AEye, Inc. (LIDR) Earnings Call Transcript & Summary

We're going to go ahead and get started as people filter in from lunch and what not, hey, it's great to be back here again this year. Of course, we're very excited about LiDAR topic, and we're glad to see some folks here in the room as well. Standard disclaimers now that you've had time to read through those, we'll jump into the presentation. Okay, what I -- how I wanted to start this since I've been a bit out there in the press about that we really need LiDAR. Is it really a thing? I want to go back to grassroot for just about 20 seconds here and show a quick video. On the right, you have a camera that you might find in a vehicle ADAS system and on the left, you have a LiDAR system. And you note that the perspectives are the similar you got a guy there in the white pants and black jacket. We're going to have him walk into the shadows of the light. And you'll notice as he leads this streetlight, whatever is out there, it just disappears, and is not very far out. And you can see on the left-hand side is very clearly visible in the LiDAR image. So there is very much still an absolute need for LiDAR. There's tons if you've been following what's going on with NHTSA and Rule 127 for safety. We're not even talking about autonomy yet. These are the kinds of cases that NHTSA's looking at for pedestrian safety, avoiding crashes and things like that. They are just scenarios where cameras and radar can't fill in. And here's another one, which I'll speak to here in a bit, but this one actually came from an OEM. This is a tunnel in Korea. I can't say who the OEM was, but tunnel in Korea, where they're having issues where a lot of customers reporting that the cars slamming on the brakes. The automatic emergency braking is triggering coming out of this tunnel certain time of year, certain time a day, and they figured out that the sun shining into the cameras was causing a problem. Again, camera image on the right, same image in LiDAR on the left. You don't even see this on LiDAR. So again, we haven't spoken about autonomy yet, but really what's driving us first and foremost in the LiDAR industry is safety. And that's really our gateway into the industry. I'm not going to go through all the bullet points, but a lot of reasons why camera and radar even when combined, can't detect certain scenarios. And then certainly, when you get into autonomous driving, especially in the space that we play in at AI, which is long-range detection of objects. This is required when you're driving at high speed, absolutely. We don't see this going away. A little bit about our company. A few things, so first and foremost, we have a very solid financial foundation for the company and a lot of very capable partners starting with GM Ventures partnering with us on the financial side and what we're going to jump into here in the next few slides is, #1, a little bit about our technology and IP that allows us to do some things that others can't. I'm going to show a little bit of show and tell here, and then our unique business model, which is a licensing or think about it like a fabless semiconductor business model. And then last but not least, since we've been here last year, we've had a very significant restructuring of the company. We have a new management team in place and one that's very experienced at bringing new technology to market. So the company's transition out of an R&D phase into a product and get the product to market face. First, on our business model, again, which is a bit different. We call it Capital-Light. And the idea here, like a fabless semiconductor or a design house, we build a reference design. And our model is we partner with a Tier 1 in the automotive space and that we provide them with their reference design, they package and customize it for the OEM. This allows us to operate under a very lean cost structure. We've seen our earnings call on our website and the material there, our cost structure is up to 10x lower. You'll see coming up here, we burn about $5 million a quarter, which is quite a bit lower than others. And our architecture and our technology foundation allows our partners to differentiate by and large with software. So it allows a very fast design cycles. It allows us to essentially rinse and repeat with different Tier 1 partners like we did with Continental last year, and now we'll talk about LITEON in a bit. And so we're not having to invest this capital by inventory to build manufacturing lines. This is a very modest number from some of these Tier 1s perspective, they can absorb this. And this allows a very high gross margin structure for us. So goal for us, automotive safety as the entry point into the market, scale this with multiple OEMs and Tier 1s in the market and then into other markets. We believe that going into automotive is necessary to achieve the scale to drive cost structure globally, a digestible cost structure across the other industries. If you know LiDAR, you probably know the old reports, and they've updated this year. I'm not going to spend a ton of time on this, but China is where it's happening right now. So somebody asked me a question earlier about, well, I heard that LiDAR is going away. Well, in the U.S. and Europe, really the reality has just slowed down a bit. It takes a while to integrate a new sensor into all the software that's running in the vehicle. You've probably read about delays and whatnot in the industry due to software integration, but China is years ahead of where we're at here. We'll get back to this point. But we believe because of the scale of the ADAS platform, this is the right entry point for the LiDAR into the various markets out there. I'm going to spend -- I'm going to be a little bit more thoughtful here because we're also here today to talk about a new product. We announced it in China about a month ago at the LiDar conference in Suzhou. It's our next-generation product called Apollo. And here's where we talk a little bit about the IP that we have that's unique. You pretty much take a LiDAR prototype of part today out there from various manufacturers. It's got like mirrors that's been inside of it. We've taken a different path. And we've built a very small microscopic type mirror structure that's virtually solid state and allows us to do a couple of things that others can't. Number 1, is we can build something that's incredibly small. This is a device that can see 1 kilometer. And I just got back from 2 weeks at Detroit showing OEMs that, right? It's not just on the PowerPoint, we actually have samples that are out there with OEMs. And we're running test cases with OEMs under ADAS live in the car. We're seeing dimly lit objects like tires and black cars and things like that at over 300 meters -- kilometer max, 300 meters. This is 1550-nanometer LiDAR. It allows because it doesn't impact your retina because of the wavelength, we can crank up the power and see very, very far. We believe the furthest in the industry. And -- what else? The other part that this enables is software definability. When you have spinning parts and mirrors in your system, there's only so much you can do with software. This thing is very, I want to say, infinitely programmable. Just to give you an example, last week, we visited 2 different OEMs, and they wanted -- they like to use the LiDAR in 2 different -- completely different ways. And it's just 2 revisions of software for us. We use the same box. Not only that, but it allows OEMs to easily scale this across multiple vehicle lines. It's very small, first of all. So it's very flexible, but it also -- we can tone down performance or look at different scenarios, urban driving versus highway, low TCO is what I'm really talking about. You don't have to have lots of variant in the hardware across your vehicle lines. So -- we're getting a very strong reception. We believe for a device that can see 300, 325 meters in ADAS and over kilometer max range. Nobody else is this small, not even close. Last but not least, we've -- this is not our first raw deal in the automotive space, supply chain-wise. We did a ton of work with Continental for 3 years up until the end of last year. We have an automotive grade sample, and this is basically reusing 80% of the supply chain, most of the changes between our prior generation with Conti and now working with LITEON, an expert in optics. They're in industry leader in automotive in disrupting camera pricing. So 10 years ago, as cameras in ADAS systems or backup cameras are very, very expensive. They're super cheap now. LITEON was a key player in this market, and they know LiDAR as well. So we're very excited about working with them. We think we're going to see another 50% cost reduction in the price point, thanks to our work with them here in the next year. Okay. Can yes.

What's the key difference between [indiscernible] LiDAR technologies and so the new Radar technologies like new companies like [ Sadar ] how would you differentiate it?

Look, the -- let me just try to up level between the questions about the difference between LiDAR and Radar. The physics underneath is quite a bit different. And even -- so with Radar to get high resolution, by the way, high resolution is super important for ADAS systems. They've been wired for years to cameras. Cameras require high-resolution images. The more you can look like a camera from a digital perspective, the easier it is for the OEM. Radars are not good at high resolution because the wavelengths coming out of the radar system is much lower resolution than a LiDAR can do. And so you see blobs. Instead of distinct objects, you tend to see blobs. So yes, I know there's some guys who have come out and say, well, we have high resolution radar, it's still at least a few orders of magnitude poor than LiDAR and its physics. It really has nothing to do with the product pace. Big difference. Moving on, I think I talked -- so look, our supply chain strategy, I think the world is familiar with the geopolitical risks that are happening right now. So we have a dual-pronged supply chain strategy. One is global through LITEON. So for example, I'm out in Detroit last week with the Head of LITEON's North America businesses from Mexico. They also have a business based in Germany and in Asia as well. And then we have ATI, Lightekton in order to do business in China. You have to have local supply chain and boots on the ground to do it. So that -- we have this dual chain inside and outside of China. Some of the OEMs are dropping hints that, hey, look, if you're going to source things through China, you can't do business with us. There's some of those hints coming. So we're ready for that from a supply chain perspective. We think it is an inflection point year for LiDAR 2024. Look, the thing that we look at and energizes me showing up to work every day is take a look at China, it's a leading indicator of what's happening, okay? I think according to the last Goldman Sachs report, they're going to do 1.5 million units shipped in automotive. It's a big growth from last year. And look, it's -- you need LiDAR. And if you look at what's happening in China, it's very clear. And yes, OEMs are still actively engaged in discussions. In fact, now that we have this NHTSA ruling, I've walked into 2 meetings in the last couple of weeks or the leader on the engineering side in the OEM and saying, "Look, if you can't see 200 meters down the road, high resolution, let's not meet because we're not interested. That's because NHTSA has elements of long-distance incident detection as part of this new Rule 127. And I'm going to say we believe that 1550nm is necessary to see that far. It's not just about seeing something, you have to be able to know what you're seeing that far, and we think that this ties to 1550nm very closely. Yes, that's it.

[indiscernible] Question about...

I can repeat the question. It's okay. I know we're on the webcast here.

So a question about the perception stack that you guys tie into for ADAS as well as in the future, of course, ADS Level 2 to Level 4. Is there ones that you work more -- ones that you work better with especially with sensor fusion or are there companies that you're already tied into that makes it easier for an OEM to adopt? Or do they have to custom adopt your sensors?

So just to put this in perspective, look, I'm an old Intel software guy. There are no standards. And going from one OEM to the next to the next to the next. They're all doing something unique. That's part of the challenge here. So the good thing is most all the perception software we've run into, except what we call point cloud. That's the format that comes out of the laser. But that's where the compatibility ends. This is an area where it feels like OEMs are doing their own thing by and large. Now you've got Mobileye out there trying to do it. And actually, we work very closely with NVIDIA. And on their platform as well, but not everybody's taking video. They're gigantic. Not everybody is using NVIDIA. It's very select. I don't think I can mention the names with very select OEMs are looking at that. Fragmentation is a huge issue. We just have to be compatible, in general in the automotive space. That's the game. And we do have a couple of partners outside of automotive, but that doesn't help us with ADAS, which is very fragmented. That's a challenge. As a company, new management team, I think we've made great strides. This is a snapshot of our quarterly earnings. It was a bit worrisome last year. I think we got feedback from investors when Conti backed out of the LiDAR space. LITEON is an amazing partner. And I can say that LIDAR is very strategic to them, which is important for the momentum that you get from working with a company like that. OEM feedback of our new product, Apollo is tremendous. We are very deeply engaged. We've got our burn rate down to a very sustainable rate with all the delays in the industry. Longevity is important, and we've been able to raise capital in the last quarter at very favorable terms. And we've got ourselves the potential for quite some runway here due to our low burn rate. We knocked it down since the management team came in, we've got it down 75%, will be somewhere about $5 million a quarter, which is incredibly competitive in the industry. We have a new equity line of credit instrument in place. I'll tell you, sometimes just lean, mean and small and focus is way faster than bigger. And being out in Detroit this week, I got some clear feedback, like, wow, you guys really started moving at warp speed, even though you're a much smaller company than you were in 2022, and it's just been that focus on automotive first, that's allowed us to cut back a lot and keep the burn rate very low. Next steps for us. Getting this guy to market, we're very busy right now. The small size is great allows folks to put it where that rearview mirror is in the windshield -- behind the windshield, it still performs well behind that. Product cost is absolutely critical around volume. LITEON is an incredible partner there. They're applying the same process they used to drive camera costs down. We're going to cut our costs -- this product compared to last year's product by half. The question may come up, what does it take? These long-range guys are only about $1,000 outside of China. And it's going to need to be half of that for the next set of quotes and that's what we have our sights set on. And we're going to hit it. We're incredibly competitive and continuing to manage and run a tight ship here. I think we're in a very strong cadence even with a modest budget. And last but not least, thank you. We're out here road testing real product today. We're not just talking about a concept. This is an actual shot from one of our road tests in Detroit last week. Sorry, you noticed I've been on the road somewhere in the same close apologies for that, I did wash them. But we're not talking about concept. This is a real product that's being demoed and road tested with OEMs today. Probably if you can't see it from there, you can actually read the writing on that dump truck that's pulling the trailer. That is like unbelievable resolution. This is what the OEMs want. They want to be able to see fine detail. That's another breakthrough that this product offers. Thanks for your time and off to Q&A.

Yes. No, thank you so much for being here. I appreciate the presentation. So just to start off, I mean, there's a perception in the industry that latter sit on them with EVs and ADAS. I mean, can you speak a little bit to the opportunity you guys are seeing on that front? And is your RFQ activity strictly limited to those programs? Are you seeing an expansion outside of ICE platforms given the delay in the ramp of EV volumes over the next couple of years?

Right? We're not seeing any connection specifically to EV. You probably already know that the technology itself is independent of the powertrain of the vehicle. And I think maybe your question is also related to some CEOs in the OEM space and say, well, the new stuff is coming in EVs. We're seeing RFIs come in, for example, that are very broadly applicable to vehicle lines. They're not specific to EV. So the pull is still definitely there not impacted by the EV.

Got you. And then maybe for some people who aren't familiar with the LiDAR space, you guys chose the 1550 nanometer laser and so the nano8. Can you kind of explain some of the benefits that you see from using that rather than a competing solution?

Right. So specifically related to the wavelength of the technology, I'm going to go straight back to NHTSA and us walking into the door and some of these guys saying, you got to see 200 meters or better, and you have to be able to identify objects at 200 meters, 1550nm allows us to turn up the power of the laser, which is very, very simple. How far you can see in some ways related to how much power you could put on the laser, 905 laser, if you turn it up too high, it will damage people's eyes. So for safety reasons, the power on 905 is limited. Looking at these new regulations, if you want to be able to see far, that needs to drive at high speed and have better safety and ultimately autonomy, we believe that 1550nm is actually going to be the game in town that's going to deliver that kind of performance.

Got you. And then maybe to switch over to how you guys measure the content per vehicle opportunity or how you guys look at it because you're taking a different approach on some of like the Tier 1 ladder suppliers. Do you view it as like a percentage of ASP or how you guys like quantify the revenue opportunity between like the latter units you ship?

Yes. Maybe, Conor, I'm going to have you come up here and talk about that a little bit. But in essence, we've looked at some of the older -- not older, but we've looked at camera deployment and how quickly the take rates have gone up. I mean, that's been one of our benchmarks that we look at -- see what the uptake is, what the price curve, resulting price curve based on the penetration rate, take rate and how many vehicle lines you are hitting. I don't know if you want to give any more detail than that, Conor. But we've looked at IHS and some of the older models and try to map them here. That's pretty much how we've done it, if you look at our financial models.

Yes. No, I think you're pretty much favor.

I'll turn it over to Josh. Conor, Our CFO, by the way.

Matt and Connor, maybe we could talk more about the capital-light model. It's something differentiated within the industry and a unique approach. So if you could just expand upon that and how that sets you apart from your competition?

Yes. So I mean I think it essence them, it's obviously a licensing model. So the big differentiator is we're designing the technology, but we're working with the Tier 1, which is, in this case, is LITEON, and they're responsible for industrializing the product and bringing it to market. So essentially, when we bring the product to market, we will earn a royalty on every unit that's sold, and that's extremely important when you think about capital expenditure and capital outlays because it's a very expensive endeavor to industrialize product and bring it to market. And so that's why it's great having a partner like a LITEON that can work with us and share that cost burden.

I just -- there is one thing I want to add to that just with like examples. So there's been a decade of work that's gone into engineering supply chain at the component level, okay? So we make underneath this, you click down one level, it's very modular. So you take a company like LITEON because it's such a modular design and basically, they're using epoxy and screws to assemble this, they can ramp up a production line or even a prototype line in a couple of months, okay? That's not something that we can pivot on so quickly as a company because you have to put the capital in to make that happen. You have to convince your component suppliers, "Hey, look, there's money coming, there's demand behind this. It is very, very fast. They can react quickly. And we're not talking about tons of money, but we don't have to invest that upfront and try to guess at when the market is going to take off. So that's super helpful for managing our costs.

The other important point is just optics when you're in the room with the OEMs and you have a partner that's brought other automotive components to the market before in the case of the LITEON, right, that's we end up with cameras. That's a very powerful message when you're talking to the procurement teams and the OEMs because it gives them a certain amount of confidence that your partner knows what they're doing and that they have the infrastructure to be able to scale to mass production.

Got it. That's helpful color. Just wanted to check if anyone in the audience had a question. I'll go on. Just maybe trying to understand a broader picture question as to what do you think are the biggest challenges that you see for broad adoption of LiDAR by the OEMs? Is it primarily cost? Or -- and do you see customer demand for this technology already.

I believe the biggest challenge in adoption is the software integration. You can read about it recently about some of the delays because ultimately ADAS, even for safety, not even talking about autonomy, there's a lot of -- there's millions of lines of code in there. You introduce a new sensor and any time you introduce a new sensor you have to retrain the brain in the car, and that's just not a matter of just writing some code. You actually have to put it out there, you have to road test it. It has to see various scenarios and retrain the AI in the car essentially. I think that's the biggest holdup quite honestly. I think that our company working with Continental and LITEON, we've been able to make leaps and balance cost improvement of this, and it's on the right trajectory for higher volume adoption, but it's really been the software integration that has kind of pushed things to the right. And I think to your question, is there going to be an adoption curve like we saw with cameras, absolutely.

Got it. Super helpful. Maybe just something I wanted to switch gears about the actual design and where you guys are integrating your product onto the vehicle. I don't know if a lot of people caught it, but you're behind the mirror on the roof line of the car. Can you maybe speak to what drove your decision to make that part of your design. And what's the feedback you're receiving from the OEMs is given that a lot of the computed solutions have to be integrated on top of the vehicle, not inside of it.

Right. Look, it's an interesting. I'm an enthusiast -- I'm in the enthusiast segment in the automotive in terms of market research. So take it for what it's worth giving that I'm enthusiasts. But look, at the end of the day, design is super important -- super important. And I think with AI, we realized that's eventually going to be a big factor. In fact, an OEM who hasn't talked to us recently, their design studio reached out to us, asked us to come to Detroit and like 30 people showed up to me like, well, wait a minute, you can put this thing behind the windshield as soon as they saw it they're going, yes. Look, we just believe it's a natural outcome the automotive industry design is king at the end of the day, and that drove the thought process here.

Got you. And then maybe kind of switching gears to L4 and L5 and the adoption, given you guys have better visibility on that. What's the time line you guys are seeing for those higher levels of autonomy. And what's like the key holdup in terms of the sensor suite that you're seeing from the OEMs?

Yes. Look, on the higher levels of autonomy, it's clearly pushed right. We are definitely seeing that, and it's software and it's training. And I look at examples of what happened with cruise in San Francisco, for example, is that it's a learning process. And you can't go unnaturally faster that learning process, otherwise bad things are going to happen. People might get hurt. And I think that was the correction that we saw here in the last year that, hey, this is going to take a bit longer because it goes back to training with a brain that operates the car even with new sensors. With that being said, I think you can read that, there are expansions happening in that, we'll call it the robotaxi space where you're seeing expansion. Cruise is coming back online, you see that Waymo is doing well and the emotional guys got a new investment from Hyundai. So it's still a very active space, but I think there's a humbleness now that's happening to say, look, you can't go too fast on the autonomy space, you've got to be careful. You need appropriate amount of time to train all of this digital intelligence on over billions of scenarios you're going to see out there on the road.

Last one for me and then I'll turn it over to the investors and Josh. But obviously, across the space, you're seeing lighter supply prices are just depressed relative to 2 or 3 years ago. What do you think it's going to take for investors to get excited about the lateral space and get going forward and for those share prices rerate higher?

Yes, I think -- I'm going to keep it to one later let Conor weigh in as well, but broader market adoption. That's what it's going to take.

I don't know if there's any questions from the audience.

The -- China seems like it's adopting using ADAS first. If that trend continues, it should drive price down, right? There's a bunch of other LiDAR players that are buying for that space, but it's their choice of hardware and software that's going to make the big difference. In your mind, you say adoption, it seems like there's a trend there. How do you see the winners emerging both from a hardware and a software standpoint?

You're talking about ADAS in general? Or is it a LiDAR specific question?

It's -- there's a sensor that's the hardware side. The software side has to be competitive. We talked about fragmentation, but...

So let me speak to the space that we operate in. Look, the first day I walked into AI. By the way, I'm not a LiDAR expert, I'm learning now, but I'm a software guy. And I sat down, I did a listening tour across the company for 30 to 45 days, and I'm going guys. It's all about the ADAS system and being compatible with the software in the best way you can. It's -- the LiDAR is not going to dictate how ADAS software gets rewritten. It's going to be how quickly can you integrate and adapt. I think one of the things that the industry struggle with is like, wow, hey, we can see really far and identify objects far we're good. No. You've got to have high resolution images that in effect, look like camera images to the software. That's the fastest fact that people are going to win are the ones who can be the best in terms of backward compatibility. I know it sounds strange for new technic. There's so much software out there. the more friction you can remove from that software integration process, that's who's going to win. That's my prediction.

Well, one more follow-up on that, which is you probably heard of Elon Musk's comments about not leading LiDAR and version 12.5 now is out there, but still, of course, when there's weather conditions, how are you seeing this Waymo method of training location versus the free-for-all Tesla version of perception and control?

Let's talk about how you train because I'm pretty sure everybody seeing in this room has had their brake slam on when you didn't expect it. Okay. I'm going to take it to a human thing that the smartest system that I know of today is the human brain. So I have an uncle. He had some weird accident in the kitchen like 10 years ago where he lost poked his eye with the broom handle, and he can only see out of one eye. What does that mean? It means he currently see in 2D really with one eye, okay? So uncle [indiscernible] here take this. He'll like go like this and he will miss it by like a foot. Okay? So he's got a really good brain, but his sensor on he sees in 2D and he's missing stuff. That's the thing we on -- I'm not going to sit here and debate him, he's brilliant, right? But you can't effectively avoid obstacles when you can only see in 2 dimensions. That's why I show up to work every day, right? And I back into a garage where there's a manhole covering the ground, the sunshines often certainly, the brakes slam on so hard in my car automatically. My wife says, "Oh, my God, we've crashed into something, No," its because camera can't distinguish a manhole cover from the ground. It can't see in 3D. I think that's a fundamental issue that most automakers maybe except one, have come to grips with. And so I think it's just a matter of time.

Great. I know we have a couple of minutes left. So I just wanted to get your thoughts around liquidity runway. You've reduced your quarterly burn rate and you have a target of reducing it by 75% through the end of the year. Just wondering if you could share thoughts as to how should we think about the liquidity runway going forward?

Yes. I mean we have enough runway right now to get us through the third quarter of 2025, which I think puts us in a good position versus our peers. I think everybody is more or less the same boat. And then I think the great thing about this equity line of credit facility we put in place is it's $50 million. And if you look at a burn rate right now, it's about $20 million on an annualized basis. So that gets us at least another 2 to 3 years of runway. Now of course, there are some conditions associated with that, right? The share price has to be at a certain price point and all that to fully maximize the facility. But nonetheless, we're feeling pretty confident that we've got a strong balance sheet, and we've done a lot on the cost reduction side and the liquidity side. So we're feeling pretty confident here. And that's obviously important. It's not about extending the runway. It's about investing in the company, bringing the product to market. And when you're going in and you're having these meetings with the OEMs, they want to be confident that you have the runway to be able to go on and execute.

Got it. Thank you so much. I think it's a good place to add. We really appreciate you as being here in exciting times of LiDAR industry.

Great. Thank you. Thanks for having us. Great discussion.