All right. Welcome everyone back. Chris McNally, Head of Global Auto and Mobility, Evercore ISI. With me, good friend, longtime friend, Dan Galves, Chief Communications Officer at Mobileye. Dan, you always do a great job of setting the stage, so why don't I kick it over to you for just a little overview of where things sit now with Mobileye, and we'll come back. We'll have a little chat.
Thanks a lot, Chris. Yeah, it's always great to support this awesome conference. Tons of great companies here and great investors. So understand that there's a lot of questions around Mobileye right now. Many of the SuperVision partnerships that we thought were lined up didn't really materialize, at least on the first round. That's not new, but it's just recognizing that. And some of our recent announcements and transactions probably brought more questions than answers. But if you dig deeper, there are a lot of reasons to be optimistic. The core ADAS business gained a lot of strength this year. We'll grow volume approximately 10%, based on our July guidance, up from our original expectation of 4% growth. Within that 10%, it's basically 3% related to China exports, 3% related to safety stock increases that happened in Q1, and 4% growth with our top 10 customers.
That last number is, I consider it very impressive because it compares against these OEMs' global production growth of -5% this year. So in a typical year, we would think that we would grow maybe 4%, 5%, 6% faster than our top 10 customers, including things like China exports. This year, we've grown nine points faster than the top 10 customers, even excluding this China export growth. So I think that's really strong performance. We're also clearly winning in the ADAS segment that's going to grow the fastest over the next several years, and that's the highway hands-free Level 2+ segment. True point-to-point, hands-free systems like Tesla FSD get all the PR, but the real volume over the next several years will be for less ambitious but much less costly systems.
These are also popular with consumers and will represent the fastest-growing segment in ADAS over the next few years, according to S&P. About 18% CAGR over the next two or three years. We're winning this segment. As for the point-to-point system, similar to Tesla FSD, again, that's what gets all the PR and what OEMs want to highlight at Analyst Days, but there's still only one real system on the roads of North America and Europe, and that's FSD. And that will be the case for the next couple of years, except for our SuperVision system with Porsche, which launches next year and creates a really strong proof point for the market, we believe. We haven't won the first round on point-to-point, and that's disappointing to us and our investors.
But until there are more systems on the road and we get the opportunity to show Porsche vehicles with SuperVision at production scale, nothing's really decided. On robotaxi, our partnerships are making progress. We now have closed user group testing with the ID. Buzz through MOIA in Hamburg and Orlando. The L.A. test fleet is expanding. Our other partner, HOLON, just announced that they will be the production AV partner for Deutsche Bahn, which is running a broad initiative to integrate AVs into the public transit system in multiple German regions. The underlying dynamics and tremendous potential of the robotaxi business has really firmed up over the last couple of years, but we still consider it really the first inning. Our technology has progressed to the point where we have high confidence in deploying a scalable driverless system very soon.
This led us to initiate our own vertically integrated robotaxi platform and service, and we're making progress there. The opportunity to define everything about the service, from the sensor stack to the customer experience, to iterate software faster, to have control of costs and pricing, to have control of when and where we can demonstrate the technology, it's the right path, and we look forward to disclosing more on progress soon.
Excellent. Dan, why don't we start with the basic building blocks, right? So on the ADAS update, we're running on, I think, always in absolute units, right? Like 39 million units for almost like the last four to six quarters. It was better, that 9 million - 9.5 million per quarter that we always talk about. We have the China exports, which has been surprisingly positive, but can you talk about some of the other dynamics that have, in this very stale, sluggish production environment, what is starting to push that number higher?
Yeah. I think that, obviously, having the domestic Chinese market weak, where we have relatively low share with our customers there, has pushed the China OEMs to get more aggressive on export units, where we have better market share, particularly with Chery and Geely. I think what's been interesting is this is not all Europe.
Yeah.
In fact, Western Europe is a small piece of this. So it's markets like Russia and Brazil and Malaysia, Australia. These are markets where we don't have strong business. That has low ADAS penetration. So I would say that this export growth is incremental volume for us. What we think is also happening is that them bringing vehicles into these markets where ADAS penetration is low.
It's stimulating.
It's kind of shaming—
Yep.
—the legacy OEMs into standardizing ADAS into these markets. The data bears it out, about 1 million, 1.5 million units of sort of emerging market ADAS adoption growth this year. According to S&P, we think that that is going to be revised higher. I think that this benefit continues into next year because the ADAS penetration in these markets is low. I think that is probably, we also have market share gains within GM, with VW, with Honda. I think the bigger factor is this kind of emerging market growth.
From a competitive set, if we analyze, okay, there are still Chinese cars. Chery and Geely, we know BYD is its own exception. Why is your position so much stronger on export? Can you talk about that safety standard once you leave China?
Yeah.
Or how you can explain it?
I think in Europe, testing criteria are super stringent, and recall rules are very challenging. So the risk of putting a system on the road that doesn't perform well to save $10, it doesn't make sense. I think that in markets like Russia, where you don't have the same type of regulatory environment, you still want a system that's not going to slam on the brakes at 60 mi an hour. Because I think you're pitching, as a Chinese OEM, you're pitching tech.
Yeah.
You're pitching these vehicles are more full of tech than what you're used to, and if that tech doesn't work right, then it's reputational risk. So we think that that's really the factor that—
Yeah.
—allows us to price at a premium.
Obviously, China is China. Design times are under 24 months. Your ability to keep that momentum, you think, in export premium markets, do you feel pretty good from what you're seeing now in terms of your conversations with customers and whatnot?
We feel pretty good because all of these models are able to be produced with a competitive system. Because the Chinese OEMs are very good at multiple sourcing, even within the same platform. The same car that is going to Brazil with our ADAS system is being sold in China with someone else's. I think it is a clear choice of them. I think as time goes by, Chinese OEMs are going to have to localize production—
Yeah.
—more in the regions where they are selling. That, I think, takes away a decent amount of the cost advantage that maybe some—
Yeah.
—of our competitors have. I think it is sustainable, but it is hard to tell. We stay relatively conservative on it.
What we hear, not just Mobileye, is that typically when we hear about a BYD localizing in Europe, it is mostly using Western parts.
Yeah.
That is great.
Yeah.
If we play that out, just to finish the thought on core ADAS, growing that to that steady state ADAS number, we always used to talk, a number theoretically in the low 40 millions for Mobileye. We should think about finding a little bit of share gains, but we should start to think about some of this 20 million or 30 million market, rest of world, that has basically very low penetration. That would be the last couple of million incremental units over the years to come.
I think that that is right. There is not a lot to get in North America or Western Europe anymore. I think also, we have faced a steady headwind of our top customers shrinking in terms of production every year. We will likely face that the next couple of years. On the tailwind side, India is a market that is not driving a lot of growth this year, but probably has the biggest opportunity in terms of number of units in the future, and we have very strong position with Mahindra and Tata there.
That should start, there are new regs in 2027 that should drive adoption higher. India will be a tailwind. Continued market share gains with some of our customers, continued ADAS adoption growth in emerging markets, counteracted by this global production headwind that we typically face. I think we can still grow, but I think you are right. Getting well beyond 40 million units is probably not that easy.
Let's walk up the staircase to Surround, Cloud- Enhanced. I think you're now at three major Surround OEM wins. But really, I think the most recent, the Stellantis Cloud-Enhanced. Wayve was discussed on one of the small point-to-point programs. But I think you've made the point that a large portion of the volume now can be this Cloud-Enhanced and can work on the EyeQ6 L. Can you just flush that out for the opportunity for more to go take basic ADAS than go to this Cloud-Enhanced and what that gets them?
Yeah. I think that supervised hands-free driving, otherwise known as Level 2+, we see it as really two different segments. One is point-to-point, like you get in your car, plug in the address, the car drives you there. Doesn't matter, most weather conditions, all road types. This is Tesla FSD. Then you have a segment that's hands-free, but really isolated to highways.
Highways.
Okay? Which is still a valuable product because that's where people do most of their driving, and that's where it's the most boring. Allowing the car to take over has value. I think the L2++ market, this point to point, there's really one product in North America and Europe. We don't see that changing much over the next two or three years. Everything that's been announced is for 2029, 2030, besides our SuperVision launch with Porsche next year. Okay, let's put that segment aside. Highway hands-free, it's about 2.5 million units in North America and Europe combined today. But what are those systems? Those are systems that were sourced five or six years ago to multiple suppliers where the OEM does the integration.
Yeah.
Mobileye is on most of those, but just with a basic front camera system. Sourcing to multiple suppliers, each of which is bringing their own compute, their own software, not really talking to each other. It leads to cost that's too high to really scale, and performance that is maybe a little lackluster.
Those are your Super Cruise, your BlueCruise that have been around for a bit, haven't. Yeah.
Exactly, and I think that they're good products, right? But if they're going to cost $1,000 or more, are you going to really put that on a $40,000 car and charge $25 a month?
Yeah.
Probably not. It's too high. Every legacy OEM is deciding where they go with this segment. I would say that there's been two separate strategies. One is to do a completely new program that has multiple cameras around the vehicle, and one is to upgrade a base ADAS system with software to get to hands-free capability. We won the first two major new programs, with VW and with a large U.S. OEM. We lost BMW. BMW has a lot of capability on their own. They've been working with Qualcomm. They're continuing with Qualcomm. Let's say we've won two out of the three of these sort of multiple camera, relatively high processing power inside the vehicle, but vertically integrated under one supplier, so relatively low cost. For example, with one of our programs, we're reducing their current cost from $1,200 to $400.
That's a huge benefit to them, and we believe the performance is going to be very strong. Then, on the upgrade side, Stellantis wanted to upgrade their current EyeQ6 Lite program. Sorry, back to the first point. This is what we call Surround ADAS. It's five or six cameras around the vehicle, multiple radars, one EyeQ6 High as the processing power, and we do all the software. We do the parking. We do the driver monitoring. So fully vertically integrated. It's 2x to 3 x the ASP of our base system, at the same gross margin. Stellantis wanted to get something on the road next year, so they decided to add our REM road intelligence as a software add to the current EyeQ6 Lite program. This is essentially—
Like a very easy add-on.
—a very easy add-on. It takes ASP from 50 to about 90. Cost stays the same. Surround ADAS versus what we call Cloud-Enhanced ADAS, ASP is lower, but gross profit per unit is pretty similar. We've won that. There's another large OEM that we feel like is also won. Now, you basically have five of the top 10 customers there. Four wins, one loss. With Ford, we added driver monitoring, but there's no big change in BlueCruise coming.
Yeah.
We'll see what happens there. That's number six. Then you've got four others that we feel like one or two of these solutions.
Awesome.
One of the two solutions is pretty likely. We'll see, but I think that this part of the business, growing 18% CAGR over the next two or three years, we are growing content within the vehicle is a big benefit to the business.
We'll see that in the ASPs.
Yeah. It's not transformative, but it's a big—
Gets the ASP up.
It's an important growth factor to solidify gross margin, grow ASP.
Yeah. To continue up the staircase, we talked a little bit about SuperVision. How do we think about, obviously, probably the window through 2028 somewhat closed. Do you think something like Porsche on the road becomes an advertisement as, hey, this is something that looks like FSD Premium product, and we always have this, when OEMs are going to get FOMO? FSD has been on the road for 10 years. It clearly hasn't really done much for driving a lot of the other OEMs.
Yeah.
When do you think that point to point, you get where OEMs realize that maybe this is a point to sell cars?
I think the FOMO is starting to come, because now Tesla's talking about 50% +—
Subscribers.
—of new sales are paying them, what is it, $100? It's $100 a month.
Yeah, $99 a month.
Yeah, $99 a month. That's a nice piece of business, and it's much better than two or three years ago. You had data points of 3% or 4% of free trials would convert. I think that that's a positive. I think that we are considered a turnkey solution. Okay? For advanced cutting-edge innovation, that's not really what OEMs want. I think that our view was, they've already tried it. Things didn't really work out that well. The next phase would be with us. I think that that's been the case with the Highway Hands- Free. With point-to-point, not, right? I think that the OEMs are favoring multiple supplier solutions where the OEM is the integrator and bring some of their own IP—
It could theoretically be an evolution to the larger play. Yeah.
—for the process. I'm not trying to make excuses. I think that this is what's happening, right? So where the volume is going to be, the OEMs are making decisions based on cost and probability of execution. With the point-to-point systems that are not high volume, they're making decisions based on, how can we develop our own IP? We can manage the risk by allocating this to one or two models a few years from now. If it works, you expand it. If it doesn't work, maybe there's a fallback.
Yep.
We've already started to get some calls about fallback programs. I'm not saying—
Which has been the history of the industry.
—which has been kind of the history. I think the Porsche launch is really huge because I think in our minds and probably in some investors' minds, it's like Mobileye is the gold standard, and they've always executed, and 250 million vehicles on the road. I think it's not a foregone conclusion to the OEMs that we'll be able to execute a kind of Surround—
Next- gen. Yep.
—next-gen system that competes with FSD. When we get the Porsche system on the road, it is going to be the ultimate proof point that we can. We do think that that is an important catalyst, not just for our revenue, but also for our reputation within the industry.
Dan, can you just level set the order of magnitude for what maybe just specifically the Porsche program could be? Because obviously the logical question is you have Audi, VW, essentially, mostly all in with Mobileye. Audi also having a Chauffeur program, which would love to just get a quick timeline update. But the ability or the numbers around SuperVision at Porsche and the ability to—
Yeah, I think the agreement, the number of models still points to 300,000 - 350,000 units per year.
In a couple of years.
In a couple of years, right? Not in 2027. I think 2027 can represent a meaningful uptick to our SuperVision volume—
Yep.
—which this year is in the mid-50s or mid-50s - 60. And a nice tailwind to revenue growth.
Okay. And then update, obviously, Level 2+ SuperVision, the dream, the autonomous, right, is at the end of the day, fall asleep in your car, right?
Yeah. Mm-hmm.
Amnon was against L3, L4, right? It's eyes truly off, OEM—
Yeah.
—is the driver. That really should be the Audi program in 2028. Can you just give a quick update on Chauffeur?
Yeah. I think, good progress, on track to launch, but we still got a lot of work to do. Year and a half or so until launch. Yeah.
Call it the beginning of 2028.
Yeah. Q2.
Perfect. All right. We've gone up that staircase. Let's talk, robotaxi. I've seen a lot of the updates about L.A., a little bit about Florida launching. MOIA's doing, I think, public rider testing with safety drivers as well in Hamburg. Orlando, right, commercial by year-end. L.A. is testing now. I want to get a definition on when we become commercial. Yeah, just let's talk about robotaxi in the launch cities.
Yeah. No, I think that the current programs are progressing well. Like you said, closed user group testing in Orlando started a couple of weeks ago, with the ID. Buzz vehicles, with safety drivers. It's been ongoing in Hamburg for a couple of months. The other target cities are Oslo, Berlin, and L.A. Yeah, L.A. is testing in Santa Monica and West Hollywood. Not a huge amount of cars, and the target there is to begin closed user group testing. Don't hold me to this. I think MOIA said by the end of the year—
Yeah.
—but I do not want to really speak for them. I think that that is going well. HOLON is a 15-passenger shuttle that people do not talk about too much. BENTELER Group, who owns HOLON, is a very well-capitalized, deeply connected European industrial player. They announced that, or really, this KIRA program, which is a kind of a multi-government entity, public transit, Deutsche Bahn-led initiative for the last two or three years that we are using our NIOs for testing.
We will start using HOLON, and they are really looking to integrate that into the public transit system in multiple German regions. That is exciting because it has got the chance for a lot of miles, but it is also something that is very much outside of the traditional ride-hailing system. I think the current programs are going well, but there is a lot of constraints around this. You are not in control of the sensor stack. You are working with hardware that is probably a year and a half old, right? You are not able to upgrade. Things get frozen early in the process when you are working with automaker—
Multiple parties.
—programs and multiple parties, and it inhibits your ability to iterate faster. It inhibits your ability to reduce costs. It inhibits your ability to price and negotiate on your own behalf. This, plus our confidence in the technology, plus the fact that robotaxi is really starting to solidify in terms of a real business, led us to—
Yep.
—our standalone system, our standalone efforts, which we are making a lot of progress. Luca from my team is attached to the working group, which is very large and working 24/7 to scope out the program, source the vehicles, make agreements with upfitters, and essentially define the full customer experience.
Yeah. Can you just flush that out for investors? I think you said 100 - 200 vehicles in one U.S. city in 2027, scaling, I think you said 17,000 over five years. How vertical does Mobileye want to go? Is that all the way to a fleet operator, meaning you are going to be your own MOIA in this? Where does that also play with, would you be a part of a 3P network like Uber—
Yeah.
—or Lyft? Where does Mobileye end and—
I think the real key change is working with someone like Volkswagen to integrate the system on the actual ID. Buzz production line, it helps a lot with scale. That means if you needed 10,000 units in a couple of months, you would get them. But it also, like all those things I talked about before, you're sort of working under the requirements and the constraints of the OEM. If we ship, we'll have vehicles shipped from an automaker without a self-driving system to an upfit facility, and then we will be the owner of that vehicle coming out of the upfit facility. That is the key change here, because if you do that, then you create a lot of optionality about how you deploy those vehicles.
Yeah.
You can sell them to a scaler. You can lease them to a scaler. You can deploy them in your own service. You can deploy them in someone else's service, as the owner—
Yep.
—of the vehicle. We're not saying that we want to go head-to-head with Waymo—
Yep.
—in a complete vertical service, but we feel like standing up a complete service, defining the customer experience, and showing it running driverless, puts us in a much better negotiating position with those scalers.
Yeah.
I think that's what we're doing, and we think we have the assets in place. Like Moovit is a division of Mobileye that has done multiple proof of concepts with European players on apps and fleet orchestration and tele-op support and things like that. That's really the plan. I think I wouldn't think of it as we're limiting ourselves to one business model.
Essentially, the vertical integration part is you're bringing the vehicle—
I think the bring—
—more to your own control.
Yeah, and you're taking the liability—
Yeah.
—for the vehicle, right? Which is a lot of risk—
Yeah.
—but it's not something that a lot of entities want to do.
Yeah.
Especially if it's not their system.
Yeah, no. Everyone wants to be asset light, but who is going to end up owning these vehicles?
Yeah.
Again, in the beginning, you could own that CapEx, but like you said, it gives you the ability to also have fleet operators and other people who can actually end up buying—
Yes.
—the vehicle, taking it on as an asset.
Yes, exactly. Even if you sell the vehicle or lease the vehicle, I am not saying we want the liability, but it is important to have a home—
Yeah, embedded.
—for the liability. You are standing behind this product, and if we are not in control of the upfit process and defining the specs of that, then who is going to take that, right?
Can we just address a little bit about the exclusivity of some of the MOIA arrangements? MOIA relationship with Uber, I think, in L.A., but obviously they are still a relatively small part of a VW organization. The ability for you to work with other fleet operators. You can just talk a little about that.
Yeah. I think even MOIA is working with Beep, for example, in Florida, right? For maintenance, logistics support, a variety of other things, they are collaborating more directly with Uber in L.A. They are collaborating directly with these public transit groups within Germany. There is multiple business models or paths there. There is no exclusivity there. Our vertically integrated robotaxi initiative has no constraints relative to Beep. We probably wouldn't enter the same market—
Sure.
—as them. There's no constraints.
Then the last question from me, then I want to open it up to Q&A. I don't know if you heard in the intro remarks, we are starting to see these real-world bottlenecks, the charging, the depots associated when we start to scale cities. Within the internal fleet, how is Mobileye thinking about that physical infrastructure layer of owning depots, charging, figuring out where the fleet has to be at a given time?
Yeah.
Not the software problem.
Yeah.
Now you're in the Uber business problem of large maintenance and things like that. How is Mobileye thinking about that?
Partnership. Yeah. I think we definitely feel like there are multiple entities that have an interest in participating in that way, and we do not have local physical assets really anywhere besides Israel. I think that that is definitely we are looking to partner there.
Okay. Excellent. Open it up to Q&A from the audience. Anyone? Yeah, Mark.
Can you talk about going through thousands of vehicles, just talk about the, what do you think of the biggest governors to the growth in terms of the number of vehicles you can get out there, and just talk about the cost curve, too?
Yeah. I think that there's a concept of an AV-ready vehicle platform, which doesn't mean that it has cameras and radars and lidars around it. It means that it has redundant braking, redundant steering, certain microphones that are required. It's not like you have to create a completely new vehicle platform as an automaker, but it's also not trivial. It's not like you can just buy 10,000 of the same vehicles that you or I buy from a dealer. The OEMs do have some engineering work to do, so you end up needing to commit some level of capital and volume to get those vehicles. So that's where we are right now.
We think we'll have vehicles by the end of the year, maybe a little bit into 2027, and have an OEM that is designing a vehicle specifically for us that could be produced in the thousands of units a year. That, again, it wouldn't be a complete separate car, but an offshoot of an existing platform. Exactly what Waymo does today. So I think that that's the vehicle supply. So we don't think volume would be constrained on an 18-month to 24-month view, but it might be constrained to hundreds of units on a one-year view, let's say. The cost curve, yeah.
I think reasonably, this is one of the advantages of being in control. I don't want to say too much, but we feel like with our own system and being in control of the design of that system, that we can come to a BOM cost well below $10,000 for the self-driving system. So then you're basically adding the cost of the car onto that. Yeah.
Which OEMs? You mentioned this, which OEMs do you think you're most likely to be working with?
We haven't said, but I think the OEMs that have been most involved in this are Toyota, Hyundai, and VW, and Ford to an extent, seem to be willing to do these sort of design tweaks to get you AV-ready cars.
Dan, it's pretty significant. I just want to make sure I understand. When you say that an OEM will be ready to produce this on the line, will it still need an upfitter? Because even Waymo is basically getting upfit by Magna. Maybe it's lighter, it's closer to the line, but—
Yeah.
—they're still using an upfitter.
Yeah. No, I should be clear about this. What I'm saying that before it even goes to the upfitter to install the sensors and the compute—
It will have the tweaks.
It has these tweaks.
Okay.
That probably cannot really be done on the exact assembly line.
Yeah.
So maybe you have a small offshoot at the end.
Yeah. It exists, right? This is what Magna is doing, dedicated facility for—
Yeah.
—Waymo. That is really exciting. Anyone else? Dan, in the last couple of minutes, I always try to ask where investors should focus the next three to six months, but I think ironically, as we went up the staircase, I think so much of the focus is on the top of the staircase for robotaxi. Let us call it over the next year, what is the city that we may see the most amount of Mobileye people are getting in rides and actually seeing the technology and then working backwards and understanding that that may also be put into the car?
Yeah. I think Orlando is exciting as a test bed sandbox for driver out, right? That is probably the first driver-out city. L.A. is more exciting from a volume potential perspective. Europe comes later because of the homologation—
Yeah.
—and certification process, but may be the best opportunity to demonstrate unique use cases for these vehicles.
Yeah. The vehicles monopolistic, yeah.
Replacing big buses with small AV shuttles. It has the potential to really be a win-win for everybody.
Okay.
I think really that's it. We feel like we'll have more to say on our partnerships on the standalone robotaxi side.
Yeah, which we don't know the cities for.
Yeah.
Yeah.
Yeah. We haven't said. We do know. We know the first one, but we haven't said. I think you'll get some updates on the October earnings call, I would expect. I think, yeah, I think robotaxi is very exciting. We'll have a Mentee Bot at CES this year doing some demonstrations. I feel like that's something that is farther away in terms of revenue generation, but there's a lot of opportunity to demonstrate the capability there. I think we have a lot going on. We do feel like expanding beyond the sort of through the OEM business that we've always had and benefited from is important. But that creates other skills that are required within the company, which I think is one of the reasons, besides just age, that—
Yeah.
—Amnon wants to find a different skill set to usher the company over the next 10 years, and we feel good about the interest that we've gotten and the shortlist that we've developed. More to come on that. It's probably closer to end of year type of thing.
Yeah.
A lot is coming up towards the end of the year that we feel good about.
I don't know what CES will be without Amnon's annual presentation.
I think he'll be there. Will it be like a handshake transition? We'll see.
Yeah.
We'll see.
Excellent. Dan Galves from Mobileye. Everyone, round of applause.
Thanks a lot, Chris. Thank you.
Thanks.