Kick it off again, now for my next guest, Dan Galves from Mobileye, Chief Communications Officer. You know, typically very little need for introduction. Long-time friend, both buy- side, sell- side corporate.
Is this working?
Is this working?
Yeah. There we go.
Great.
So Dan and I go way, way back in the business in many forms. Dan, maybe.
You were my first client.
There you go.
First, buy-sider, that would speak to me.
First answer and best.
Yeah. Okay.
I'm not gonna go that far.
Okay, perfect. Dan, maybe to kick it off, 'cause I think you always give, you know, great updates. I know you have slides, but to frame a little bit of the discussion, the questions that I get from investors on Mobileye, typically in the two camps, what's going on with core ADAS, right? Changing OEM and maybe a little bit on the competitive landscape. And then second, the new, you know, the sort of the future, everything beyond Level 2, SuperVision, a lot of exciting stuff happening in Chauffeur, you know, vehicles that we can buy soon for Level 3 and Level 4, and then Robotaxi. Maybe we could start to, you know, talk about some of those opportunities.
Yeah. I mean, that's kind of how I think about the business too. It's really three businesses. One is generating a lot of revenue and cash flow today. The other two are in markets that barely exist today, so you know, the core business, like, you know, Mobileye is maybe $0.30-$0.40 of EPS, but there's $0.60-$0.65 of EPS that is R&D for advanced products that are not really generating revenue today. So, you know, the core business is very strong. It's $1 billion of free cash flow if you strip out that R&D. It's $1 billion of free cash flow, $1 of EPS. You know, we've been kind of outperforming our core customers in terms of their production because there's still growth here. There's still kind of market share gains within certain customers. You know, there is opportunity for new customers.
There's new markets like India and Latin America that have pretty low penetration today. And I think that there's a, you know, and the driver of that ADAS business, which today is essentially single front camera, maybe a couple of radars. Our product is essentially the software to interpret the camera and the radar data, as well as the silicon, the edge processing that it goes on. So we design both software and hardware in-house. The key driver is safety ratings, right? T hose safety ratings have specific criteria that you have to meet to get certain scores. And those criteria expand every couple of years. And we're getting to the point where you can really no longer get to a five-star safety rating in North America and Europe without side cameras as well.
That pushes the need for more processing, and it pushes higher selling prices for us. So we see, you know, a good kind of growth trajectory for the core business, like, you know, over the next, you know, five to 10 years for sure. The other two businesses, one, it's really split into, like, how are the vehicles gonna be deployed, right? I mean, it's all about, like, the car taking over more and more of the driving, from the human, but there's really two segments. One is vehicles that you or I would buy. So vehicles that are produced in hundreds of thousands a year, you know, they're sold through dealerships to regular people. We have two products there. We have an Eyes-on product called SuperVision.
It's very kind of targeted to be very similar to Tesla FSD, hopefully better, but, you know, a system that takes over the driving, you know, in most scenarios, urban, suburban, highway, but the driver needs to pay attention, and be essentially the backup for the system. This is a camera-only perception system, you know, with crowd-source mapping, decision-making on two EyeQ6 Highs. We sell it for about $1,300 per unit. We have a design win with Porsche that launches late in 2026 and will cut across Porsche and Audi models. We have an Eyes-off product called Chauffeur, which is targeted for highways.
The idea is, you know, you get in your driveway, you plug in the address that where you're going, and, you know, as you kind of move out of your driveway through the city streets, you have the SuperVision system, so you have to pay attention, but the car's driving for you. As soon as you hit the highway system on-ramp, the system is able to engage and allow you to do other things, while you're driving until you get off the highway. We have a production program here with Audi, that will launch in the middle of 2027. It's about $2,500-$3,000 per unit for us. W e see that, you know, we and I think, like, the industry sees this as kind of the killer app for this type of technology because the value proposition is clear.
Like, you can get your time back, in kind of the area where you're doing most of your driving, which is commute on the highway. SuperVision, I think the Eyes-on, there's, it's not clear how the consumer will value this, right, which has created some level of kind of delay in decision-making. With Chauffeur, I think it's very clear to everyone that there will be a strong value proposition, but it's so hard to do, no one's been able to do it. So kind of questions around the technical maturity, and that's why the Audi program is so important for us, is because we have kind of a very strong engineering team, you know, inside Mobileye and with Audi.
Really the first of its kind, the Level 4.
Really, first of its kind, type of product coming out in mid-2027. So that's the second business. And then the third business is Robotaxi. And, I think here it's not really about, you know, who you're teamed up with or kind of what, you know, like, everybody's kind of announcing this city, that city, this kind of value chain collaboration. And that's important, but what's really important is, like, can you commercialize, you know? And you can't commercialize until you remove safety drivers, and you can't do that until the precision of the system is at super high levels. And, you know, there's been one company that's done that, Waymo, and, kind of amazing technology. And we believe we're gonna be the second to remove safety drivers in the middle of next year.
And I think in that way we can, you know, hopefully separate ourselves from kind of all of the other, you know, players. 'Cause this has started, this is a business that no one asked us about a year ago.
Yeah.
Now it's kind of the main focus, but that's also driven sort of like PRs and announcements across the industry. Well, I think we'll have to see how kind of reality plays out.
Excellent. So why don't we dive into each one of those? Let's maybe we'll start on the core ADAS. You know, this is some numbers that you've put out publicly, you know, 33-35 million chips. It's been running like just a little bit under, not, you know, 8.5 million-9 million a quarter. You talked about growth in that business. How do I think about terminal, right? 'Cause you have U.S. and Europe that will need to grow, but we have this large, you know, portion outside of China that is still relatively low of ADAS. So is there a way to think about how to grow that business for a normalized end state on just Level 2?
Yeah. No, I mean, I think you're right. Like, that's a big question. Like, I don't see ADAS as like a huge swing factor in our valuation, but I think, like, what is the growth trajectory? Is this still a growth business?
Yeah.
Or is it not? Like, I think that's an important question. I think this year we've been growing faster than our customer production, right? T hat's, that's been kind of, you know, mostly consistent for the last 10 years. W hy is that happening? You know, we still have customers where, you know, we're 50% of their kind of ADAS, you know, portfolio, and, but we have design wins that take us closer to 100%. Honda is a good example there. You have markets like India that you mentioned that, you know, maybe, you know, ADAS penetration is 5%. There, you know, have been, you know, Mahindra was the first company to put ADAS on a domestically produced Indian vehicle. They thought, you know, people, you know, it was an option for pay. They thought people would, you know, maybe 10% of people would pay for it.
It ended up being like 50% or 60% of people paid for it 'cause safety is a real priority in India. There's new regulations coming into that market. And so, you know, a company like Mahindra has told us, you know, maybe we do a hundred thousand units with them now. Like, that looks like a million unit, you know, business within the next few years. And there's other automakers in India. So that creates a good growth trajectory. New customers like Volvo, where we don't have any volume now, but we start to ramp up, you know, starting in the second half of next year. Yeah. So I think that there, you know, and then surround ADAS, right?
Yeah.
So I think you're, you know, you're getting to the point, like I said, where, you know, you can't really score highly on European and North American safety tests without having multiple cameras around the vehicle. That makes the software much more sophisticated, much more complex. It requires more processing power. So the average selling price of our system will go up. We have one design win with Volkswagen so far for that, you know, but now we have a kind of a pipeline of six or seven other automakers that are kind of looking to, you know, to do, you know, to, to kind of install this type of system in, you know, big portions of their portfolio starting in like, you know, late 2027, early 2028.
Can we just review the 101 on the geographic differences within ADAS? You know, you have 80% plus share in U.S. and Europe. Just a reminder to people how many of the 35 million units are in China?
Right. So yeah, our guidance for this year is $34.0-$35.5, just to be super clear. And, you know, obviously we'll report at the end of October and update, you know, and kind of make comments on that. About 6 million units this year roughly will be shipped into China. It's split half and half between Chinese OEMs and then shipments to foreign automakers that produce cars in China. So kind of three and three approximately. The Chinese OEMs want to kind of have multiple sources for ADAS. So, you know, at best we're, you know, 25%-30% of these companies' kind of ADAS portfolio. I think that there's a strong kind of push in China to use local suppliers, but it's very challenging to meet these European standards. You know, the standards are not as high in China, so they really need our help.
You know, we can be a big ally for their kind of expansion into Europe, so that creates, I think, a good relationship for us with those customers and, you know, stability in the volumes. The foreign automakers in China, they're losing market share in general.
It's also a sales issue. Yeah.
Yeah, exactly. So that's been a declining business for us and probably will continue to be a headwind, but we've been able to kind of grow past that. So yeah, we, we feel like it's, you know, it's, the Chinese business is smaller than it used to be, which is probably good.
Yeah. So let's turn to SuperVision 'cause I think, you know, it's some version of chicken and egg, like you said, with the OEM and the consumer. Can you talk about, you know, in December of last year, you know, you did the infamous sort of, I call it the flag slide. How many of sort of those conversations are ongoing and just maybe there's a decision about which models to attach it to, how the OEM's going to price, whereas maybe for some of them they may have now be waiting, they're either gonna downgrade to surround or waiting for Chauffeur. Could you kind of give an update on that front?
Yeah. So I mean, I think that's kind of what we've seen over the last couple of years is, you know, basically like a kind of a long-term evaluation of our technology and kind of getting to the point where kind of the engineering team within the OEM is like, this is the solution we should go with, but then kind of upper management pulling the trigger to do sort of an aspirational program that you don't have to do.
Yeah.
Right? Now, do you have to do it to be competitive in the future? Like, it's an open question 'cause Tesla is really the only one out there.
Yep.
This is kind of what's been happening. I think with the flag slide, you know, if I remember correctly, there was like four Japanese flags. There was a Korean flag and maybe a European flag.
Yep.
The two that were closest to kind of decision, one was Nissan. They kind of famously decided to go with kind of a multi-supplier, like kind of down scoped system, that we're not involved with. We'll see how that plays out. The Korean automaker, I would say, is more like in limbo and, you know, maybe transitioning to kind of engagements around kind of other pieces of the portfolio.
Yep.
The rest have progressed, right? Especially the one that was all, that was the farthest to the left, you know, has moved very quickly. You know, we have, you know, three or four of these engagements that are targeted for decision this year, but we've been there before.
Yeah.
Right? So I really don't wanna predict kind of the likelihood of decisions. Like, we still see every OEM in the world having a strategic priority to create a strategy in this field of kind of more autonomy. I think it's just, you know, when do you pull the trigger on it is kind of unclear. We do think that there, you know, one of the things that delayed decisions was, you know, kind of the changes in automaker portfolios when it became clear that EVs weren't gonna be.
Yep.
You know, 80% of the market, you know, in the next few years, the OEMs had to go back into their portfolio and make changes.
How do we upsell some of these ICE vehicles?
How do we upsell some of the ICE vehicles? And so that, I mean, that created a period where they weren't making decisions, but it also led to a lot of kind of modern software-defined architectures that are launching in 2028. And we are seeing like quite a bit of sort of urgency around kind of what ADAS, what advanced ADAS system should be in these vehicles because if these are our highest tech vehicles, like we don't want them to have just a simple ADAS system.
Yeah. So the age-old question is it's a lot more about when are we launching our software-defined vehicles and not whether it's ICE or EV, which is relevant to you.
Yeah, exactly.
And the timeline for a lot of the SuperVision type decisions, is it still, you know, the programs that you talked about at the end of the year, they're 2027, 2028, whereas a lot of them are 2028 or 2029?
Yeah. So you're still, you know, the most near-term like decision points are for second half 2027 launches. But you have some that are for 2028 as well.
Okay.
I think like the thing that, you know, one thing that helps us greatly is these Volkswagen programs.
Yeah.
I mean, for us, it's like critical because you can't really develop a technology into a product like in a vacuum by yourself. You know, you really need kind of the collaboration of the automakers and kind of the integration into vehicles, validation of kind of the performance. And so, you know, the first programs are the hardest. So these are taking basically three years to develop, but once you have those in production, you can reuse a lot of that technology for other customers. And for the OEMs, it creates this kind of lower risk where, hey, I'm not the first, you know, Volkswagen's gonna launch before me.
And before we move to surround and Robotaxi, I just think it's 'cause it's super important for everyone in the room when I think of, you know, my investors, you know, a lot of them are purchasing those Audi Q whatever vehicles that are gonna be launched SuperVision to Chauffeur. Can you talk about how you're gonna launch some of those vehicles where it's, they'll have the Chauffeur technology, but, you know, you're gonna validate it in that 2027 'cause, you know, you have a lot of buyers here who'll be, you know, sort of interested in that first Level 4 vehicle.
Yeah. So, you know, the SuperVision can launch, you know, without significant kind of oversight by regulators because it's an Eyes-on system. So there's, there are regulations in Europe, but they're, they're very clear and it's a lot about driver monitoring to make sure that people are paying attention. With Chauffeur, like you need to certify the vehicle based on kind of real-world data. So we will launch in mid-2027 with all of the hardware required for, you know, Eyes-off autonomy on highways, but the system will launch as Eyes-on. And, you know, if you have, you know, 20,000-30,000 units in market and you can monitor those, the performance in the background, like very quickly you can have enough data to, you know, to certify the vehicle. So that's the plan. I don't think Audi's talked about a specific timeline, but.
It's like a shadow validation.
Right. It's not like what Tesla's doing where, you know, the system doesn't perform anywhere close to what you would need for Eyes-off. And over time you'll improve it. This is like the system launches where we in the lab kind of know what the failure rate is.
And you're trying to.
You know, and then you're just validating it through kind of on-the-road data.
You gave public numbers around SuperVision, which is, you know, to paraphrase sort of, like $150-$200 of content versus your typical $40-$50 to hit those 2028, 2029 U.S. European regulations. But some of the numbers that you've discussed of the programs you're interested are almost, you know, I think 700,000-800,000 units if I remember, from the December Analyst Day, because those are the major platforms, right? These are significant, this is where the OEMs make a lot of money. So.
Yeah.
We should expect those timelines are almost more set because I know the OEMs have to hit those launch dates.
I think that this is, you know, overall, like in general, just easier, right? Because it's, there is kind of a have to do it component related to the safety ratings. So, you know, if you want to continue to get kind of the highest safety ratings for your vehicles, there's some level of kind of have to do this. On the side benefit you get is because you have the extra processing power and you have the extra sensors, you can also do highway hands-free driving, right? So, you know, not in urban areas or suburban roads, but like on the highway, you'll have like a nice Level 2+ system. T he cost of surround ADAS is probably half of what automakers are putting in their current like Level 2+ systems, which haven't really scaled.
I think that makes it an easier sell as well because you know, you're taking a theme that the automaker wants to participate in, GM Super Cruise, Ford BlueCruise, BMW Drive Pilot, and you're basically cutting the cost in half. I think the purchasing process is more normal. It doesn't require like a lot of kind of strategic, you know, thinking. It's going on many different vehicles. It kind of increases the focus on execution, which is kind of a benefit to us because it's like if you need this to happen, you know, Mobileye's already done, you know, a dozen program executions for you. So you trust us, and yeah, so what we're seeing is for most automakers, this is being targeted for about a third of their annual volumes.
Wow.
and for a couple, it's being targeted at close to 100% of their annual volumes because they wanna make a splash and have kind of, you know, highway hands-free standard on all vehicles, similar to what BYD is doing.
Wow. And then last one for me, and then I'll open up to the audience for questions. We have to end on, you know, full Level 4 and Robotaxi. To your point, the acceleration here, you know, has felt very, very real. Launching 2026 into 2027, the two public programs, correct me if I'm wrong, VW, MOIA, and Uber, you can talk about those cities, Lyft, Japanese Marubeni, and an undisclosed OEM. Could you just put some detail to, you know, when, what cities will we see in? I think it's five or six. It's not, it's some of the traditional ones, but some ones that no one else is in.
Mm-hmm.
But you know, this is, we're gonna get to see you on the road in a year.
Yes. Yeah. So I think that we, you know, we have about 150 vehicles kind of testing on the road today in various cities, and those are used for kind of data validation and also to kind of really understand like where failures happen. So if you can kind of get to a bell curve where, you know, 80% of your failures are in four different categories, you can really start to kind of tailor your software to attack those issues. And that's what we've been doing for the last two or three years. So we feel really good about, you know, our ability to get to the precision levels to allow us to remove the safety driver. Our main kind of partner in this business is Volkswagen.
So for the last, you know, couple years, we've been working on integrating our self-driving system into the VW ID. Buzz vehicle, and this vehicle will come with redundant steering and braking. So it's, you know, they're creating like an AV ready.
From the start.
Car from the start. There have been kind of a glide path of kind of mean time between failure performance, you know, every couple of months you're kind of needing to hit new milestones. We've been on track with that. And essentially it gets to the point where you're comfortable removing safety drivers in the middle of next year. Our kind of business model strategy is we would, you know, we would like to isolate ourselves as the self-driving system provider 'cause we feel like there's only gonna be two or three that can actually.
Yep.
You know, stand this up and there will be value there. We don't see ourselves as really having the skillset of like owning cars or maintaining cars.
The network.
The network. So what we're trying to do is basically put together supply and demand. You know, supply we would like to come from normal automotive factories to create that sort of automotive-grade validation and kind of scalability of the product.
Because you wouldn't be an upfit, right? We talk about upfitting in trucking. Like the idea is for some of this to come off the line.
Yeah. The VW is basically gonna be produced on the same line as like a normal ID. Buzz that you could buy. So, you know, like I said, so you can scale up and down sort of at will, more efficient, you know, easier to control the manufacturing process. And then the demand side, you know, I think it's like obviously like a ride hail network has a lot of users.
And so, you know, Uber and Lyft are, you know, good customers for demand generation, you know, but you also got other use cases. There are like our tens of thousands of kind of shuttle buses running around the U.S. There's, you know, humongous public transit systems in Europe that are facing driver shortages, high cost of drivers. And so there, there's just a tremendous business out there, you know, four or five trillion miles driven in those two markets.
Particularly in Europe, they seem very open to the idea of, like, the, you know, the addition to public transit.
Yeah, exactly. And so I think that there's still a lot to be kind of figured out with the value chain. Like, you know, is the asset management, you know, VW would like to be that asset manager. You know, do they have the right skills to do that? They have a car share group called MOIA that has a lot of experience there. So I think VW wants to take sort of a somewhat end-to-end position.
Yep.
But there, yeah, there's an engagement with Uber where in LA we will, you know, the VW vehicles will insert into the Uber network in LA. You know, the target is, you know, kind of late in 2026. We'll do kind of a standalone market, you know, as kind of a test bed in the middle of 2026, driver out. There's, you know, multiple kind of European, you know, engagements, for, you know, launch late 2026, early 2027. And then with Lyft, you know, we have a relationship with this Japanese fleet manager called Marubeni. They've kind of made a deal to kind of own and operate the vehicles, insert them into the Lyft network in Dallas. And we're looking to kind of pick an automaker for that soon.
So they'll do both. So they'll do both the ownership of the fleet and the maintenance.
Yeah. That's the idea so far, but like I said, I think we're really early stage in terms of kind of business model and value chain development, but our main focus is the technology and, you know, kind of separating ourselves from the pack by removing the driver and commercializing. I think once we do that, we have a lot of scalability advantages. Like the vehicles are being produced on kind of a normal assembly line. We have the kind of crowdsourced map. So we have, you know, more data than kind of anyone in the industry has. So our ability to kind of expand geographically is very high. You know, yeah. So I think that's. We feel good about kind of our positioning there.
I wanna open it up to questions from Mark.
Where do you think we are in terms of regulatory willingness, local municipalities, you know, willingness to enable, allow to go out, outdoor? Do you feel like that's easier and easier every year? Do you still make the road less?
Yeah. No, it's, I think it's a really good question. I think in general, there's like a lot of benefits you can point to here to get kind of local municipalities like engaged, right? You know, better access for lower income people or people with disabilities, you know, more safety on the roads. Like, you know, there's clear kind of benefits to this technology. And, you know, I think local municipalities wanna be innovative. And so, yeah, we're seeing kind of an open playing field. I think the permitting and certification in the U.S. is somewhat unclear. It's sort of state by state. So certain states are, you know, basically kind of come in and do whatever you want. Certain states are very strict about it.
But we're seeing like kind of really good signals from the Federal Government about creating like a federal framework that would kind of preempt all of that. And, and so we're kind of working with our partners to, to try to foster that. So, you know, the regulatory environment in the U.S. looks good. In Europe, it's tougher, right? Because you actually have to kind of certify these vehicles through,
Wow.
You know, through a regulatory body. And you do it with real data and validation where in the U.S. everything is sort of self-certified. But we kind of like that because you know, then you know what you are targeting, right? Like you know what the hurdle is you have to cross. And if you cross that, it gives you some protection, you know, when inevitably there is some issue or some crash, right? Because you can say, you know, I kind of did everything that I was supposed to do. And it makes it more difficult. So for example, we won't be able to remove safety drivers in Europe in 2026 because this certification process is gonna take about a year. And VW believes that they'll be ready to start doing that at the beginning of 2026.
But like once we get to the end of 2026, you know, we don't expect anybody else has done that work. You know, VW has great contacts there. You know, we're ready to go. So, hopefully that answers the question, but you know, we see it as pretty supportive.
Any other questions from the audience? Dan, maybe you could just go over. You didn't hit the cities that you'll be sort of targeting for initial deployment over the next, you know, year and change.
Yeah. So the initial kind of driver out, what I would call like kind of a small scale, will be in Florida. We haven't talked about the exact c ity yet. We're testing in Austin today. The second commercial deployment will be in Los Angeles, with Uber. Dallas with Lyft should follow that. In Germany, we have, you know, good sort of support for Hamburg, Berlin, Munich.
Mm-hmm.
You know, well, Oslo is another city that's kind of very aggressive in kind of wanting to be like a, you know, a hub for this type of technology, so there, there's a lot of opportunities, you know, and, and I think, you know, with kind of the removal of the safety driver and the commercialization, you know, lots of other opportunities will come, but I think that that's probably enough on our plate for now.
And I believe some of the unit numbers for the size have been out there. You've mentioned before for, you know, VW has talked about targets and things like that.
Yeah. So VW, like the initial engagement is for 12,000 units over, you know, kind of four or five years. You know.
Not a small number.
It's not a small number when you're talking about, you know, $40,000-$50,000 of kind of revenue per unit and, you know, around $0.20 a mile, from kind of the operating revenue of the, the vehicle. Now we also have, you know, form factor is gonna be important. So like the VW is a four passenger, right? Like we're hearing of kind of big demand pools for more like 12- 15 passengers, like maybe think.
In Europe.
Yeah, in Europe, thinking of like replacement of kind of fixed route buses.
Yes.
But even thinking of, you know, airport shuttles.
Shuttles.
Like resort shuttles. And so, you know, Holon is a division of Benteler that we're doing kind of the same thing with VW. They're creating a 12-person, 12-passenger van with no kind of human driving controls. And that'll launch maybe six months after VW, but there's a lot of opportunity there as well. Yeah. So.
Could definitely use that for the Ryder Cup over the last, last weekend coming in the U.S. Time for one more. Hey, going, going.
Thank you. When the Chauffeur launches, it's like a highway only for the Chauffeur part. Do you intend to expand that to decide what kind of system they're gonna launch?
Yeah, that's definitely the plan. We don't really think it's necessary for the first generation of the product, but yeah, I think that that's, that's definitely part of the plan. You know, whether people use it in urban areas is like kind of an open question. Like in China, there's some urban, it's, they're not L3, but they're L2+ , but it doesn't look like people are really using it 'cause it's, you know, the, you know, there's a question of like, you know, are, are you comfortable enough like in a chaotic environment to use it? So we, we feel like the, the highway use case is, is the right place to start.
Okay. Well, with that, everyone, let's give a round of applause for Dan.
Thanks press .
For Mobileye. We'll be back in about 10 minutes, for Aurora.