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

Good morning, everyone. I'm Stacy Feit, Vice President of Investor Relations, and welcome to Aurora's 2024 Analyst and Investor Day. It's so great to see so many familiar faces here in Pittsburgh. We really appreciate you spending the day with us. And thanks to all of you on the live webcast for joining us as well. We're really excited to share our plans to scale the autonomy in the trucking industry to commercialize the Aurora Driver, and provide a little more detail on our commercial opportunity, our customer value proposition and our financial ambitions. Before I turn it over to Chris, one housekeeping item I need to note, which you're all super familiar with, but bear with me as I have to read this. Our statements regarding the company's future business, economic models, achievement of milestones, anticipated time frame, financial performance and future prospects are forward-looking statements we make pursuant to the safe harbor provisions of the Private Securities Litigation Reform Act of 1995. These statements are based on assumptions and beliefs as of today and assumptions about future events, involve risks and uncertainties that could cause actual results to differ materially from our expectations. In providing projections and forward-looking statements, Aurora disclaims any obligation to update them, except as required by law. For additional information on important factors that could affect these expectations, please see the company's Form 8-K, issued this morning, and our financial report on Form 10-K filed with the SEC for the fiscal year ended December 31, 2023. Our discussion today may also include non-GAAP financial measures. These non-GAAP measures should be considered in addition to and not as a substitute for or in isolation from our GAAP results. A replay of the webcast is going to be available for 30 days following the live keynote. And with that, I would now like to introduce our CEO and Co-Founder, Chris Urmson.

Thank you, Stacy. You always get the funniest part of our presentations. So good morning, everyone. Welcome to Pittsburgh. Thank you for joining us today. Aurora was founded in 2017. And since then, we've been executing against our mission to deliver the benefits of self-driving technology safely, quickly and broadly. I feel incredibly lucky to be leading a company that has such an incredible opportunity to have a huge impact in the world. Trucking is a massive market. It's a $1 trillion market in the U.S., driven by the 200 billion miles of truck drive on the road. The Aurora Driver is going to be essential to our customers. It's going to allow them to build more revenue while having lower costs. The competitive landscape, if you look at what we would have thought of as competitors a year ago, most of them don't exist anymore, or certainly on -- operating in the U.S. market. We're the only company with the necessary partnerships, both in the OEM space and in the Tier 1 space, to be able to deliver at scale. And importantly, we have liquidity to make it well through our commercial launch, thanks to an incredible set of blue-chip investors, who've been with us since very early, and many of you are here today. So thank you very much. This is going to allow us to build a hell of a business that's asset-light, has really nice margins, is going to create massive shareholder value we expect. Said simply, we look at this market, and we just see it as [ Aurora's ] to lose. Over the next 2 hours, we're going to share our view on the steps Aurora's taking to hit commercial launch, the revenue drivers and cost levers that we have from there, how we're going to get to scale, and where we'll see that huge potential of what we're doing and what that will mean economically to Aurora as we get into a self-funding -- become a self-funding company. As Stacy mentioned, we're going to share with you our view of the future. There are risks. I think you're all aware of that. We see those risks to commercial launch is measured in months, not quarters, not years at this point. Once again, we're confident that as we execute against what our plan is, we're going to deliver an immense amount of shareholder value. The core of our business is the Aurora Driver. It's a combination of cutting-edge hardware, software and the off-board services that enable vehicles to drive through the world. The first place we're going to bring that product to market is going to be in the trucking space. We're designing this product to meet the needs of our customers and address some of their biggest pain points. So let me walk you through a few of them. First, our roads are dangerous. In the U.S. alone, 6,000 people die in accidents with trucks on our freeways. The Aurora Driver can drive that number to 0. Safety is fundamental to what we do at Aurora. It's core to our mission. And importantly, it is the top priority for our customers. The trucking industry has faced and continues to face a structural challenge with not having access to enough drivers. The American Trucking Association expects that over the next decade, we're going to have to hire 1 million new truck drivers, and we just don't know where they're going to come from. The industry also struggles with incredible degrees of turnover, for the large fleets, you see between 60% and 90% turnover year-on-year. Can you imagine how hard it is to run a business in that space. It's amazing what these companies do. The Aurora Driver is going to be able to provide a stable and scalable supply of drivers for our customers. Human drivers are limited to 11 hours a day. It's a good reason. The Aurora Driver should be able to almost double that, if not more. This means that our technology is going to importantly allow our customers to generate almost twice as much revenue or more per truck, while moving their customers or our customers' customers freight more quickly to where it needs to be. And from a sustainability perspective, we expect the Aurora Driver to be able to reduce fuel consumption by about 10%. Fuel is currently, after drivers, our customers' second largest cost. And so this will have a meaningful benefit to their total cost of ownership while also being good for the world. And lastly, we're all very familiar with the high cost of insurance in the trucking space. We expect the Aurora Driver not just to be safer on the road, but to provide an unprecedented level of visibility into the collisions that do occur. We expect these 2 things together will help drive the cost of insurance down and further help our customers. Trucking, as I said, is a $1 trillion market in the U.S. It offers really attractive unit economics, and it has real need for this technology. Globally, trucking is a $4 trillion market. This is a massive, massive market. To maximize penetration into what we see as this huge market, we know it takes a village. And we've built what we think is the premier ecosystem in all of automated vehicles. Our trucking partnership is just unmatched and gives us incredible competitive advantage. First and foremost, we need to work with the OEMs. It's absolutely critical that we work with them to deeply integrate the technology into their vehicles, not just to be able to launch 1 or 2 of these, of course, you can bolt something on and do that, but to be able to have safety as we deploy and to be able to do so at scale. That's why we've partnered with Volvo and PACCAR because without them, you're just not going to be able to build a real business here at all. We're excited by the way to have Volvo here with us today, who's going to share more with you about the partnership and how we're working together. We have a truck, we have a driver. Now we need somebody who needs to use it. That's why we've been autonomously pulling loads for some of the leaders in the logistics space for years now. We work with -- we pull loads for FedEx, Hirschbach, Schneider, Uber Freight, Werner, Volvo Autonomous Solutions and others. These pilots are important because they allow us to learn what matters most to our customers. So the product we deliver is actually useful and valuable to them. We're really excited to have Hirschbach and Werner with us today, and they're going to share with you what -- how they see automated vehicles impacting their business and how we're working together. So look forward to that conversation. At some point, all trucks are going to break down. You need to have a partner that can keep them on the road. That's why we've put in place a partnership with Ryder to do on-site maintenance to make sure that we get as much uptime as possible from these vehicles. And finally, last but not least, we have a first-of-its-kind partnership with Continental, to manufacture and support our hardware kit at scale. I'm thrilled that Continental is going to be with us today and share with you how we're working together to deliver AV trucking at massive scale. So let's dive into our path to commercial launch. As we look at the path to commercial launch, we see 4 things that need to be true. First, you need to have a technology that's safe and proficient on the roadway. That, we expect to be in place by the end of the year, and we see the timing risk in that as months. The second is that you need to have customers that are ready to bring this technology into the market. We're excited to have customers who are on this journey with us, and we expect as we execute, they're going to be excited to bring the product to market. You need to have a regulatory environment that's accepting of the product, and what you hear today is that there are no barriers from a regulatory perspective to launching a product. And finally, you need to have a fully redundant and OEM-validated and certified truck. We're excited that later this year, we're going to be able to receive one of those trucks, the initial platform we intend to launch with. And then from that point forward, we expect to have -- be able to solidify timing as we get into final validation and on-road approvals with our OEM partners. Over the course of the morning, we're going to show you why we're so confident in this ecosystem, our plan to commercially launch and then to scale the business. So with that, I'd love to welcome Yanbing Li on the stage. She's our Senior Vice President of Software Engineering, and she's going to give you a little bit of insight on some of our technology.

Thank you, Chris. Thank you. I'm excited to provide a quick update on our technical progress. Since becoming feature-complete about a year ago, we've been focused on refining and improving Aurora Driver's safety and performance and also validating to get us ready for a commercial launch. And I'm excited to share that the Aurora Driver has been performing very well in both nominal driving conditions as well as navigating complex scenarios. So can I see a show of hand, who's taken a ride in one of our trucks? Looks like many of you have done so. How was your experience? Okay. It's probably very exciting, riding in an autonomous truck down the highway at a very high speed. But a few minutes into the right, you likely find out there's nothing robotic about that experience. It actually felt quite natural. For those of you who have not done so, this is a hyperlapse video showing the Aurora Driver pulling a commercial load from Dallas to Houston. As you can see, this feels like a pretty natural ride. Actually, most of our rides today look like this, fairly natural and even boring, in the best way. This is also an example where we're pulling the commercial load in 100% end-to-end autonomy. Of course, the Aurora Driver also encounters complex scenarios. In this video, what we're showing here is the Aurora Driver detects a pedestrian on the side of the road on the Dallas to Houston [ lane ] at night, 300 meters away. On the top, this is what's shown in our camera view. As you can see, it's really difficult to perceive much of anything at night, let alone a very dark, small pedestrian on the side of the road. But thanks to our proprietary FirstLight Lidar technology, the Aurora Driver was able to detect it and also proactively performs a lane change, so that we create a safe passing buffer, in a whole 11 second. 11 seconds have passed between the Aurora Driver's detection and the human eye detection. And that is the power of technology. And this is just one of the many examples where the Aurora Driver confidently handles those complex scenarios. In order for us to focus on our commercial performance and profitability, one of the things we pay attention to is our operating efficiency, especially how often the Aurora Driver requires on-site support. We introduced this measurement of the autonomy's performance indicator, or API to measure just that. So this metric [ penalize ] anytime we use on-site support, which is the most expensive form of support. As you can see from the chart, we've been making very nice and consistent progress in the past few quarters along on this metric, and now we're standing at 99% in aggregate across all the different loads that we've put in Q4. Myself, having grown up in an Asian family, 99% is simply not acceptable. So we share that similar mentality at Aurora. So rather than focusing on that 99% aggregate, we actually measure how often we achieve 100%. And this is the chart that shows that progress. And we also heavily scrutinized every time we don't hit that 100%. Again, very positive progression on this measurement in the past few quarters and now at about 62%. Within engineering, we continue to work, improving our commercial performance so that we -- by the time of us getting ready for commercial launch, we expect this number to be at around 90%. So in addition to measuring the autonomy performance on road with the supervision of our vehicle operators, we've also been conducting close track, high-speed driverless testing. This is an example of video showing just that. For those of you attending in person today, you're also in for a treat because we will invite you to experience a driverless truck ride and also to observe how the Aurora Driver handle really complicated scenario. There is an extra bonus, that is Chris, may be your personal onboard narrator to take you through that experience. So with that, teaser, let me hand it back to Chris.

Thank you, Yanbing. Yanbing is not just our Head of Software Engineering, but also my personal PR leader. So thanks, Yanbing. To convince ourselves and our partners that we're ready, we use a safety case. Safety case is widely used in other safety-critical industries, like rail, aviation and medical devices. It spells out why we have confidence that not just the Aurora Driver, but Aurora is going to operate safely when we have these vehicles on the road with nobody in them. Our safety case has 5 core pillars. The first is proficiency. This means that when the truck's on the road, it drives well. Second is fail-safety. This means that if something breaks, we're able to detect that it broke and mitigate the risk associated with it. Third is that we're continuously improving, that we take the lessons learned as a company, but also from what the Aurora Driver sees on the road, and we use that to make the product better, and we make sure we can close those loops. We have to be resilient, which means that we have to think about foreseeable misuse. You have to think about things like cyber-attacks and make sure we mitigate the risk associated with those. And then finally, we need to be trustworthy. That means we have a culture where people feel confident to speak up, raise concerns and that they're going to be addressed appropriately, and that we're proactive in the way that we engage with our regulators and policymakers, so they can make informed decisions as we move forward. These pillars translate into about 460 claims or work items that ultimately kind of make the argument that we do all those 5 things, and ultimately, we're safe enough to be on the road. We've been quantifying our progress towards this through the Autonomy Readiness Measure, which is a way to estimate the amount of work in person hours that needed to be done to close that safety case. And we're excited about the progress we're making. As of mid-January, we've done about 93% of that work for launching on Dallas to Houston. A key element of that safety case, of course, is that first pillar of proficiency. And that means how do we -- and to close that part of the case, we need to validate the Aurora Driver. When validating a self-driving system, one of the challenges is that many of the interesting things that happen on the road, the dangerous things are thankfully rare and unlikely to occur. And so we were clear from day 1 that on-road testing was not going to be enough, that validating required a multimodal approach, a combination of on-road testing and virtual testing. And this is why we made such foundational investments in simulation. By utilizing the Aurora Virtual Testing Suite as part of our validation strategy, we've been able to tame the long tail. The Aurora Virtual Testing Suite allows us to both simulate and replay the data that our trucks see on the road, and we can create the equivalent of about 50,000 trucks on the road, which is just an unimaginable scale to be doing for testing and development. To tame the long tail, we break down kind of the scenarios that the truck would see on the road into 3 buckets. The first bucket is the stuff we see every day. The second is the things that we might see once in a while. And the third is the stuff that we don't really expect to see any time soon or maybe never see. So for the first bucket, this is easy. We got lots of data for this. We build some tests. If something doesn't work here, we're going to learn about it very quickly and be able to address it. For bucket B, this is the stuff that we see once in a while. Here, we take that on-road event. We create a simulation of it. We make sure the Aurora Driver can solve that problem and drive safely through that scenario, but we also importantly create variations of that scenario, where we adjust the speed or the timing or the offsets to ensure that we didn't just get lucky that onetime, that faced with a similar situation, we're going to be able to handle that as well and be confident that the Aurora Driver will be safe there. And then finally, for Bucket C, these are the scenarios that we basically haven't seen on the road. We synthetically generate simulations using NHTSA's collision taxonomy. This is the way NHTSA's broken down all the way vehicles get into crashes. And we create -- and we expect, sorry, the Aurora Driver to respond similar to a person. If the collision is avoidable, the Aurora Driver should avoid it. If the collision is unavoidable, then the Aurora Driver should do what a good human driver does, which is take the energy out of the system, hit the brakes, try to minimize the negative outcome in that scenario. Success across these sets of buckets will give us the conviction the Aurora Driver is ready to go. And we do that through building tens of thousands of simulations that we're running regularly. In addition to how we validate, we've actually architected our perception system to further tame the long tail and ensure that it isn't a barrier to launch for us. The Aurora Driver is designed to identify vehicles, people and signs on the road. This is almost everything we see on the road, of course. But we also understand that almost is not good enough. And so we've built a system that is a backstop for this. And what it does, it looks for matter in the world and reports them as what we call generics. And this gives us redundancy in our perception system that's equivalent to -- or like the redundancy we have in our multimodal sensor system. So in the video you're looking at here, the blue boxes are what our primary perception system sees. And you can see we're seeing all of the same stuff with our backup system as well. It just doesn't understand, in this case, the backup system, that these are vehicles. It's just stuff moving through the world. Our [ redundant ] perception system enables the Aurora Driver to identify and respond to unusual objects. So on the image on the right here -- on the left, I'm sorry, there's a helicopter being towed down the road on a truck. I think we've seen this exactly once. It doesn't occur very often. And what you can see is that the primary perception system identifies this and sees it as it should, as a truck and a trailer. And we're able to track and that's the blue box. But we also have a backup system detecting it. And so if this has been a low flying helicopter traveling down the road or a helicopter that had landed in the road, we would have confidence that we'd be able to see that and respond to that appropriately as well. On the right-hand side, what you're seeing is we're driving down the freeway, and there's a mattress in the middle of the road. In this case, this is not a person, a car or a sign, but the generic system is able to detect it, and the motion planning system is then able to make a decision in response to that generic object in the world and react to it. And so part of the testing we do is to make sure, even in the case where we're using the generic backup system, our motion planning system is responding appropriately as well. Beyond the core validation work that we've done, we wanted to ask what is the real-world impact of this. And so we looked on our launch corridor in Texas, I-45, between Dallas and Houston, and we pulled all of the reports of fatal collisions that involved a tractor trailer. It turns out there were 29 of them where the Aurora Driver could have been operating in the vehicle that you initiated or you can think of as cause of collision. We then simulated them and ask what would happen. And it turns out that none of those collisions would have occurred. Said simply, the people involved in those crashes would be alive today. That's huge. When you take a step back and think about the 6,000 people that died in truck accidents across the U.S. and the 500,000 truck accidents that happen every year, there's an incredible opportunity that we see with this technology, making the world and the road safer. As I mentioned earlier, safety is absolutely critical and top priority for our customers. So I think now is a great time to transition to Zac Andreoni, our Vice President of Business Development, who will talk a little bit about our customers.

Thanks. Thanks, Chris. Hi, folks. As Chris mentioned, I'm Zac Andreoni. I head Business Development here at Aurora. You just heard some great updates from Chris and Yanbing about the status of our technology and the progress we're making, but we know it's going to take more than great tech to be successful. We have to launch a product that's valuable to our customers. One of the ways that we do that is by ensuring that we're partnered with the most innovative, forward-thinking carriers in the space, just like our relationships with OEMs, our partnerships with suppliers and basically every other facet of our go-to market, I believe that our customer relationships are the strongest and most mature in the space. You guys can come up. Here are a couple of stats to service some proof. Through January, we've hauled over 4,300 loads for our partners commercially. That represents over 1 million miles on highway. And all at nearly 100% on-time performance, which I know our customers really appreciate. Currently, we schedule over 100 loads a week for our customers. Throughout 2024, we'll continue to demonstrate quarter-over-quarter progress in this metric as we ready ourselves for scale. But we don't just haul these loads for the hell of it. Each of them allows us an opportunity to test different use cases with our partners, to make sure that our systems are integrated appropriately in a variety of other things. But I have a bias, so we've invited a few customers here today to tell you a little bit about why they think pilots are valuable, why they are invested in autonomy and more importantly, why they're partnered with Aurora. So gentlemen, can you come to the stage. Joining me today are Richard Stocking, Co-CEO of Hirschbach Motor Lines, one of the largest refrigerated carriers in the country; and Daragh Mahon, Chief Information Officer of Werner Enterprises, one of the largest FTL carriers in this space.

Gentlemen, thank you for joining me. Let's get started. My intro on you was very brief intentionally. I wanted to allow each of you to maybe share a bit about your respective firms. Daragh, can we start with you?

Yes. So I'm with Werner Enterprises. Werner was founded in -- 68 years ago. So we're 68 years old. We love to tell the story. We had -- our founder started the company with one truck. And today, we are one of the largest full truckload carriers in the U.S. So we have approximately 8,000 trucks, 30,000 trailers and most importantly 10,000 drivers. As you would expect, in our asset division, we do dedicated van one-way freight, but we also have a large brokerage operation. We cover Mobile, Power Only. So pretty much every mode that you could imagine, we're involved in some sense. So we -- as we'll talk today, we put safety first on everything we do. We have a saying at Werner where we say, nothing we do is worth getting hurt or hurting others. And we also put a huge focus on our drivers and our driver community. So that's Werner in a nutshell.

Great. Thank you. Richard?

Yes. I'm working with Hirschbach, and I absolutely love this organization. It has been around a while as well, over 85 years. It's had a handful of owners. And currently, Brad Pinchuk, who is with us today, is sole owner of Hirschbach. The reason I love this company, it's very focused on delivering better lives to our country. We feel like we help feed the country and hauling protein and produce and groceries and so on and so forth. So we feel like we're involved in very noble purpose, and we do that through an OTR business unit as well as dedicated. We also really focus on an expedited division that we have that I think lends itself to autonomous. A couple of unique situations. We also have a solutions logistics arm as well. But we also haul singers around the country, country music singers. And then we're veteran owned. And we are -- we just received that distinction last year, very proud of that. And we're one of the largest out there, and we have some pretty aggressive growth goals.

Great. Do you get that Taylor Swift contract or still working on it?

No. Not yet.

Okay. Great. Cool, I feel like I know the answer to this because we work together quite frequently. But why or how do you guys, as executives of carriers believe that AV is going to transform the industry? What are you excited about?

So I think the first thing we want to say is -- and Chris touched on this earlier, but first of all, drivers, right? We focus on drivers, we focus on our employee community, the experience the drivers have. And we feel like AV can add to that experience. There's no doubt we want to say this loud and clear that in the next decade, ATA and others have given us the stat that we need somewhere between 800,000 and 1 million drivers to enter the market on top of the 3.5 million we have today. So it's really important that we consider that as we talk about AV and how AV fits into it. I think for us, as an organization, where we feel excited about this is how AV can help us attract drivers because we feel like it will allow us to create better jobs. The driver community today, it's not the 1980s, 1990s where there was a -- teams were a great thing and going out the road for 3 or 4 weeks was somehow romantic. I mean, that doesn't happen anymore. So what we feel like is our driver community, they want to be home at night time or home 2 or 3 times a week. They want to have short length of haul. They don't want to necessarily drive from Dallas to Atlanta anymore and spend hours and hours on lonely roads. So that's where we feel like is one of the most important things that will attract drivers, that we feel like we can attract drivers because we can offer better jobs. We also feel like it has a very -- we'll have a very unique way, AV, of fitting into our network. And again, what I've given away, the farm, as you said earlier. We feel like there's a definite benefit to how our network would operate, especially with utility of vehicles. So we feel like there's a lot it has to offer. We just have to overcome some things to get there.

Thank you.

Yes. And I would echo that. It's all about the driver and delivering a better life to them. They want predictability in their home time, and they want predictability in their paycheck, just like all of us. And so we feel like this will help develop a better job, a better career for our drivers. We also believe that it's going to improve safety, and maybe down the road, reduce insurance costs and so on and so forth. So we think it will transform that part of the equation. And I think Chris mentioned this, which is very important as well, is utilization on these trucks. I mean if you look at just one driver, driving these trucks, you're only utilizing the power unit, 35-ish percent of the time. And if we can double that, we feel like that's really good. It helps take trucks off the road because you have so much utilization. And we believe that it will help the supply chain, could take the days off of, let's say, strawberries for instance. And that's going to be a benefit to the customer community. And really, I think the service that we're able to provide, you can haul some of the solo-type freight with team-type service and kind of changed maybe the amount of product that they have to store and the inventories could change, too. So there's a lot of different ways I think it really could revolutionize this industry, which is much needed.

Can I go one level deeper on that? I think there are 40 or so people in the room super eager to hear how you're going to utilize this in your network. I don't want you to give away any secret sauce. I know Daragh sort of already touched on this. But you, in particular...

Yes. I think how we look at this is we want to start with our power lanes. So those lanes that create that overnight or 2 or 3 nights away from home-type scenario with our drivers and start there and really utilize that in those lanes that we have huge density that we could obviously improve that service, maybe lanes where it requires some team-ish-type service. I think that's where we'll focus first.

Awesome. Thank you. Okay. So when it comes to selecting an AV partner, what do you look for? And what was it about Aurora that stood out to you as a sort of differentiator?

Yes, let's say, in spite of Zac, we think Aurora has been a great partner. We're working on that. But now, honestly, the reality is that when you look for we've looked at all the AV vendors out there. We've talked to every single one of them. I think you start with -- what I'd fill in their buckets too much, but when you have Chris and Sterling at the helm, right? So great reputation. You know these guys are going to build good tech at a good company. So that helps. When we think about safety, which is the first thing we think about every day when we wake up, we believe that Aurora have exactly the same emphasis on safety that we do. So that helps a lot. Like I said, we know the tech is good. And we feel like, in general, the partnership has been fantastic. And this is where it's really important to me as a technologist, I don't believe in just buying a product and handing it off to me and let me go [ run ] with it. So we feel like we've had a very open, transparent, very clear conversation about what we expect and about what you guys are delivering, and for the most part, we're on the same page. So I think that helps tremendously as well. So great partnership, and we're really happy with where we're going.

We'll see what we can do about the Zac problem. I don't know that it's going away, but I'll ask Chris.

You'll ask Chris. The top 3 things is a safety case, safety case, safety case. It's all about safety. And I've worked with a couple of other groups that, as Chris said, aren't around anymore. This is a real deal. They listen, they're humble. I think in meetings we've had with Chris and Sterling, they're probably the smartest people in the room, and they learn the most because they're humble, and they listen. So I think the improvement that they -- or the process that they've gone through looking at this and the reasons why and it can help. I mean, think about those lives saved. That's real. And I think it's kind of a character and competence thing. They've got great character, they've got great intentions and the capabilities, as you just discussed, are just awesome. Being in this truck as many of you have, it's just simply impressive. And we're just extremely excited about that. But one that is well funded, which Aurora has been, to get this technology where it needs to be, it takes a lot. And I think because of the track record and the vision, the leadership, the structure, the processes they have in place is second to none. I mean, it's just not even close. And so I believe that, that's helpful. And one that will be involved in the business with this and really build this together. They've not done this in a vacuum. It's -- you've reached out, and I think that's part of your success.

Thank you. Okay. Let's talk about readiness. When we are ready, how will you know that you're ready? How are you assessing the readiness of an AV company who says, "Hey, it's time to pull the driver"?

Yes. I think the truthful answer to that is we don't yet know, right? You said it and Chris, we're on this journey together. And I think, there are barriers to entry, but we're working on all of them, right? So for us, safety is number one, the safety case has to be proven out. And again, we feel like you guys have the same emphasis on safety that we do. So that's important. But we're getting there. I think the cost, we've got to figure out the cost equation and how does that work out? And again, we've been working hard on that with you guys, and we feel like we will get there, but it has to get there before. It has to be cost effective for us to do this, and we have to be able to maintain a level of margin that we're used to. And then I think an awful lot of it comes down to regulation. And there's not a lot of barriers right now that we know about, but there was also an awful lot of regulation that has been written yet. So I think it's very hard to commit without knowing what that regulation is going to be. And I think the bottom line is we'll know when we're ready, and we know that we're working with you guys to get ready. And it sounds like not much of an answer, but I think that's the point, right? We just have to make sure that all those box are ticked. But we also don't have an exhaustive list that says, "Hey, if we tick these 5 boxes, we know we're ready." I think some of those boxes haven't even been created yet. So we feel like we're on the right path. We feel like the technology is getting there. I mean, I've looked at -- we've been looking at Aurora as a technology for 3 years. And every single time we get in the truck, it's gotten a little bit better. And that's important. So we got 4 or 5 things in mind. We get over those hurdles, we know that we're ready. But we've already committed as soon as the safety case is passed and you take the driver out, we're going to put 2 drivers in, for example. We're going to take 2 of our most experienced, million plus mile drivers, and let them sit there for a couple of weeks, months, years, whatever it would take until we feel like they're satisfied. So we'll get there. We're confident of that. We just got to work through all of those hurdles.

Great.

Yes. There's not much I can add to that. I think we're ready. We have a contract with you guys. We're ready to go when you're ready. We trust your process and the things that Chris talked about, the risk, and it's not quarters, it's months. And we feel like we're getting really close. But when you're ready, we're ready. And I just think we're trusting in that process and working together with you.

Awesome. Okay. I think we have time for one more question. I guess what I'll ask is, is there an unlock on our road map that you guys really feel like creates an opportunity for scale, something you're really excited about, whether it be a particular lane or anything like that?

Yes. I mean, I think, again, the unlock for us is safety. And we feel like once we get over that barrier, we feel comfortable, then we can start to look at our network and start to -- we already know where we think it will fit in and we already know what gaps we have or where we have trouble filling those gaps. And I think, again, once we get through -- we have the contract with you guys, and we're about to expand, and we know that we're going to move more freight with you over the coming year. And once we do that, it's just a matter of getting more and more comfortable as we move through the next year or so. But we already know. I mean, believe me, there's a lot of conversation around where we can slot in AV, where it fills the gaps in our network and how it helps us out. So I don't think scale will be a problem once we get to where we have our boxes. The boxes that we don't know yet are checked.

Perfect.

Yes. And we're same. We're aligned in a couple of areas, and we think we can scale big time in those lanes. They're ripe for growth. They're aligned for AV and aligned to help our drivers get that predictability and home time and paycheck that we talked about. And over and above that, we have a few other areas in lanes that we think that this will be great. So...

Awesome. Thank you. That brings us to time. Folks, please join me in thanking our guests for being here.

Okay. Before I hand the mic off, I have a couple of other brief updates. Daragh talked about readiness. And I think what you heard him say is like we still have some boxes that we haven't checked yet. I want to talk about a program that you're going to hear more about as the year progresses. Later this year, we're going to launch something called Commercial Readiness. We believe that removal of the driver includes all of our stakeholders. And so this program is going to allow our customers and other third parties to put folks in the truck for an extended period of time to evaluate the driver themselves. In addition to the ride alongs, this program will include training about how the system operates, how we safely launch and land missions, and how we will deal with an incident should one arise. This is a direct result of feedback that we've gotten during our pilots, another reason why those are super valuable. And to me, I don't think you're going to see much more transparency than this from any of our peers. This should represent the most transparent program that exists. So more on that in the months to come, but we're really excited about it. The next thing is that -- sorry, I missed the slide. Okay. And finally, I think, as I mentioned in my opening, we know it's going to take more than just good technology. We have a whole team of people working on solutions and tools to make sure that our product integrates seamlessly with our customers' businesses and meets them where they are. So you can see a demo of that in the back, and here's a little bit more in video form on what the Aurora services team is building. [Presentation]

Good morning. I'm Gerardo Interiano. I lead our Government Relations Team here at Aurora. As you just heard from our customers, they are ready whenever we're ready, and our partners and government and our regulators are no different. We get to work closely with government at federal, state, local levels with law enforcement. And one of the most common questions that we get is, what is our regulatory framework? What does it look like? As you can see from this map behind me, we can deploy in the vast majority of states. In the states where we've needed to change the laws, we've work to build coalitions, we work to get bipartisan support, and we pass the laws that are necessary to explicitly allow us to deploy in those states. There are some states where we haven't had a need to change the laws because the laws already allow us to do so. But the most important of those states to us today is Texas. So what does Texas look like? We get to work with the Texas Department of Transportation with the Texas Department of Public Safety, with the legislature, the governor's office. And the one thing that I can tell you is that just like our customers, whenever we're ready, Texas is ready.

Specifically with the autonomous vehicle industry, Texas was really at the forefront of some of the legislation to support that. And the legislation that really kicked that off here in Texas was passed all the way back in 2017, and it was Senate Bill 2205 that created the legal framework, if you will, that would allow for autonomous vehicles, both passenger and truck to operate on public roads in Texas. And as long as those vehicles met certain requirements relative to licensing and safety standards, those vehicles were permitted under that legislation to be able to, again, operate without a driver on public roads here in Texas. And we've seen that occurring in a few areas. Senate Bill 2205 explicitly allows for automated motor vehicles to operate in the state of Texas today. They have to abide by the same requirements that we have that nonautomated vehicles have to abide by. They have to be registered with the Texas Department of Motor Vehicles, and they have to be able to enabled the Texas Department of Public Safety and other law enforcements to be able to interact with and enforce the laws of the state. But 2205 really cleared the way for autonomous vehicles to be able to move forward.

As you heard from Marc, they're ready whenever we are. Whenever we close the safety case, we can deploy these vehicles in Texas. We work closely with the Texas Department of Transportation and their entire team with the Texas Department of Public Safety. It's about collaboration. It's about trust. It's about transparency. We have to build those relationships, and we have over the last several years. They trust us, right? At the end of the day, their mission is no different than ours. It's to make the roads safer. It's to reduce the number of fatalities on the roads. And we are mission aligned in that work, which is why this work and what we're doing with them is so important. Earlier this year, Chris was invited to testify before the House Transportation and Infrastructure Committee. We were the only company that was invited to testify because that trust is established, because they know that we are doing things the right way, because they know that we put safety at the forefront of everything that we do. I was in D.C. earlier this week. And what I can tell you is the federal government, the federal agencies that we work with understand the importance of this technology. They understand that the benefits that we are delivering are real, and they're looking forward to seeing that technology on the roads. We also work closely with law enforcement. We are working with them to make sure that they understand what -- how do they interact with our vehicles? How do we learn from them? What is it that they need from us, in order to be able to interact with our vehicles? We've been working with the Frisco Police Department. And just this week, we did over 200 pullovers with motorcycles, so that we can understand not just what does a truck or a motorcycle look like when they're pulling us over. I think, we want to understand and have that relationship and that transparency because law enforcement is key to a lot of the work that we do. So with that, I'm going to invite Sterling Anderson, our Co-Founder and Chief Product Officer to the stage.

All right. Thanks, Gerardo. As we've been saying since we started Aurora, close partnerships with OEMs and developers are essential, delivering a commercially-viable self-driving product. One simply can't reach the levels of safety, of reliability, of cost and of scale without them. And carriers and fleets expect that the major OEMs who produce the vehicles on which their fleets are so dependent, fully stand behind the product that they've entered into their operations. It is a key differentiator for us and an area of significant investment for us, for Volvo and for PACCAR over the last several years. Today, we'd like to talk a little bit more about our work with Volvo. Over the last few years, our teams have been working together to develop our first commercial launch platform with Volvo based on the revolutionary new VNL 6700. So it was just announced a couple of weeks back, Nils? And just last year, we built and began testing a late-stage prototype of this platform with safety-critical redundant systems built directly into the truck by Volvo. We've got a number of exciting milestones coming, but rather than tell you about those myself, I'm delighted to welcome Nils Jaeger, President of Volvo Autonomous Solutions, to speak a little bit more about the company and our plans together. With that, Nils?

Thank you, Sterling. So again, my name is Nils Jaeger, representing Volvo Group, specifically Volvo Autonomous Solutions. Great to be with you here in the room and also with you online. I will start with a quick disclaimer also. I also want to point you towards the last sentence here, which says that Volvo Autonomous Solutions is a minority shareholder in Aurora Innovation. I think that's important for you to know. This was written by lawyers. I would have stated Volvo Autonomous Solutions is an important shareholder in Aurora, but make out of it what you want. Before I get started, I want to talk a little bit about Volvo Group and what we do. And I think it is important to know we were founded in 1927. We are committed to shaping the future landscape of sustainable transport and infrastructure solutions, and it's every day that our customers, they use our products, they use our service, to transport, deliver, build, ferry goods, objects and people around the world. And it's our products and it's our services, which help to put food on our tables, to carry people to work, to bring our children to school, to build roads, keep our cities clean and much, much more. Effectively, we are helping to meet the growing demands of our worlds and our society. Now in the Volvo Group, we have 10 business areas. You will all know Volvo Trucks, you know Mack Trucks, you know Volvo Construction Equipment. In total, we have 10. I represent Volvo Autonomous Solutions. We were founded in 2020. And I will talk to you a little bit what we do at Volvo Autonomous Solutions. So at Volvo Autonomous Solutions, we're not only focusing on the development of autonomous technology, but we go further and we provide our customers with autonomous transport solution, and we're providing this also through new business models. Now the way we do it is different per industry vertical or per segment. But important is, it's not conflicting and it's not competing. What you see here on the slide is 2 different ways, how we operate on -- this side here on the right side, you see a full stack solution provided by Volvo. You see here an application in a quarry, in a mine, actually in Northern Norway. And for this segment, the confined segment, we've actually built our own virtual driver. And we are offering a full stack solution that means we're integrating our virtual driver into our truck, then we equip that truck with additional soft and hardware to then allow autonomous operations. And it is in this industry, where we're actually already today revenue generating. And we're doing this with true autonomous applications. So that means we have actually removed for this segment, already the safety driver we did that last year. But I'm not here today to speak about that segment. I will speak about the hub-to-hub segment today, which is also the single largest opportunity we see when we speak about autonomous technology. So for this more complex use case, on-highway transportation, there we have partnered with Aurora, and I think it is important to state that we have actually cooperated and collaborated with Aurora since the early days of the existence of Aurora. For me and for Volvo, this partnership is strategic. It is vital. It is very strong. And both parties act with big commitment and engagement as we have this objective to introduce to North America a new form of movement of goods into North America. Sterling and I worked very close together. I used to say we're that close together, the only thing which fits in us is an autonomy-enabled Volvo truck. What is Volvo contributing to this partnership? It is the autonomy-enabled Volvo truck, and then Aurora is contributing with the Aurora Driver, as you can tell. Now the integration, however, which is a crucial part, which is very important for the safety case, that's where we're working very close together. And we also have together a very well-defined commercialization strategy. This next slide here is a very important one. These are the elements, the cornerstones for both parties towards market success, and I will go through all of those 6 points one by one. I want to start with the first one, the purpose-built autonomous platform because it is very important that you also understand that, of course, it is very complex to build the autonomous driving system from Aurora, but it is also a very complex undertaking to build the autonomy-enabled truck. Everyone can do a proof of concept. Everyone can do a demonstration. But to put a truck on the road with redundancy built in is a completely different ball game. So I think this slide here, I need to go back to the 6 elements, they actually show why Volvo and Aurora together are the emerging leaders in autonomous transportation space. Jointly, we have the highest credibility. We have clear progress. And together, we're building the strongest ecosystem in the EV space. And to begin, I want to talk here, as I said, about our purpose-built autonomous truck platform. So at Volvo, we're building a very strong AV platform. We actually call it the Standardized Vehicle for Autonomous Transportation. The short cut is SVAT, and is built using the Volvo CAST system. So another acronym, CAST stands for Common Architecture & Shared Technology. And why is that important? It is important because this technology can be applied not only at the U.S. VNL truck, which you see in the middle here, but it can be scaled across the other truck brands we have within the Volvo Group. So it can be scaled to a Mack Truck, to Renault, but also to the European Volvo truck range. This allows us to scale them across geographies. Aurora and Volvo can scale across use cases, where we can also scale across different drivetrain options. Today, we start with the internal combustion engine. Tomorrow, it will be BEV trucks, battery electric trucks. The day after, it will be fuel cell and hydrogen. We will also be able to scale across different business models. So it is this platform approach, which brings us the scale, which brings us the leverage. With this, we have competitiveness, and this will enable us to establish a strong market relevance. In other words, we're really thrilled and excited to start building a new autonomous transport system for the U.S. hub-to-hub industry, but we know that the future holds more for us. Now Aurora and Volvo are a good match, as safety is in the core of both of our companies. Now in the case of Volvo, I could argue that safety is also part of our heritage. It was Nils Bohlin, a Volvo engineer, who invented in 1959 the 3-point safety belt, and I hope you've all used it today. And we've given this patent for free to the world, and it is estimated that it has saved more than 1 million lives. Why am I saying that? I'm saying that because I believe, together with Aurora, that autonomous technology can be the next big thing, the next big element to increase safety on public roads, but also on work sites. We have just introduced a new truck platform, which is built to change everything. It has new levels of aerodynamics, as you can see on this slide. It has new safety levels. It also has new safety features. It has improved fuel consumption and important, it's designed from the ground for autonomous use. Actually, requirements and capabilities needed for automation were also coming from Aurora. What you see behind me is not yet the autonomous version, obviously. We will launch it at ACT Expo in May. But to be not only in the business of PowerPoint, we've also brought to you a camouflaged version of it. Actually, Aurora has been using this truck since last year. And important about this truck, again, is the redundancy. But what we do with this new truck is that we integrate and we centralize the controls for braking, steering, propulsion, by that, we make it standardized, but we make it very efficient for the integration of the Aurora Driver, and again, purposely designed for the autonomous use case. The new truck does include redundant safety systems to ensure safety, even if a [ fault ] would occur in the vehicle. We're monitoring the status of the critical system, and we're having a backup ready to take over and that backup can be enabled in a moment's notice. The same applies for the commands being issued by the Aurora Driver, that generated and transmitted via 2 independent channels to ensure we have a fast and safe switch to the backup system. We're also extending the existing Volvo diagnostic capabilities for automation use, specific use cases. So here, we are ensuring a quick and reliable detection of any safety relevant failure within the vehicle and then we can ensure an appropriate reaction. Now we are in an industry, which, in the past, has received a lot of merits and a lot of goodwill. What I want to point out is that at Volvo and Aurora, we are laser-focused to develop our autonomous truck solution to the market, and we will convince with value realization. We have a very focused road map to deliver the first trucks for commercialization. And more important, we jointly have a clear understanding of what needs to be in place and delivered for production at scale from the main production line. You see the truck here. Again, officially, we will launch it in May. But now Aurora and Volvo are operating with safety driver. And we then will move also into testing with -- into commercial operations, again with safety driver. And of course, the significant milestone ahead of us is the removal of the safety driver, and again, we will do this with safety and redundancy in place and in mind. And we have a clear road map. So we know when and why we can launch, and we are proactively taking all the necessary steps to secure that the critical enablers are in place. Important is, of course, that we are undertaking already now the necessary steps to allow them the industrialization. So we work with our necessary partners within Volvo. We have created privileged access to scaling opportunities. We have all the required resources from Volvo in place. We will deploy our most modern factory. And from that factory, the truck will come from the line, with the Aurora Driver equipped, and I will come back to that. But what is important about that slide is Aurora and Volvo, we share a clear road map towards launch and towards scalable production, which then allows for both of us a real purposeful market introduction. But talking about the market, of course, it is important that we get the technology in place, but being sharp about the commercialization is as critical. And we believe that autonomous transportation is not just about replacing the human driver. It's something different. It's something new. It's a new form of transportation, which complements today's transportation modes. Today, we can transport by air, we can transport by train, we can transport by truck. Tomorrow, we will add autonomous transportation. But it is something different. It is something new. Now at Volvo, we have almost 100 years' of experience when it comes to transportation. And still, we realize we must diligently prepare for that. For that reason, we did establish a transport company. We have a motor carrier license. We've opened an operation center in Texas, and we are today deploying transportation. So we are, in fact, mimicking already today the future autonomous transport flow. We're pulling loads with standard trucks, with safety drivers for key customers. But when we're ready, when Volvo and Aurora together are ready, then we can really start operating. So what we're doing today is a very crucial step, but this step we've taken already. On the next slide, I want to talk a little bit about how we are doing the go-to-market strategy and how we approach our customers. We have realized in discussions with customers that this technology, particularly in its initial period will be something different. It will be new. And by that, it will be more challenging for our customers. So our approach is we will help them deploying it without them having the hassle of operating in themselves. So one way for us to launch is with Transport as a Service, where we will be a solution provider for the transport industry. We will make it simple, we'll make it transparent in how we create the value for the industry. The cornerstones for Transport as a Service is, of course, the autonomy-enabled truck and the Aurora Driver, but integrated into an ecosystem of autonomous trucking. We understand that not everyone will want to have that comprehensive solution. So we will also provide solutions like Truck as a Service in order to be able to respond to different demands, different needs from different segments and customers. But being Volvo, we have, of course, established customer relationships, and we have those since decades. So that allows us to operate from a good starting point. The next one is, of course, very important, manufacturing at scale. We see, we hear from the market that there is a strong demand for autonomous transport solutions, autonomous trucking solutions, and that demand is growing. And you all know the reasoning for it. I think Chris has talked about it at the beginning, and I think the overall economic benefits are clear to everyone in the room. We will make this happen. We'll bring this to light, but what we're after is scale. We don't think we need to work towards small volume production. So we are in process to prepare ourselves to build and manufacture VNL trucks with redundancy and the Aurora Driver integrated on our mainline, in our most modern factory, which is in NRV, New River Valley, that is Dublin, Virginia, you see it here on the slide. We have privileged access to the best-in-class manufacturing capacity. We can utilize top-tier hardware. We can integrate the Aurora Driver in this plant where we already today build from this plant, trucks with conventional diesel drivetrain, but also with BEV drivetrains. Then we have an ecosystem, which is supported by the strong Volvo dealer network with more than 1,000 service points in the country. But to add, to achieve all this, to get to all this, we, of course, need to have access to talent, and we believe on both sides that we have access to the best talent in the industry. We are on a very strong path. We're supporting each other, and we both know that this business we're in is actually a people's business. So it is the amount of talent we deploy on both sides, on both companies, on both sides of the ocean. But very important is how we collaborate, and this is where we believe we are making the difference. Our skills are not conflicting. They're complementary. But one thing we all share, we are fueled by making the next big step, and that is bringing autonomy to life, and we want to do that for real, and it will happen. Thank you.

Thank you. All right. And with that concludes our program. We'll finish with Nils. It was great. Thank you, Nils. On a personal note, we've been working with Volvo since before the world knew we were even working on trucks. We've been working on this platform that the audience here will actually see off to my right, maybe the cameras can, too, if you guys can pan that far. But this is effectively -- this is the [ VNL 6700 ] built with those redundancies. They began testing last year for us. We expect to receive the launch capable version of this platform this year. And with it, the final piece in this 4-part arrangement that we've outlined for delivery of commercial launch. Now as we're preparing for commercial launch, we're also preparing for the next chapter in our product story, which is achieving positive gross profit margins in 2026. Now the path to positive unit economics passes largely through 3 primary revenue drivers and 3 key cost levers. On the revenue side, it's really about the growth in scale, the depth and penetration, the utilization of the trucks on the lanes that we've opened. And on the cost side, it's really about driving down the cost of 2 primary operational drivers in remote assistance and on-site support, and finally, driving down the amortized cost of the hardware that we're paying for on these trucks. And I'll talk a little bit through each of these and some of the key factors you should consider as it relates to how much you [ ought to ] risk discount, kind of the ability to do some of these things. So first, our R&D investments to date and the self-similarity of the road network that we operate. I say road network, not highway operate -- not highway network only because our trucks do drive off-highway. They drive through industrial urban-type settings, they drive on surface streets to get to, for instance, our terminal in Houston. Now the self-similarity of this and the R&D investments we make have allowed us to bring online new lanes in a much more rapid way. And to give you a specific example of this, Dallas to Houston, which we started operating on first, when we went to expand that to Fort Worth, El Paso, a lane of approximately 3x the size in terms of length or mileage, that required bringing online only 2 new capabilities, driving on grades with the engine brake and passing through inland border control crossings. That was it. Turning that crank the first time all the way around is a tremendous effort. Subsequent turns for additional lanes is far less complex, largely facilitated by the safety case framework we developed and its specific focus on enabling incremental validation of new capabilities required for a new lane. A couple of other things. So through 2026, our lane expansion will pass through the Sun Belt, where approximately 50% of the -- or over 50% of the vehicle miles that our trucks travel will be on lanes that allow operations that exceed the hours of service limitations for a human driver. And you've heard several references made today, I think, Richard mentioned it, Daragh mentioned it, but utilization of these assets is a critical value unlock that autonomy provides. On these lanes, when we're driving over 50% of our miles at longer than 11 hours, suddenly we're unlocking more than double the utilization per asset that our customers have historically seen on their trucks. And that's a big deal. Return on asset is a key metric that this industry cares about for obvious reasons. Further utilization growth beyond just that enabled by these long lanes and exceeding the hours of service of the human driver will come as the driver operates in increasingly inclement weather. When we launch, we will have validated driverless operation in the majority of the weather conditions expected on the Dallas-to-Houston lane. Now as we expand, particularly into the Southeast corridor, where severe weather is more prevalent and severe, we'll be validating driverless operation in more inclement weather required to do that. Finally, as we expand breadth-wise, we'll also be deepening our penetration of open lanes. And one of the ways we do that is by bringing online new trailer types. To give you an idea of kind of what kind of effect this has on our business, on Dallas-Houston, the dry vans that we'll launch will make up approximately 40% of loads that are traversing that lane. When we bring online -- if we pick up refrigerated, intermodal and dual 28s, that increases the serviceable addressable market on that lane alone to over 200 million vehicle miles traveled per year. You extend that to the operational domain that we'll be operating through 2026 and now we're talking over 20 billion vehicle miles traveled every year on the trailer types we expect to support in the weather conditions we expect to allow for driverless operation. If you want a sense for like, okay, so how hard is it to bring online a new trailer we're already operating. Richard mentioned some of the food. I'm not sure he ever mentioned refrigerated trailers. It turns out refrigerated trailers are required for a lot of that, and that's one of the things we've been working on with Hirschbach and we're actively hauling those today. We're also hauling intermodal trailers for some of our pilot customers. So we expect the incremental lift to validate those for driverless operation to be very [ tractable ]. Finally, on the revenue side, it's important that long term we drive directly to our customers' locations. Today, we drive between these market consolidation points we refer to as terminals. These are convenient in a number of ways, primarily because they allow us to aggregate volume of loads from our customers in one location with relatively simple operational complexity, so that we can turn these trucks and turn these goods at a rapid pace. But they also require us to have shuttles on either end of our routes to get them to and from various customer locations. Now we're putting in place today the pieces, particularly with some of our larger customers that have natural consolidation points already in the network, be that a large sortation facility, a big distribution center, their own terminals where they're doing a lot of their operations out of to allow our trucks to drive and ultimately their trucks to drive directly to the location. To give you an example of, well, okay, how complex is this one? When we drive to Houston today, our trucks drive right past the entrance to one such location. We're already driving to many of these regions. We expect that, that rapid incremental capability, the ability to bring that online will allow an increased growth in our network. It will allow us to make our network more capillary, augment our terminal coverage and ultimately increase the value of our product to our customers, which is a key lever in the pricing power equation for being able to increase our top line. Now as it comes to -- as it relates to our bottom line and some of our key cost drivers, there are a couple of big ones that we're really focused on. The first one is the operating cost of remote assistance. If you recall, and we talked about a little bit earlier, you can see active remote assistance happening right now by 2 of our literal remote assistance operators back here in the left corner. But remote assistance is the operational mechanism by which we allow Aurora-driven vehicles to request high-level guidance from a remote human to deal with a situation they haven't encountered before. That could be something like an unusual road closure or something else. Now as the Aurora Driver's experience continues to increase, the frequency with which they encounter situations they haven't seen before or have no idea how to deal with will decline, and commensurately, so will the cost to provide that remote assistance service. Now the key cost driver here is the number of remote -- command center specialists supporting remote assistance relative to the size of the commercial fleet that they support. We expect to be at the breakpoint in this curve shortly after commercial launch and well into the diminishing returns of a one-to-many type remote assistance to fleet operation to support our gross profit objective in 2026. Finally, if you recall in our discussion -- Yanbing's discussion about the Autonomy Performance Indicator metric, another key cost component of driving is the rescue required for any operational issues encountered that we couldn't resolve remotely, be that a blown tire or a mechanical issue. Now in Q4, as Yanbing noted, over 60% of the miles driven in representative commercial pilots didn't require any on-site support. And we expect that number to be around 90% when we launch. But keep in mind a couple of things. First, as our product -- we're never going to be 100%, no matter how mature we are, because we're still subject to the same mechanical issues and the same onboard faults with the rest of these fleets are with one exception: the Aurora Driver has access to far more data about the base truck than a human ever does. And by virtue of that, we expect that, in time, as our ability to predict, to diagnose, to preemptively address issues with the base truck platform, these trucks will be even more reliable on the road than their human-driven counterparts further driving down the rate at which they have to be rescued on the road. Now final piece in that cost story is really about the hardware. How -- getting to the next generation of hardware and reducing the BOM costs. And to talk about some of the hardware strategy, I'd like to invite Sandor Barna, Senior Vice President of Hardware Engineering at Aurora, to talk about it.

Thanks, Sterling. So today, I'm going to talk about our hardware generations, and in particular, how they impact our cost story. Our hardware is a collection of sensors, radar, off-the-shelf LiDAR cameras plus our FirstLight Lidar, which we believe is really important to making this safe, and a server class computer and all of the bits and pieces that are required to connect it all together. For those of you here today, you can see the commercial-ready launch generation of our hardware mounted on this truck. It's actually the pieces that are not camouflaged and are in black. This is actually live hardware and this truck, as was mentioned earlier, has been driven in our -- with Aurora. The hardware here has been hardened to automotive standards and that allows us to operate with confidence knowing that the hardware won't be damaged if we hit a pothole or something like that. Next year, we're going to be introducing our next-generation hardware kit. And this is a critical step towards both increased reliability and further cost reduction, and we'll talk about that in a minute. And looking further ahead, we're going to introduce the Hardware as a Service generation, which we're developing with Continental for launch in 2027. So let's dive into that next-generation hardware. In 2025, we plan to introduce this generation of hardware. It's designed for manufacturing beyond our internal capabilities to include OEM requirements for reliability, durability and vehicle line integration, and importantly, to drive the step reduction in hardware costs. It will be manufactured by a contract manufacturer for integration with autonomy-enabled truck platforms produced by our OEM partners. And it also brings us some exciting performance. It's powered by a new fully redundant computer that delivers a 40% reduction in power and weight at roughly half the cost of our existing computer. And the next generation of our proprietary FirstLight Lidar, which gets even better: larger vertical field of view and increased range while coming in at nearly 40% less cost than the current generation. These performance enhancements will allow us to detect smaller objects at even greater distances than before, enabling improved reaction time for road hazards. We extensively test our hardware components using our in-house reliability lab as shown here. Key to reducing the hardware cost is building a kit that lasts as long as the truck, which can be 1 million miles or more. Cost per mile is directly tied to this reliability. So if you extend the life of the hardware by a factor of 2, that cuts your cost per mile in half. And so this also helps reduce the overhead involved in servicing the hardware. These are just some of the examples that will result in material costs that we'll realize in 2025 and 2026 to achieve our cost reductions and positive gross margin. Our path to positive gross margin in 2026 builds towards the ultimate goal of scale and self-funding. In parallel with this, we're working closely with Continental develop the Hardware as a Service generation. The scalable Hardware as a Service generation that we plan to introduce in 2027 with Continental gives us yet another step function reduction in hardware costs. The kit is being codeveloped with Continental leveraging their industry, engineering and manufacturing prowess to drive down costs and ensure smooth integration at the OEM manufacturers' assembly lines. It will continue to feature our FirstLight Lidar, but this time, implemented as LiDAR on a chip. We leverage our silicon photonics capabilities to take discrete optical components and convert them into a chip scale, which reduces cost significantly and enables automated manufacturing. It will also feature a Continental-developed fallback system, further reducing costs by eliminating the need for that redundancy in the primary computer. Under the Hardware as a Service business model, Aurora will pay for the hardware on a per-mile basis. As a result, we expect that the initial cost of the customer for an autonomous truck will be very similar to a standard truck and that limits the capital investment required to adopt this technology. We're the only AV company with a Tier 1 hardware partnership to commercially scale an autonomous vehicle hardware platform with this Hardware as a Service model. So first-of-a-kind structure for the industry. So along with Volvo and PACCAR, the partnership we have in place in Continental will position us to fully scale up the road driver as we look ahead to 2027 and beyond. Continental's legacy in development and industrialization of automated driving systems, combined with their commitment to transform transportation, make them an ideal partner for Aurora. So it's my pleasure to introduce Jeremy McClain, Head of the System & Software Business for Autonomous Mobility at Continental, to share a bit more about our exclusive partnership and why they are invested in this.

Sandor, thanks. And good morning to you all. I'm thrilled to be here on behalf of Continental, and we really appreciate the opportunity to talk to you today about our truly unique partnership together with Aurora. Aurora and Continental are collaborating to develop and industrialize a fully commercially scalable version of the Aurora Driver. And with that, we're preparing to revolutionize the mobility industry as we know it today. Continental always has the future of mobility in its focus. In fact, we've been developing mobility solutions for 150 years as a company. And for the last 25 of those years, we've been delivering adaptive driver assistance and automated driving technologies to the industry. Our business area, autonomous mobility, develops sensors, systems and software solutions at large scale, at automotive-grade quality and at cost levels that meet the tough demands of the automotive industry. Our track record, in fact, speaks for itself. We've delivered to date more than 200 million autonomous mobility products from our global network of manufacturing facilities. That includes 6 generations of automotive radar, the components just as an example; and then imaging radar component, which you can find in almost all of the top Level 2+ and Level 3 automated driving systems that are on the market today. Back in 2012, so 14 years ago -- or 12 years ago, we were the first automotive Tier 1 supplier to be licensed with the test vehicle in the state of Nevada. It was the first place at the time, which allowed such a thing. And since then, so over the last 12 years, we've developed a worldwide network of engineers and experts who specialize in the area of autonomous mobility. And that specialization has been proven in numerous projects around the globe at various levels of automation from Level 2 to Level 4 and in a wide range of vehicle categories from passenger vehicles, of course, to heavy commercial trucks. And now it's time to leverage all of those years of expertise and success to bring fully automated driving vehicles to the road and to business success and a commercial success. Continental's history and experience, coupled together with Aurora's unquestioned leadership in autonomous trucking, makes this partnership a perfect match. Our competencies complement each other, in fact, quite well. Both of us share a deep passion for bringing fully autonomous vehicles to the road first, to do that at commercial scale, but while also ensuring that transportation gets safer. Now we heard it today already many times, safety is very, very important and the passion for commercial scale, safety and bringing autonomy to the roads first we share very closely as well as remedying the extreme driver shortage that we see together in the industry. We've been detailing this partnership, in fact, for quite some time. having made it official in April 2023, just under 1 year ago. And there's substantial traction and commitment from both sides towards our mutually set goals that we put together. In just September of last year, September 2023, less than 6 months after we formed the partnership, we reached the first most important goal in that development. And that was to complete the design and the architecture for the future Aurora Driver, hardware and the fallback system, which Sandor mentioned as well just a moment ago. In the next 2 years, so this year and in 2025, we will build initial versions of that particular generation of the autonomous hardware kit and test them at and near our plant in New Braunfels, Texas, and you saw some images of that particular manufacturing facility just a moment ago and as well across our global engineering footprint and as well our global manufacturing footprint. In 2026, we will industrialize and validate those kits, those systems, and start ramping our production, and as well, finalize the integration into the final vehicles. Finally, starting in 2027, thousands of trucks with that generation of the Aurora Driver system will hit the road right here in the United States. At Continental, we're very proud to contribute key items within the partnership, and I'm going to describe briefly some of those items, which were mentioned already by Sandor. First, we're responsible for manufacturing and delivering all of the hardware in those particular systems, which enable the operation of the Aurora Driver. That includes LiDAR sensors, radar sensors, camera sensors, high-performance computers, telematics units and even sensor cleaning systems. All of those components are then assembled into modules, so-called pods, and then shipped together to the manufacturing facility of our truck manufacturing partners to be assembled onto the trucks. We are as well developing a cost-reduced and completely independent fallback system, providing an additional value to the Aurora Driver. That is a parallel vehicle control system with own sensors, own hardware and own software that is running completely in parallel to the primary driving system, which is driving the vehicle under normal operating conditions. In a very, very unlikely event of a failure or malfunction of that primary system, the job of the fallback system is to take over the control of the truck and bring it safely to the side of the motorway, therefore, ensuring an additional layer of safety. We, as well, operate as the primary interface to the truck manufacturers. This partnership is, of course, therefore, leveraging our extensive experience in working with vehicle manufacturers to deliver vehicle technologies according to automotive standards, of course, and tailored to the specifications of those particular vehicle manufacturers' needs. We're providing as well service and maintenance over lifetime of these particular systems. We're responsible for servicing the autonomous technology over its lifetime, including planned maintenance, warranty and repairs and as well developing service manuals and training and certifying technicians to ensure the technology is in proper working condition. And all of those elements of the partnership, of course, are designed to maximize uptime. Maximize uptime of those systems, of those trucks and so on because only when we can maximize uptime do we really unlock the value of autonomous trucks of this particular partnership because, after all, that's how we make our revenue and that's how the entire value chain, in fact, is making the revenue. And that, of course, brings us as well to the business model, which also follows a completely new approach and it's very different than the business models of the traditional Tier 1 and manufacturer in the automotive industry. Continental and Aurora have agreed upon a Hardware as a Service business model, and therefore, our revenue comes on a per-mile basis. So for every mile that these trucks drive autonomously, Aurora is charging a per-mile fee and Continental is earning its revenue via a share of that per-mile fee. You see, with this partnership, the conditions have never been better. The transfer of years of research and development experience into driverless mobility is now becoming for us a palpable reality, something that we can really see coming to fruition and coming to the roads, made possible in this particular case by the perfect interaction between our 2 companies, and of course, the other partners that were up here and spoke today. And that allows trucks to soon be part of our everyday life and the everyday experience of driving on the nation's motorways not as a vision, not as a research project or an experiment, but as a substantial contribution to the prosperity of our society. And therefore, Continental is very proud to be a part of this partnership and of that story. And with that, I would conclude and hand back to the colleagues at Aurora.

Thanks, Jeremy. It is awesome to be on another adventure with both Continental and with you. All of us at Aurora are incredibly excited about working with Continental and the path that we have to bring this technology to scale. In addition to this important Hardware as a Service partnership with Continental, our path to scale and self-funding leverages that ability to scale rapidly that Sterling talked about, and of course, our OEM partnerships with Volvo Trucks and with PACCAR. I really can't stress enough how important that partnership with the OEMs is. It's not just to get the first thing out the door, but it's about being able to get them out the door safely and get them out the door at scale. We're proud to be working with Volvo and PACCAR. Collectively, they represent about 50% of the trucks in the U.S. market every year. When I look at a few years, this how I see the ecosystem working. Continental is going to manufacture the Aurora Driver hardware. They're going to ship it to our OEM partners. They're going to manufacture their autonomy-enabled trucks and line side install the Driver hardware. That's then going to get shipped to our customers and partners. And then these customers are going to use the Aurora Driver out on the road to grow and build their business. The best part about this whole thing is that the ecosystem is fully aligned. The more our customers drive the trucks, the more revenue they generate. The more the trucks drive, the more revenue we generate, the more revenue Continental generates. That alignment is pretty magical and powerful. Everyone benefits. This ecosystem positions us at Aurora to deliver on our vision to start first in Texas then cross the Southern freight corridor and then ultimately throughout the whole United States and beyond. And as Gerardo laid out, the regulatory environment supports this. Autonomous vehicles can be put on the road in the majority of states today. So to summarize, we're focused on 3 primary goals: being ready to launch commercially, achieving a product with positive unit economics and the scaling and ultimately self-funding our business. This updated road map shares how these things tie together over the next few years. And they talk about our ownership and delivery models, how we're going to -- the product capability expansion plans and that's planned for scaling and what enables that. We're convinced this plan is going to create immense amount of value for society, for our partners, for our customers, and of course, our shareholders. So with that, I'm going to turn it over to our CFO, Dave Maday. He's going to give you a little dive into our financials and what this all means economically. Dave?

Thanks, Chris. All right. This is a lot to follow. That was a great review from everybody, and I appreciate our partners and our customers being here with us. Honestly, I couldn't be more excited about the progress and the path that our team is on. And really, I hope this morning demonstrates to all of you the remarkable opportunity that's in front of us for us to deliver rapid capital-efficient revenue growth, gross profit and free cash flow. But before we get into some of the future stuff, let me just recap and dive into where we ended last year. Operationally, and as Zac pointed out, and this is just through 2023, we delivered roughly 3,000 loads over 820,000 miles through those loads with our 7 announced partners plus a few others. From a financial perspective, we had about $2 million of pilot revenue. Now as a reminder, prior to commercial launch, we recognize that as a contra-R&D expense. From a cash perspective, we spent a little over $600 million while demonstrating really strong fiscal discipline. Okay. And notably, in the second half of 2023, our cash spend was actually quite a bit under our targeted guidance of $175 million to $185 million per quarter. And we ended the year with $1.3 billion in liquidity, which I now can say we expect that to fund our operations not just well past commercial launch, but into the fourth quarter of 2025 based on the savings that we've recognized in our strong fiscal discipline. Now let's look ahead a little bit. And as Sterling did an outstanding job of summarizing, our key drivers to our business in the near term are pretty simple, right? It's -- the first thing is the vehicle miles traveled, which we'll describe as the utilization of the truck, ensuring that our trucks are driving by lane expansion, capability expansion and market share capture. And then on the other side, it's the gross profit per truck. Increasing our revenue per mile by delivering benefits to our customers, driving down their costs and then creating value such as unlocking customer end points. And then I think the team did a great job of outlining the reducing of the cost that we really control: the hardware, the remote assistance, the on-site support and other costs. So I also would point out that -- and Sterling talked about through 2026, Chris had mentioned the trucking market is a massive opportunity. This map right here reflects at the start of 2028 we expect the Aurora Driver to be operating on lanes that represent approximately 50 billion vehicle miles traveled annually. For context, the entire U.S. market is about 200 billion. And notably, over 60% of the 50 billion VMT are associated with these longer routes, greater than 600 miles, that exceeds the traditional hours of service limitations that exist today for human drivers. But we also plan on continuing to serve these shorter routes like Dallas to Houston where unlocking customer end points for LTL customers is incredibly valuable. Now before we get into our highly capital-efficient Driver-as-a-Service business model, which relies on all of our partnerships, let me just briefly explain the revenue potential at early commercialization. And our focus in early commercialization is about adoption. We expect to own and operate a very small fleet of vehicles carrying them directly for our customers or supporting our customers like V.A.S. The chart on the right really provides some benchmark data from DAT, which looks at both contract and spot rates, including fuel surcharge, and gives you an idea of where the market stands today. You can see that the U.S. average and you can see the variability both for our 2 launch -- our first 2 launch lanes, which is Dallas to Houston, followed shortly thereafter by Fort Worth to El Paso. All right. But today's focus is really about the long-term business model. Again, our Driver as a Service business model, which is capital efficient and really targeted to meet the customer needs. Okay. This model will charge a price per mile for driving the customer-owned trucks, inclusive of the entire Aurora Driver and associated services that we've outlined earlier. Now to clarify, just so everybody's in line, the Aurora Driver hardware, sometimes this gets a little confused, that's part of our cost structure. That's included in our per-mile fee that we intend to charge when we operate both the small fleet at the beginning, but also when we transition to the DaaS business model. We expect this model to deliver not only positive profit, but really high SaaS-like gross margins. Now I think it's really important for us to recognize the importance of providing value to our customers, and you heard Richard and Daragh explained it a lot. We have to deliver total cost of ownership benefits, otherwise, we're just creating a solution and leaving a portion of the value just on the side and that's not what we're about. We're trying to create customer value. And you do that by providing more efficient, less variable driver cost, increasing their revenue-per-truck opportunity, providing better fuel economy, and of course, lowering insurance costs. These are the same ones we can't help, but reiterate these several times. And our product and pricing strategy is really designed to provide this compelling value proposition based on those elements. And it's really designed to compare it against existing alternatives. On the chart that you can see on the screen here, this is an ATRI estimate and they're kind of the benchmark when it comes to estimating costs out there, they expect the total cost on average to be about $2.25 per mile with the driver wages and benefits being roughly $0.91 per mile. That's the DaaS market opportunity that we are going to go after and also deliver value for our customer. But if we take a deeper look into this, it's not just that, we need to look at the trends and where we're headed. And you can see on this chart, over the past decade, roughly, the cost for drivers, including their benefits consistently raises and has consistently risen and is expected to in the future where there's been some large contracts in the future about continuing driver wage costs. Our indicative DaaS pricing range, and again, recognizing we're a few years away from operating our DaaS business model, is $0.65 to $0.85, okay. And that's going to vary based on the value that we deliver for our customers, okay. We believe this pricing environment will allow us to do 2 things: it will deliver total cost of ownership benefit for our customers, which is critical; and it will deliver high gross margins for us in our business. From a customer perspective, and I almost feel like this isn't so helpful given the fact you heard from our customers, but it's not just about the driver wages and benefits. It's about the other indirect costs that we can provide value to. We estimate those to be about $0.15 per mile, but this is about the driver sourcing and -- reducing driver sourcing and turnover costs, eliminating workers' compensation, ongoing driver training and there's other services and overhead related to driver management that we think we can create real value for our partners. Let's take this analysis just a step further, right? So here is an illustrative case study that demonstrates both how the customer revenue and profit can meaningfully improve with the Aurora Driver-powered trucks, right. This first example is a terminal-to-terminal example, driving between Fort Worth and Phoenix. The distance is roughly 1,000 miles, right? That exceeds the traditional hours of service limitations by a single driver. In this particular example, even if we factored in incremental cost for the -- getting the haul to the final destination, what we'll call the local haul drayage cost. This creates high utilization and lower cost and can deliver nearly 600% increased profit per truck per week on this lane. And it's because we can deliver more loads at a lower cost. Let's take another example. This time, it's a short-haul route between Dallas and Houston, which is roughly 200 miles. Again, as we mentioned before, these trucks aren't associated with drivers. You have an opportunity to continue to utilize these trucks and substantially increase the asset utilization. I think Richard mentioned that a lot of times it's maybe 30%, 35%, something along that, that we're actually using that asset. By us being able to drive the truck more often and drop the price down, in this case, it's an end to end, which is really important for our LTL partners, we're able to, again, higher utilization, lower cost, deliver 300% increased profit potential per truck per lane -- or per week on this lane. So let me just turn back to the business model real quick. So you saw a lot of stuff from Sandor and Sterling showing how we're expecting to drive the costs down to achieve these high gross margins. And I think we have a really good scaling and strategy for reducing our cost structure, right? And it's all the elements that we talked about before. We expect that we can deliver positive gross profit in 2026 and we can get to these high SaaS-like margins over time. With the strategies and plans that we have in place today, we estimate gross margin could achieve 70% by 2028. Let me summarize this. we expect our customer-centric strategy to generate overall more than $1.3 billion of revenue in 2028 with our key drivers being gross margin improvement, which I just highlighted; controlled spend, which I can now confidently say we're going to estimate that the total cash use will be between $175 million and $185 million per quarter until we achieve free cash flow positive, okay; and then as we had identified before with a lot of our great investors that were with us in July, we expect we'll need another $850 million over the course of time before we get to free cash flow positive and that will be our final step in our funding plan to that point. So in closing, our plan supports the rapid capital efficient revenue growth, the high gross margins and the positive free cash flow. With that, I'm going to turn it back over to our CEO, Chris.

So many of you know that I was a scientist and engineer before becoming appointed here boss. So I obviously love the technology we're working on. But when I talk with Dave and we get into the opportunity with our business, it is incredibly exciting for me. So thank you, Dave. As you've heard, we're putting in place -- or have put in place all of the essential components for commercial launch. We have confidence the technology is going to be ready at the end of the year. We're confident that our customers are ready to take this journey with us and bring this technology to market. We know that the regulators are in a place where, if we are good actors, we're going to be on the road. And we expect at the end of this year to receive a platform that will be the platform we launch with from our OEM partners. Beyond this initial launch, the road to scaling and self-funding is clear. It's, in fact, never been more clear for Aurora that we're on the right path. At this point, we started the day with a video in Dallas. I'm going to show you a quick video in Houston, and then we'll get folks up here for Q&A. [Presentation]

So with that, I'd like to invite Sterling, Yanbing, Sandor, Gerardo, Dave, and our Chief Safety Officer, Nat Beuse, up to the stage, please. And we'll turn it over to Q&A. Thanks, guys. And I think Stacy and [ Colin ] are going to be around with mics for folks with questions.

George Gianarikas from Canaccord Genuity. I'd like to ask a little bit about your confidence in scalability. You spent a lot of time talking about the different scenarios you're in and your first specific lane. And how confident are you that over time that as you launch additional lanes, that, that's easy to do? I mean, you mentioned the 2 additional test cases you had to do in I think your Fort Worth lane. But as you expand across the Southeast, how much more work do you have to do on each incremental lane?

Yes. And maybe I'll start. Are we missing someone? Okay. Well, we've got an extra chair if anyone would like to join. So maybe I'll start and then, Yanbing, you take it. So we are convinced that the hard nut to crack is getting it the first time, that if you just think about your own experience in driving that doesn't really matter a whole lot where you drive once you've got those base skills. For us, the way we've architected our validation is those common skills are going to continue to be there. They get tested any time we make a change. And then we're like, we'll look at, as Sterling laid out, the customer [ work control ] station that we have to go through. We'll create a test suite around that. We'll make sure that we're confident in the performance and we'll go there. And then it turns out the other customer [ work control ] stations, we're going to be able to operate through as well. The other question you might have is around mapping and that, of course, we use maps as part of the technology that enables us to drive throughout the world. And really, that's leaning into what computers do best, right? Computers are really good at storing information and then recall that information and using it. And so we see the opportunity to provide the truck the experience that's in the world of this is what the road looks like is a real powerful -- a really powerful tool. And the cost of doing it is inexpensive. And once we have it in place, maintaining it, it becomes almost automatic. And it becomes this appreciating asset for us and that we will have the best maps of anyone in the world. But Yanbing, I'm sorry, I prattled on. Do you have anything to add?

I think Chris said it really well. I do believe the technical requirements is very, very marginally incremental. Today, we've been driving on a highway confidently. We've been navigating surface streets. And the good news is the highway and surface street look awfully alike in this country. I do think the main investment we need to look into is really how we expand operationally. Maybe Sterling can add some additional color on the operational front.

Sure. Yes. I mean we talked a little bit about bringing online terminals. Now our business model really remains asset-light. So while the first couple of these we built, we expect to bring on third parties to do that as we also start moving to customer locations. I mentioned big customers with natural consolidation -- or the natural kind of consolidation points. LTLs, private fleets, they've all got them, right? You look at a massive private fleet and they've got massive sortation facilities that are sources and syncs for tremendous volumes of trailers. And so that will also help operationally as our customers will actually start to take on more of that burden than we do just by nature of facilitation of kind of the incoming and outgoing trucks.

Nick Amicucci from TD Cowen. Just wanted to drill a little bit deeper into kind of the path to positive free cash flow and the kind of trajectory of getting there. So if we look at kind of the time line and then the projections that you have done on the last -- on the final slide there, it would imply like 8,000 to 10,000 trucks. So I just want to kind of frame that and see if that's kind of the way to think about it. And then the trajectory to get to that, like what type of inflection we see and when?

Do you want to speak to that, Dave?

Sure. So yes, I think we talk about it more in miles than in trucks because I think that's how our customers think about us delivering loads, in the miles. So I'd center a little bit more on that, but obviously if you want to drive a lot of miles, you need a lot of trucks. For us, we're in this crawl-walk-run approach in terms of how we think about -- and this is important because that's how our customers think about it as well, about how we're going to roll out our technology. For us, we feel like once we get to the Continental Hardware as a Service opportunity and business model, that's when we plan to scale substantially and start going to thousands of trucks and even exceeding that measure. So that partnership comes in, in 2027. And so as soon as we achieve that partnership, that's when we really plan to pick up the volume and we need to have the operating domains that we work in. So that's why it's important to have all the areas that we're going to operate in ready beforehand so that the trucks have somewhere to drive. So for us, when we look at it, we think that the path is still reasonable. We do believe that there's some flexibility, right, in terms of free cash flow, in terms of the number of trucks. What we've provided to you in terms of some of the guidance is if we get relatively close to those mileage-per-year numbers, we're pretty confident we will get to free cash flow. And sometime towards the end of '27 through '28, that's when we expect to have positive free cash flow. Importantly for us, we're trying to manage the business in the incremental capital request of the $850 million additional that we need. That's our benchmark in terms of how much funding that was a commitment we tried to make for our investors. That's how much additional funding we'll need before we get to positive free cash flow.

David Vernon from Bernstein. I have a question for you on the financial side, Dave. If you look at $1.3 billion in revenue and 2 billion miles, that works out to something less than $1 a mile, call it, $0.60, $0.70 a mile. Your partners are talking about using these trucks and miles where they're having to pay 2 drivers, which would be much higher than your average $2.50 a mile. How do we think about the $0.65 a mile that's underlying the assumption there? And then how do you think about maybe pricing higher if you're going to be using these autonomous trucks in super high-value lanes that are enabling customers to reduce inventory, run further distances, all that kind of fun stuff. I guess, the question in the question there is like are you underpricing this because of its unique characteristics if you're just pricing at the price of a normal truck?

Thanks, David. Great question. A couple of things. If I'm going to be a good steward of the financials, we're going to be a little bit more conservative in terms of some of the assumptions and not put every aggressive assumption out there. So I think what we had mentioned was the indicative pricing range is somewhere between $0.65 and $0.85. It's important for us to recognize that the value proposition that we deliver and being able to deliver this cost of ownership reduction and the increased value in other areas is what's going to guide our pricing envelope, right? Certainly, there's an element of taking 2 drivers and shifting it to an autonomous truck, there's tremendous value in that. I think roughly our customers could do a better job, I think that's an incremental $0.10 per mile on average for that second driver. But we are intentionally trying to focus on making sure we deliver value to our customers and we're able to deliver the margins that we want to do. Again, it's an indicative range. If we're delivering only -- and there's short-haul lanes there, there's some terminal mix in there where they're going to have to do drayage. So it is a combination element of it, but we think that this is a very achievable, if not slightly conservative outlook in terms of like our revenue potential.

Ravi Shanker, Morgan Stanley. Two questions from me. One is, it's great to hear from your early partner customers that they are ready to go. But can you talk about conversations with like the future pipeline of customers, kind of are they waiting to see if the first few guys can monetize this, commercialize this before they get onboard or are you already lining up a pipeline of kind of new business beyond that? And the second question is on the regulatory side. We have seen the passenger robo-taxi space have a little bit of a setback in California. Some operators have been decertified, if you will. And even on the trucking side, California has had its issues. Do you kind of -- to your point, regulation's ready already, but do you see a risk of that potentially taking a step back?

Sterling, do you want to speak to the customer part then I'll start with the regulatory?

Yes. As it relates to the pipeline, we haven't announced all the customers that we have. So we're currently operating pilots with as-yet-unannounced folks. Further, the conversations that we're having with others are quite positive. I think there is a real -- on the backs of years of promises that they heard, in particular from some others who are no longer with us, I think there is some skepticism about, well, I'd like to see it. I'd like to see it be real. But I will say that the visits that we've hosted from leadership of many of these companies, including most recently, I think, Werner's leadership team was on site with us at our [indiscernible] location. It is a very typical reaction to get out of that truck after driving on the Dallas-Houston lane and just say, holy cow, that is much closer than I had anticipated. It is much more smooth, much more confident, much more boring than I'd realized and in a good way, as Yanbing mentioned earlier.

The other thing that I would add is that we are not super aggressively going out of the market right now. We're in a period of time, one, where we're trying to learn to make sure the product we bring to market makes the most value for customers. And the other is that, frankly, we're going to be limited in how many customers we can serve in the first year or so. And so the folks who've been with us on the journey, we want to make sure that those customers start to see the value of this and be great partners to them. So I agree with everything Sterling said, and I would just add that other lens to it in that we're not kind of -- we haven't really deployed a high kind of...

A wide net, yes.

Yes. Wide net, yes, there we go. Thank you for getting my foot out of my mouth. On the regulatory side, I would actually say, just on California, it's a little different. I would say it's been a step back on the regulatory front. I'd say it's the regulation doing what it's supposed to do, right, in that it is protecting the public and it is -- should actually give confidence to policymakers and the public at large that the regulatory framework that we have in place is sufficient right? Now in California, of course, there's a limitation that they don't yet have regulations that enable trucking. And you may be referring to the driver in a driverless truck bills that we talked about. Maybe Gerardo, that would be a good thing for you to...

Yes. Happy to unlock that. So to put that in context for you. Last year, so just in 2023, there were 8 bills throughout the country that were introduced that were similar to California. Every single one of those bills was either stalled or defeated. There was not a single driver and bill that passed. I think if you go and read even Governor Newsom's veto in California of that bill, he signals to the fact that, look, they're not looking ban these driverless vehicles from operating in the state. And more importantly, they're working on the regulations to be able to allow for autonomous trucking to be in California. We are regularly working with the California DMV, with Cal STO. We just recently hosted the Cal STO Secretary in our offices in 2 different occasions because they're working to understand exactly what does California have to do to enable autonomous trucking in the state.

Jeff Kauffman from Vertical Research Partners. A quick question for Sterling and then for David. A lot of truck fleets will tell me, we need to see about 200,000 miles on something new before we really understand what we have. And that's when we're talking about new oil or new brakes or new diesel fuel. We're talking about a whole different truck. And I guess my question is, as you're engaging customers, everybody's watching this, everybody's tried something with somebody, and to your point, there are some bad experiences out there. Can you get to this implied 8,000 to 10,000 truck number out there? And I know, David, you said focus on miles, not trucks, right, but trucks drive miles with your existing base of customers without adding new customers because it may take some people a few years to get comfortable with everything. And then the second question for David is, you had $675 million in OpEx last year. How does that scale as we move to 1,000 vehicles, 10,000 vehicles or 500 million miles, 2 billion miles, however you want to think about it. How does that change? Not the above gross margin, but below gross margin OpEx.

Take it away.

All right. Well, a couple of quick things. First, as Chris mentioned, we are heavily supply-constrained in the early days, which is to say like we have far more demand than we can satisfy in the early years. And so a couple of ways that we're dealing with that. First, with all of those who we are working with today, we share a great deal of information about everything that's happening on the road. I think Daragh highlighted this in the Werner case, but we've been operating pilots with these folks for over a year now. With our trucks driving 25,000 miles a week at this point, they're getting a tremendous amount of experience and data and information about what's coming off of them. That said, it's not enough, right? And as Zac highlighted earlier, we are preparing for what we're calling the commercial readiness program wherein a some of the million milers or some of these -- the driver trainers or more experienced drivers from our customers are free to ride along in the truck and judge it just as they would a human driver. And we'll do that for a period of time until they have comfort in that operation. The other thing that we expect to do is with many of these customers, the data that we're accruing is at least in a kind of post facto basis we expect to be valuable to many. So when we bring online a new customer, we'd say, look, our trucks have collected -- the Aurora Driver has collectively driven like it's not a million-mile driver. It's 10 million-mile driver at this point and here's the data, right? Why don't you put your driver trainer in it? Why don't you put some of your more experienced drivers in it and spend a week with it. Chances are they'll be bored after the first 3 minutes, but have them spend a week. And I expect that as we do that, they will get confidence more quickly on an incremental basis because of the fact that they've got so much data accrued. Again, keep in mind, and this is an important distinction from human drivers, this is one driver. The experience gained on any truck anywhere with any customer accrues the benefit of the entire fleet, everyone who's using that driver. And I think that's a really powerful part of the message with these new customers. Now you also ask, can you get to thousands of trucks with a small stable of customers that you've got? The answer is no, not unless we penetrate kind of a significant portion of their business. And so yes, we will be widening that net as we go. But as Chris mentioned, it doesn't behoove us to cast a giant net at the moment when we've got a handful of trucks expanding into hundreds, that just wouldn't make sense to have a bunch of customers waiting with bated breath for a truck they can't buy yet.

And just to add to what Sterling said, part of the reason why we have that intermediary step of Transportation as a Service rather than Driver as a Service is really to reduce the barrier for customers. So they don't have to buy the asset, that they can get the experience with it, they can see the truck operating adjacent to their fleet, as part of their fleet, but not having to put the capital outlay and be able to kind of be semi arm's length while they're warming up to it. And then as Sterling said, all of the experience that they're going to be able to garner from this, I think, is going to be pretty profound. Dave, you want to talk about the financials?

No. Yes. Look, I think it's important to understand we've made some significant investments in R&D over the past several years. And the really beneficial thing about this is, as Sterling and Chris have both talked about as well as Yanbing, going for that first driverless is technically really hard and it takes a long time and we have to build a safety case. Incrementally adding lanes and capabilities is, by comparison, relatively easy. And so we're very confident that the head count that we have in place operationally from our software and our hardware teams to deliver that scale, we have what we need today. We don't need to ramp it up to be able to do it. We just will refocus some of the priorities and we will restructure a little bit on how the team operates. Importantly, from a hardware perspective, they're already working with Continental on our third generation essentially of hardware. So we aren't just working on launching a truck. We are concurrently running a business that's setting up for scale. So I think, for us, we're very confident that we can manage to the number that we had -- which I had mentioned before, which was the $175 million to $185 million per quarter until we get to free cash flow positive. There will be some element of the SG&A that we will increase a little bit, right, if we have to add more sales force to get to the numbers, if we have to add more marketing spend to promote our product, but that is pretty de minimis relative to the R&D expenses that we have.

Mark Delaney with Goldman Sachs. Manufacturing has been a challenge for several of the early-stage auto and industrial tech companies that we've been seeing in the public markets. You had Volvo Autonomous Solutions give a presentation. They talked about their path to getting the redundant truck ready to go and you spoke a bit about that yourselves in your portion of the presentation. But could you double-click a little bit more and talk about your confidence of having a broader set of suppliers being ready to support your time lines and then your ability to work with some of the contract manufacturers, getting the outfitting ready and not only in terms of timing, but in terms of some of the cost and coming in on target.

Yes. So maybe, Sandor, I can start a little bit and then...

Yes.

This comes back to our core philosophy as a company, which is do the things we know we can be best in world at and then go find the people who are really good at the other stuff that's needed. That's why I never want to be a Werner or a Hirschbach. They're amazing companies. They're well-run. There's not much we can bring to that. And similarly, that's why we don't want to be manufacturing hardware at scale. That's what Continental does, it's best in the world. So we -- the reason why we have confidence that we can hit the price points, the scales that we expect is because we're working with folks who know how to do that and have a very proven track record of delivering in that way. But Sandor, you, I'm sure, can add more color.

Sure. And what I can talk about mostly is that next generation of hardware and we already have that computer and that LiDAR up and running today in-house. So it's really a question of hardening it to the final mass production stage and taking that into the contract manufacturer. Now we don't want to just staff up our army of technicians to build that and then eventually transition to Continental. That doesn't make any sense. So of course, we're going to use outside contract manufacturer, but we're already well down that path with them and working closely on how we actually transfer our production line to them.

We do a weekly bill of materials review, where we understand who's -- what's sourced, what's not, what the cost risks are. So this is managed. I'm not going to say there's no risk in it. But at this point, we're in a really good place with it.

We do quite well understand those costs.

Chris and team, I'm Ken Hoexter from BofA. I have 3 questions. One, what happens between now and commercial ready? Second, to clarify, this is standard hardware is a set installed during manufacturing or is it aftermarket? And then I don't know if you gave the cost of the fully loaded tractor, did you provide that?

Okay. So maybe I'll do those in reverse order, if that's okay. Because there -- she took your microphone away.

Will you give Ken back the microphone, please.

Okay. So -- and I'm sorry, and I'm glad you got the microphone back. So third one was -- what was the last question?

Cost of truck.

Cost of the truck. So cost of truck. So one, our OEM partners are going to decide that. But the way to think about it is there's minimal kind of new hardware components in there, right? And the driver parts, the things that you would think of as the expensive parts are part of that Hardware as a Service. So they're on the Aurora side of this and our customers will pay us for that over time on a per-mile basis. So the base truck, we expect to be roughly in line with what it costs to buy a truck today. And I would expect that our partners, again, over time as this becomes more broadly adopted, there's an opportunity to depopulate some of the content in the truck and create an even better offering from our OEM partners to their customers.

She was taking it because she wants to be quick, so I'll be really. The mapping -- you use maps. Do you have to train each route before you go live? Or does it learn as it can go on a new route? So do you see constraints if you have to -- on the training there?

Yes. So we think of it as a data set. So we go out and build a map. We have to go drive to get the data to do that. But then the kind of the behaviors take that as one of the inputs along with whatever we're seeing in the world from a sensor. So no, there's not any kind of specific ML stuff per trip -- or per lane per se.

Just important building a map is one of the most misconceived things out there. We drive it a couple of times, we have the basis of the map. Like this is not even a short pull in scaling, right. This is just kind of everyday use we...

Turn the crank.

And it's really based on the investments that the organization made early on in terms of how we thought about mapping. We eliminated this as a constraint or a barrier to scaling based on how we really constructed the entire system.

What if they stopped the traffic on that road and reroute? Your truck has to learn?

Yes. So we are developing the systems to be able to handle construction. And there's different types of construction, there's a lane closed, there's intermediate kind of re-striping or something that's happening. And then, yes, there's certainly be routing. It turns out the rerouting doesn't happen instantly, right? There's a sign about it. There's a bunch of people. So there's an ability to be able to prepare for that. And then, yes, of course, our system has to be able to react on first contact with that situation to what it perceives in the world in real-time and that's part of the capabilities we're developing is to handle that.

And just to add to that, we have a mechanism to allow us to rapidly update our map. But as we perceive the first contact new map, we also have a way to rapidly update our map and we have a way to push over the air a really lightweight map to enable that rapid update.

Yes. And so before Stacy has to come up here and rip my lapel mic off, thank you. We're in a really interesting moment of, I guess, [ brilliant ] timing. So it turns out that 20 years ago yesterday was the first DARPA Grand Challenge, which I was fortunate to be part of and competed in here. For those of you here with us live in Pittsburgh, we're going to get you down to our test track, which is exactly the place where we were developing those systems 20 years ago. And so we're looking forward to giving you a chance to both see the trucks in action doing some interesting things and get a ride in a truck without a person on board in the seat. So love to have you come down with us. I think do we have housekeeping notes, Stacy? No. We're done. So I think we're on our way in 15 minutes to head down there. So please grab some food. Use the men's room if you wanted to and we'll head on out. Thank you.