Aurora Innovation, Inc. (AUR) Earnings Call Transcript & Summary

Thanks so much for coming back. Chris McNally, Evercore ISI. It's our pleasure to have Dave Maday, Chief Financial Officer of Aurora. Aurora is one of the companies under our public coverage. And let's -- look, let's get right to it. We have a lot to talk about. Dave, I know you brought slides. If you can give a brief overview for those who may be less familiar with Aurora, and what you've been up to for the last couple of months?

Awesome. Awesome. Thanks, Chris. I appreciate it. I appreciate the invite. This is a great event. This year is even more amazing. You can see there's a lot of excitement over the AV industry overall, and it's really a pleasure to be here with all of you. Some of you may not be super familiar with Aurora, so I'll do a quick presentation, real quick presentation, and then I'd rather just get into the Q&A. With that, and to kind of keep the lawyers happy, this is our statement in terms of forward-looking statements. So please take a quick gander 1, 2, 3, it feels like it's about right. All right. With that, let me just talk a little bit about Aurora, right? So we are a mission-driven company. Our mission is to deliver the benefits of self-driving technology safely, quickly and broadly. We are operating driverless on our launch lane, which is Dallas to Houston. We surpassed 50,000 miles in early September. We have a pleasure of working with several really great customers as we start to roll out this technology throughout the Sun Belt by the end of next year. And everything that we do is really focused in on trying to create value. Let me talk a little bit about Aurora. We feel like we're in the pole position when it comes to autonomous trucking, right? And so let's talk a little bit about, first off, what's our product? Our product is the Aurora Driver. That's a combination of hardware, software and data services, which allows us to operate autonomously on public roads. I had mentioned we are the only company that actually operates driverless trucking on public roads today. Trucking is our very first market. Why is that? It's a $1 trillion market. We -- it drives 200 billion vehicle miles traveled each year. And we believe that autonomous trucking can create great value for the industry. If you think about both the safety, if you think about the utilization of assets and the total cost of ownership. And I'll talk a little bit more about those in the future. In terms of an ecosystem, right, we feel like we've got great strategic partnerships, both on the OEM space with PACCAR and Volvo. They represent about 50% of the market, but also in terms of -- and all these partnerships are not to get to a driverless launch, we're there, we're past that point. We're all focused on trying to build the business and to scale a business together. And so that's what our partners do. From a hardware side, we've got a partnership with Aumovio, used to be Continental, the recently spun off automotive arm. And that's what's going to allow us to actually scale our hardware kit into the tens of thousands. So we're really excited about that. From a financial perspective, at the end of Q2, we had $1.3 billion in liquidity that allows us to get in the second quarter of 2027. We continue to be pretty opportunistic about our fundraising. We don't want the concerns of liquidity to impact our customers, our partners, our investors like yourselves and our employees. And so that's our focus. And then what I would say is, if you think about the next couple of years, 2025 is all about really unlocking the technology promise. It's great to have a driverless product. But to be relevant, you have to actually be able to do certain things really well, and you need to make that product performing enough that customers can create real value from it. '26 is all about scaling to hundreds of trucks, bringing on our second-gen hardware, operating across the Sun Belt. And then '27 is what we'll start our transition to our Driver as a Service. That's our long-term business model, right? That supports our capital-efficient approach to the market. And we're really excited about that. That also is when we bring on Aumovio and our third-generation hardware kit. And again, our goal is to accelerate what we believe is our first-mover advantage and reinforce our leadership position. A little bit more about 2025, and this is important for us. When you're in a safety critical industry and you launch a product, it's incumbent upon you to do it responsibly. And we have this crawl, walk, run approach for 2025. And so for 2025, our goals are really simple. First off, we're going to continue to increase the driverless miles that we operate every day. The second thing is we're going to improve the capability of our product. So one of the things when we launched, we were operating in climate conditions in the daytime. We unlock nighttime. We pulled that ahead of schedule. Just based on the investments and things that we've done, we are actually able to accelerate our development approach. What's next is unlocking rain and heavy wind and then operating on additional lanes, whether it be Fort Worth to El Paso and then extending out to Phoenix. We've been operating autonomously on these lanes for quite some time, and we're making excellent progress to unlock these before the end of the year. And in fact, I'm even excited to say that our first 2 driverless launch customers on this lane are going to be Hirschbach and Werner, so we're very excited about continuing to build on that momentum. Okay. Why is it valuable? And we even have a colleague in the back that would challenge us on this all the time. We know there's value in providing this technology to customers, right? And it's really about having a higher level of safety than what's out there on the road today and then increasing the financial performance of carriers because that's kind of our primary customer cohort, right? And today, if you look at the human-driven fleet, there are some restrictions that all autonomous trucking companies can deliver upon. One is increased revenue utilization. The trucks can operate without hours of service limitations, and that's a real benefit. And you also have to provide lower total cost of ownership. You combine those 2 things together, and you're creating a margin improvement for customers. We can argue about what the dollar amount is, but there is no doubt that this can provide a higher level of safety, increased utilization and lower cost of ownership. Okay. The last thing before we go to Q&A is, look, maybe a little bit of a plug. Like Aurora is a company that prides themselves on being transparent. We've been a public company for quite some time. We're kind of leading in a lot of the transparent things, whether it's our safety case approach, how we measure performance and progress. The latest one is how we operate driverlessly on the roads. So at any point in time, any of you can go out to our YouTube channel on @AuroraDriver and watch our driverless trucks operating on the roads. They operate daily. The live stream goes between 8 and 5 p.m. Central Time every single -- or Monday through Friday. So go check it out. It's pretty cool. I know a lot of investors watch it all the time. I happen to watch it quite a bit. And with that, maybe really excited, let's get into some questions. I'll cruise over there.

Yes. Really appreciate it. David, I really appreciate the detail, particularly on -- I think you've been very consistent about that walk into -- crawl to the walk milestones. I want to maybe start in reverse order and sort of zoom out to whatever the time period be in years or long term, what Aurora truly wants to be when it grows up, I do a phrase from Mark Mahaney. You're starting with hub to hub. Should we still think about something beyond that, meaning to the customer last mile? How do we think about the true evolution of Aurora long term?

It's a great question. And it is a bit of a perspective on time horizon, right, and everybody knows it. Like the first thing is, like if you think about today in the world, almost everything that you touch feel you get delivered, operates through trucking. Like trucking is essential part of our economy and about society. If you ask me like in the next 10 years what I feel like we should be? We should be an integral part of that, right? We should be operating integrated with a human-driven fleet, we should be operating with a vast amount of carriers across the U.S. and creating efficiency, right? And giving customers an opportunity to kind of rethink about how they think about their network, really increase the utilization and the performance of the product. And also just improve the safety generally of trucking across the industry. Now how do you get there? So if you think about where we are today, and what do we need to do, I mentioned some of the things in '25, but there's a couple of other core components that you need to get to, to be valuable. You have to go to the customer endpoints, like it's great, yes, we have operated terminal to terminal, and that's great, and it demonstrates that it's done, it's capable, but you have to go to end points. For us, starting in '26, the vast majority of where we're going to go to are customer endpoints, right? And I think that's a really important element to be able to drive customer adoption and build confidence in the product that you have. So I think that's a really important one. I think the second 1 is you got to kind of align your business model with what works with customers, right? Customers buy trucks today. So ultimately, at the end of the day, again, we're transitioning to a Driver as a Service, where they get to buy trucks and be able to place these trucks where they want them and be able to operate how they want to do it. And so for us, we need to enable that. And so 1 of the most important things for us is to make sure that we transform our business model. That is why we have long-term partnerships with our OEMs to build scalable platforms. That's why we have a Conti strategy or Aumovio strategy in terms of building out our hardware sets. We need to be able to provide the product where they need it at the time they needed to. And so I think those are a couple of the key enablers that are going to get us to where we need to get to. But ultimately, at the end of the day, by the end of '26, I fully expect us to be operating throughout the vast majority of the Sun Belt. By 2030, I expect us to be operating in most of the U.S. And again, we have to integrate it within the customer networks. And we have to integrate it within the human-driven fleet, right? Like this is not an autonomous take all sort of thing. This is an integration element that you have to work with individual customers.

So if we focus on that business model, do you think it's fair to say, David, this conversation of -- there's an industry discussion of hub-to-hub versus customer endpoints, there's actually probably a lot more of an overlap there? Because what I'm thinking about who is a conversation you're going to have, maybe hub-to-hub, you're talking to someone like Werner work, a customer endpoint, maybe you're dealing directly trying to get coke for example, to work from a production to a distribution line. Can you talk about who that end customer is that drives decision to start an AV route?

Yes. Well, I would say it this way, like your 2 primary customer groups in this type of example are your FTLs, and your LTLs, right, and maybe your private fleets. We need to go where they want us to go. But autonomous trucking is predominantly where we create the most value is in these over-the-road routes that get between DCs or warehouses. That's that's where the greatest value is. It's not getting to driving into city centers and delivering end points like you have a whole new challenger. What we think is the real value proposition is getting to DCs, getting to warehouses, getting to customer endpoints and each 1 is different. So if you're FedEx, you want to be able to get to the regional sort facilities, and to be able to do that, you need to be able to do that reliably and on time because you can't miss a sort, right? For a company like Schneider, you need to deliver goods from 1 of their DCs to 1 of their customer endpoints. Maybe it's a Lowe's or a Home Depot, the same thing with Werner and everybody else. So every 1 of them is going to be a little bit different in terms of where they see the most value. But generally speaking, it's DCs or warehouses to warehouses, distribution centers to distribution centers, sort facilities to sort facilities, things like that.

And if we had Chris here talking about the AI challenges of doing the side streets, right, again, these are not usually in urban centers, but the side streets to do some of those customer end facilities. Is there an AI challenge? Because clearly, there's less of a business model challenges, less drayage cost and things like that for them. But is there a little bit more of an AI aspect on side streets?

I'm not sure you want the CFO talking about AI. But what I would tell you is I think the short answer to that is no. I think there's a little bit of a misconception, right? Even today, when we deliver to our terminals, so when we're going to Houston, we traverse 5 miles of surface streams, right? Like this is, if you look at the vast majority of distribution centers, over 80% of them are within 5 miles of a highway, right? So you're not talking about going into dense urban city centers. We're talking about meeting the customers where they want us to be as opposed to trying to drive something different. So I don't think it's an AI challenge at all. There is some technical differences, right? In some cases, you got to go through school zones or maybe you got to navigate in railroad crossings and things like that. So I don't want to say it's -- there is no challenge. There is some challenge, but it's not the same challenge as what you experienced in kind of ride hailing.

And then a little bit on the timeline for, I think, the one-on-one for people. When you think about that first generation to third generation, today versus 2027, can you just remind people how the Software as a Service model and hardware as a service model through Conti or Aumovio will sort of work high level? You don't have to go into specific numbers. But you are sort of a first of a kind in a Hardware as a Service model, so I think it's really important to go that.

Yes. So there's a couple of things. Number one, I think it's important if your perspective is -- if you think about our mission, which is delivering the benefits of self-driving technology safely, quickly, broadly. To be able to achieve that message -- that mission, you have to think a little bit longer term. So you have to think about, well, if it's broadly, that means I need to get it everywhere. I need to be able to produce enough volume. What do I need to be able to do that? You need OEM partners again. You need hardware partners, but you then also have to think about, well, how does the business operate? The majority of the business operates on kind of a pay-per-mile type of structure, right? And you want to have unwind incentives. Like our carrier partners would like to pay us when we deliver a truck. I would like to pay Aumovio when they enable us to deliver a truck. And so this Hardware as a Service model was a really interesting concept that everybody seem to get aligned with, right, because it creates the right incentives for everybody to behave the right way. And it focuses on having a reliable product for the future. Now how do you get there? So today, we've got a -- we have our small fleet of trucks that's with our first-generation hardware kit. Admittedly, it's a great hardware kit, but it's not designed for a long-term potential. Our second-generation hardware kit is going to come online next year. And that one, we've designed it, but we're going to use -- we've designed and engineered it, but we're going to use Fabrinet to help build it, that allows us to get to hundreds to maybe thousands of trucks. And that's a great step, but that's hardware that we're going to own and pay for it. The third generation, that's when Continental is essentially going to finance it, manufacture it, service it, and they're going to get paid a cost per mile. Now the 1 really important thing for a lot of -- especially some of the new investors is -- an important clarification here, right? In the traditional OEM model, an OEM buys hardware components, they put it on the truck. That's not what happens with the AuroraDriver. The AuroraDriver is a combination of hardware, software and data services. That means all the hardware that is on the truck, actually is not purchased by the OEM, it's included on the truck, and it's part of our per mile fee, right? So there's no big upfront cost for the AuroraDriver hardware. There will be some incremental costs associated with the work that the OEMs do, but the AuroraDriver hardware itself, that's just part of the per mile fee that we have.

And can you just review that evolution '25 to 2027? You mentioned it, but I think it's important. We're in the last disclosed 3 or maybe it's more trucks, you said hundreds next year into thousands, 2027 on generation 1, 2, 3, is that good paraphrase?

Yes. Yes, that's how we think about like where we're at 10 to hundreds of thousands. We've got a small fleet of vehicles that we operate today, that we use for both commercial driverless miles as well as for development, right? And you need these same trucks to do both purposes. Next year, we'll increase our fleet to hundreds. And then the following year after that, it's thousands. And then at the end of '27, you're at a run rate that gets you to tens of thousands, right? So that's how we think about the stair step approach.

And then I started with the zoom out, Dave. So now start to zoom back in. When I think about the next couple of years, you put out targets for 2028, this idea of a TAM or a SAM, right? 800 billion, it's a big, big number. But even just making a small portion of that over the next couple of years, is the difference between thousands of trucks to your point or tens of thousands of trucks. I think you put out a 50 billion SAM for your initial market of what you just discussed. Can you help us frame that, how you came to that discussion? How do you think about customers or the size of lanes that you need to address that sort of SAM?

Yes. So to achieve the SAM, and this is like at the end of -- this is like in 2028. So think about it as we've done the Sun Belt. We win a little bit north of that and kind of above like the Mason-Dixon Line, and we're operating in all those areas. Now how do you get to a TAM of 50 billion VMT, which is an opportunity for us to then go play a part of, right? So you have 200-ish billion vehicle miles traveled in the U.S. today. How do you bring that down? So first, you look at the lanes you're not operating in. The second thing is you exclude routes that are like less than 100 miles, right? Just conservatively, does that mean AV can't handle it? No, it can. And we may, in certain use cases, do really short routes that people want us to do. But let's just -- this is a conservative approach to business. I'm a finance guy, so I'm a little bit more conservative. And if you look at just operating on those lanes, it gets you to 50 billion vehicle miles traveled. The average length of haul is over 600 miles, right? So we break it down by length of haul, but you're averaging more than 600 miles per trip based on that 50 billion VMT. Every 1 of those lanes that I talked about, like they all -- those all support autonomous trucking today, right? So it doesn't -- actually, that number doesn't include California, which today doesn't allow for autonomous trucking. So we only include the states where AVs either are explicitly or implicitly allowed today. So it's not a real big challenge to get to that kind of number. And then, again, we think we're going to be a portion of that like we're not going to be the majority of that. There's still a big human-driven fleet that's out there, and we're going to start to supplement that in 2028. And so 50 billion seems like a pretty reasonable estimate. I think there's a little bit of mischaracterization out there in terms of like autonomous trucks can only go 500-plus miles, and that's where the use case is valid. That's actually just not true. I mean you can talk to customers. I know that that's actually not true. But if you go to endpoints, and endpoints, again, being relatively easy to unlock. Again, we drive 5 miles surface street for 5 miles even in our Houston terminal, 80% of the routes are under 5 miles from the highway. It's not hard to get to end points. Customers want us to get to endpoints, but customers also want to see us have a valuable product that can operate in the day, it can operate in night, it can operate long hours and it's reliable. And as soon as we continue to demonstrate those things, the ability to operate in endpoints is going to slowly mature and like we're going to start unlocking endpoints at the beginning of next year, right? And we're really excited about that. I think most of our customers would like us to go to end points. Some terminals are valuable or handoff spots and things like that. But generally speaking, endpoints is going to be our primary mode of distribution.

And strictly to sort of your CFO's virtue, as I think about that 50 billion SAM to get to some of your targets, you're talking about 10,000 trucks, which operate, I don't know, 200,000 miles plus for years. So you're targeting within that 50 billion over the next couple of years, something like 2 billion to 3 billion of that 50 billion. Is that a fair way to think about what you're initially going to target within that 50 billion of SAM?

Yes, that's a fair way to do the math on it. I mean we have a lot of work to do to prove that out. But I'm very confident that the value proposition is there. As we demonstrate the safety and the reliability and address some of the customer pain points, that's a great opportunity. That's kind of a target out there for us. And over the next couple of years, we'll continue to deliver towards it, not guidance.

Yes. And then 1 or 2 more for me that I want to definitely give enough time for questions. You talked about the near-term milestones, closing the validation on night driving, but over the course of the next couple of months, you're going to close it on some version of inclement weather with rain, but also your new lanes. I think there's always a little bit of confusion of whether you are solving that or whether you're closing a validation case, making sure that it's at your level of acceptance to roll out publicly. Can you talk about what needs to be done internally to satisfy the check -- the box for rain and the new lanes?

Look, I think it's incumbent on AV technology providers to hold ourselves to an incredibly high bar for safety. That's the value proposition, right, and you have to be able to do that. And so for us, unlocking a rain, I mean, if any of you have driven our trucks and several of you have, like we operate great in rain, right? You even question whether we're ready to go? We need to hold ourselves to an incredibly high bar for safety. And for us, it's all about validating the various scenarios to ensure that we can close our safety case. So there's not a technology unlock that needs to get done. It's more about validating the confidence in just closing the safety case itself. And we have a very robust process. We spent a lot of time right in the validation and refining the validation elements. So that's why we're able to close nighttime faster than what we had planned before, and so for us it's really about having the confidence that we will be safe on public roads. I mean all of our families drive on public roads. We have to hold the same bar as everybody else. So like we really think it's important to make sure that like you don't compromise safety. We don't do this for demos or experiments and things like that. We're doing this to drive a commercial business, and we're never going to sacrifice safety, like we have the long-term vision in mind. And I think we've demonstrated that quite a few times, and we'll continue to behave that way.

And then the last 1 on that transparency. I mean, obviously, seeing is believing, so something like the YouTube channel is fantastic. But also, we've seen now audited data sets, self audited from Waymo. There's insurance companies that are adding on to that. Do you think that's something that we could or should see in the in the trucking industry, just so that you have that added level of regulatory consumer customer buying?

Look, I think we are objective, I think that we're the most transparent AV trucking company out there, that's somewhat by necessity, we've also been a publicly traded company for quite some time. So there are certain things you have to disclose. And we've tried to provide insights into this. I think it's -- I do think you'll see more in the future. I'm not going to tell you what it is, but I do think you'll see more. But I also think that even if somebody else told us, hey, this is great, like we've done a lot of stuff too soon, right? And they love the processes. We did -- many of you and our customers, we had J.J. Keller evaluate our trucking product. We do a lot of stuff. The most important thing, though, is that we have to stand behind it. And you can't rely upon somebody else to justify the safety. I think this is still a technology that's pretty complicated to close the safety case. I think all the companies that work in this space, like it's amazing some of the accomplishments that the industry is making, and we're really excited about it. And I think there is a level of assurance that we need to continue to provide. I believe seeing is believing, that's why we do the AuroraDriver live. I do think we'll look at other modes. But I will tell you that we allow and encourage all of our customers to put their best drivers or their driver trainings into our product before they ever go driverless with us. And I think that, that's actually a pretty good testimonial to a lot of stuff that happens.

Excellent. I want to open it up to Q&A to the room. Don't be shy.

Awesome.

No, maybe I can do -- I always have a follow-up question. It's always a trick, it's always a trick. I always have a follow-up question. Stack AV presented this morning, and when they talked about their timeline they mentioned, and I think this is a discussion in the industry, that of all of those multiple components of a trucking ecosystem that the OEM still seems to be 1 of the most hesitant because they have to put out AV product that is actually different from their existing hardware, drive by wire, redundant steering. How are you seeing OEM acceptance and timelines because it's not probably a very difficult issue, but it's one where they want to make sure that their hardware is working even if you say that software is ready to go. So how do you think about OEM landscape for partners like PACCAR and Volvo rolling out over the next 18 months?

I think in some ways, they're not wrong, right? I do think there's a bit of an under appreciation on the complexity that maybe AV companies have relative to either like the customers' business like running a carrier is really fri***** complicated, being an automotive manufacturer is really complicated. And I think it's prospective, right? Like so for us, at Aurora, 100% of our business is on autonomous technology. If you're an OEM partner, I think it's it's like a tiny portion of somebody's brain space at the time. And I think they recognize the opportunity and benefit. Competition tends to breed a little bit of innovation. There's not quite as much competition in the large trucking space as there is on passenger cars, which is why you see passenger cars -- more passenger cars able to support autonomy than you do in the large trucks, but they'll get there. And again, if you take a perspective, a longer-term view, and have the appreciation for what they're trying to achieve, not every partnership goes great, right? Like -- and sometimes, you got to be patient, sometimes you got to pivot and sometimes you got to keep the end in mind. I think they will get there. And when they get there, you're going to have all the benefits of the OEM partnership that you started up for. And yes, it might be a little bit slower than what you want, but such is life. I mean, like we're -- they could build 10,000 trucks today. Our customers, and we are not ready to deploy 10,000 trucks either. So like have some perspective in it. So I think it's a little bit true, but this is new technology, not everything goes great.

And it's still on pace.

Thank you. [ Rakel ] over there will squash me.

[indiscernible]

So yes, what I would -- I would describe it a little bit different. So we never want like our runway to be a concern, right? Our employees don't like it, our customers don't want to invest a lot of time and energy in a company that they're not sure is going to be around. Our partners don't like it. Investors like have a tendency to breathe together, and so if they're concerned about it, then they don't want to continue to invest. We had the benefit of having an extraordinary long-term investor base to support us. And we've been very fortunate in that, and we worked really hard to build the confidence with them. We will continue to be pretty opportunistic. If I bias towards anything, I bias towards having pretty long runway just for the near term, I think that will be a pretty big focus of ours. So I would expect us to continue to be opportunistic. We have a -- I think most of you know, we have an ATM that's accessible to us at any time. We've been pretty opportunistic on fundraising in the last couple of years. And again, our goal is to make sure it's not an issue for our partners, customers, employees or investors, that's our goal.

[indiscernible]

Yes. The last -- we talked about this, I don't really have any updates so that we think we'll achieve positive free cash flow in 2028.

[indiscernible]

Our expectation is, and it's probably a better question for some of the customers, but our expectation based on those discussions, is that carriers like to own their assets, right? Like they like to deploy them where they want to deploy them, there's still a service and maintenance that goes on with these. Remember, the difference in trucking is these trucks can last up to 1 million miles, right? They operate today 100,000-plus miles per year. In future, we think, with AVs, it's 200,000, 250,000 plus miles a year. And so they want to be able to deploy them on whatever route they can, at whatever time they need to send them and having the flexibility to do that and keep them at their end points is really important to them. So generally speaking, we believe that most of our customers really like that approach, and that's what we're going to enable.

Last one.

[indiscernible]

We don't really use them, to be honest. I think they're great tools. I think they're inefficient to have a safe autonomous driving product. So we have our own hardware suite with cameras and everything. We do integrate with a vehicle through our ACI. So we're able to connect and do things like tire pressure monitoring, that gives us a good signal of vehicle behaviors and stuff like that. But when it comes to some of the other telematic devices, we don't really use them. We have probably way more -- well, definitely way more substantial tools to use. And so for us, it's not super helpful.

And with that, round of applause for Dave and Aurora.