Good morning everyone, I'm Brian Drab.
Welcome.
Thank you for joining us. Delighted to have Mobileye here to take us through their story.
It's an exciting time.
We've got amazing amounts of innovation in autonomous driving functionality.
We've got cross currents and tariffs.
We have a really big interesting product cycle that's up and coming for the company. To take us through the story, we have Dan Galves, Chief Communications Officer, who knows the story as well as anybody and really can help us get to the bottom of it, so Dan, thank you for joining us.
Thanks, Brian. Great conference.
We're going to do this in a fireside format. Should you have any questions, raise your hand and we'll try to get to you as we go through the presentation. Dan, why don't you level set the audience. A lot of people are familiar with Mobileye but maybe take us through a brief history where you're different than others in this space and how you go to market and win.
Yeah, so the core of our business is software that's deployed on an in-house developed chip, so it's basically a SoC that we sell into the auto industry to support what's called Advanced Driver Assistance Systems. Kind of the main function of these systems is to create an extra layer of safety to avoid collisions. The main driver of the business from the automaker perspective is that in order to get 5-star safety ratings you need to meet specific requirements for this. The type of technology that we produce under real tests in Europe, North America and a variety of other markets, and we've shown the ability to kind of get the best results on these tests and that's what's kind of helped us to maintain a really strong position in the industry.
Like if you look at our top 10 customers which generate, you know, most of our volume, we, you know, our systems are on about 70%-75% of that volume that's produced every year. Those top 10 customers are about half the auto industry. We do over 30 million chips a year.
You know, 70% gross margins, about $50 selling price. Over the last like six or seven years we made the decision to develop autonomous driving technology. So really the goal is to create kind of an overall tech stack that would support a system that doesn't need any intervention from a driver. Could be no driver at all for both the kind of the robo taxi market which would be a set of cars deployed in fleets or also for privately owned vehicles. And the way that we've built the tech stack it gives us the ability to use certain building blocks to create lower functioning systems that are not fully autonomous. And that's really kind of where the opportunity comes in. The business is going from a $50 per unit single chip system to Surround ADAS, which is between $150 and $200.
Per chip across large volumes, to the SuperVision system, which is, you know, a hands-free system in all environments. But the driver needs to keep their eyes on the road, which generates about $1,200-$1,300 per unit for us, all the way up to, you know, eyes-off systems where the driver can fully disengage, which you know, is significantly higher revenue per unit. And I think that that's the, where the debate and the stock comes is kind of how quickly will these systems be adopted? When will we be able to announce new customers? So far it's really the Volkswagen Group that's endorsed this product set. We'll be launching SuperVision and Chauffeur, which is the eyes-off product in 2026 across a number of different vehicles. So that gives us a kind of a great starting point.
But we're working on, you know, many other design wins with other OEMs. And I think that that's really where kind of the opportunity of the debate on the stock comes in.
Completely agree. Maybe for this audience who hears a lot about other approaches, whether it's a Waymo or what Tesla is doing or some of their own OEM in-house solutions. How would you describe technologically where you're different and how your approach differs from those companies and then maybe also just weave in a thought on how artificial intelligence either creates a bigger moat or what that means for your business as well?
Yeah, so I think, you know, at the end of the day, like it's really all about accuracy, and this is kind of it goes back to the core business, like the reason why we're, you know, for most of our customers, like all of their ADAS business instead of, you know, in the auto industry, like multi sourcing to different suppliers is kind of the way they create, you know, the way they avoid suppliers having leverage. It's not the case for us because, you know, because of how difficult the performance requirements are, we're really the only company that's been able to demonstrate kind of consistent accuracy across these requirements. And that's based on kind of the accuracy of our computer vision software, the ability to interpret data through a camera.
We also have been, you know, through by being kind of the main player in this market. We've been able to generate like massive amounts of data that are used to retrain this computer vision network and technology, improve it over time. So this kind of data set that we have is a big competitive advantage. We've also kind of used our position in ADAS to create a situation where we have four million vehicles like you or me could be driving one of them that is consistently sending data to the cloud that we use to build a map. This high definition map becomes like sort of the memory of the autonomous driving system. Like when you come to an intersection, you know, if you are only kind of analyzing that kind of the structure of that intersection in real time becomes very challenging.
But if, you know, the last 10 or 15 vehicles to go through that intersection, what path that they took, which traffic light is relevant to the path you're taking, these types of things, it creates, you know, more robustness in the system so that we use that for the higher level systems. So, you know, I think we have a variety of technology moats that, you know, make us kind of the best, you know, the best choice for an OEM for advanced products. In terms of new AI.
We were the first to deploy a deep neural network, you know, on a production vehicle in 2014. It was actually with Tesla in Tesla's first Autopilot system. So we've shown an ability to integrate new AI technology into our system, you know, very rapidly in the past and that's kind of what we're doing now. So our EyeQ6 based system, which is what will support the Porsche and Audi launches in 2026, you know, makes significant use of transformer based architectures and you know, has, you know, much, you know.
The inference chip is kind of 10 times the processing power of the prior chip at the same cost. So we've also shown an ability to have like very kind of efficient, you know, inference chips. So yeah, I think we, you know, new AI really helps us to get to the accuracy levels that we need for these systems.
That's great. Before we talk about some of the newer products, which I'm really excited about, I just want to make sure that we're understanding the visibility in that base product line, that $50 chipset where you have such good market share, how much visibility do you have going out over the next few years in terms of unit volumes and design wins on that roughly $1.7 billion of revenue or so?
Yeah, so I think like I said before, like our kind of the bulk of our volume, you know, call it 90% of our volume is across 10 OEMs that are about half of the auto industry production. We still, with three or four of those companies, are not, you know, fully penetrated, you know, into their business. We've won some programs, you know, in the past that, you know, took competitors out. And so, you know, we're gaining market share within some of these customers. And that's part of the, you know, there's a clear tailwind this year related to that. We have new markets like India, which, you know, for reference, you know, in North America and Europe, maybe 90% of all cars sold have some kind of ADAS system on them. In India, that's 10%.
But we launched the first ADAS system in India a couple of years ago. It was sold as an option. The adoption rate was much higher than the OEM expected. And that created sort of a groundswell of interest in ADAS in India. So that becomes, you know, two, three, four million unit opportunity over the next several years. That's creating other opportunities with kind of, you know, other new customers. So, yeah, so I think the visibility from the perspective of kind of what percentage of each OEM's business will include products from us is very high. What's not so visible is kind of what the overall production levels of those 10 customers will be for the next few years.
Because that has been a headwind that these kind of legacy OEMs, particularly in China, have lost a lot of market share over the last few years.
Absolutely. I think the biggest shift in the auto industry since the 1970s when the Japanese OEMs came on scene and took massive amounts of market share.
Yeah, exactly. I mean, just to put it in perspective, in 2023, those top 10 OEMs produced 47 million units. This year, we expect them to produce around 41 million units. Numbers in China are getting much smaller. So I think that the headwind becomes less over time, but I think it's something that we've had to face as a business.
Absolutely. Want to talk about your advanced products now, but maybe set the stage here. BYD was in the news recently talking about God's Eye and bringing Level 2 plus autonomous driving functionality to a broad line of their vehicles. What does that mean for Mobileye?
Yeah, no, it's, I mean, I think, like we're in interesting times in many different ways. So kind of explaining this BYD announcement, because we do think it was really important. The product that they're talking about is something where once you get on the highway, you can engage the system and the car will basically steer for you, it'll control the speed for you and it will make some automatic lane changes.
It's not going to position your car for the next exit. It's not going to plan your route for you. It's comparable to a product that we call Surround ADAS. What BYD said is that we're going to put this kind of hardware and provide this system to customers for free, you know, even on vehicles that are kind of $15,000. It reminds me of an announcement that Toyota made in 2015 where they said that ADAS technology would become standard at that point. You know, the products that we were selling, like as the consumer, you would have to pay extra to get them. But Toyota basically said, we're standardizing this. I think we were like a three or four million unit a year company in 2014, 2015.
I think that that announcement, even though Toyota was not our customer, was kind of the core reason why our business is 30 million units and above now. At that time, like the OEMs, they had ADAS technology. All they needed to do was say, okay, it's going to be free now, instead of an option. It took a while for that to happen, but it happened. Surround ADAS or kind of. Today.
We look at this kind of highway autopilot, highway assist, whatever you want to call it, as like the next ADAS. And most OEMs do have a product like this, but the way it was sourced is that they sourced it to kind of five different, five or six different suppliers, each who would have a different role and kind of the integration would be done by the OEM. And so what you ended up with is decently performing systems. Like, I don't want to mention names.
Blue Cruise, Super Cruise, DRIVE PILOT . Like there's all these products out there, but most of them are four, five, six different ECUs, all from different suppliers that aren't really talking to each other. So you end up with too costly, not very efficient systems. This Surround ADAS product is a single ECU, you know, with parking software, with driver monitoring, you know, with, you know, analyzing five cameras around the vehicle, four or five different radars, you know, sensor fusion. It comes kind of all in one as a consolidated ECU product. It's very competitive, cost wise, and I think it hits the market perfectly. So we're, you know, this is a product that most OEMs, you know, we don't have any design wins that we've announced yet, but we have a lot of RFQs that are looking at you know, maybe 30% of an OEM's volume over time.
You know, this kind of an announcement by BYD could lead to this being, you know, much higher percentages of their volume. So I think it's a real seminal moment for our, our type of business.
That's interesting because when you first kind of reemerged as a public company, SuperVision was the primary product that people were looking for that was more of a discretionary where consumers were going to have to pull that. But with BYD kind of making this more normalized across their portfolio. And now also the regulatory backdrop, driving Surround ADAS, it seems to have a much more brisk kind of timing, tailwind component.
Yeah. And I think like we have to think in terms of, you know, SuperVision is a product where you get in your car in the driveway, you plug in the address and the intent is the car drives you there with very little intervention from the driver itself. Right. So it's a much different functionality than Surround ADAS. And I think it's always been kind of thought of that way by the OEMs where SuperVision is a system that would be on premium vehicles, you know, maybe 5%-10% of the volume, aspirational, you know, the consumer would need to pay extra for it. So it's a different function and it keeps you on the path to eyes off. The other thing about Surround ADAS is.
It doesn't really have the scalability to eyes-off that SuperVision does. So we see them as two separate products. But I would say that it's kind of a more clear path to volume. It's a more clear path to a decision to an award when you have a product that it does have a regulatory component because it's going to help you kind of keep getting five-star crash ratings. It's something where there seems to be building up a lot of competitive pressure to have a more efficient system. And so in that manner it's kind of a more of a standard purchasing driven RFQ process. So yeah, I think you're right.
Maybe. Let's talk a little bit about the macro and your guide. I thought on the fourth quarter call a month or so ago, you did a really nice job de-risking the guidance across, you know, production China. A bunch of other areas that I'd like you to explain. But with Tariff Tuesday in mind here today, also want to get your kind of thoughts on how tariffs initially impact that. It feels like you've left some buffer in the guide relative to peers. So maybe just unpack those things.
Yeah, so, I mean, we've had some negative surprises for sure the last couple of years since we were public company again, so you know, kind of a high priority for us as a company is to rebuild credibility on our kind of forecasting ability and kind of make, you know, take that off the table as a concern for investors as much as we can.
You know, essentially what we did was, you know, look at kind of the production expectations, you know, driven by third parties.
Do they make sense to us? Like, one area where it didn't really make sense is that the third parties were assuming these kind of legacy OEMs would be down like 7% or 8% in China, when the last couple of years they've been down more like 20%. So we made that adjustment and assumed that they would be down 20% again. That created basically like a million unit kind of reduction. And then, you know, compared to what the customers were telling us, like maybe we were 2 million units under in terms of kind of the midpoint of our guidance. So it does create a bit of a buffer and kind of understanding all the uncertainties that were out in the world. Like we wanted to have that buffer.
I think we're, you know, we're fortunate in some ways where we don't manufacture our products. Our product is, or we have a contract manufacturer. Our product is part of a kind of an ADAS subsystem.
So I think we are fortunate that we don't have to devote a lot of resource to figuring out, you know, what is kind of the path, the supply chain path of our chips. We know that. I think where tariffs will impact the entire industry is clearly in terms of incremental cost.
Potentially driving lower volume. So far we have not really gotten any queries from our customers about can you change where your product is produced. I think we're low cost enough that it doesn't really matter. There's bigger priorities. But I think to the extent that there's just general disruption in the North American industry, which I expect that there will be, you know, it could result in lower production volumes. We definitely have some buffer there, but we'll see what happens.
This is a market where the obstacle is the path.
Let's maybe talk about Surround ADAS. You have a handful of bids out there that you're working on. Maybe quantify the number. You talked about the penetration into those OEMs as being potentially significant. What could this mean for volumes in the out years? And when could we see this product begin to adopt.
Yeah. So I think what we're seeing, you know, yeah, we have five or six kind of RFQs we're responding to now. On average, they represent, you know, around a million units a year. Like, once the program gets to scale, like in the auto industry things, you start shipping product to specific vehicles when that vehicle launches. So if this is going to be on 20 or 30 different vehicles, it could take two or three years before all of those vehicles, like, get to their launch phase. So it doesn't happen overnight. That's basically the average that we're looking at. It represents about 30% of each particular OEM volumes.
So that's kind of the level, you know, we're pricing at, you know, $150-$200 per unit, you know, kind of similar gross margin to what we have today.
Yeah, we are hoping to be able to announce some of these deals in the near term. They would launch, you know, call it. Call it like mid 27 type of time frame.
These are roughly a million units each.
Yeah, for each OEM.
Okay.
Exactly.
Okay.
You know, that makes a lot of sense. Is there a cutoff point, time-wise, where if you don't have design wins, you know, it becomes a 28 or 29 thing?
I think it depends on the OEM. You know, there may be RFQs right now that would launch in 2028. So I think that, you know, I think in a lot of ways, like, we're trying to set ourselves up to create much higher content per vehicle and kind of much higher revenue. You know, towards the end of the decade is when we would expect things to be, you know, more at scale to really impact the business. It would be impacted before then. Like, you know, with Porsche and Audi launching SuperVision in 2026, you'd start to see the inflection point then.
Yep.
Speaking of, it's a good, good segue into SuperVision. How is the design work going for VW and how should we think about kind of initial volumes or units or whatever perspective you want to provide for us on that?
Yeah, I mean, the execution is going great, and I think it's like, it's generally underestimated by the market that the actual process of getting a product into production, you know, at scale, validated by the OEM is not trivial at all, right, and it will create benefits for other customers as we kind of go through this process, because you develop something for your own test fleet and, you know, it's, you don't have the same type of validation intensity and kind of, you know, basically getting put through your paces by the automaker that you're working with. The design really needs to be frozen by the end of this year to kind of meet the start of production date, and we're on track to do that.
So I think the other thing that, you know, we're key for us was like to kind of not be a black box to the OEM. Right, so to give the OEM control of, you know, the consumer-facing parts of the system. So, for example, it's like we're doing all the heavy lifting in terms of, you know, where are all the vehicles, where are the lane markings, where are the pedestrians, you know, integrating the map. So you have that kind of memory doing the decision-making, like in terms of safety, creating all of the safety layers within the system, but the OEM, and you know, that would benefit the OEM like from our expertise. But they don't want a system where it's going to act the same for them as for every other automaker.
So giving the OEM the ability to tweak and control and decide like, how the car reacts to different situations, the ability to get feedback from consumers and be able to change that, like, based on the feedback after the fact was critical. So we created this application layer called DXP and that's been really positive in terms of this execution phase with Porsche and Audi. So I think everything's on track and feeling good.
Nice to see some cars on the road with the technology.
Yeah, I mean, Polestar 4 is a system with our SuperVision technology. It's EyeQ5 based. So, you know, the system is not going to be as good, you know, as kind of the EyeQ6 what we're targeting for EyeQ6.
But that vehicle is being sold in Europe. It's really up to the OEM when they kind of turn on the system. That should be fairly soon. Polestar is trying to or has a plan to start producing that car in Korea. Right now it's produced in China and so they basically can't sell it in the U.S. like, but if they start producing it in Korea, it can be shipped into the U.S. and so, you know, I think we're. One thing we've learned is like, we got to get more investors into these cars.
Because when we did our analyst day in Munich, I think people were expecting, you know, kind of one of these, like, you know, western OEM type of systems that they've driven in and kind of what they found was, you know, our system acts a lot like Tesla FSD, which people have experienced a lot. So operating kind of on any road type, you know, kind of very assertive driving, like confident. So I think getting more people into the cars is something that we'll be able to do now that Polestar is being produced.
Excellent. Maybe let's peel out and look a little into the horizon. There's a lot of enthusiasm about kind of your fleet and robotaxi opportunity. Lyft's been in the news recently with Mobileye. Maybe just frame that opportunity in some of the mile markers investors should look for.
Yeah, so I think from a technology perspective, we really believe we're on a path to kind of a high functioning, high precision driverless system. But the robotaxi business, I mean in the consumer business, we need customers to adopt it, right? We need OEMs to adopt it. In the robotaxi business you need to kind of build an ecosystem because we don't want to own these cars. We're not experts in, you know, operating fleets of vehicles. So you really need partners. The Lyft deal is a good kind of example of what we see as kind of four layers. You need somebody to generate demand, right? And that would be Lyft. They've got millions of users. Like you could plug a robotaxi onto their network and generate demand very quickly. You don't have to build a brand name or it's et cetera.
You want someone to be the owner and the operator of the vehicles. That announcement mentioned a company called Marubeni, which is a Japanese fleet operator that, you know, is very kind of eager to participate in this business. So they would be the ones that would own the vehicle. You know, we would provide the self driving system so we would be the technology behind it. And then, but then you also need someone to produce the vehicle and that was not in the announcement. So that's still to be determined, you know, what vehicles will be used. The key there is like, you know, we're a company that's always looking for scale and making sure that we're able to scale.
In order to scale this business, you need to be able to have vehicles kind of coming out of normal factories with your self-driving system integrated into the vehicle. You don't want to be buying vehicles and then like sending them to some workshop and like manually putting the like all the sensors on. So we have a deal with Volkswagen to kind of do that. We have several other deals and kind of this Lyft agreement, it's still to be determined who will have the vehicle. But yeah, there's a ton of activity in this space. Waymo is really kind of the leader here. Like, they, they've focused on precision within geofenced areas. Like, you know, we feel like we have a big cost advantage over a company like Waymo based on kind of differences in the tech stack.
But we have not proven that we can remove safety drivers and actually like operate in this environment like we think that we will. But we're behind them in terms of the precision. So we'll see what happens. You know, it's. We feel like the TAM is definitely big enough to.
Justify the resources we're spending on that business.
Absolutely. Last one here and then we're going to have to go to the breakout session. If your business were to double in the next several years, right now you're at about $50 content per vehicle. How does your content change and what's the mix of products in this scenario where you double your business?
Business. So I mean, I'll answer it a little bit differently. At our analyst day in December, we showed a slide. It was based on Volkswagen because Volkswagen is kind of the farthest along in terms of adopting our portfolio. We do see the potential for, in terms of volume, for kind of 30% of the volume of an OEM like Volkswagen to be Surround ADAS somewhere around 10% to be SuperVision. The rest would be still basic ADAS like towards the end of the decade. What that would do was take our average selling price with any particular OEM from around $50 to like well above $200 per unit. So even if you're kind of, you know, delivering the same amount of volume, you're quadrupling your revenue with a particular customer.
That's really the goal is to kind of get adoption of, you know, to get all of our customers to adopt our technology and create kind of a portfolio within their own kind of set of volume.
That sounds great, Dan. Thank you so much for sharing with us. We'll adjourn to the breakout session. Thank you everyone in the audience for joining us.
Thanks everybody.