All right, so we're going to go ahead and get started. In case you were wondering or in case you were lost, this is Track Two, and we're here with David Maday. He is the CFO of Aurora Innovation, which is a very exciting company and industry leader in self-driving technology for commercial vehicles. Before we get into questions, I'm going to let David—I know you have a few prepared remarks for us.
Awesome. Thank you. Thank you for having me. Really excited to be here. Maybe just a bit of background for those of you who may not be as familiar with Aurora. Let me just give you a quick couple of highlights. First off, for us, everything we have and do starts with our mission. Our mission is to deliver the benefits of self-driving technology safely, quickly, and broadly. That is what guides us.
Our core product is the Aurora Driver. That is a combination of cutting-edge hardware, our verifiable AI software approach, and data services. We have designed the Aurora Driver to be able to operate in multiple use cases to support multiple industries. For us, that is really important. Our first product introduction will be in trucking. Why trucking? First off, the market is massive. It is a trillion-dollar market. There are over 200 billion vehicle miles traveled every year.
The second thing is we can provide incredible value right at the start, whether it be increased revenue through higher utilization per truck asset or lower total cost of ownership, whether it be fuel efficiency, driver wages, insurance costs, etc. Of course, and paramount to everything, is the safety we can provide to the industry. For us, when we started, we were founded in 2017 by three outstanding founders: Chris Urmson, Sterling Anderson, Drew Bagnell. We have collectively over 50 years of experience. One of the things they set up from the beginning is we are really good at the self-driving and autonomous technology, but we need a partnership ecosystem second to none to be able to deliver upon our mission and to be able to scale our business.
For us, from the very onset, we've been very focused on establishing what we think is kind of an industry-leading partnership ecosystem. A couple of examples of that is our OEM partnership. We partner with both Volvo and PACCAR to develop autonomy-enabled trucks. Those trucks will allow us to commercialize at scale. The second thing is—Volvo and PACCAR, by the way, represent about 50% of the truck market today. The second thing is our partnership with Continental. Continental is a major Tier 1 OEM partner. They have the ability to build our hardware kit at scale. That allows us to build the necessary hardware kit to coincide with the vehicle platforms that PACCAR and Volvo are building. That will allow us to do tens of thousands of trucks. We expect that partnership to come to fruition in 2027 with our third-generation hardware.
Another example is NVIDIA. One of the core technologies is the processing that we can use in our compute. We are going to be utilizing NVIDIA's DRIVE Thor system-on-a-chip for our third-generation hardware in collaboration with NVIDIA and Continental. That is just a couple of things that allow us to be commercially successful. When I think about commercially, I pivot to kind of our approach to commercialization. The first thing is we have a tremendous list of what we call pilot customers today. These are the FedExes, the Werners, the Schneiders, Uber Freight, Hirschbach, just to name a few. It is important for us to remember that we today, even in development, are delivering autonomous loads today with vehicle safety operators in them for these customers. We are really practicing like we play. We are delivering loads on routes that we will commercialize driverlessly.
We've done over 9,500 loads, over 2.6 million miles, and I'd like to say with almost a perfect 100% on-time performance metric. For us, we're really focused in on the commercial elements. I think if you look at the regulatory standpoint, we have an excellent regulatory team, but our launch market is in Texas. If you think about Texas, Texas is a great market to launch autonomous technology from. In their own words, they've said, "When Aurora is ready to launch driverless, we are ready to launch driverless." There are no additional regulatory challenges that need to be overcome. That's actually true in 39 of the states and all throughout our rollout throughout the next several years. Maybe finally, from a financial perspective, we closed 2024 with $1.2 billion in cash and short-term investments.
That is sufficient to get us through commercial launch and well into the second half of 2026. We got an amazing group of investors, like some of you in the room, that support us, and we're really excited about our progress. Maybe with that, I'll turn it back over to you, and we can get into the Q&A.
Thank you, David. That was very thorough and very helpful. I appreciate you going into that. Certainly, a lot of questions I want to touch on. Maybe just to take a step back, I'm wondering if you could just paint a picture for us as to what does the self-driving market look like today, and maybe particularly for commercial vehicles. How would you characterize where we are today?
You know, it's a great question. You know, if you think about autonomous technology generally, I think we're really at an inflection point. If you think about autonomous technology, we have robotaxis out there from Waymo that are really starting to permeate in the marketplace. They're rolling out in additional cities. For Aurora, we're on the cusp of launching our driverless product in our first launch market on the Dallas to Houston launch lane. There has been a lot of discussion, a lot of hype, a lot of challenges maybe in the past. The biggest question is, is it ever going to come to fruition? I think we are now at that point. I'm really excited about the opportunity of where we are today.
I think if you think about the market generally in trucking, which is a great question, the other thing is you can start to see whether it's the customers, the regulators, or our partners, everybody's starting to get excited about the value proposition. Us delivering on the technology promise and demonstrating that for the world. I think it's going to be a great year. There's lots of talk, lots of buzz, and we hope to deliver on the promise this year.
Got it. You know, I think recently we've been hearing a lot of news about Waymo and about Tesla's upcoming robotaxi segment. I'm curious if you could maybe help us understand why does self-driving technology, why is it way more practical for commercial vehicles than for passenger vehicles?
This is a great question. I think there's a couple of things that are first and foremost. If you think about the trucking market today, again, it's a trillion-dollar market. They do 200 billion vehicle miles traveled per year. It's a huge market, and there are driver shortages. You don't need to create this unnatural demand, more people to take rides, let's say, in the robotaxi space, to just succeed in the marketplace. I think the second thing is if you look at the economics, just the raw economics, driver wages for a CDL versus a gig economy driver, three times as high. If you look at the various economics in there, I think those two things in particular make it very opportune. We don't need to grow the market to generate the business. The second thing is the driver economics are just a lot more viable.
It allows us to start off slow and commercialize and then drive down the cost structure. I think the other thing is we're actually creating real value. I mean, there's an element of robotaxiing where it's cool and the technology, and it eventually will create more mobility for everybody. I mean, we do believe in the space, just not as the first entry point. We think it's going to be a great opportunity for us to go into once we've driven down our cost structure, once we're producing tens of thousands of hardware kits and we're on the road. We then believe that we'll be in a better opportunity with the right cost structure to go into robotaxi. We can create tremendous value today in the trucking market.
I think to your point, our studies have found that last year alone, there was a shortage of approximately 80,000 truck drivers, approximately. You are driving longer distances, you are on highways, there is a bigger need for these trucks to drive themselves, and particularly given the truck driver shortage. I think that is interesting. You touched on this a little bit in your prepared remarks, but just maybe for folks that are not as familiar, what does the regulatory environment look like for self-driving technology in the United States? Is it done at a federal level? Is it done at a state level? How does it work? You touched on Texas, obviously, being your initial market.
Yeah. The regulatory regime, if you think about the U.S. and the federal regulatory system, it's really a self-certification process, which is kind of monitored by NHTSA. All of the regulation that happens happens at the state level. What we look at is continuing to support adoption of innovation. If I look across all the states, there's either explicit or implicit, where it's not restricted, regulation in 39 of the states that say that you can operate autonomously without a vehicle operator. That's very consistent throughout. It's very supportive of an innovation approach. It's grounded in the fact that this could be and will be a safer product, ultimately, and that's what we're all intending to deliver. I think that's really great. I do think there are some advantages of a federal framework to provide some consistency.
If you look at the recent administration, there's been a lot of positive talk about autonomous technology and the benefits it can provide in supporting innovation. I think even Secretary Duffy in his confirmation hearing even talked about supporting an AV adoption for regulation. I think there's some positive momentum that would make it easier for states to continue to support the adoption. Right now, it's state by state, and again, 39 states. Every state in our rollout plan that we've outlined through 2028 is supportive of the technology. The one state that always comes up is California. California, which is kind of interesting, I'll stay out of the politics of it, but California in the ride-hailing space, kind of the first market for everybody, but it's restricted in trucking.
There are some reasons for that, but we've been working with the state, and I would expect that the state will be supportive of this in the future. Right now, our rollout plans actually don't even count California. If California comes, then great. It's just another great market and another huge opportunity.
The current administration has talked about or has hinted about potentially implementing a federal mandate for self-driving technology in the U.S. I think the point that you're making is whether that does get, in fact, implemented or not, legislation is already favorable, particularly in the markets that you are targeting, such as Texas, such as Arizona.
Yep, 100%.
I think that's important to mention that. As recent as February, you had Aurora-equipped trucks had driven, I think if I have my numbers right, a little over 2.6 million miles with supervision and have already delivered about over 9,500 loads. What lessons have you learned? Where is the technology now? Maybe what's left? Because since you're going to be commercializing in, what is it now, 30 days, roughly, or next month? I guess what lessons have you learned thus far, and what's left between now and commercialization?
OK, let me start with what's left first, and then I'll get into the lessons learned, because I think it's a little bit more interesting and maybe longer. What's left for us is continuing the final bit of validation and verification of our software to close our safety case. We measure our progress two ways. Number one, it's with our Autonomy Readiness Measure, our ARM. That's just measuring the safety case and closure of the safety case. We need to be at 100%. That will indicate we feel we are acceptably safe to operate on public roads. This is a very methodical approach. It's utilized in most, if not all, safety-critical industries as an objective and very quantitative way to make sure that we're going to be safe when we operate. It takes some of the debate about it.
People talk about other measures that really do not indicate safety. The first thing is ARM. We are close. We are 99%. We expect to burn down the last few items of that and be launching in April. At this point, we measure our error bar in weeks, not even months. We are really confident in our approach. The second thing is our Autonomy Performance Indicator, or API. That is where we measure how often we are driving in autonomy. We do not measure it in terms of how often. We measure it in how often are we getting from our between our terminals, which is where we will start off, how often are we doing that with 100% API, meaning we do not need on-site support to get to our location. There, our target for our commercial launch is 90%. We have been operating since mid-October, around 88%.
We're consistently kind of getting at that threshold week over week or beyond that threshold. For us, we think we're really close. There's nothing that we need to do, or there's nothing we need to understand in terms of technology to go to commercial launch. We just need to continue the work. What have we learned? To your point, why do you do pilots? Why do you run 9,500-plus loads? If you think about our strategy, we've picked a diverse group of customers to work with, some LTLs, some FTLs, some brokers. We've done private fleets. We've done reefers, et cetera. For us, it's about learning all the different customer nuances that are going to allow us not to be safe to operate, but to actually scale a trucking business. We've always had the end in mind.
Everything we do is to learn from the end. What are some of the things we've learned? We've learned and been reinforced that trucking is really freaking hard. You have to have a lot of appreciation and respect for the trucking companies. I think on-time performance is sacrosanct. It is super important, especially if you're a FedEx, and you have to get your loads delivered over to the regional sort facility before the date. Otherwise, the packages are going to be delayed. On-time performance is super critical. For us, because we're operating first to start off on a terminal and a terminal, but ultimately, we'll go end to end, trailer quality is really important. Trailer quality is kind of mixed within the industry.
One of the things we realize is we're not going to change the trailer quality of all the customers and all the trailers, the hundreds of thousands of trailers out there. We need to have on-site support for repair and part storage so that if a trailer is not up to the standards, we can repair it so we can have an autonomous truck going down the road. Those are some of the things that are really important. Some other ones is that customer demand and density varies by lane. You have to work with a multitude of customers to be able to prioritize kind of your lane rollout strategy. I think the last thing for us is if you're going to scale, every customer has specific systems and processes that they use to integrate within their network.
They are not all the same. We need to make sure that we have a robust, we call it Aurora Services, but we have a robust platform with APIs that can integrate into multiple systems to be able to integrate into their network. The goal of adoption is for us to integrate within their network, not for them to integrate within our process. Those are some of the big things we have learned and uncovered over the last couple of years.
That's great. That's super helpful. Actually, you just made me think of something.
Uh-oh.
No, no, in a good way. When truck drivers, they are, and please correct me if I'm wrong, they're limited in the time period that they can actually be behind the wheel. At some point, they have to legally stop and rest. That presents a use case for, say, the Aurora Driver, which doesn't need to rest. The idea is that you can potentially deliver these payloads faster, more efficiently, without having to, without increased stops along the way. Is that a fair assessment?
Yeah. Truck drivers are limited to 11 hours of service today or 14 hours on the time, but 11 hours of driving. What happens is when customers plan their network, they have to determine how I want to get goods from point A to point B, what time frame do I need to deliver it on? Do I want to do a team driver, which is really expensive, and there are not as many team drivers as there are solo drivers, et cetera? What autonomous technology unlocks is a whole new way to reshape your network. Before, what would have taken two-plus days to, let's say, delivering from Los Angeles to Dallas, just because I know the route, that would take two-plus days to deliver goods, you can do it in a single day with an autonomous truck.
We are actually going to, with our customers, create unique AV shipping lanes, which allows them to offer a different level of service. It is good for society. Your produce gets there faster. That is something that is really valuable. It really allows the network to be more efficient. The trucking industry is highly complex and really efficient, and it has been around forever. Sometimes, if you really want to have a step change in kind of how the network works and the efficiency, you need technology. We think AV technology is one of those enablers that will allow us to transform the industry generally. Really excited about it and a great point.
Got it. I want to maybe turn to commercialization. You are entering into service next month. Very exciting. I am wondering if you can maybe share with us your vision for that commercialization strategy. What would that look like, and maybe what are your plans to ramp up for the rest of the year?
Yeah, I think we think about it more in terms of, I mean, we're super excited about a commercial launch in April. I think that'll be a great milestone. What we really think about is 2025 and what's the objective of 2025. For us, the objective of 2025 is fulfilling the technology promise. We are going to start off in a crawl, walk, run approach for 2025. We are going to launch a small fleet of up to 10 vehicles that operate driverlessly on our launch lane, which is Dallas to Houston. It will be on our first-generation hardware. It is hardware that we've designed, developed, and manufactured and built up internally. It will be on a platform that we can operate autonomously.
What happens is we'll start off, we'll maybe do one truck the first day and then gradually build up into a level of trucks. We will then start to expand our capability. We'll start off in daytime. Then we'll add nighttime. Then we'll add rain or maybe the inverse of that, depending on which order we go. We will start to operate on a second lane, which is our Fort Worth to El Paso lane. We will expand that lane into Fort Worth all the way to Phoenix. Towards the end of 2025, we will introduce our second generation of hardware. This has also been designed by us, but it'll be manufactured by our contract manufacturer, Fabrinet, based in Thailand. That's what allows us to go from tens of trucks to hundreds to over 1,000 trucks.
Twenty twenty-five is all about delivering on the technology promise. Then twenty twenty-six will be about focusing on the scaling and the financial aspects, like getting a positive gross margin.
I would just maybe add, particularly for those folks who are maybe not as familiar with the industry, your plan, I know you know this, is not to manufacture trucks over the long term. I just want to make sure everyone understands that point. You are, in fact, installing your hardware and software system into these newer trucks.
Yeah. In the early stages, we're basically upfitting our hardware and software onto a truck, ensuring that the truck has got the necessary redundancies. In the long term, and this is part of our partnership ecosystem, to be able to go to a scalable kind of production schedule where you're building tens of thousands of trucks, those need to be built by the OEMs. The hardware kit that would be supplied from Continental needs to be delivered to the OEMs and built line- side, like any other option that you would get in the trucking business. They would just be, and in the long term, we have a Driver-as-a-Service business model. That means that the OEMs sell the trucks equipped with the Aurora Driver hardware to customers, and then they would subscribe to our Driver-as-a-Service.
At that point in time, we do not own any of the assets. We have a very asset-light business model in the long term. Essentially, what the customers are doing is paying for a service while they own a truck. The difference to them will be instead of a human driver driving the vehicle and them paying a human driver and having to deal with all the challenges of a human-driven fleet, they operate a mixed fleet, some with human-driven and then some autonomously driven.
Got it. That is one of the main themes of the business, to have an asset-light and high-margin business model. You touched on this a little bit in your prepared remarks, but I am wondering if you can maybe expand on your partnerships with Continental and also with NVIDIA. The Continental one, maybe we start there, 2027 high-scale manufacturing. This is a big catalyst, big inflection point. Just help us understand what that means and how you are going to leverage that partnership.
Yeah. So again, if you think about it, we have the first generation of hardware, which is what we're going to launch our trucks in in April. We go to our second generation of hardware. Ultimately, to fulfill on the mission, we need to be able to deliver hardware at automotive-grade quality in high volume. When I say automotive-grade quality, this is for the trucking industry. The trucking industry, these trucks operate more than a million miles. We need a hardware kit that lasts at least a million miles. We need to be able to build it at scale. We are not an upfitter. We do it today in small volumes. To be able to build at scale and to have serial production, you need to actually deliver it line- side to the OEMs. We chose Continental because they're a great Tier 1 supplier.
They have a strong reputation. Also, the other thing is that we have an aligned interest. For us, we've developed a business model. It's a hardware-as-a-service business model. It's kind of an industry first. We're co-designing the next generation or the third-generation hardware kit. They will manufacture it. They will finance that kit. They will service that kit. They will warranty that kit. Ultimately, at the end of the day, instead of us buying the parts or even the OEMs, the traditional model is an OEM buys a part from a customer. Instead of us buying the hardware kit, we're going to pay a portion of the per-mile fee that we receive from the carriers. We're going to pay cents to Continental. They're getting a cents per mile kind of payment for this. It's the first of its kind. It hasn't been done.
There's benefits to that because it's reoccurring revenue for them. It's more predictable. It allows them to plan their business a little bit more reliably. They're making a big investment because they believe in the future of autonomous technology. As a matter of fact, all of the upfront NRE that they are spending is going to be amortized in the cents per mile that we are paying them. They're making a big strategic bet because they believe in the future of autonomous technology. We're really excited. This partnership is going on today. We are continuing to develop and work together. We're getting ready to receive A- sample parts for this. It won't come to fruition in terms of putting their generation kits onto trucks until 2027.
Got it. Maybe switching to financials, I'd be doing us a disservice if I didn't touch on that. Can you just maybe just remind us, what is your liquidity status and what are your capital needs?
Yep. We have $1.2 billion at the end of 2024. We expect that that will take us well past commercial launch into the second half of 2024. We've said that we will need.
Of 2026.
I mean, sorry, 2026. We've said that we will need to raise again. We expect to achieve positive free cash flow in 2028. We've been pretty consistent about that over the last, I don't know, two years or so. We're going to continue to think about fundraising opportunistically. Our goal is to ensure that we have sufficient liquidity so that our employees, our customers, our partners are never concerned about our liquidity status. At the same time, we also want to capitalize and raise money when it's the right time. We think we have a lot of catalysts that will be coming forward this year that will give us plenty of opportunities to do an additional fundraise within the marketplace. We're going to continue to be opportunistic about it. No firm plans on when we might raise.
I think if you look at our track record, again, we want liquidity not to be a concern for any of our constituents.
Makes sense. We have about five minutes left. With that, I'd like to maybe turn it over to the audience to see if we have any brave souls that want to ask a question. Otherwise, we'll call out. I'm kidding. Any questions from the audience? Yes.
Thanks for the presentation. I was wondering if you could refresh us on your path to commercialization, the specific timeline. With your closest competitor going back to Asia, how's the competitive landscape looking?
Yeah. If I think about the timeline, 2025, we launch. In April, we will launch a small fleet of driverless trucks. 2025 is all about the technology promise and operating exactly as we intended, so in more conditions to prove that it's a scalable business from a technology deployment standpoint. 2026 is about growing the fleet and focusing on positive gross margins, growing the fleet into the hundreds to over 1,000. 2027, 2028 is all about landing on that serial production where we've got the Continental arrangement delivering line side to the OEM partners. That is when we start to go into the tens of thousands of trucks. We think in 2026 we'll be operating in a variety of states across the lower U.S. I think in 2027, we should cover the vast majority of the southern U.S. in terms of technology deployment.
In terms of competitiveness, I'll focus in on the U.S. because that's where we intend to operate. I won't dive into the other markets because I think the opportunity is enormous for us, even in the U.S. We will talk about global at some other stage, but maybe not today. If I think about the U.S., I kind of think we're in a pretty clear leadership position today, and it's kind of ours to lose. There has been a lot of falloff of people who may have been perceived as leaders in the past and for whatever the reason they've fallen off. We think there's opportunity for more than just Aurora to be successful. Frankly, we think we have a pretty big lead in the marketplace, and we're going to focus in on the execution.
There are companies like Torc out there that have talked about 2027, and they've got a great partnership with Daimler. We think they might be a credible partner in the future. There are other ones that have been talked about, but we still think we're pretty separated from them for a variety of reasons, whether it's the leadership team, our liquidity position, our partnership ecosystem, our technological approach, or the fact that we've been just doing it for a long time. I think we've got a good opportunity to really lead here. Again, I think the market is enormous, and there's room for others.
I would just add, isn't it true that Waymo exited the self-driving business for trucking? I think they used to be in it initially, and now they've almost pivoted entirely to passenger, if I'm not mistaken.
Yeah. This is maybe over a year ago. They've talked about a pivot where they want to focus on ride-hailing, and they paused. I think the term they use is pause all of their trucking aspirations. I think there's an opportunity for them to come back into it at some point in the future. I think they've got a great opportunity in front of them in the ride-hailing space, and I think they're focused in on that for now.
Other questions? Yes, in the back.
Just one more. I guess I don't understand the business model of the OEMs. You'll be selling a kit essentially for autonomous driving to ultimately to the OEM. It's going to be a certain price that will be paid through Continental, or basically, I know there's a subscription you'll have afterwards, but will be free. Then basically, the assumption is basically you'll be getting the money over X number of years through SaaS.
Yep. The long-term business model is Driver-as-a-Service, which is just like SaaS. In this particular case, the OEMs would not purchase the Aurora Driver hardware kit. They would just install it onto the truck. They would not own the kit. That would be owned by us. That vehicle with the Aurora Driver hardware kit would be purchased by the customer, just like customers buy trucks today. They would not be paying a markup for the Aurora Driver kit because they would be paying that down in terms of a dollar per mile basis. I am not going to say in the future we would not look at some upfront payment for a portion of that. The long-term plan is that the OEMs never buy the kit. We own the kit. We recover that in our Driver-as-a-Service fee.
Huge working capital.
It is. One of the benefits of the partnership with Continental is they're actually financing the hardware kits. We don't actually own them. They will be financed by Continental. For us, super asset-light business model, no working capital adjustments that are needed for them. I think it's important to understand.
Do you recognize the revenues from Continental then?
We will recognize the revenues of what we get. We will pay Continental just like as a cost of goods sold, like any other cost that gets us for every mile we drive. We would pay them cents on the mile for the hardware kit.
That's right. The CapEx is quite small as well on an annual basis.
It is important to understand the reason why this not only is it attractive, but it is really the way the market operates generally. The market today, a shipper pays a carrier a certain cents per mile to deliver a load. That is basically what we are doing. We are trying to replicate that system. It is important for us. It is important for our partnership with Continental. We should make money together when we are actually providing a service. For us, there are aligned incentives all the way across the ecosystem, which is really helpful, actually. It avoids a lot of debates. It is also super useful. Now, for us, when we first launch, we are going to own a small fleet of vehicles that we will operate, and we will operate more like a carrier for our customers. The reason we do that is for customer adoption.
It allows the customers to see, experience, and start to integrate the AV fleet into their business without having to put the upfront capital in.
Wonderful. I think we have to leave it there, unfortunately. You heard it here, self-driving trucks to commercialize next month. David, thank you so much for your time.
I really appreciate it.
David from Aurora.