Ouster, Inc. (OUST) Earnings Call Transcript & Summary

All right. Good afternoon, everybody. We're good to go. Thank you so much for joining us for this session with Ouster as part of Deutsche Bank's Global Automotive Conference. My name is Emmanuel Rosner. I'm the lead U.S. autos and auto technology analyst at Deutsche Bank. We're very pleased to have here with us today, Anna Brunelle, who's the CFO of Ouster. As you may know, Ouster invented its digital LiDAR in 2015. It is a leading manufacturer of high-resolution digital LiDAR sensors, used throughout the industrial automation, smart infrastructure, robotics and automotive industries. The company is looking to be a major mobility disruptor with a larger focus on non-autos end market than many of its LiDAR competitors, but also with a focus on auto. So -- the format, I think, will be -- Anna, you will be presenting one slide.

Sure. Brief introduction.

Brief introduction of your company and then I have prepared some questions, and we can do that through a fireside chat.

Perfect. Sounds great. Yes. So I think a little bit about Ouster. We're a leading digital LiDAR company today, as you said. And that's based on really our CMOS digital LiDAR. And so it's a very differentiated technology. I'm holding an example of it in my hand. It's a very performant system that's based on digital CMOS flash architecture. And what that means is that much like the end state of radar or camera technology, CMOS silicon is the end-state of the sensing technology. And so what that means is off of 1 CMOS silicon chip through firmware, you can customize a product to get 75 different SKUs. And what that opens up to us is a very diversified strategy, meaning that our products work across multiple verticals, whether it's automotive, industrial, robotics or smart infrastructure applications, and it's all based on the same single CMOS digital architecture LiDAR chip. And so through that, we're able to have a product that has absolutely no moving parts and is highly performant. And every time we take out a chip much like Moore's Law, the product basically doubles in performance. So we have the highest resolution sensors on the market today at the lowest cost, about 30% margins today at our kind of emerging company volumes. And because we're first to market with this kind of technology, we also have 50 granted patents, 200 pending patents and 250 exclusively licensed patents. But more than that, I think what our diversified business strategy means is that we're able to go after about an $8.6 billion projected total addressable market. And that's again because of the flexible architecture that suits so many use cases. But more than that, because we continue to perform to improve our performance every time we do a chipset tape-out, and we've released 3 generations of products today with the fourth generation expected later this year. We've developed over 600 customers in over 50 countries. We have 30% gross margins. We've scaled production, which we outsource with Benchmark manufacturing in Thailand. We've shipped over 10,000 sensors. And as I say, this architecture is kind of proven out by even at these early volumes with that 30% gross margin performance. And so I think I just wanted to kind of give a brief overview of that so that you understand why Ouster is different than other LiDAR companies you may be talking to.

That's great overview. Good place to start. Thank you.

So maybe just to start. So my understanding is Ouster is much more involved in nonautomotive end markets relative to some of your other competitors. Can you go over the motivations for that? What are some of the large opportunities that you're going after there?

Yes, sure. So we sell sensors across markets, like I said. So in 2021, 34% of our units shipped were in automotive, which is actually more than most of our peer group combined. And we shipped about 25% of our units in robotics, about 25% in industrial and about 15% in smart infrastructure. And so in terms of where we're seeing the growth this year, we're seeing a lot of growth in warehouse automation, where we signed additional strategic customer agreements in the first quarter. And when you think about that, I think it's really important to talk a bit about the other markets, not just automotive because the other markets are really developing more rapidly. And so when you think about warehouse automation for one Fortune 500 company, that's enough LiDAR sensors in kind of warehousing and robotics applications to -- frankly, it's more than an automotive production run. And so I think just only talking about automotive, you missed the opportunity near term. And so in 2021, we about doubled our business off of the 2020 fiscal year. And we've signed with our 600 customers, 72 of them have signed strategic customer agreements where they're giving us 3 or 5-year forecast into their business. And so we can understand that our customers' business through those projections are also about doubling year-over-year-over-year into the future. And so in addition to warehouse automation that I talked about late last year, we announced that we'd signed about 110 deals in the smart infrastructure space. And we are currently also a leader in robo trucking, which we see is contributing strongly to our growth this year, where we have the largest binding deal in automotive ever with Plus.ai, where Amazon ordered 1,000 robo trucks from them, and we're providing LiDAR sensing for that. And so I think there are many other markets that are developing in the near term that are real catalysts to driving the growth in our business. And so yes, we are planning to win the automotive market. But in the meantime, we're also winning in other markets, and we truly are a diversified business.

Understood. I think you addressed some of this, but what does the time line from mass adoption or I guess, adoption look like for the various end markets? Where do you ultimately see the largest opportunity at scale?

Yes. I mean I do think that the automotive opportunity at scale is very significant. I think we all know that. That's certainly something that's been talked about a lot today amongst automotive manufacturers as well as in the press. And I think that's where we have a really unique and differentiated strategy. So this one single architecture, we can do long grain, short range, mid-range, live field of view, various different applications off of. And so what that means to an automaker is that you can move towards full hands-free, eyes-free autonomy with CMOS digital LiDAR. So if you believe that the end state of LiDAR is no moving parts, surround view like it is for cameras or radar, and a CMOS architecture like it is for cameras or radar. What you get out of that is an order of magnitude fewer parts, no moving parts, much higher reliability. And you get the ability to seamlessly have 1 point cloud that surrounds the vehicle off of kind of a one-stop shop for LiDAR. And so what surround view LiDAR means is if you just have one LiDAR on the vehicle facing forward, what you're really doing is tracking the car in front of you. So you would drive to the highway, on the highway, you would follow the car in front of you and then you would have to take over driving the vehicle to leave the highway, and you would stay in the slow lane the whole time. With surround-view LiDAR, now you can move through an intersection because you can see to the side, you can merge on the highway and change lanes because you can see around you. You can park the car. And so suddenly, you have full hands-free eyes-free driving. And that's really what auto OEMs want. They want to move towards full hands-free eyes-free driving, and our application allows them to do that. We're the only ones out there that are creating surround-view LiDAR, long, medium, short range off of one platform with no moving parts, no FMCW, no mirrors, no spinning scanning mirrors, no nothing like that truly solid state. And then you have a system on chip that is doing all the compute power for all of that data that you're bringing in. So if you think about our sensors, you're using a digital flash technology, so you're getting an order of magnitude, larger resolution at a lower cost point, which creates a ton of data, which then is a system-on-chip to create a single point cloud for automakers to work off of. And so whether it's an automaker or a tractor company or a farming company or a mining company, it doesn't matter, that's what the end state of LiDAR is, and that's what Ouster is providing to our customers.

Let's talk about your margins. So you mentioned a few times gross margin of 30%. I think the official guidance is 25% to 30% for this year. How does this vary across your end market application?

Yes. I mean we've talked a little bit about our margin expectations longer term. So right now, those margins are off the blended unit sales that we have today. We've had 30% gross margins for the last 2 quarters in spite of some supply chain constraints that have put extra cost burdens on the business. But that being said, I think what's so exciting is that this architecture much like a digital camera at volume, it becomes very efficient to produce. So we sold 6,500 units last year. On average across the year, we did 27% gross margins. You're absorbing a ton of fixed costs at that volume level. So when you start thinking about 100,000 units, you're at like high 50% margins, right? Like it's just because of the amount of fixed cost you are absorbing per unit. And so finest days with our customers, that gives us the ability to drive our ASPs down as volumes increase while preserving margins at the rates that we intend to over that churn. And so I expect that in nonconsumer ADAS, we'll see kind of high 50% margins across our business on hardware only, not software ancillary services, which would be an additive. But in automotive, we've said that we expect to see about 25% margins on hardware because that's just in line with their purchasing power and the margins that other hardware companies see in automotive, which is you're selling a lot of units but at a lower margin.

And I think your average selling prices have been rising.

Yes. In Q1, interestingly, we added 90 new customers into our funnel in Q1, and so those 90 new customers were buying at higher ASPs than we've seen previously because they're buying closer to list price, and there were so many of them. And so the top of the funnel kind of strength is strong in our business. But generally, we've been working to trend ASPs down as customers order more and more significant volumes.

Understood. Now how should we think about your cash burn going forward? I guess let's start there.

Yes. So we ended Q1 with about $163 million in cash. We used about $23 million for EBITDA that quarter. In the fourth quarter of last year, we acquired the only other CMOS silicon-based performant LiDAR company out there, which was Sense Photonics. And with that, we acquired about 75 employees who are primarily hardware engineers related to this VCSEL-stat approach and a commercial sales team 100% focused on automotive. And so the reason why I bring that up is this year in 2022, we kind of accelerated some of the hiring that we planned into the fourth quarter of last year. So this year, we'll probably continue to build our sales team across the world because that is -- that's been a strong benefit to us so far. That's one of the areas that we've been investing in our proceeds since we went public about a year ago. And we are currently growing the business across all 4 verticals in all 3 regions. So Asia Pacific, Europe, EMEA really and the U.S. And so as I mentioned earlier, we saw 90 new customers via LiDAR sensor in Q1. And so we want to continue to expand that because we are literally working with over 600 customers in 50 countries, but we expect LiDAR ultimately to power every moving object on earth as well as the smart infrastructure that monitors those moving objects. And so there are still many, many more customers and additional total addressable market to go after and continuing to build out that commercial sales team will help us do that. So just kind of circle back to your question, in addition to the $163 million and kind of talking a little bit about our projected burn rate and because we have 30% gross margins, I think we're on a good trajectory to get to cash flow positive. But we've also crossed the 1-year mark in terms of being public for over a year. And so that's allowed us to also put out an S3 and an ATM that will allow us for the next 3 years to draw down as much as $150 million in our business. And we also signed -- we released an 8-K that we've signed a $50 million term loan with no warrants. So no equity dilution. We're very cognizant of equity dilution in this [ spot ] market. And so I think, along with our trajectory and growth trajectory and positive gross margin trajectory and those sources of cash. I think we'll have plenty of cash to fund the business now and into the future.

Should that take you towards your projection of breakeven free cash flow?

I mean, yes, I think that's the goal, right, is to ensure that we have enough access to capital for that and for any future needs.

Okay. So let's maybe take a step back on your technology and obviously, you previewed some of it at the beginning. But there are so many types of LiDAR available on the market. Now what are the key differences with your digital architecture and road map compared to the other competitors?

Yes. I mean, I think what I was talking about earlier about Digital Flash LIDAR with CMOS silicon is the cheapest, most reliable, most performant solution out there. And so you're getting, again, an order of magnitude, more resolution. You're getting compute in a system on chip. You're getting the ability to customize one single architecture to meet multiple use cases across verticals for short, medium and long-range LiDAR, all off a single architecture. And that gives you the ability to offer a 5-LiDAR suite in automotive, for example, for about $1,000, whereas our competitor sensors are in a prototype stage based on older analog technology. And they -- there's no iteration path for that technology, right? Like analog technology at some point, you just can't continue to iterate and make it better. Whereas we're following a Moore's Law curve of every time we take out a new chip, again, we're about to release our fourth-generation chip. This year, we've had products in customers' hands for about 3 years. That just allows us to basically double our performance each time. So much like digital cameras taking over analog cameras, when the digital technology first enters the market, it's not as performance to take over like the high end, like LI-COR or Hasselblad cameras. But over time, it is. And what happens then is it opens up not just -- it doesn't just take over the analog market, but it opens up a newer, larger market because it's so cheap. It's so small. It's so performant. Suddenly, it can go anywhere and the price point just is no longer an impediment. And I think that's what's so exciting about digital CMOS technologies, it just eats industries wherever -- any industry that it enters, it is the end state. And there's no one else doing this technology that creates performant digital LiDAR sensors except Ouster.

And then when you're speaking about these 5 LiDAR solutions for [ $1,000 ], which would be basically price of one LiDAR from some of the competitors. Does that make you more focused on surround-view rather than long-range forward [ side ] facing LiDAR. Is your LiDAR for a different application than some of the other automotive LiDAR?

Yes. I mean I think we're really focused on winning series production with auto OEMs in the '25 and '26 time frame because those are the contracts that are being decided in the next 12-plus months. And in order to win series production with automotive makers, we -- different automakers are on a different path to hands-free eyes-free, right? So some automakers are trying out a single sensor to start. Others are going with surround-view LiDAR to start. From my perspective, it doesn't matter which path they're taking. They all want to get to hands-free eyes-free. And the only way to do that is with LiDAR. Like we know that cameras make great L2 systems, but they also crash, right? And so if you want full automation, you have to go with CMOS digital LiDAR, and it has to be surround view. That being said, we're bidding on RFI/RFP/RFQ, work with OEMs today for single LiDAR sensors that is a first step in their path, and we're also bidding on it for surround-view LiDAR sensors that is a first step in their path. But what I think is important is automakers have already been through this with digital technology for cameras and digital technology for radar. And both of them, the end state is surround view, multiple sensors on a vehicle and it's all CMOS and that's what we're doing. So I don't think there's any question if you go and talk to auto OEMs, that that's the end state of LiDAR. And we're happy to help them work towards that via a single sensor or a surround-view sensor, but the end state is going to be surround-view CMOS digital LiDAR.

So I understand the cost advantage and certainly the trajectory advantage. What about the performance, so let's say, range. Are there any -- are these comparable to some of the other solutions there or especially the front-facing LiDAR in automotive?

Yes, absolutely. So we have a digital flash sensor that we just delivered a samples to with our -- we have a strategic customer in automotive that is a major global auto OEM, but we are not allowed to name due to NDA reasons. And we just delivered a sample. So we meet the automotive spec today and that means range. It means everything else, too, but it means range to answer your question. So yes, we meet that today with our automotive prototypes that we can say as an example too, yes.

Okay. And then just I wanted to come back to your -- the timing comment around the 2025 '26 because obviously, some of the initial autos, LiDAR production contracts actually about to launch, right, where there's already been an allocation of Volvo, Mercedes, Nissan, like all these things have essentially already been allocated sort of like auto contracts. So is it a function of where you are in your development process, where you're not quite there yet, either in terms of cost or performance to be able to have seen that you've already won and launched in the near term? Or is it a function of -- yes, I guess, how would you explain?

Yes. I mean I think that we have to break that down a little bit. So when you talk about some automakers who have announced deals for short-term software production contracts that allow the automaker to better understand how to use LiDAR in their software stack. It's important to note that there are no serious auto production wins behind those. So I think that sometimes you'll talk to LiDAR companies who kind of infer that there are. And if you don't ask directly, they won't necessarily be as transparent with you about that. So in order to win series production in automotive, like we had -- I mentioned we have a strategic customer agreement with a large auto OEM, and that's helpful to be a part of their development program. But you still have to actually go through the RFI/RFP/RFQ process and win series production. And so some of those brand names that you just mentioned, those contracts are actually for short-term software production -- software agreements. And they -- what I'm seeing is that many LiDAR companies are not able to sell hardware. They're still in prototype stages. They don't have significant hardware sales. They certainly don't have 600 customers and 10,000 units shipped. They don't have outsourced manufacturing. They're not scalable. They're still shipping prototypes. And I think what that means is they've had to pivot their dialogue to software. But I think differently, I think that if you want to win the software business, you first have to have the hardware, and then you can build a software business on top of that. It's not possible to build a software business on hardware that doesn't meet spec that cost $10,000 per unit. It's just not possible. And so I think that -- we have to be careful that some of those brand name OEMs that you discuss don't have any serious production behind them. So I guess what I'm saying is the field is wide open.

I feel this. Okay. I'm just going to follow up on this, just what about in China? Just feels like literally just over the next 6 to 12 months in China, half of all EV makers will be launching vehicles that have LIDAR in there as standards. Is this an opportunity for you as a market? Is this something that as the Chinese customers requesting different level of pricing or performance than what you're offering?

Yes. I mean I think that our commercial sales teams are involved with discussions with all auto OEMs, regardless of what country they're in, right? I mean there's a limited number of them that are meaningful and also with EV companies and Tier 1. And so I think we understand very clearly where each of them is at in their process. But in terms of do I separate the China market from another market, No, I mean I think we just are working with all OEMs to make sure that we have products that suit their needs. And that's really where we're at. I mean we meet back for auto OEM projects that we're bidding on and we're bidding on sub...

Okay. And I guess would you expect major automotive LiDAR production contracts to be awarded in the next 12 months?

Yes, I think that in order to be in production in 2025, 2026, the automakers, I mean, it's not my time line, but the automakers time line is 12 to 18 months. Yes, they pretty have 2.

Okay. How should we think about the 72 customer agreements signed this year in terms of size and scope and what it does for your revenue trajectory?

Yes. So last year, we started really maybe a little more than a year ago. We started emphasizing this idea that we wanted to get more customers on long-term agreements, where we understand better what their volume requirements are and what their long-term pricing needs are. And this was important to me because as we move more and more customers to SCAs, I better understand what volumes I'm going to be selling at. And so I better understand what volume pricing I can give those customers to lower ASPs, but at the same time, lower COGS more rapidly than I'm lowering ASPs, which then allows me to open up the market. And then what that allows for our customers is for them to have pricing predictability when they're in production and their unit volumes are scaling. And so I think this is just a great thing for us all to understand each other better and to have better predictability into my business and my customer to have better predictability into their business. And so last year, we kind of set out to do an initiative around this. We started last year with 10 strategic customer agreements, and we ended with 68%, and that was about 20% of our revenue last year. This year, we have -- I think we're up to about 72% as of Q1, and that's $550 million in contracted revenue opportunity. So as we move more and more customers on to these kinds of agreements, it just gives us better and better visibility and predictability into our business that we can then share with investors and where we can have confidence in out-year pricing with customers.

Then maybe finally for me, how is the process different in terms of bidding, quoting, bidding, winning in the non-autos market versus autos. Is there shorter lead times? Is there fewer competitors? Is there more competitors? What does that environment look like?

Yes, like I talked about that we were seeing a lot of growth this year from smart infrastructure. I've mentioned earlier that we've signed -- we put out a press release that we signed 110 agreements late last year. And I think smart infrastructure is really interesting because it's a greenfield opportunity for LiDAR. So right now, there's a $32 billion camera market in like CCTVs, infrared cameras, things used for kind of infrastructure applications. And this is an opportunity where a single LiDAR sensor covers a much bigger territory. So your hardware costs are actually less -- and then with LIDAR because you're getting a 3D point cloud, not just a camera image, you have actionable data. Like is that a human being moving towards the perimeter of the building or is it a dog? Is it moving towards you or away from you? Is it human being stepping off the sidewalk at the curb at an intersection, are customers running through a casino. Those kinds of applications, LIDAR gives you data that's actionable. So it's a better solution, and it's actually cheaper to install because you need much less hardware to run a better solution. So we see this as a greenfield opportunity for us this year. And why I bring this up in response to your question is there's a very small time frame like you can go from purchase implementation in a few months because in smart infrastructure, it's not being built into an end application like the LiDAR isn't being built into a forklift or a dockyard crane or a mining -- piece of mining equipment and underground vehicle or something of that nature. And so -- it's a much shorter time to market. And it's just an absolutely huge market that we've been investing in our commercial team for over the last year.

Awesome, let me open it up for the room. If there's any questions?

What is the long-range capability, the actual range?

Yes. I mean, so I think that most people focus on the automotive spec. And so I think that's what everyone talks about in our industry is meeting a 200-yard range with less than 10% reflectivity. -- but our sensors will actually see further than that, yes.

Any other questions? Yes. I think you were very clear.

Okay. Great.

Anna, thank you so much for interest in us to Ouster and for your time and participating here. I invite everyone of you to join us for drinks on the rooftop terrace of Deutsche Bank on the other side of this floor.

Thank you. Thank you for attending.

Thank you.