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

George Gianarikas, one of Canaccord Genuity's sustainability analyst. Thank you for coming to day 2 of our 45th Annual Growth Conference. And we're incredibly excited to have Aurora with us today, CFO, Dave Maday. Short presentation and then some questions. So Dave, please go ahead.

Awesome. Thanks. Thanks, George. It's great to be here again. It's a great event, and I look forward to talking with all of you. I think maybe just before we get into the Q&A, just for those who may not be super familiar with Aurora, I thought we would just cover a couple of things. Of course, before I do that and to keep the lawyers happy, this is our forward-looking statements. So any claims related to that. So just take a quick look. All right. Let's get into it, right? So first and foremost, we're excited to be operating driverless trucks on public roads. And so this was our first quarter. This is a video of one of our trucks operating on the lane that we deliver goods on. And so it's been really exciting in terms of the progress that we've made through the second quarter of the year. So we launched at the end of April. And then through the end of June, we operated 20,000 miles. We were operating first at daytime, and we started with 1 truck. We ended up going up to 3 trucks. We also just recently launched nighttime as well. So now we're operating both day and night. And we work with a lot of partners. Our launch partners for driverless were Hirschbach and Uber Freight, but we also have other partners, FedEx, Werner, Schneider, Volvo Autonomous Solutions, others that will be rolling out driverless lanes too in the future. So we are very excited. Look, we feel like we're in a really good position as a leader in the autonomous trucking space. Obviously, we are the only company that's operating driverless on public roads today. We think the -- it's a massive market opportunity. It's a $1 trillion market opportunity in trucking. And so we believe that the Aurora Driver can unlock tremendous value, not only safety, which is obviously paramount to our mission, which is to deliver the benefits of self-driving technology safely, quickly and broadly. But also we're able to drive down total cost of ownership, and we're able to increase asset utilization for our customers. We're the only player that actually has the strategic partnerships in place to scale the business. So you need -- to be able to scale at the tens of thousands, you need OEM partners who are willing to build trucks, scalable platforms integrated with your hardware at high volume. We have a partnership with Continental, which supports building our hardware kits in the tens of thousands. So we're really excited about that. From a financial perspective, we started to recognize revenue, which is for a CFO, that's pretty exciting, instead of just talking about R&D expenses. And -- but importantly for us, we've got a really strong balance sheet, right? In this last quarter, we announced that we have $1.3 billion in cash and short-term investments, which allows us to fund operations into the second quarter of '27. We have a Driver as a Service business model. We launched as a Transportation as a Service, and that's what really for development and initial adoption that will run through '26. And then starting in '27, coinciding with the production for our Continental kit, we will be able to switch to a Driver as a Service business model, which will be great for both significant revenue growth as well as SaaS-like margins. So we're really excited. And then obviously, from a competitive thing, we're first. We think we want -- we have that first-mover advantage, and we want to continue to take advantage of this in our leadership position. What's left for '25, right? So we've launched -- our focus is increasing the number of driverless miles. Obviously, we had a couple of technology unlocks. And the point of this is that we're trying to -- '25 is all about demonstrating the technology promise. So we talked about what the Aurora Driver can do for people. Now we are going to deliver it. And to be able to deliver that, you need to operate in all kinds of different conditions. day, night, which we pulled ahead and actually launched. Next up for us is rain and wind. And then coinciding with that, once you're able to operate day and night in rain, you're actually able to open up longer lanes as well. So in '25, we expect to be able to operate in all these conditions. That sets us up for the following year to be able to really pick up and expand our operations and expand our customer base and expand our fleet. If you're a carrier, like this is great, right? If you can deliver this technology promise and do what happens, like my simple takeaway, and this is an example, an illustrative example of Fort Worth and Phoenix, it's 1,000 miles. What does driverless operations give you? It gives you better utilizations of your assets. So that same truck can now operate 20-plus hours a day when today, it's operating less than, let's say, 8 hours a day. And then the other thing is we can drive a lower cost of ownership. This is in addition to the safety benefits that you get, the fuel efficiency benefits that you get and the insurance costs that you get. So really exciting opportunity for both the top end revenue growth and the bottom line margin. And then maybe before I close, I'm super excited about this myself. We tend to want to be the standard bearer in terms of like safety and transparency and things like this. And I'm proud to announce that we have opened up a viewing site. It's -- if you go to YouTube, it's @AuroraDriver, and it allows you to take a look at our trucks operating on the roads, driverless every day between 8:00 a.m. and 5:00 p.m. Central Time. And so it's not too monotonous. We switch between different views, but these are our trucks operating each and every day. And we're really excited to be the kind of the standard bearer in the industry. And so with that, I will turn it over to George, and we can talk about questions.

There we go. Also, we watch the video often. It's very relaxing, by the way. ASMR is a word that I've learned from my daughter. So I suggest to the audience that take some time to watch if you need to calm down. So I'd like to first ask about the relationship with PACCAR. And you've sort of had a little bit of a change in strategy since your first autonomous truck launch, you have what's called an observer in the vehicle. I'd like to ask, first, can you update us as to how many autonomous miles you've driven and whether or not this observer has had to intervene at all in any of those?

Yes. First and foremost, the Aurora Driver operates the vehicle. It's designed to operate the vehicle at all times and in all conditions, even if we had to pull over to the side of the road for a tire blowout, an engine failure or the Aurora Driver Lidar was struck by a bird. Like we have to operate safely at all times. That's how it's designed. We actually operate in what we would refer to it as driverless mode. So it operates all the time. We did 20,000 miles through June. I'm not going to disclose where we're at today, but I will tell you, we will be higher than second quarter without a doubt. And no, the front seat observer has never needed to intervene, and we don't expect them to have to intervene in the future. And I might also say, and while our relationship with PACCAR, even though that optically, it looks a little bit weird, we can respect their opinion and their risk profile versus like what we think is an incredibly safe product, and still have a really great relationship going forward. Like I've been married 30 years. Every once in a while, you don't agree with your partner, but that doesn't mean that you throw away the marriage. And so in this particular case, we had an observer in the rear seat a lot of the time, not all the time, but a lot of the time. And I think there's a benefit of that. If you think about an early rollout, having this observer doesn't prohibit us from any of our development or expansion or anything else. It is not prohibiting us from moving forward and developing our plan. And you do get feedback, right, on the behavior. So the analog, I'd say is like Waymo. Waymo is a very successful ride-hailing company, but they also get feedback all the time from their customers. Every ride, they get input on the ride and the behavior. Unfortunately, for us, we carry goods. So toilet paper can't fill out a survey. So front-seat observers, it's actually not the worst thing in the world. And when you think about the crawl, walk, run approach, we really think about the future. This front-seat observer is just on this initial launch truck of vehicles. It is not relevant for any other vehicles that we're going to be launching when we get their scalable platforms or any other vehicle that we launch. So this is just with this initial launch fleet of trucks. So we think it's okay, and we continue to move along, and we're really excited because they are all driverless. And again, check out Aurora Driver live, and I think you'll get a good view of what an observer does, they observe. So they're not hovering over the wheel or anything like that. They're sitting back. I mean sometimes you'll see them twiddling their thumbs, but they are actually observing.

You mentioned some of the signposts to scale that you have this year, slide presentation. Can you just help us understand how that -- what that looks like maybe in 2026? What can we look for additional lanes, et cetera, weather conditions?

Yes. So '25, again, all about the technology promise, right? So do the things that we can operate. And if you think about like customer adoption, you kind of have to do those things before customers are going to go all in on anything anyways. If you can only operate in daytime dry conditions, that's great. And it's a tremendous progress because we can operate in all the conditions, construction, heavy congestion, things like that, but it's not enough to be an integral part of somebody's business because sometimes you end up being late because of construction. And so what happens if you go into nighttime. Sometimes you get rain, and we need to be able to operate in enough conditions. So '25 is all about being able to operate in a sufficient set of conditions. This sets us up for '26 when we have a couple of things going on. The first thing we have is our second-generation hardware kit will be available in early '26, right? We can't build a quantum of trucks until we get that second-generation hardware kit. This kit will be cheaper. It will be -- it will last longer. It's designed to 1 million miles. We've designed and engineered it, but it's being manufactured by Fabrinet in Thailand. So therefore, we can build into the hundreds and thousands of trucks with this particular kit. The second thing is we got to drive in more spots. It's great to be able to drive in Texas and in Arizona and New Mexico, but we want to be able to operate over the Sunbelt. So we will open up lanes throughout the entire Sunbelt in 2026, right? And I think those are 2 really important things. The last thing is we are going to start going to customer endpoints. So yes, we operate through terminals today. That's great for development. And it's great for some coast-to-coast operations where drivers couldn't make the whole route anyways. But ultimately, you want to get to a customer's distribution center or warehouse. So those are the 3 things that we're going to do that are going to lock the customer adoption, and we'll expect to be operating trucks -- significantly more trucks in '26 than we are in '25.

You have a relationship with Volvo. You talked about getting more trucks from them by the end of this year. What's the time line to production-ready trucks from Volvo? And when they do get on the road, do you expect them to be fully driverless? Or will they be an observer there as well?

Okay. So we are getting trucks right now as we speak and throughout this year, they refer to as [ SAE-builds ]. These are essentially the scalable platform trucks that they intend to -- that we intend together to operate driverless with no front-seat observer. So there's no requirement for a front seat observer. These trucks are not ready yet. They still have validation and work to do on their side for the components on the software side, in particular, to validate. But these trucks that we'll have and we'll start operating on, we'll be using them for development. We'll be using them for bring up for our second-generation hardware kits, and they will be using them to carry commercial loads with vehicle operators. So they'll be driven autonomously with vehicle operators. Once they've completed their validation efforts and their work, these same trucks will then just transition over to driverless trucks. And so we're really excited about the progress. We'll let them define when the timing is, but I would expect that '25, you will not see them driverless, and we're working really hard to have them driverless in '26.

There's a big player in trucking in the United States, Daimler. They currently have an internal solution called Torc. I think many people in the audience know who that is. What's the potential to maybe become a second source there over time?

So I would say this. We believe in choice. We believe in competitive pressures. I think our customers want to have options. And so today, we have 2 partners that represent roughly 50% of the market. And we think that that's exciting in terms of development. We would expect at some point, the Aurora Driver is the essential and the best Aurora -- self-driving solution out there, and we would expect it to be on all trucks. And so I definitely can see a future where that happens. I mean that's obviously something that they have to work on with their contractual commitments of what they have, but we would be open to putting the Aurora Driver on that. And I can see a future where the Aurora Driver is on all the trucks in the U.S.

It's an enormous market. We completely agree with you that there's a huge potential, about 200 billion miles plus just in this country that are driven by Class 8 trucks. There's Torc that's a competitor. There are others who are vying for a position in the space just because of its -- the huge economic potential. What do you expect the competitive environment to look like if you fast forward 3 to 5 years? Waymo was a competitor at one point. They dropped out. They're focused on robotaxis. So do you expect that to come in? Long-winded question, but like how many competitors do you expect over the long term in this business?

It's a $1 trillion market, there's going to be competitors. There's been a lot of competitors. And everybody likes to predict the positioning of all these competitors, and it used to be -- TuSimple was the leader and they're no longer around, and then they've reinvented themselves into a different company, and you've had Embark and things like that. It's a tough business. Like the folks that say that it's easy and that you can do it really cheap and I'm going to be so much more efficient. I probably don't have the right appreciation for how challenging and difficult this is. But it's a $1 trillion market. So there's always going to be competitors. I would expect a handful will offer some sort of self-driving service at some point. I think we have a multiyear lead, just to be honest. And I also think we're one of the few or the only that's really set up to scale a business. And I think there's a difference between operating on one lane and getting yourself ready to scale business. So I think there'll be some competitors out there and competition breeds creativity, innovation and competition is good for the industry. So we welcome competition. I would tell you, I am probably more concerned with people trying to catch up than maybe doing things unsafe that could impact the industry. But there's a lot of really talented people out there. And I give a lot of credit to everybody for going down this endeavor. Autonomous vehicles in trucking is needed, and we're excited about the future.

There are a couple of phases to your business approach versus proving the technology this year, which I believe you've done. Next is expansion, but also bringing the cost down, right, to make sure you can get to these SaaS-like margins. So can you talk about your relationship with Continental in that context? How confident you are that once you begin to scale with them, I think it's in 2027 that you can get to the right level of hardware costs such that we can see this margin improvement over time?

So our relationship with Continental has been fabulous. It's really been exciting. And I think they would equally say the same thing. They're always excited for us to come be at their events, and we're always excited for them to talk about the partnership. That doesn't mean that you don't have conflict. That doesn't mean that it's not hard. But when you have shared beliefs in kind of the vision and you're working on something that has shared benefits and you're working on the aligned goals and you have a mutual respect, that's a framework for a great partnership. And so our partnership is great. Continental is obviously going to be spinning off their automotive division to Aumovio right, in September. And when that happens, this project is one of their cornerstone projects, which they talk about. So excitement level is there. In terms of cost targets and our path to get there, we're really excited. We think we're on track. And it is not like a big source of concern for me in terms of like our risk profile in terms of the margins. There's always going to be $0.01 or $0.02 like delta that might be in there. But like overall, this is -- we've got a pretty good track. We have a target. We're starting to get bids and sourcing for each of those that Continental is doing, and it looks to be well on track. So we're excited.

You have a really interesting relationship and economic model with Continental. For those in the audience who may not know, can you just sort of describe that relationship?

Yes. So the traditional Tier 1 to OEM relationship is the OEM ask them to build a component, they sell them that component. The OEM buys the component, they put it on the truck and then they sell the complete truck. It's kind of a point-in-sale model, and our model is different. We kind of believe in the shared benefits and the economic profile. So if you think about the trucking industry, most of the trucking industry, shippers pay carriers a price per mile. They're going to pay us a price per mile to drive the trucks. We're going to pay Continental a price per mile for that hardware kit. And so instead of this point-in-sale, we're going to be paying them a per mile -- a cents per mile charge every time we drive a vehicle, so the Aurora Driver drives a vehicle. And they're making a significant upfront investment. So they're spending over $300 million of upfront NRE, so nonrecurring engineering that will be paid back in this price per mile over time. So it's the first time that it's been done in the industry. for sure. And for them, there's this reoccurring predictable revenue, which is really exciting for them, and they're making an investment in the future. One of the other things is they're making investments and they're going to develop their -- what they call a minimal risk maneuver system, a lane B, a secondary system to pull over in the event that the primary operating system had a fault or something like that. So that's the part of the redundancies. They're able to develop that and then use that technology to deploy in other markets. So they're looking at it as a great way to get into new businesses. They've got a reoccurring revenue stream. And so for us -- and they're also financing the vehicles, all the components. So that's part of the charge. They're also doing the service and warranty of all the vehicles, and that goes into that charge as well. So for us, we have a fixed cost that we pay cents per mile that's very predictable over time, and they get incentivized and they're incentivized to make sure those trucks are always on the road.

Any questions from the audience? My friend, Keith.

Didn't think that one through, did you?

I guess when you walk through the economics, can you walk through what the cost is to outfit a truck? And then when you look at it from a trucker perspective, obviously, they're saving on the driver, but how much of those economics are saving of the driver versus higher like hours of service is a big issue for the truckers? You're going to basically drive this thing a lot more utilization. Can you just kind of walk through some of the assumptions behind those numbers?

Yes. So I can't give you the specific cost because we don't disclose them. But what I can tell you, I can try to answer your question this way. So a traditional truck driver's hours of service limitation's 11, they operate well less than that, maybe on average, like 8 hours of productivity a day would be a high number. Our trucks can go conceivably 24 hours, let's say, then you conservatively, you can double that utilization. So every truck can be operating double the time. So that's just from an asset utilization perspective, there's a ton of top end growth. It does a couple of things. It means that they have more productive assets and maybe they even need fewer assets in a given time, but they also can deliver goods faster than they could before. It creates new opportunities for like if you're FedEx and you had a lane that you had to do like a flight to get to a certain spot in a certain time. Now you can do it with trucking. So those are interesting opportunities on the revenue side. On the cost side, what we've said before is the driver costs continue to increase. The latest ATRI number for driver cost alone is $1 plus you have indirect costs associated with that. We've given some indicative pricing of $0.65 to $0.85 for our Driver as a Service model. So that's our long-term model. So there's a substantial opportunity to get lower costs just for driver costs. Then you have fuel efficiency. Our trucks versus like traditional truck drivers, at least 15% more fuel efficient. You have reduced frequency and severity of incidents, and that's kind of the primary point of this. So there's a ton of opportunity to drive the cost down. So we do expect it to be able to do both top and bottom. We own -- we are paying -- so when a customer like buys a truck equipped with the Aurora Driver, they're just buying a truck. They're not buying our hardware. Everything, all the costs, the hardware costs, our software costs, our data services costs, what we pay Continental, those are all included in the per mile fee. So for them, they have a very predictable cost model, and that should drive substantial savings for them. So that's what I can tell you.

[indiscernible]

Sure. The easiest compare and contrast is we are actually operating driverless and they're not. So I'm not convinced. I don't know, George might disagree with me on that. We believe -- and everything for us is guided on safety. We believe in a multimodal sensing approach where Lidars, cameras and radars all provide value and input. Certainly, for the vast majority of times, a camera could probably do the majority of the jobs that need to be done, but we don't think they're sufficient for all the times. And if you're Tesla, you have an FSD system that has a driver in the vehicle, right? And so technically, they're supposed to be able to take over at any individual time. It's unclear to me that Tesla is ever going to get to a world where they truly are driverless in all these scenarios. And we'll see, right? Like they're going to be able to do it. They're piloting it out now on passenger cars. Trucking is different, right? Trucking is an 8,000 -- 80,000-pound tractor trailer driving 75 miles an hour. And so for me, I feel like that's a different solution set. You need to be able to see longer, you need to be able to predict better than what you really need to do in passenger cars. So for us, we believe this is the right technological approach. We think Waymo has a very similar technological approach. They're the leader in the ride-hailing space. We believe we're the leader in the trucking space. So if there's ever a world where we need less components, we would do that. But frankly, we had to develop our own proprietary FirstLight Lidar because we did not think that even the existing Lidars on the shelf would allow us to safely operate trucks. And so we now have our own proprietary FirstLight Lidar that we purchased in 2018 and developed. It allows us to see 9 seconds further than a traditional Lidar. Imagine 9 seconds with a tractor trailer and being able to plan your route and see things in advance 9 seconds faster. It just makes you more safe. And there are plenty of instances where in the nighttime, a camera cannot see something that a Lidar can see. Like it's just a fact. So we're going to focus on our approach. And hopefully, we'll continue to learn and adapt accordingly.

Thank you, Dave.

Thanks. Appreciate it.