever control where the stock price goes, right? But they tend to have a vision of how they're going to get there, and the vision they had when they granted these, along with the way the awards are set—

—very much plays out with the vision you're describing, right? So that's nice.

I guess my pushback on what you just said—and we can wrap it up after this—is that a lot of what I've heard has been at the places where I've made the greatest amounts of money and the biggest mistakes, right? It's, “Hey, if you talk to an investor who's been here for 10 years, they hate this management team,” right?

Yeah.

And every time I invest in a company, it's, “Oh, this company has been misunderstood and mispriced for 10 years, but now I'm going to invest in it and the market's suddenly going to change.”

You know, this management team—they've been there for 10 years, they've got gray hair, and they're pulling their hair out, but I'm going to come in and the management team seems nice to me. You kind of learn after 10 years, and it's like, “Oh yes, these guys are really good talkers, but they never deliver.”

A lot of what you've said has some of that flavor to it, where it's like, “Hey, these guys can never quite land the contracts on time because it takes a while.” It's like, well, yeah, but I'm thinking about one company that has been a pain in my side for years. At some point, it's not me. It is actually them, right?

20. Why the stock reacts slowly, and where the risk really sits

They keep saying, “Oh, this big refi is on the come. Oh, this big performance is on the come.” I'm sure you believe that, but now I've got 10 years of you overpromising. It worked for Elon Musk, but Elon Musk is the true outlier. It doesn't work for most of the people who do that. So that would be the last pushback I'd end with.

I completely agree, and it shows in the price action. There aren't a lot of institutions in the stock; most of it is traded by retail. It's like a small cap on the London market.

The most interesting thing here is that it usually reacts really slowly. So, for those listening, probably when they get the refinancing done, it will move slowly. You'll have some time to look into it.

In my opinion, this will probably start to move up quickly when they get their fleet segment—you know, they do a turnaround there and start to deliver on that. My view is that I have a decent margin of safety with the DMS ramp-up and those royalties.

If they execute on fleet, and if any of the next regulatory legs comes, that's huge upside. But it's definitely one of the riskiest stocks in my portfolio.

You know, it's funny you say that because it goes back and forth: the business is more than covered by the European ramp-up, right? That would tend to suggest, “Hey, everything else is a cherry on top.” This is the least risky. But then it's like, “Oh, the fleet is on the come-up.” It's just really interesting.

This has been great. I had tons of questions because this is a super interesting company, and you did a great job of answering all of them. I will include a link to whatever write-up you want me to put in the show notes, so people can go check it out there. This has been great—your second appearance—and I'm looking forward to having you on for a third time.

Okay, perfect. Thanks, Andrew.