Channel number one. Thank you very much indeed. I think that's all, last few people walking through. Thank you. For the attendees online, please feel free to submit your questions just using the Q&A tab situated on the right-hand corner of your screen. Of course, management will take time to look at those questions at the Q&A point. When it does come to the Q&A session, for those in the room, I will hand you the microphone. Please bear with me as I make my way up and down the room. Paul, if I may, over to you, sir. Thank you.
Thank you very much. Welcome everybody, and thanks very much for taking the time to attend this event in person. We're also recording the event and hosting online as well. We have a full house here today, 60-70 people, and we have over 220 people online, so reasonably well attended. The format for today is, Martin and I will go through a presentation that covers the results, but I think this time around we're gonna place an emphasis on outlook and some of the more recent events that we have talked about, but I think important to consolidate, whilst everybody's here. We're then gonna have a panel discussion, and we have Richard Schram from Euro NCAP, Colin Brandon from Semicast Research, Michael Inay from Seeing Machines, hosted by Peter McNally from Stifel.
I'm sure everybody will find that a really interesting and frankly quite unique experience because it's not often you get a group of people like that who are eminently qualified to talk about the regulatory landscape and how it impacts the industry that we're in. Please enjoy that and save up your questions either for the end of the session or afterwards. We'll then have a Q&A, and that Q&A can be live. We'll invite questions online as well. We have a demo set up next door, which Damien, one of our key engineers here, will take groups of around 15 at a time through, so 10 or 15 minutes, 15 at a time. We'll be looking there at the technology that we've deployed through Magnet in the mirror.
You will see the actual mirror, in place, slightly modified so that we can put a TV screen in it. You will see that it is real. It's not fake. It's in cars. It's in production, and it performs with a high degree of capability. You'll also see some new tech that we're developing and have recently announced around 3D, quite important for our future, bringing depth into the equation and enriches features that quite profoundly change the safety case for our customers going forward. You'll also get to see what I know everybody, including me, has been waiting for, for some time, which is the new Guardian 3 unit for our aftermarket business. I will pass this around the crowd. It's essentially a black plastic box.
I don't know what you'll actually do when you touch it, but you will get a chance to touch a product that is run off our full production line in the last few weeks. This is it. As we say in Australia, it is a boomerang, so please make sure it comes back. Just with regard to questions, we've been asked 50 or 60 questions already online, and I've sort of structured those in broad themes that I think this presentation should cover most of. There are some that are a bit more egregious that I won't cover in an open forum. I'm quite happy to take those questions one-on-one later. No problem at all. We should leave no rock unturned today. That's my commitment to you.
Before I get into the material, there's really a four-point message for this presentation and for all of you rather patient shareholders. The first is, you know, what is going on with automotive royalties? What are we thinking about the July 26th GSR legislation and the impact that will have on the growth rate of automotive royalties between now and then? We all know that we had an unannounced and disappointing reduction in our royalty flow in Q4 last year, December quarter last year. Prior to that, growing on average 20-25% a quarter, and we thought we'd have a linear pathway to July 26th. That theory was debunked in December. What we do know is that in July 26th, the cars that are sold in Europe will have a camera-based DMS system, and we'll cover that in a bit more detail shortly.
The first point we wanna get across, not in order necessarily, but there's a strong degree of comfort and conviction around the growth rate in automotive royalties, and that is the primary factor that moves us from where we are to profitability and growth into the future. Very confident about that. Generation Three is the second of the four points. Lots of problems. I'll talk about what those problems were and how they've been resolved. I'll just start passing this around now, and you can touch it, if you wish. We are now in production with Generation Three, and we've started shipping product. The constraint that we've had that's created the volatility in revenue for our aftermarket business has largely been supply, not demand.
Our intention is that, or our expectation is that that constraint is now removed, and we should see a solid path forward. The third area of the four key areas I wanna talk about is the cost, the operating cost of our business. There's a recent announcement where we mentioned $12 million annualized cost reduction. I think there was a large amount of confusion about what that meant. Was it in last half's results? No, because it only happened two weeks ago. Will it be in next results? Yes. Is it planned or done? It's done. Martin will explain in a little more detail the impact of that on our numbers going forward.
Those three things combined, confidence around royalties, finally getting to a point where we're shipping Gen 3 and a materially reduced cost base, lead us to confidence around profitability. That's really important. All of the investors that we talk to, institutional and otherwise, and importantly, those potential investors that have closely followed the story but are mandate prohibited from investing in loss-making companies are very interested in when we're gonna be profitable. That is our core foc When I'm asked questions about what are we doing about the share price, the primary fundamental thing that we're doing is being very focused on getting to profitability. The final thing, which is really accumulation of those three and the profit outcome, is what we do, what are the options around the convertible note that is due in October 2026.
We have a pretty basic plan driven by those cash flows to repay the convertible note. We have a number of options should we require them to supplement cash and whatever additional debt that we need to do that. We have a high degree of confidence around the note being repaid. They are the four key themes that I hope that you'll take away with some confidence today. Please feel free to ask any questions about any of those. On that note, I'll move into the content and provide a little bit of color around each of them. Firstly, Christmas Eve last year, we announced our transaction with Mitsubishi Electric Mobility Corporation. We refer to them as MelMo, a wholly owned subsidiary of MelCo, which is the giant Japanese electronics company. They've invested directly in us.
In addition to that, they purchased some shares from shareholders, to bring their total shareholding to 19.9%. Why 19.9%? That is the limit under the Australian rules. Over that number, you have to make a bid for all shares on issue. The point about the shareholding is kind of subtle but important. It is not really an issue for Mitsubishi about the money, the quantum, the price. They needed a meaningful enough equity investment so that they could demonstrate publicly to their customer base that their strategy to engage more directly with global benchmark or best practice software providers was real. Rather than them just say, "We've entered into a collaboration agreement with Seeing Machines," which many do, they are doing that, but also as a significant shareholder. For them, the shareholding was very important for the communication with their own customers.
Now, why did they do this at all? They entered into this arrangement with us because they're changing strategy. They've recognized in Japan that, after decades of kind of Japanese-only contributors to the supply chain in automotive, the OEMs are starting to move, beyond, Japan's borders for best-in-class capability. Rather than wait for that to happen to them, Mitsubishi Electric Mobility have adjusted their strategy. This is really one of the first building blocks in that strategic change to help them shift their business to retain their leadership position in Japan. This is a major strategic change from their perspective that has resulted in a relationship with us that is underpinned by almost 20% shareholding. For us, it's profoundly important, and we're starting to see some really interesting benefits already, and they're really pushing us to get cracking.
I'll talk about some more of that in a minute. Just back to royalties and cars on road and this, you know, there's lots of talk always about who's better, who's winning. I'm not really fussed about any of that at all, to be honest. The fact is there are two primary competitors, ourselves and the other guys. You know, we've just under 3 million cars on road. I think on any measure from a royalty perspective, certainly up to now, we have a significantly, a significantly higher share. One can debate all day long pipeline and all of those things, but where the rubber hits the road, pardon the pun, it's all about royalties and royalty flow.
Now, with GSR arriving in July 26th, I'm very happy for both of us to be wildly successful, driven by that regulated increase in volume. I think having two quite successful, viable competitors in the automotive landscape is really important because automotive companies demand at least two. They prefer many more, but at least two capable suppliers. If there isn't two, they'll make one. I think this is great news. GSR 26 and having 17 million vehicles in Europe with a camera-based DMS system is awesome for both of us and anyone else that participates in that business. Now, we'll cover off that in a little more detail. On the cost front, a million dollars a month of OpEx out, that's material to us. I think everyone would appreciate that, that we don't do that as a stroke of a red pen.
It's not an Excel exercise. There are a large number of people that are impacted by this kind of effort, and in our case, almost 130. We have to manage those things very carefully and very, very sensitively. In a forum like this, we talk about the benefit, the $12 million annualized saving, the $1 million a month. Back home, where we have to deal with this important decision and remain sustainable and continue to perform and outperform going forward, it's a completely different way in which we describe it, obviously. Guardian, as I've said, and you'll see the unit is now in production, thankfully and finally, and we're really encouraged by what Mitsubishi can bring to the table here.
On that point, this is a wonderful photo of the Mitsubishi team and our team, Kaji-san in the center, who came to our office to execute a recent reseller agreement. Now, Mitsubishi Electric have been working in the US for decades. They have a very, very strong network of customers that serve more than a million trucks in the US alone. Of course, they have similar channel businesses in other countries around the world, Europe and obviously, Japan. This is a very unique and very important channel partner opportunity for us. They're really engaged, buying into our purpose and really engaged with the technology, and they've hit the ground running. It's only been going a few weeks, so I've had quite a few questions in the early question, you know, where are the sales from Mitsubishi?
We only announced it four weeks ago, but we have a sales pipeline building. We've already attended shows together and have built pipelines of leads. This is the fastest I've seen any new channel partner get engaged. I think it's really strong. We're collaborating in a whole range of different ways, though. Not just, you know, how can they sell these units for fleet, but also they're helping us in a pretty serious way around bill of materials costs, advice around design, design for manufacturing cost out, all of these things that they're obviously world experts at. I expect in time, their assistance will be evident in a gradually reducing cost of goods as well. This is really important. I'll just go to auto.
Complicated picture, but if we just look at the black bars along the bottom just for a minute, this is a combination of business that we've already won and that which we're highly confident of winning. It's not very bullish, right? It's pretty realistic, and it accords with our sort of long-stated position of holding a share of around 35%. Okay? But if we look at the numbers out to say 2028, from 2023 to 2028, I've had a lot of questions around where's the $390 million of royalties, you know, why aren't you doing $100 million a year? It is because it doesn't work that way. I think everyone in this room would understand that. We win a program. It's typically six or seven years. It has a year or so of ramp, stabilizers, a year or so of deramp.
Then you win the next one. It builds on top of that and on and on and on it goes. If you were to draw a circle from 2023 through 2028 around those black bars, you will see roughly 40-odd million units, which is roughly $390 million of license revenue, roughly. That is where it is if, if, if anyone's really concerned about that. The key point here is that we were expecting a linear movement between now and July 26th, debunked in December with a 30% reduction. That 30% reduction was driven by a mix shift primarily in our European OEMs, but consistent across all of them. The reduction was roughly the same in all of the programs that we're running. In that quarter, they produced cars with the absolute lowest feature count, essentially in order to combat price competition.
Now, as we get closer to GSR, of course, the fitment rate is no longer optional. That short-term problem that created the volatility that I've talked about goes away. On the competitive front, I mentioned just a moment ago that I think if we've got a, if there are a couple of strong competitors, I think that's good for everybody. What's really interesting, though, is there isn't six, and there isn't, you know, there are probably four. If you exclude China, there are probably four competitors that are, let's say, in our sort of circle.
What I can say with some confidence is that when we look at the RFQs that we're involved in, and at the moment we've got eight or more RFQs in our pipeline, and last year it was next to zero, so they've come back, which is great. Startup production for that cohort of RFQs is, say, from 2027 onwards. These are not GSR RFQs. That work's already done. As I look at that group, there are essentially two primary competitors bidding with the capability to actually execute. I'm, as I say, really happy with a two-horse race. We talk about and will continue to talk about because we don't know what comes out of left field, but a level of comfort with a 35% share in the long run could well be higher.
That's the kind of number that we use to plan our long-run sort of infrastructure. Gen 3 in production now, thankfully, we do have a very strong pipeline. We always have. Demand has not been the problem. Supply has been the problem. We've got trials that are running in most of the major markets with some, and those trials, by the way, are all with what we would call enterprise customers. It's not like the typical sale that we would make through our existing distribution network, which is the bulk of our sales historically. These are enterprise sales direct. As a consequence of them being corporate customers, on average, they're a lot larger. What happened? What was the problem? Three primary problems. They didn't all appear on the same day, of course.
They appear as you test, as you field test, things come up, you analyze them, you find the root cause, you fix them, you retrial them, you retest them, and away you go. Really, just so that we're fully transparent on what these issues were, there were three. The first was the EMI certification, primarily in North America. We had a range of fairly, well, highly technical issues that caused a very extensive delay in that certification. Those issues are now resolved. We are now certified. That problem is solved. The second problem, which was a very big surprise, given that it was in the plastic, the polymer cover on the top of that unit that you've just seen. We designed that cover to spec.
We've manufactured 60,000 of them before, but in a different context, of course, designed to spec, passed the desk testing, oven testing, vibration testing. We trialed them during the Australian summer, in cars, not trucks. We had a really hot summer, and we had a 9% fail rate in the polymer. The heat buckled the polymer, and that left a gap in the unit, and that unit could then potentially be unsafe. We had to, a 9% fail rate is obviously non-viable. We had to step back from that, understand the problem. We hadn't seen it before. Was it just the heat generated by the slope of a car windshield being of a different angle to a truck? Therefore, was it really gonna be a real issue in a truck? Yes or no?
All of those theories were tested, and we deemed that we had to change it. We increased the, essentially the glass content of the polymer, fixed the cap, oven tested it, field tested it, and we're now good to go. Second problem fixed. The third problem, and most important of all, given the kind of product this is, was we had a series of issues with the over-the-air updates. Now, you test units, it works. You put them into a few trucks, it works. You start, you know, jamming megabytes of OTA data into the box, it stops. This is the problem that we found. Essentially what would happen is as we sent an update into the unit, and this impacted about 23% of units, so obviously non-viable, it would go into a thinking loop and never reset.
Software issue now resolved. There were the three issues in Gen 3. I think with our current distributors, the trials that we've got going, the increasing level of interest in Europe, the new relationship with Mitsubishi starting in the US but moving into other countries, you know, for now, and unlike the last sort of six to nine months where we've been dealing with these issues, we have a high degree of confidence that we'll be able to move this product. Martin will go through some numbers and what those numbers mean for us in terms of getting to profitability. One thing I would like to say, this product is now embedded automotive-grade algorithms. This is the difference between Gen 2 and Gen 3, and this might shock a lot of people.
The Gen two product was a market-leading product in terms of its technical performance by some distance. It was twice as expensive as anything else, and we still managed to sell it, and we still have less than a 2% churn rate. It's a great product. The performance of the new product in terms of true positives or reduction in false positives is orders of magnitude better, and it's consistent with that which we deliver to our automotive customers. Now, why is this so important?
From a user acceptance point of view, if you're a truck driver, and I can tell you that if you take almost any competing product and you ask a driver what they think about it, there's lots of expletives because if you've got a product that is alerting you to an issue, say, 20 times more than it should, then the issue that it alerts you, it's the cry wolf thing. It'll alert you to something, and you won't pay any attention, or you'll turn it off, or you'll put a hat over it, or a bucket over it, or a bit of sticky tape, but it won't be considered serious. It'll seriously annoy the driver, creates all kinds of industrial relations issues. It's a big problem.
Having an environment where the alert to the driver is consistent with a real event is very, very important. It is also very important because we are actually addressing the specific issues that the products are designed to address. Fatigue and distraction, we all know that they are major contributors to road deaths and insurance claims costs and all of those things. You get a far better safety outcome because those things you are being alerted for are real. You pay attention, action gets taken, risk removed. Finally, and for us as a business, as opposed to looking at this from our, with our safety hat on, this means that the number of events that come into our Guardian center is materially reduced by orders of magnitude difference. What does that mean?
That means as we look forward, today we've got, let's say, 60,000 Gen twos performing at this level, you know, whatever that means, it doesn't matter, but performing at that level, and we have a cost to serve associated with that. As we increase the proportion of Generation 3 Guardian units to Generation 2, that will have a profound impact on the number of events on average coming into the center and therefore the cost to serve. We are highly confident that we will see growing services margin. Growing re margin from automotive royalties, dramatically improved margin from a Gen 3 product that is at a materially lower cost to produce and a materially improving over time services margin driven by less events needing to be reviewed. Aviation, look, I have no exciting news on the progress of aviation.
Unfortunately, we still have an excellent relationship with Collins. It's rock solid. Collins customers are in a different position, certainly those that are non-military. We have work that's continuing. We're continuing to get paid. We're continuing to develop the product. That is all good, and it's all moving forward. We are really reliant on Collins winning an award with their customers to really put a fire under aviation. I think we are looking at that now in probably the 2027 timeframe. The main game for us in terms of moving the dial is auto royalties, Gen 3 sales of hardware and services, and cost out. On the cost out, this is a really important point. There are a couple of ways to do an exercise like this.
I know I think everybody would be familiar with it one way, either directly or indirectly. We have taken cost out in the past and really, each time we've said, you know, let's just reduce by a number. We were at a point today where we were really more focused on efficiency. Our objective is to produce features at a far lower cost and a far shorter time to market. That's really our core. If we can do that, we know we'll materially reduce cost. We have changed the shape of the executive team. The big change is that John Noble, one of our important team members that's been with us 20 years, has taken over the new role of Chief Technology Officer. In that role, we've combined innovation, R&D, and engineering, and we have reorganized that group to deliver in a profoundly different way.
Firstly, under John's leadership, there'll be an increased deployment of machine learning tools to crunch or increase the pace at which we develop software. We've reconsidered the way we test and quality certify our software, and we've moved from a milestone check into a continuous inline process. We've dramatically increased the amount of synthetic data or digital data as opposed to the physical data. We still use physical data, of course, and we collect it from all kinds of sources. If you think of the difference of moving from $1.50 from a physical image to $0.20 in a digital image, it makes sense to do that. The digitization of data collection and engineering. There are just three or four areas that John has brought together in reshaping that technical team.
Of course, we were previously organized in kind of a matrix or functional-based organization, and he has reset them to focus on customer platform. I think those things are gonna make our engineering output far more efficient. That is evident in the cost that we've just taken out. I think it will become even more evident as we begin in this next phase of RFQ awards. We will see a profound improvement in productivity as that goes forward. Mike, who's here today, has taken on a different role, Chief Safety Officer, quite important. He's directly involved in the regulatory landscape and helping to shape outcomes there. He's also deeply engaged in this new role with customers.
I think what we're gonna see there is a far greater voice of Seeing Machines in the dialogue that we have around each of the markets in which we operate. They're the two big changes. We've also got, in aftermarket, I've just combined aviation and aftermarket. Aviation's on the slow ball. There's one or two resources there. I've asked Pat to take responsibility for moving aftermarket forward. That's reduced the executive team by two. Underneath that, there's roughly, over the course of the last quarter or so, a 120-odd difference in the number of FTEs that we have delivering in our business. We are quite confident that in changing the structure in this way, we'll have a very sustainable outcome. Thanks for that.
I'm gonna hand over now to Martin, who'll take you through some of the numbers that support those, those items. Thanks, Martin.
Thank you, Paul. CFO at Seeing Machines. I'm finishing. Testing. One, two. Facts of those automotive royalties, the benefit of Gen 3 in aftermarket. The impact of the, the cost reductions that we've done, and then kind of wrapping that all together as to what does that mean for cash burn and then eventually cash flow generation. How does that impact on the repayment of the, the convertible note for Magna? There'll be the, the key things that I kind of focus on. There'll be some things that we just kind of skip through, and feel free just to ask questions on those as we get through to the end of the session.
The performance in the first half, as Paul mentioned, was disappointing for us in terms of a revenue perspective, flat in terms of automotive royalties and aftermarket revenues down because of the availability of hardware. We have continued to see improvements in the profit margin, or the GP margin that we see, and we have had benefit from some of the ongoing cost management initiatives that we have had. However, we have had minimal impacts on the level of cash burn. This is something that I just kind of want to focus on in terms of cash burn, and we will kind of return to this at the end to see how all of the elements impact where we are at from a cash burn perspective.
Over the last few years, we've continued to average around $3 million of cash burn per month, an increase when we got to the first half of FY 2024, which has moderated somewhat over the last 12 months, but still in that $3 million per month range. The cost reductions that we did at the end of the first half and continued throughout the March quarter lead to a million dollars per month benefit in cash burn. We're looking at starting the current quarter from the 1st of April with a reduced cash burn rate of $2 million. A pretty significant change there from what we've seen previously, extending the runway that we have with the cash balance largely from the equity injection from Mitsubishi.
In addition to this, there'll be margin benefits from auto and also from aftermarket, which we'll run through over the rest of the session. In terms of automotive revenue, this is largely underpinned by royalty growth. You can see on the right-hand side the light blue line representing royalty revenue. That continues to increase as we have increased volumes. Importantly, on the left-hand side is the mix of automotive revenue. This is really important in terms of what drives the cash flow generation as we move forward. Royalty revenue largely flows directly through from revenue straight through to the bottom line and to cash. We can see over the period of time that certainly I've been here, that proportion of royalty revenue has increased. We're getting a benefit from that, from an automotive perspective.
Just thinking, what does this mean from an outlook perspective? I think this is particularly important to understand the scale of change that we are gonna see over the next 15 months. On the left-hand side, we've got a view of the quarterly production volumes that we report in our KPIs. There's a somewhat of a linear scale that you can see there from left to right. As Paul mentioned, in December, things reduced, and Paul went through the issues that we had in relation to fitment rate. One of the things that I'd like to talk about here is the split that we had here. It's important when we talk about where we'll be in June 2026. December quarter, 266,000 units, around 160,000 of that was related to Europe, and 100,000 related to North America.
Now, when we get to June 26, the fitment rate in Europe is gonna be at 100%. Where we're at now, we're less than 10% in terms of the OEMs that we're with. Over the period of time between now and the end of June 26, we're gonna see roughly a 10 times increase in the volume of royalties on a quarterly basis. Based on that 160,000 units in the December quarter, we'll be looking to increase by the time we get to June of 2026 to around 1.6 million units, just for Europe, in addition, adding on whatever we get from North America. We're at a run rate at that point of close to 2 million units per quarter.
Now, in terms of the lead into that period, in discussions with tier one customers and OEMs, they're planning that they will need a one quarter or potentially two quarter lead-in to achieve that run rate by the end of June. When the regulations start in July of 2026, a car is compliant when it's sold to an end user or to a customer. From our perspective, we get paid when the car is produced, and there's somewhere between a six-week or a three-month lead time from production to get to a car yard and then into the hands of a customer. Our expectation is that the June quarter will be significant in terms of increase, and we'll probably see over a million units in the March quarter as we have that lead-in.
The important thing for us, just in terms of what the rest of the calendar year looks like, is will that lead-in start slightly earlier so that we have a significant increase at some point in the calendar year, or will it really only happen in the early part of 2026? Now, there's a couple of things, you know, three new programs going into production during calendar year 2025, which will be pretty significant, one of which is a new production date for VW. Because of the scale of the VW program, it's been split into a number of start of production dates. There's a new production starting for VW in the latter part of 2025, which will add a significant amount of volume on a quarterly basis.
We also have two new OEMs going into production over the summer with a different tier one. They will add significant volume as we go through the calendar year, probably hitting more so with the ramp-up in production by the time we get to the December quarter. The unique thing about those programs is that they come with a minimum volume guarantee. Now, some of those have already started, and the pattern of the revenue guarantee or the volume guarantee matches with the expected production schedule. It starts off relatively slow, it picks up, gets to a plateau, and then tails off after about three or four years. We had a small, less than $100,000 benefit from that in the first half of this financial year.
It'll increase slightly in the second half, but the benefit that we get in FY 2026 is the equivalent of just over 1 million units of production. I think just in terms of thinking of the scale of change that we'll be going through over the next 15 months, we just did over 1 million units in the last 12 months. This is a minimum volume guarantee underpinning our FY 2026 production volume just from one OEM, two OEMs, one tier one over three programs. It's a pretty significant change. That represents around $7 million of revenue. In terms of near-term production, Paul talked about the issues that we had in December. We only receive official reports on a quarterly basis.
The March quarter that just finished, we'll start to see production reports from our tier one customers in a couple of weeks, and that'll go on through to the end of April. Early indications that we've had from a couple of the OEMs and the OEM programs, which saw significant declines in December, are that we do have growth. We expect that the December numbers will be around 300,000, slightly over, not up to the level that we saw in the September quarter, but certainly above where we were in December. They'll start to continue to increase as we go through the rest of the calendar year.
I think the important thing just to take from an auto perspective is EU regulation driving additional volume, getting to a run rate by the time we get to the end of our FY 2026 year of around 2 million units per quarter. That is really what takes us from a cash burn rate through to generating cash. By the time we are entering FY 2027 and getting to the point when we have the repayment of the Magna note, we will be generating significant cash flow on a quarterly basis. In terms of aftermarket, to the right-hand side, you can see the revenue here reducing from what we saw in the first half of FY 2024, largely because of the availability of hardware. That was the reason for the reduction in the aftermarket revenue.
You can see again with the, the line there, this is the recurring revenue stream. Recurring revenue continuing to increase. It's slightly plateaued from an ARR perspective. You can see on the left-hand side chart. That's largely because the majority of revenue that we have from services is in AUD for Australian and New Zealand customers. In terms of the connected units, it's around 75-80% of those units are in Australia and New Zealand. As the AUD has depreciated against the USD, it's just led to a USD reduction in ARR. As we start to sell through more Gen 3, we again expect that ARR balance to start to increase.
Just looking in terms of outlook for aftermarket revenue, now having the supply of Gen 3 available, I think it's important just to look at the left-hand side chart first of all. This just matches Gen 2 sales of hardware on a quarterly basis over the last few years. You can see that there are these kind of plateau or spikes and troughs of volumes. This is largely because of the availability issues that we've had with Gen 2 over the last few years, post-COVID supply chain issues. As we've got through to now just finishing production and not having Gen 3 available to sell in its place, that's really what's led to the reduction in the Q1 and the Q2 numbers.
The end of production was in Q4 of 2024, and we had a number of delivery commitments through Q1, Q2, and Q3, which we retained stock for to meet those obligations. That is really what has driven any hardware sales for this first part of the financial year. In terms of the line there, that represents the average 12-month quarterly sales of hardware. It is around 4,000 units per quarter, and that has had a cap on it, one because of availability, and then the other is because of the real inability for us to be able to sell in North America. That has largely been because part of the unit, the Gen 2 unit, is manufactured in China, and so therefore has attracted tariffs when we have sold into the U.S.
It just makes us uncompetitive from a pricing point of view. Now with Gen 3, I don't know if you've looked as it's been passed round, but it's manufactured in Indonesia, so it won't attract any of the China tariffs going into the U.S. If there are any other Trump tariffs added to non-Chinese production, then it's likely to impact all telematics producers equally. There's no telematics, or I'd say very minimal telematics, that's actually produced in the US. We're not concerned from a competitive pricing position point of view as we were with Gen 2. Just going into some of the economics of Gen 3 compared to Gen 2. Paul talked about the service performance or the performance difference between Gen 2 and Gen 3. We also have a margin difference.
The cost of a unit for us to produce is around half of what Gen 2 costs. We have a significant benefit from a margin perspective. Gen 2 was delivering a margin of single digits. We'd expect with Gen 3 to be aiming for around 45%-50% margin. In addition to that, we have the benefit from a services perspective. I just kind of want to talk a little bit about the production capacity because this is really what's going to be driving the benefit from a cash flow perspective over the next six to nine months. Near-term production capacity is around 6,000 units per quarter, slightly higher than that. It's probably going to take us the rest of this quarter to reach that, to reach that run rate.
We won't quite have 6,000 units in the June quarter, but as we go through September and December, we'll be aiming at just over 6,000 units and selling those through. We have an expanded distribution channel in North America. Paul talked to the benefit that we get from Mitsubishi Electric and the access to their customer base and the joint leads that we create. We've also got our own expanded sales team in North America as well, which has also been replicated in Europe. We have much more capability to be able to sell outside of Australia and New Zealand. 6,000 units per quarter, selling that through for us will generate somewhere around $2 million worth of ARR. In combination with the additional hardware margin, that'll be benefiting us to the tune of around $1 million per month in additional margin and cash flow.
In summary, I think there from an aftermarket perspective, over the next three to six months, getting to a run rate of production of just over 6,000 units per quarter, selling that through, delivering on the ARR benefits and generating an additional $1 million a month in margin and cash for the business. Paul talked about aviation. We do have some continued activities here, albeit at a slightly lower, slower rate than what we had been doing previously. We do expect in the second half that the revenue generation will pick up slightly, but it is not gonna be material until we start to get to FY 2027 in terms of a major contribution to revenue and earnings.
In terms of gross margin, you know, as we talked about with auto, we've had a change in revenue mix over the last few years, a higher proportion of higher margin revenue streams being royalties, largely from auto, services from the aftermarket business, licenses generally at a high margin rate. What we're likely to see as we go through the rest of the calendar year is hardware, which has been a low margin contributor over the last few years, turning into somewhat of a high margin contribution. On the left-hand side, you can see with that line that the overall trend is that margins have been increasing from what we had a few years ago, around 35% up now to close to 60%. By the time we get to June of 2026, we'll be at a run rate of around 70% gross margin.
That means more revenue falling through to the bottom line and to cash. In terms of cost management, this is just a look at what has happened in terms of employees and contractors over the last few years. Paul talked a little bit about the impact of some of the structural changes within the organization. Just thinking about this from an overall employee numbers perspective and what it means going forward. We will talk about the cost impact on the next slide. We have been slowly reducing the number of people within the organization and also the external contractors that we use to undertake development work.
We had a pretty significant increase, from 2022 through to the end of 2023, and that was largely done to meet the commitments that we had for some of the larger programs that we'd won, particularly with VW, where there were 50 new features to develop. A significant amount of external resources added in order to meet those commitments. As we've rolled through the start of production dates, particularly the production at the end of December 2023, we've been able to slowly reduce the number of headcount within the organization. In December, in the first half of 2024, we had a small increase from ACE because of the acquisition there and taking on an additional 30 people. As Paul mentioned, we've been through a restructuring over the last few months.
We had a few people, around 20 people that left at the end of December, and then around 100 people that have left over the last few months. After the end of the first half, most of those were finalized actually last week. The ongoing benefit of those changes from the first half is around $1 million per month in operating costs. This is what we can see in terms of demonstrating that from a cost perspective. I just kind of want to talk a little bit about how we think about costs. When we're talking operating costs, we're excluding cost of goods. We're not talking about cost of goods there. We're talking about operating expenses that you'll see in the P& L.
We strip out depreciation and amortization not being cash costs, but we include the R&D that's capitalized, given that that is a cash cost that we have. That is the internal metric that we use and drive for cost management purposes within the organization. That is what's reflected here on this chart. The darker bars are what you'd see in the P&L as operating expenses, and then the lighter blue stack at the top is the capitalized R&D. Over time, again, since the period of peak cost in the first half of FY 2024, we've had a significant reduction in costs, which has happened kind of slowly over the last 12 months and then accelerated over the last six months. The impact here coming from largely from headcount reductions, a scale down of the external R&D outsourcing.
The impact of the, the changes that we've made will start to be felt in this half. The second half of 2025, there is an element of exit costs associated with the changes that we've made that you'll see in the second half numbers. As we go into FY 2026, we'll see the full impact of the reduction of $1 million a month. What I want to do here is just demonstrate the impact of those changes that we've made on the cash burn rate over the next six to nine months. We use adjusted EBITDA as our proxy for cash, and use that as a cash burn metric. The left-hand column is the actual numbers for the first half of FY 2025.
The cash burn just under $18 million for the six-month period, that equates to just under $3 million of cash burn per month. If we include the impact of the operating cost changes that we've made over the last couple of months, that's around just over $1 million per month. That puts us on a run rate as of the first of April of just under $2 million per month of cash burn. Incorporating the other changes that we've talked about, in terms of auto and Gen 3, they're the things that take us from a $2 million per month cash burn to a break-even point and then to generating cash as we move forward. We talked about with Gen 3, just over 6,000 units per month in terms of production capacity, additional ARR of around $2 million a year for each 6,000 units.
We expect by end of September quarter, that is gonna be generating us an additional $1 million per month in margin, reducing cash burn down to $1 million per month. In terms of auto, it's a little bit more difficult to forecast when this is gonna happen. Obviously, we've talked about what the run rate will be by the time we get to June 2026, around 2 million units per quarter, probably around 1 million units per quarter in the early part of calendar year 2026. We estimate to get to a $1 million a month benefit from where we're currently at will take around 750,000 units per quarter.
That is something that we expect to be able to achieve by the time we get to December through a combination of current programs increasing in terms of the fitment rate, so increasing volume in current OEM programs, and then the contribution from the new programs that will go into production over the summer in 2025. In terms of auto, it does not stop at that point though. If we have 750,000 units per quarter, increasing up to 2 million units per quarter by the time we get to June 2026, that is really what then takes us into the position where we are generating cash. By the time we get to the first quarter of FY 2027, we will be generating somewhere in the region of $10 million per quarter.
This is what really gives us confidence in terms of being able to repay the convertible note from Magna and generate significant cash as we move forward, from an auto perspective and then as Gen 3 continues to increase. Just moving that through into the implications for the convertible note. We have a cash balance at the end of December of just under $40 million. We're expecting to get to a cash flow break-even run rate during calendar year 2025. That will be dependent on the timing of really seeing that significant uptick in automotive royalties. Calendar year 2026 will be cash flow generative and we'll be doing around $10 million per quarter by the time we get to the mid part of 2026. The convertible note matures at the end of October 2026.
The quantum of that initially was $47.5 million. We've been capitalizing the interest on that such that $60 million approximately will be repayable by the end of October 2026. Our expectation is that at that time we'll have around $40 million of cash and be generating $10 million per quarter. Our primary plan is that we'll be repaying $30 million of cash and then looking for additional debt on a short-term basis to repay the other $30 million. I'd expect around $10 million of that debt would be related to receivables funding and $20 million from other sources.
Now, in terms of other possibilities or other options, if we do not actually get to that level, we would be looking to do something similar to what we have done with some of the licensing deals in the past, similar to what we did with Caterpillar in order to raise additional cash, to meet that obligation. From that point of view, just to summarize, currently a run rate of $2 million per month of cash burn, looking to get to break-even by the end of the calendar year, additional cash flow generation from the expansion of auto royalties as we get to June 2026, and a cash flow generation position of $10 million per month. That enables us to be able to easily repay the note from Magna per quarter. Sorry, glad someone is listening. Thank you. Over to Paul.
Can you go down and get him to check that? Thanks very much, Martin. We're now gonna cut over to the panel discussion. I'm gonna ask Peter, Peter McNally from Stifel, to run the show from here on. If I could ask Mike, Richard, and Colin to pop up, up front here.
Are we on? Okay. Thanks everybody for joining the panel, and thanks for those presentations, to both. My name's Peter McNally. I'm an equity research analyst here at Stifel. Work on the European team. Myself and Freddie Hindley, who's at the back there, we lead the charge on U.K. equities. I think just before we start, just some overall thoughts from us. If we look at DMS software, it's, it's always been, you know, it's been in development for well over 20 years.
It's always been something who's, you know, its time was gonna come, eventually, due to the passenger benefits. But it, it, it's taken a long time. Today we think it's really starting to happen. There are two main drivers behind that, and that, that's safety ratings and regulation. From a regulation perspective, GSR, as you probably know, started in 2024 in July, for all new vehicle models in Europe. As we approach, and that was to detect fatigue. As we approach the July 26th deadline, regardless of if it's a new model or an old one, all cars that are produced in Europe need to have DMS. That requirement also extends to detecting distraction as well. That to us means that it needs to be camera-based. You know, all new vehicles are gonna need DMS camera-based.
We also think that increasingly consumers are paying much more attention to safety ratings. In fact, around 65% of people believe that safety should be the top priority when buying a car, according to Consumer Reports. Europe is taking a leading role on safety globally. This is what we think is changing the broad automotive landscape. Joining on the panel today, we have three very experienced guests, with automotive industry insights, each with a unique perspective. From Seeing Machines in the middle here, we have Chief Safety Officer, Professor Mike Lenné. Mike has been leading the global safety discussion across transport for 25 years. He leverages his experience and depth of knowledge of both human factors and safety science.
He's been building global partnerships with Seeing Machines customers, research and technology partners, and safety groups to drive future growth for the last 10 and a half years. We also have Richard Schram at the end. Richard Schram has worked for the European New Car Assessment Programme, also known as Euro NCAP, since 2010, where his current role is the Technical Director. Richard's primary focus is the technical execution of the Euro NCAP roadmap, but developing the rating scheme, and all the underlying test and assessment protocols for 2026 and beyond. Welcome. Finally, we have Colin Barndon to my left, to your right, the Principal Analyst at Semicast Research. Colin is a fiercely independent analyst with over 30 years of experience, 25 of which has focused on the automotive electronic sector.
Over the past five years, Colin's focused a lot on the DMS market, landscape, including suppliers such as Seeing Machines, SmartEye, SIPIA, Tobii AutoSense, amongst others. Thank you all for joining. If we may start off, Colin, and by the way, anyone chip in to any of these questions? Colin, the GSR regulation is already live in Europe. What can you tell us about the current impact of the regulation, and what the deadline in July 2026 means or, or might mean for fitment, for, for expectations?
Okay. Can anybody hear me? No. I guess your mic's about to fit. Thank you. Peter, good afternoon, everybody. Okay. GSR, the General Safety Regulation. That is live.
I still don't think that's on. Yeah.
Testing. One, two, one, two. Sorry. Thank you. Cheers. This is what technology troubles look like, folks. This is real time, managing these things. Okay.
The GSR, the general safety regulation, is live in Europe. You know what European regulations look like. You know, if we think about airbags, anti-lock brakes, seat belts, we know how it works. You know, if a vehicle is sold in Europe, 100% fitment rate is in the technology that's mandated by the general safety regulation. We're in the second phase of this now, so it's called GSR 2. Those folk who've been around for a while, following this stuff like myself, this was proposed back in May 2018, and it covers a raft of safety systems. There's intelligent speed adaptation, automatic emergency braking, lane departure warning. There's also a technology called DDIW, driver drowsiness attention warning. These technologies, they're actually already in, 100% fitment rate. The introduction phase for those technologies was 2022 to 2024.
There were some technologies which are much newer and were much less understood by the industry and indeed by the European regulators, that are on a two-year grace period, if you like. Peter's just touched on that. ADDW is the name, advanced driver distraction warning. That is vision-based. That requires a driver monitoring system. That's the phase that we're in at the moment from the 7th of July, 2024, through to the 7th of July, 2026. The fitment rate of these systems is rising from close to zero to 100% of vehicles sold in Europe. Now, it applies to class M and class N vehicles as well. That means goods and passenger.
We've seen it essentially, it covers anything with four or more wheels that's sold in the European Union has got to have a driver monitoring system fitted by the 7th of July, 2026. We may come onto a question later about whether that will move. There were some people asking me that over lunch, but I'll give you my answer to that if that question comes up. Essentially, about 20 million vehicles will be sold in Europe, light, medium duty, heavy duty that will require a driver monitoring system either embedded in the vehicle or after manufacture, that is in the trucks. That's where we are for the legislation as it stands. Fantastic. Thank you.
Colin.
Yes.
Thank you, Colin. Maybe Richard, maybe you could give us a bit of a view from the consumer side of the world.
What approaches to safety in Europe are you working on, and where does DMS fit into the conversation?
Maybe one step back to give a bit of perspective for where Euro NCAP stands within the whole regulatory framework as well. Just to be a bit proud for myself and the organization I work for is that most of the General Safety Regulation is based upon what Euro NCAP has done. Also, the introduction of General Safety Regulations typically becomes a lot easier because common factors are already doing it. I think driver monitoring systems is a specific area where especially Euro NCAP is really focusing on as nearly a prerequisite to even have higher performing safety systems.
I would nearly argue that we're at the limit of what ADAS systems can do without annoying the hell out of you during driving. We really need this additional element that the vehicle knows what the driver is doing in terms of warning later or earlier. The introduction of driver monitoring systems at this point in time in our rating scheme is, well, you can ignore it to get a five-star vehicle, but some manufacturers already pick it up. Certainly from our new rating scheme in 2026, it is such an important item, not only in one of our new pillars in the rating scheme, but also in the link towards other elements in the rating scheme. You cannot do without it. We are obviously also basing ourselves on what the regulation will prescribe us that will help us.
But driver monitoring because for us becomes a key element, not just for monitoring the driver, but specifically to have higher performing ADAS systems available in the vehicles because you need to know the status of the driver to, to do this.
Very good. Thank you. Richard, I wondered you if you could, or sorry, Mike, maybe you could give us a view on, you know, how Seeing Machines software fits into this picture. And, and maybe a view of, you know, what are the OEMs asking for at this stage?
Is it working without a Mic? Okay. I just thought I'd give the opportunity. So, we've been working with our customers for years demonstrating concepts around distraction and drowsiness. And this has been driven, you know, in, in part, large part through what Euro NCAP has led, introducing, forcing the introduction and discussion around distraction and drowsiness.
There are five reasons why, behavioral reasons why, drivers injure themselves and others. They speed, they do not wear a seat belt, they are distracted, they are drowsy, or they are drunk or impaired by drugs. Speed, eyes-off systems, intelligent speed adaptation taken care of. Unbelted drivers is part of the Euro NCAP roadmap, which I will comment on in a sec. Distracted driving, visual distraction, we have covered. Cognitive distraction, part of the roadmap. Drowsiness covered. Alcohol and drug impaired driving is being introduced in 2026. I think these conversations around what are the, what do the accident data tell us are the risks to the public?
Yeah.
Then how does our technology roadmap evolve in ways that address that? Then how do we communicate that to our customers, our safety partners, you know, such as Euro NCAP, to make them aware of what is actually available?
You know, what's the maturity? What's the performance? And, and so I think detecting those things is absolutely critical. Every OEM that we are dealing with has asked for GSR for drowsiness and distraction. The OEMs are asking us for the NCAP 26 features, which is a critical tipping point between a narrow focus driver-only view and a wider cabin view. For Seeing Machines, if I look at the mirror, this can be one and the same. Right? We provide the accuracy of a narrow field of view on a driver for distraction and drowsiness and the wider field of view of the cabin to detect seat belts, out of position, et cetera. From a safety viewpoint, this is super exciting because as Richard said, it's not just about detecting that someone is distracted or drowsy or drunk.
It's how do you use that information with the other vehicle systems? Airbags is the one that's coming in now. How, you know, if the vehicle knows that I'm like this, tilted out of position, or I'm leaning very close to the airbag, we can provide that signal to the OEM. We provide additional value by adding value to the airbag sensors. If I could just say, we did have an announcement yesterday around 3D depth sensing technology, because the question is from a safety viewpoint, we know these are the issues in the roadmap. From a consumer policy, we know where it's likely to head, and that's a collaborative discussion. It becomes a matter of how do we technologically achieve that?
Exploring technologies such as depth sensing for 3D, as we announced yesterday, is a key part of positioning ourselves with our customers to meet not just NCAP 26, but also as we participate in the discussions towards NCAP 29, that we are technologically ready for, for how that may evolve.
Very good. I think we want to, that point you make about airbag deployment, I think we want to come back to that a little bit later. Before we get there, maybe Richard, you could talk a little bit about how the OEMs are achieving safety ratings today and where are we currently and what do they need to do to get a good safety rating from you?
Yeah, the way our system works, we sort of have four major topics and it will completely reorganize in 2026, but the same principle holds. There are four main pillars with different elements in there, that in every pillar you need to have a minimum score to be able to get to five-star levels. Typical manufacturers would aim for a certain target, five stars, four stars, sometimes even three stars or lower. It's a clear choice by a manufacturer to go for it. Luckily, most of them would go to five stars. I would say nearly all of them, the vast majority would go at least for four stars. We have some budget vehicles that go for three stars. Now, obviously they all need to meet regulatory requirements.
As Mike said, you might as well make the most use out of the sensor that is already in there from a regulatory perspective, to get your cost down and get as many Euro NCAP points as you need for the topics that you want to cover and economically cover in your vehicles. Manufacturers, to be fair, will wait until they actually have to. There are very few of them that will do and install a lot more than what they have to, to achieve a certain target. That's what we saw in current rating schemes. We see an increase of take up of standard fit Driver Monitoring Systems already today, even though it's not mandated. They choose to maybe have those points available in Euro NCAP.
Mostly, I would argue to show off a high score, at a, I would argue it's not required. That counts for every single element we have. You don't have to ace every single test we have to get into five stars, but you need a minimum performance. Again, OEMs will wait until the very last moment until they have to do it. We saw, if I take a completely different example, like AEB or like rear seat pretensioners or rear seat load limiters, we saw nearly 0% standard fit at one point for rear seat pretensioners. The year we introduced a dummy in the rear, it went up to 100% fit. Industry just follows exactly what they have to do. There are only very few of them that will do more than what is required.
Now, again, the good thing is that regulation requires this, but this is the world we work in. What I think the biggest value for Euro NCAP in this case is, yes, there must be a driver monitoring system in your vehicles. It doesn't mean you have to have a very good one if you want to meet the regulatory requirements only. That's where Euro NCAP comes in to really push manufacturers to do significantly more than what the regulatory base minimum is. That's how we cooperate with industry. That's also where I believe Euro NCAP and regulation should work together, where regulation was set at the bare minimum. We will progress and then regulation will raise the bar themselves and close the gap a bit so that manufacturers have some time to introduce these systems, develop them.
I think what Mike mentioned is the relation to ADAS and other systems is very new, and also very scary for them. You do not want to wrongly trigger your airbag. That is probably even worse than doing something else. They really want to be confident and learn a bit more over time on what these systems can offer and how accurate they are, how robust they are in terms of the information given before you link them to other key safety systems.
I think as we look over the next few years, for the income rating requirements, my understanding is that as we go forward in years, the amount of weight that DMS is given within the safety rating increases. Is that right? Is it Mike or Richard, either of you want to talk about that?
Yeah, especially this, this 2026 is gonna be a massive jump, also because we knew obviously that regulation came in, and the systems were fitted. Even, and that's Brent, something we haven't done before is put a prerequisite in general on a safety rating that if you offer a, I think what the vast majority knows as level two systems, assisted driving systems are available to you, and they are claimed to do more than what they are. I think the key element missing there is that the driver is not monitored and kept in its role as a driver. Therefore, if the manufacturer doesn't couple the driver monitoring system to its assisted driving system, we will effectively penalize you in the rating scheme.
There is a lot of links in the rating scheme next to just the single element of driver monitoring, which is around 30% in one of the pillars where you need to get up to 80%. Without it, it's impossible to get a five-star rating in the end. That's even left aside all of the other couplings that we have to other elements in the rating scheme. For me, driver monitoring is really the next step in the S curve to be able to go for higher performance and safer vehicles. Without it, it sort of becomes impossible without doing things that are unacceptable to consumers in terms of warnings and interventions because you cannot do more if you don't know the real state of the driver.
Mike, you wanna
just to, just to add, add a, a perspective here. I mean, I think if we look, so income 2020, if we just look at the number of features that are recognized in the income protocols over time so that require DMS. 2023 was distract, you know, distraction, visual distraction and, and drowsiness. 2023, did I say 2023? I meant 2023. 2026, it's, distraction, it's drowsiness, it's non-drowsy related impairment, bracket alcohol. It's the seating position of the driver. Are they in the seated position or not? It's, is does, is the driver wearing, does, is the driver using their seat belt appropriately or not? What is the height, of the driver? There are a lot more features now that are recognized, in the point scoring in the Euro NCAP protocols going from 2023 to 2026. That's a fact.
The question is, where does it go over the next three years? You know, what are the, what are the safety, how do we join the safety minds with the technology minds to go, what problems can we solve? How do we evolve? Visual distraction into cognitive distraction being one example. Alcohol impairment to perhaps other forms of drug impairment, right? If we, when we talk to people in the U.S., they're more concerned about cannabis, equally concerned about cannabis as they are alcohol. I think these are the sort of discussions, and for those that don't know, you know, Seeing Machines does represent the tier two suppliers on the Euro NCAP occupant status and monitoring working group. You know, the model is about collaboration and bouncing back. Where does Euro NCAP want to go? Where do the regulations want to go?
Where do we wanna go and how do we support that technologically?
Great. I didn't really appreciate how much the two were really intertwined in the past. Colin, I'm gonna ask you a tough one. The million pound question, which of the DMS providers in your independent view have a product that can accommodate these rising requirements?
Shall I sit closer?
Just check it.
Just check it. Could you hear me all right?
Yeah.
Okay. Great. Okay. I wasn't sure if this is gonna work. Okay. Everybody claims they can do it. What's really interesting about this industry is there's so much PR and there is such a lack of hard data. I'm speaking entirely as an analyst, you know, looking at the claims that are being made and the, I'm getting a change to my thing. So yeah, maybe let's try that again.
There you go. I'll just say that again. There is so much PR and there is so little hard data. Again, some of the questions that we go, we might come to that, but everybody claims that they can do it. The really important thing to remember is, and we've touched on it already, that there are two sets of guidelines that are coming through here: ADDW, which is part of GSR for homologation. That's pass fail. You know, that's written down, what is required to be done there. The key part really is the 14 attention zones that the DMS has to be able to identify. That's really an eye gaze fidelity issue there.
You either pass that and the vehicle can be sold in Europe, or you fail it and the vehicle can't be sold in Europe. That's homologation. Now, Euro NCAP is not a pass fail scoring. It's a star rating. As Richard's touched on, you can go for five stars, you can go for four stars. For some of them, you can go for three stars. This is really where the innovation comes in. If we look at what's happened in the last several years, Euro NCAP 2023, on offer for driver monitoring was three points, only three points. Essentially what could happen, and largely what did happen is that many of the OEMs just ignored DMS completely.
They did not even put it in because they could score more highly on ADAS and concentrate more highly on ADAS, and score the requisite number of points to get a five-star rating. Now, what has happened in CAP 26, and Mike's touched on this and Richard's touched on this, is this introduction of four different categories. One of them is called safe driving, and that is where DMS fits in. You have to get a certain % score within that category in order to get a five-star rating. Now you have to get a certain % score across all categories to get a five-star rating. What this has done, essentially Euro NCAP has held the automakers' feet to the fire and said, take driver monitoring seriously or surrender your five-star rating. Richard might use other words. Those are my words.
They've come outta my mouth, but that's essentially what the new legislation means. You can no longer just ignore DMS and expect to get a five-star rating. That, I think, really is one of the issues, the core issues that maybe a lot of people haven't quite picked up on, is the way that the new protocols work. For 2026, I think you've gotta score 60%. For 2027, you've gotta score 70%. This is specifically in that category. For 2028, I think it's 80%. What in 2026 has done is massively increased the importance of DMS. The automakers can no longer just say, we're gonna ignore it and focus on ADAS.
They've now got to put in a vision-based driver monitoring system that will score points within this category, safe driving, in order to get enough points to aim for that five-star rating. Really the summary of all of that that I've just said is that good enough DMS is out. You've now got to focus on what is required in order to score those points in order to get a five-star rating. You can come below that, you can get a four-star rating. You can come below that, you get a three-star rating. What it means is, the earthquake, which has really gone through the industry since the protocols were published in November, is causing people to stop and rethink about which providers they can work with, and which companies they need to be partnered with, in order to hit those ratings for income 2026.
It's a big question. It's a big answer. As I say, I don't really think the consequences of what Euro NCAP has done have really landed, certainly with the investor community, and even some of the automakers.
Any thoughts on any of the providers?
Damn. Okay. I mean, I'm put on the spot here. This is my personal opinion. Other opinions are available. People can seek elsewhere if they want. Seeing Machines is the technology leader by an absolute mile. That is really, really unusual. If Seeing Machines can't do it, it can't be done. As Paul alluded to earlier, you know, really in the industry, the automakers are looking for two suppliers to be able to do something.
Seeing Machines, in my mind, as an independent analyst, is the technology leader by a long way. We might come onto some of the issues about that. The two other suppliers that I look at, SmartEye and Cipia, I actually rank them about equal in terms of their technology, and they are both behind Seeing Machines. I'll come onto that, something to do with implementation. I think we can probably fit into this, but I rank them about second equal. SmartEye say an awful lot in their PR. You will have seen stuff, no doubt, on LinkedIn about my interactions with their CEO. That's a separate matter. Cipia, I think, is probably the most underappreciated supplier in the industry. Tobii AutoSense is a really interesting one.
At the moment, they're in production in aftermarket systems, for heavy vehicles in Japan. They have some production, for cabin monitoring. Their data says at the moment, they're not actually in production with the driver monitoring systems, for series production light vehicles, I should add. We'll have to wait and see how those systems get scored when we see in 2026 and some of their systems come to production. At the moment, Tobii's a very competent player on the cabin monitoring side. I do not rate them at the moment as a DMS provider because, as far as I can see, they have no programs currently in production that are light vehicles with the software embedded in the system. We'll see how that goes.
A lot of this really will become clear once income 26 comes into effect, and once the income scoring system starts to come through for new vehicles starting next year.
Excellent. Thank you for being so brave with your answers. I think it's not just about software, right? It's also about implementation. I wonder if Mike or any of the panelists really have any thoughts on, you know, implementation and integration in the process. I mean, how does it impact programs getting successfully launched? In addition to, you know, the consumer experience as well. Any thoughts on those topics?
I have a perspective. I think the implementation, so I look at implementation from the viewpoint of people sitting in this room who are gonna buy a car with DMS fitted in it.
The implementation hits, hits the, is effectively the interaction between the driver and the vehicle systems, the human machine interface. This is the, I think it, this is one of the make or break of DMS. You know, there's been talk earlier about systems that beep and bop for random reasons. Interior Sensing, DMS, OMS, these are technologies that interact with you as a vehicle occupant in ways that no other technology has before. The interaction between occupant and vehicle is just going up and up and up. The importance of design and correct design is just skyrocketing in parallel. Colin mentioned, you know, good DMS, or, or sorry, just, you know, good, good's not good enough anymore.
This is really where years of R& D and thinking and testing and evolution really come to the fore because it's not just about putting an algorithm into someone else's system. It's much more than that. It's putting a compute efficient feature onto a platform of a customer's choice, into a location of their choice, et cetera, into an HMI warning system of their choice. Right? There is this chain, but you know, one of the things that we do is we invest in that whole chain because ultimately, if the end user doesn't have a good experience, acceptance drives sales. That implementation around you as consumers, as well as investors, but having a good experience with the technology is absolutely fundamental.
This is where having a false positive rate, as Paul said for Guardian earlier, driving down false positives is equally as important as driving up true positives. In another way, being right when you should be right is just as important as not warning when you should not warn. That is the key to acceptance. I think that is a massive implementation issue.
Very good.
Maybe to add a bit of what Mike said, I think consumer acceptance has been there. We saw the first implementations of, arguably, some mostly Chinese manufacturers coming to Europe showing that, well, look, we have DMS there. You can quickly see that the complaint rate of those vehicles went up skyrocketing simply because they did, I would argue, the dumb thing. You just warn whenever people look away and that is it.
That's where typically your income does not want to head for. Where we are completely different than regulation, as I mentioned earlier, we want this clever link with ADAS. If you look away, but nothing critical happens around you, the car will prepare for a critical event, but you will never know it. That's really the implementation where we want to go as your anchor, but we believe consumers could have an acceptance of these systems when you have a longer duration, but also critical events are distracted. That's where the system should adapt and only use it and apply it where it is. I would argue it's even more than false positives and true positives. It's a clever implementation. On top of that, that requires high performing systems in terms of DMS. You're not gonna change your airbag setting.
You're not gonna change your AB by a signal that might be right or wrong. So purely fitment of DMS is not enough. You need high quality DMS systems. And there, I do agree, with Colin, and obviously with Mike, he would argue that, Seeing Machines will be one of them that can deliver these high quality systems that are that good that they can also then tune and rely on when you are adjusting ADAS or restraint system design.
Okay. Fantastic. We, okay, we are getting limited on time. Okay. And I've got a couple more I wanna ask you guys before we go.
Okay. This is really, but it, you don't have to be a market analyst with 30 years experience to get this or an electronic engineer, but it sure helps. This is the critical issue.
A DMS is a driver monitoring system. Okay. The S in DMS means system. Okay. It's not just software. Now, you know, when people talk about hardware agnostic design, what they basically mean is we're just gonna focus on the software and we're gonna kick the can of system implementation over the fence to somebody else. You know, so Seeing Machines, in my view, is unique in its overall system focus. You know, and the term that's often used is tier 1.5. So the things that I've written about in the past that really matter, co-design, training data, you know, synthetic and naturalistic and human factors expertise, you know, they're the things that come together to develop an overall system, not just the software. Now one of the issues, and this is so important, is that the tier ones, they're experts in ADAS.
You know, we look at some of these companies, you know, Aptiv, Bosch, Conti, Magna, Valeo, you know, they're companies with, with long experience in ADAS. What these companies don't have is, is detailed understanding and knowledge, of driver monitoring, and particularly not working in a nine, 40 nanometer optical path. If the tier two is just simply focused on the software, you know, the question is, where does the system level expertise come from? If the, the tier one doesn't have it, in depth itself. The answer is it has to come from the tier two. So the tier two, if they're gonna be successful in this market, they have to have a system level approach.
That really is what underpins almost all of what I write about and all of the positions I hold is that Seeing Machines is the only company that I can see that is a true driver monitoring systems provider. That underpins all of my comments on LinkedIn. All of the stuff that you'll see me do publicly comes down to that one important issue. You know, if a tier two then is gonna cut out the tier one and become a software tier one, then the question I've got is, where is the system expertise coming from if the tier two doesn't have the system expertise and they're trying to work with the automaker directly? It is such a fundamental part of the argument about how this market develops and who the suppliers are going forward.
Is this point of an issue with that DMS is a system, it's not just software. Thank you for giving me the time to say that.
Of course. That was great. Thank you. Here we go, we gotta do one more last quick question. I think it's an important one to get in, if you don't mind, and then we'll wrap it up. Look, I think there's some perception out there that the GSR deadline, which is only 15 months away, you know, there's some perception that some of the OEMs might not be ready out there. Do any of the panelists have a view on whether that's true or not and whether or not this deadline might move?
Just in terms of the OEMs being ready, I mean, we're actively working with a number of OEMs on the features required to support GSRs.
We're also actively working on new business and expanding current business. I guess the great thing is that there are a number of options available for OEMs who may not be as prepared as others. There's different packaging locations. Yes, the mirror is an option, but the great thing about the mirror, as I said earlier, is that it's got Seeing Machines into narrow field of view and wide field of view. So whether it's a single camera, dual camera, any location, you know, we've been doing the R& D to support any of these locations. Whether it's a new program or whether it's an expanded current program, you know, we're seeing both of those pathways to support the OEMs being ready. From our viewpoint, we're very ably supporting.
Great.
Very short to just put things in perspective. You're in GEP and it's always the case. Just before a new rating scheme launch, we have over 100 new car models coming through our program this year, and I'm sure that none of them have invested a lot of money to develop these vehicles for just half a year. They're all ready and have made their precautions and the pre-developments to fulfill regulation, half a year later.
We don't think that deadline'll move.
No, that will not move.
Thank you very much to all the panelists. That was fantastic. Thank you.
Okay. Thanks. Richard, thank you. Colin, thank you. Mike. And, expertly, led by Peter. Thank you. Thank you very much. That's quite a long session and quite a lot of content.
I really do appreciate you, you've been able to hold your attention for that amount of time. Hopefully you'll have got a lot out of it. I mean, we're trying to be as open as possible to the extent that we're able to be on the outlook. The outlook that we've put forward today is related to some things that we've actually done and reported and some things that we are highly confident of happening, auto royalties being one. You've just heard from some people external to us who are entirely independent that that pathway is set and 15 months isn't far. Thank you again for being able to have that wide ranging discussion. It's quite unique. You may never get the opportunity to hear that from the horse's mouths again. Thank you.
Now what we're gonna do is go into some Q and A and, after the Q and A, we'll have an opportunity to take a look at the technology that we've just been referring to. We can't fit you all in the room, so I'm not sure if some can stay or have to go or what have you. Feel free to hang around. There's an opportunity to talk with us here at will, but we'll be moving in, say, groups of 15 for, yeah, 15 minutes or so. Okay. Are there any questions?
Any questions in the room? Yep.
Sure.
Sure.
Thank you all for your presentations. Paul and Martin, it was really nice to see the detail under break even plan. You know, I think the core savings you've made, really difficult decisions to make, I'm sure.
You know, you save a million, a million dollars and it's a million dollars on the bottom line. From a shareholder point of view, that's really good. I understand why you've got a plan for the Magna loan, but I don't need to tell you that it's their option to take the shares at 11p. It was a great deal when it was done, and it still remains a great deal today. On market share, I think Colin said about six years ago, 65%, 35%, 40%, it doesn't matter what it is. If you look at the lowest end of those projections, find me another company with 35% market share. I think Apple has about 30% of the smartphone market.
If you look at it from that point of view, my question is to Colin, if that's okay. Colin, in terms of what you've seen, looked at recently, whether at CES or in your own research, what's the most exciting thing you're seeing kind of coming through automotive at the moment?
Thanks, James. I mean, the key thing that I'm seeing really is just a pace of innovation around driver monitoring. You know, and I'm gonna leave things to do with self-driving cars out of it, generative AI out of it. If I'm looking at what is possible, in order to save lives, make roadways safer, it is the pace of innovation of driver monitoring systems technology that's coming through. It doesn't get written about. You know, I'm still pretty much the only analyst who's talking about these areas.
You know, there's so much focus on AI, generative AI, in the car. Qualcomm is talking about that. NVIDIA's talking about that. The software-defined vehicle is a really key area. One of the things that seems to be happening, I wrote about this last week, is maybe the automakers that were trying to do quite a lot of stuff by themselves, particularly around integrating software themselves, might actually already be looking at moving back to working with tier one, competent tier one partners, and the implications that that's gonna have across the industry and across the driver monitoring supplier base. That is very much an emerging trend there. Really, you know, the ability for this technology to save lives is just incredible. As NCAP 26 plays out, and we start to see it, that will be the technology driver.
GSL will drive the volumes. The feature set, desire that, or designs that have been driven through in the last couple of years, that's started to come through now. That's quite a long answer. I'll stop there.
Can I just add one piece to that? From my point of view, there's lots of talk about full autonomy, and I get that question all the time. I'll, we'll be fully autonomous. There'll be no need to monitor anything. We disagree.
I think what's really clear, despite the investment in full autonomy and despite the fact that it's still going, the zig of the zag is that it appears to me that level two plus capability, semi-autonomous driving, will go into every single vehicle, regardless of whether it's a high level vehicle or low level vehicle, before we get anywhere near full autonomy. If you think about that and the implications of that, I think level two plus outpaces full autonomy by orders of magnitude and we'll get to full penetration of level two plus before we get anywhere near a meaningful shift in full autonomy. That's a really big thing that I see, read about, hear about, and talk about. I think for us, that's pretty good. Thank you all.
Thank you, Paul.
The recent redundancies, the fucking Seeing Machines, I'm sure that's really hard for you and everybody else in the business 'cause these are people you know and that you know their families and so on. Can you reassure us about the purpose of these redundancies? How Seeing Machines is able to do that without impacting long-term capacity, without impacting ability to do the engineering needed on potentially eight new RFQs coming through later in the year, potentially winning those and being able to engineer those without impact, impacting the ability to design and build the Gen four. How much does our external partners, Mitsubishi, other partners, how do, how does that impact on how we do it? How much is it that there is engineering work that's been done that doesn't need to be redone in the future?
Can you reassure us that we have not damaged our capacity and our ability to move forward as a business?
Thank you for that question because you are absolutely right. It is very difficult. There are two things I would like to say. We talk internally about purpose all the time, and that is to get everybody home safely. In order to get everybody home safely, we have to be here. In order to be here, we need a business and a business model that is sustainable. This is the conversation I have at every town hall, and this is linking purpose to what we actually do in our day job. Sustainability includes financial sustainability. We are operating in a market where we are not in control of everything, but for those things that we are in control of, we actually have to be responsible.
When I have a conversation internally about being sustainable, I'm not talking about the environment or pollution. I'm talking about our ability to be in business to enable our purpose of getting everyone home safely. We link purpose to the actions that we take, and we are very stand up and straightforward about that. That's the first thing. The second thing is just looking at the numbers. We had increasing volatility and uncertainty. Would I have preferred to wait a little longer to make the changes that we've made? Of course I would. In fact, as a human being, I'd rather avoid it altogether. Being accountable, we can't do that. We look at the numbers, we look at the gap, and then we try to maximize or minimize the cost to give us the maximum runway.
That's the math of getting to a sustainable business through purpose. Now, one of the things I said in my presentation, in particular in relation to John and his role as CTO, is we haven't just taken a pool of, you know, 450 people and put red lines through it, 130 of them. I mean, in effect, that has happened, of course, but we've actually sat back and with great consideration, rethought and redesigned the way we develop software, the way we interact with customers, moving from, you know, functional capability with lots of handovers, lots of checks, lots of balances to continuous inline development and testing process. We've moved into an environment where we are restructuring teams with more machine learning expertise to reduce the manual effort in coding.
They're the things that give me great confidence that not only can we continue to support the customers that we are commercially obliged to support, but we will actually come out of this and operate at a level of productivity that prior to this we'd only imagined because we've reset the way we do business, the way we're organized, and the methods. It's much more sophisticated than just chopping off a bunch of cost. I hope that answers your question.
Thank you very much.
Is that working? Yeah. Great stuff. My name's Graham Herndon. I did have a badge. I dunno where it is. It's all in out. Firstly, when I first studied engineering, we had things called valves, but none of you were alive then, I'm sure. And then we had these new fancy things called thyristors.
The DMS was a diploma in management studies. I love your jargon. I've got most of it, just a couple of bits I'll catch with later to avoid embarrassment. Brilliant presentation. I bought shares in your company, and of course they're worth half, half what I paid for them. Doesn't matter because the tax man will take 40% when I go in any event. I want to make a point that you've been in business since 2020, 25 years. Life's a journey, not a destination. I set up my company in 1993. We are still there. We're still making profits. Now my question is very simple. When are you going to be able to introduce your product to old cars? Because July the 5th last year, I was interviewing a fund manager.
The next thing I knew, it was two days later and I was in Watford General Hospital having had a stroke. So I've got an old Porsche, which I still go to the office and back in. I'd like to have your product in that Porsche question. Thank you.
I was, I was just going to go right out on a limb here, and say, we'll come and put one in for you, right? And it'll look like that box that we've just handed around. And whatever that takes, if you're into it, I will do that. I think your question, I think your question is, targeting a, an addressable market of greater than one.
Look, the new cars is the focus 'cause it's driven by regulation, but in trucks, of course, there is regulation emerging certainly in Europe, but in the rest of the markets in which we operate, there are no such regulation. There are just organizations that are completely committed to safety and keeping their people safe and getting them home safely. There is a product, an aftermarket product that we have today, and you've just seen it. It's dedicated to essentially heavy transport. Now, could we get to a position in what we would call, let's say, a Gen 4 or a Gen 5 product where that, essentially that camera and processing unit can be shrunk enough and have its cost reduced enough that it can perform the same set of safety features in a truck? Absolutely, it can.
I do believe that that is a very, very real and very significant market because there will always be more cars on the road than get manufactured each year. I think that future is coming. I don't think we are gonna get the product as we see it at a, the level of economics that would make it consumer viable probably inside five years. It's certainly on our roadmap. The first step for us is to deliver what's immediately in front of us in order to be profitable, in order to be sustainable. At that point, we have quite an array of plans around broadening our current product and service base, and that would certainly be one of them.
Perfect. I think Sophie is very kindly. We've got a number of questions that have come in online, Paul.
You had 70, as you know, that were pre-submitted. Thank you to everybody for your engagement ahead of today's event. I hope Paul has managed to address a number of those topics. There is a question here for you, Colin, as well. I will come to you, if I may. Giuseppe asks, Chinese market update, please.
Thanks, Giuseppe. That is a big question. I assume you are referring to our position in the Chinese market. In that case, I will give you our view. Several years ago, we made it really clear that we did not have any desire to have a direct market entry into China, and that really has not changed.
If we consider the reasons why, apart from the kind of populist reasons, if we just look at the economics, 80 million cars, rest of world, 20 million cars, China, okay, that's changing rapidly. I get that. If we look at DMS volumes, which is the business that we are in, and let's say we were to look at an average price of $10 for the 80 million, you know, that you're getting an addressable market of 800. If you look at the 20 million in China, it's probably a dollar. The real question for us is, how do we operate in a market with a completely different price point and a completely different set of behaviors, understanding, and attitude towards performance, right? We are not unique in this case.
The way that we've elected to do that is to work with our current tier one customers. With some of them, we are working on, let's say, a platform standard dedicated to China through a tier one customer. What that would do is we would take all of the work that we have in our broader platform, add it to, let's say, the China specific platform. As we continue to develop features, and if we think of feature development in a European context or a US context, tier ones wanna know when you have an idea, they say, oh, that's lovely. Can you make it work on this? You say, yes, you make it work on a platform. Then they say, can you make it cheaper? Yes. Can you increase the level of performance at an even cheaper cost? Yes.
That's a two-year exercise, a two-year conversation. In the Chinese market, they want your new feature idea the minute you've thought of it. They want the proof of concept first on a spreadsheet with a GoPro. Yeah, that's not true, but just by way of comparison. They want features really quick that don't necessarily work but have prospect that they can then, you know, move forward with themselves. They're about speed of entry, and the rest of the world is about performance. That's completely different in terms of an operating business. I think there's limited money in it, for us. The way we are addressing it is to, as we develop features and ideas, but before they get to a level of performance that a European OEM or tier one would accept, we pile them into the Chinese platform, right?
It's derivative works that we don't put any extra effort into, and then we allow the tier one and the Chinese OEM to develop them however they will. That we can sell for $1 or $2 because it has very, very limited effort. That's the strategy, very high level, and that's how we are addressing a market which is a materially lower ASP.
Cool. Thank you very much indeed, Giuseppe. I hope that answers your question. Colin, a question from Jonathan. Can we ask Colin, if any of the large telematics providers that provide driver monitoring for insurance purposes, such as PowerFleet, have any interest in targeting GSR requirements, will they have to OEM this capability?
When I talk to my customers, I say to them, if you can find anybody else with more knowledge on this market than me, then go and work with them.
That's how confident I am about being a, an authority commentator on this industry. Okay? My expertise is inch wide, mile deep. It's unusual actually for me to be talking in a public forum because my intellectual property, my expertise is chargeable. What that means, however, is that you step out of that inch wide, mile deep area of knowledge. I have no expertise whatsoever. When it comes to telematics, I don't know anything about that market whatsoever. Thank you.
Okay. Look, I think, Paul, yes, of course, sir. Let me just bring the microphone.
I was looking at the kind of projections for AR, and I just wanted to check because my understanding is a bit ropey on this. The Gen 3 Guardian, does that, would that work even if you didn't have 24/7 monitoring?
Because I'm hearing from some people that I've spoken to that, some of the, well, they want closed loop systems in aftermarket. I just wanted to check, you know, how that accords with plans and projections.
Our service offer is a closed loop 24/7, that 24/7 service offer. That's what we sell. Okay? The Guardian Gen three, as the Guardian Gen two, picks up the driver condition, relays that through the cloud. There are in-cabin messaging, alerting, vibration. If there's a micro slip, all that information goes back through the cloud into the Guardian center, and that closes the loop between the supervisor of the fleet and the driver in the truck. It's that closed loop and the intervention that delivers the 90+% reduction in fatigue and distraction risks. That's what we offer.
To the question of can it, it can, of course. You can have a system, our system even, that just provides alerts in the cabin. You can. We don't, look, if a customer said, look, we wanna start with you with cabin alerts only and no, no, can, you know, we may or may not do that depending on who they are, we might. If it was a, if we felt that there was a path to getting to the services angle, because that's about deriving the highest safety outcome. We're not really in the business of selling a beeping box. That's not what we wanna do.
That's great, Paul, Martin, and to the entire panel, thank you very much indeed for your time today.
I know we've got some very exciting stuff to show the guys in the room next door, and the air conditioning's a little bit chillier in that room. Thank you once again to everybody for attending today. To those online, thank you very much indeed. We'll redirect you now so you can provide your feedback in order that the management team can really better understand your views and expectations. Paul and Martin and the rest of the team will be available to take any other questions that you may have over the next hour or so. Thank you once again, and good afternoon to you all. Thank you.