Thank everyone for making it here, particularly on UN Week, for what is the seventh annual ADAS, AV, and AI Forum. Chris McNally, head of automotive and mobility research for Evercore ISI. And really, over the course of the day, you're gonna hear, you know, from the experts. You know, we like to joke that we are the referees, we are research. Like, if anything, I've found that over the years, one of the most important things that I can do for clients and companies is, you know, put you in direct contact. And so I'm just gonna do a couple of, you know, literally, five overview slides, about how this kind of forum came to be.
Because, you know, we look at the opportunity within ADAS, and then we're gonna be able to, you know, dive into the details of that journey from ADAS into full AV. So really quickly, we're gonna talk about what has changed and what has sort of, you know, stayed the same. So first, we'll talk about the opportunities. When we started this forum in 2016, it really was about ADAS, right? It was just about Level 2 braking. This was 10 million units that we saw an opportunity for it to be essentially 100% penetration at some point in the future, and as you know, in auto, we like to make things complicated, so there's a lot of acronyms.
So we started, you know, adding our own sort of lexicon, and Level 2 plus, Level 2 plus plus is sort of what the focus is today. It's sort of an easy way to describe. You know, Tesla coined the term Autopilot, but essentially a convenience, the next level of cruise control. And we'll go, you know, go into why that's sort of most important, but you know, there's no argument that we're up and to the right. What was 10 million units. By you know, most estimates we're looking at 2030, something like 60%-70% penetration. And again, this is production cars.
These are cars on the road, thinking about your sort of, you know, your Audis, your premium vehicles, down to sort of the Detroit Three, and this is still driver in the loop. I just stole this from Valeo. You've seen this slide from every Tier 1, Tier 2. This is the opportunity why the ecosystem is so excited about rising forms of ADAS. For $400-$500, essentially, you know, your car with AEB, what is known as Level 2, can brake before hitting an object. By itself, different estimates are 30%-50% reduction in fatalities and injuries. Like, this is, from a societal standpoint, what we're really excited about in terms of saving lives.
As we move up that curve, you know, $500 with a couple of extra sensors, more compute, gets you some version of Level 2 plus. And then cars that aren't really around today but are coming, Level 3, Level 4, where you can actually take your eyes off the road, where the OEMs would assume liability, the dream of falling asleep in your car. You have extra levels of redundancy, extra levels of sensors, and this is where we see heavier penetration of imaging radar, of lidar. And there, estimates vary from 4,000 in the future, but probably closer to 8,000-10,000 today. We move from production cars into sort of the dream of full autonomous.
And, you know, we call it the dream, but, you know, we have players here, like we're gonna speak very shortly with Aurora, that are doing it in very short order. But one of the things, every bank, every VC firm has some version of this slide. It's a very big number, right? I think there is no argument that the end game of autonomous is sort of the curing cancer of mobility. It is probably the ultimate manifestation of physical AI. We put that number at $2- $3 trillion, even if I start sort of material miles in the mid-2030s, and I do that to sort of make the point, that as long as we are making progress toward here, this is the, this is the end game.
And probably what has not been discussed yet, great books have been written in the past, you know, Zero Dollar Car. That one hour that you are sort of captive in a vehicle, we think the data monetization could be a $1 trillion. And that is something that is probably very, very nascent. It's been obviously addressed, but it's one that's gonna get a lot more attention as we see people using these services, like Waymo, more and more. So again, over the next month, particularly with Tesla's reveal, this part, I don't think there is a real discussion, as we can all agree that the value is there. What we're all gonna debate is the how, the where, and the when of reaching certain levels of autonomous.
So I went through the opportunities, so let's just do two. You know, I said play referee, so we'll do two slides on sort of the other side. And this is really why is there a debate around the timeline of autonomous. And I love using this slide. We've used it for physical sort of components, but you know, if I translate this into Gen AI, my simplest one-liner is that ChatGPT, an error is not gonna be the same as one that will be made in a three-ton vehicle made of steel traveling at 70 miles per hour. We're gonna have a different level of safety cases. We're gonna hear about and a different acceptance level of failure.
The only thing close to automotive grade, you know, you would argue is aero or, you know, a military grade. And so I'll end on the risks. I'm an auto guy by background, so we always focus on the negative of what can go wrong. And so here, you know, just, you know, bringing up some of the risks that will be brought up over the course of the day. The first, legacy auto is slow. You know, four-to-five years is probably for a production car, actually quite fast to bring something to market. And sometimes the market has completely moved from the design cycle to when it is ready to be delivered to consumers. Delays, one-to-two years is sort of easy standard.
Everything from the technical side to what could be just the consumer adoption of that vehicle. We are at a time of seismic change in OEMs. I think we all know that the sort of we've seen the rise of China, we're seeing the rise of companies like BYD. If you wanna talk about AI exponential growth, look at BYD units. They were 400,000 units in 2021. They're 4 million units today. That 10X is the sort of level of exponential growth we're seeing in China. This topic of insourcing versus outsourcing been a two-decade-long debate, particularly in ADAS.
But actually, if I think about, you know, moving into AV tech, legacy OEMs have probably started and abandoned internal R&D programs at least two to three times, and that's very different than what we're seeing in China, where we're seeing much quicker adoption cycles and a lot of vertical integration. BYD is literally going down to the chip level. You know, this idea I call chasing the shiny new object is sort of related to that insource versus outsource. There are lots of advancements in tech and AI that have, you know, sort of required a changing architecture, but what you've also seen is that friction cost has caused a lot of the R&D programs at the OEMs to look like stop-start, and they have sort of gone about faced several times.
We can go into, you know, some of those, and a lot of the people who will be sort of presenting today are in that new generation of outsourcing for legacy auto. And finally, I think this one is the most interesting. It's sort of the chicken and egg of poor consumer awareness. McKinsey has done several studies across many countries in the Western world, and consumer preference for ADAS and assisted driving is just really low. It's about 10% in terms of a buy factor. Compare that to China, where it is the factor. It's almost more important to have the tech inside the car than EV. Now, we can ask: Is this chicken or egg? Is it because there's not good systems yet?
You know, is it because outside of maybe Tesla, who has almost defined the genre? You know, we call it. Alex Roy calls it narrative command. Basically, like most of the terms that we use come from Tesla in this space, for the good and for the bad, but this is an OEM issue because ultimately, you're gonna have to price, right? When I showed you that Level 3 technology at $3,000-$5,000 of cost, you're gonna have to be able to charge $10,000 for it. You're gonna have to be able to charge $100-$200 a month, so that's it, from me. Today is really about balancing that opportunity set of the enormous TAM with the risks along the way.
What I really want to do more than anything else is I hope you enjoy yourself, and I hope you learn a lot today, so we have a fun-filled lineup. I think it's the best agenda I've had in the seven years, and with that, why don't we welcome Aurora to the stage for the first presentation?
All right, why don't you take this one right here?
Here you go.
We'll just have the hand mic. They were saying you're really loud.
I am. And that was a great, great overview. Being a legacy car guy for over twenty years, a lot of that stuff really...
Excellent.
Thank you very much. Okay, I might stand for this first part. Chris, and thank you. Thank you guys for inviting me. Really pleased to be here, talk to all of our investors, but also just be able to share a little bit more about Aurora. I'm gonna just cover a couple of slides for those who may not be quite as familiar with Aurora, and then we can get into the Q and A, the meat of it. So anyways, unfortunately, and to keep my lawyers happy, I do want to remind everybody that I'm gonna provide forward-looking statements and answering questions in the presentation. And obviously, it's inherent with appropriate risk. So please, take that under consideration. All right, first off, let's start with Aurora. Okay, for us, we are a mission-driven company.
Our mission is to deliver the benefits of self-driving technology safely, quickly, and broadly. We were founded in 2017. We have about 1,800 employees today, and every single one of them can tell you our mission, that's what drives us, and what also drives us is the ability to transform the industry, and so we're really excited about that. We think we're in a really tremendous position to be successful, right? If you think about it, trucking is a massive market, over $1 trillion of revenue in the U.S. I'll talk a little bit more about that. We are confident that the Aurora Driver will unlock value for our customers, both on the revenue, utilization, increasing asset utilization, and on lowering total cost of ownership.
We are the only player that has the strategic partnerships to truly scale the business, where you need both OEM partners as well as Tier 1 providers to deliver the tens of thousands of trucks that are required to be successful. The competitive landscape, always changing. It's probably cleared in the last year and a half or so with a lot of people who used to be presumed to be leading no longer in this marketplace, so we think we have a clear playing field, but there's always new entrants, and we're always focused on what we're delivering. We have a tremendous group of investors, including some of you today, and you know, we have sufficient liquidity to not only go past our commercial launch, but really fund ourselves well into 2026.
Then obviously, we think we have a really attractive business model. This Driver-as-a-Service long-term business model, which is highly capital efficient. What are we building? The Aurora Driver, for those of you who don't know, Aurora Driver is a combination of hardware, including our proprietary FirstLight, long-range hardware. It's our verifiable AI approach to software, and then data services, things like our fleet intelligence and our remote assist to ensure that the trucks are consistently operating and highly utilized. This technology, from its inception, was always designed to be agnostic to both platforms and use cases, meaning, again, if you go back to our mission, part of our mission is to ensure that this technology goes broadly. The trucking market, as I mentioned, it's massive.
It's got great unit economics, $1 trillion, $4 trillion globally, over 200 billion vehicle miles traveled in the US, so it's a great opportunity. But to capitalize on that opportunity, you must deliver value for the customer, right? And for us, we look at the industry pain points for the customer and how are we going to approach that. First off, and it's the obvious and the most important one, there's about 6,000 fatalities in the trucking industry each year. That's just not acceptable. That's something we all have to strive to do better of, and our goal is to take these incidents down to zero in terms of fatalities. In addition, we also know that there's a tremendous shortfall of drivers, and you can-- there's all kinds of statistics on this.
ATA estimates that you'll be a million drivers short in the next decade. The reason for this is simple: the aging population of the driver workforce today, the fact that there's hurdles for younger, drivers to actually get into this market. It's predominantly a male-driven market, so you're missing out on a significant portion of the population, and frankly, preferences for what people want to do. They want to get home more on a daily basis, and these over-the-road jobs are hard, and they do a great job. These drivers are amazing, but the Aurora Driver will provide the scalable, stable supply for customers. Obviously, another pain point is, hours of service limitations. There's a reason for it for drivers, right? There's fatigue, there's distraction that happens along the way. It's eleven hours of service. Most people don't even average that today.
Our vehicle doesn't have these restrictions, so our vehicle can operate up to, you know, close to 24 hours a day, aside from, you know, refueling and service and things like that that happen. So we're allowed to increase the asset utilization for the trucks, and we can also deliver freight faster, right? Another industry pain point is fuel costs. Everybody can see what fuel costs happen. It happens to be the second highest cost in the overall cost structure for these carriers. And so we have the ability to reduce fuel, and we estimate that it could be up to 32%. One of the major drivers of this is just driving at a slightly slower speed, like 65 miles an hour, which has a remarkable benefit for all of the fleets.
And then, lastly, everybody understands higher insurance costs. It's not just the insurance cost per truck, but it's these nuclear verdicts that come on. We anticipate that the Aurora Driver will deliver a safer operation, and we have tremendous data for fault attribution, and this allows us and provides our customers confidence that we're able to drive down overall insurance costs. So pretty exciting. But you can't do any of this unless you have a partnership ecosystem, right? We're great at self-driving technology. We need a strong partners ecosystem to be able to deliver the benefits. So that includes our OEM partners like PACCAR and Volvo, that represent more than 50% or roughly 50% of the U.S. market. We have industry-leading customers, and they're helping us shape and form what this product solution needs to be.
There's gonna be times when you need to service your trucks. We have a partnership with Ryder to make sure that we're servicing and keeping these assets on the road, and then finally, we have kind of this industry-first, Hardware-as-a-Service partnership with Continental, and that's what allows us to be able to scale and build. We design and manufacture our hardware today, but we can't do it at the scale that's necessary or the reliability. Chris had a really good stat in there in talking about reliability as being a very important measure. We take that seriously. That's why we use folks like Continental. One of the most common questions we get, and we'll probably get it in all the investor meetings, is the regulatory landscape, and this is just because it's regulation, and it's constantly changing.
What we have today is the blue represents all of the markets where we can deploy autonomous technology without any further input from regulators. So they're either explicit, that we can operate driverlessly, or it's implicit, where there's no restrictions from it. This includes our launch lane in Texas, right? So there's a tremendous opportunity here, and we're really excited about this. There are a couple of markets that we hope to unlock, and we're working with them closely, like California, and so that's something that we continue to work on. The reason why we are confident in this approach is also the fact that we work really well with regulators, both at the federal, state, and local levels. It's core to what we do in terms of our partnerships. Okay, and then finally, maybe I'll just close with this, and then we can go to questions.
And, you know, Chris, you did a good job of describing the market opportunity. Here's how we think about the market opportunity, and this is just in the U.S. If you look at this map right here and just take the solid blue lines, those are all the routes that we expect to be operating on by the start of 2028. That represents about fifty billion vehicle miles traveled, of the two hundred billion overall. And of those lanes, about 60% of them are over 600 miles, which exceeds the hours of service limitation. So we think this is a tremendous opportunity. There's a lot of self-similarity, so we're very confident in our approach in terms of being able to unlock new lanes, and I think this is a great stepping stone for us to have an impact into the trucking business.
With that, maybe we can do some questions.
Excellent. Well, look, we really appreciate the overview, and as my intro remarks, you know, sort of hit the point of, we're gonna know very quickly, right, about you know, Aurora on the road and in the public. But maybe we can start. You mentioned that core mission. Every company, right, over the course of the day talks about employing AI. Some do it on process, some are doing it more on the front side of the software stack.
Can you talk about, and obviously, you know, you're gonna be parroting what Chris says on the tech side, but a little bit, what makes Aurora different about your approach versus, you know, everything else that we see where people say, "Look, we can skip steps and be there quicker?
Yeah, it's, it's a great question. I'm not sure you want the CFO answering that when we have three founders who are actually considered some of the world's most foremost experts in this space, including Drew Bagnell, who essentially is, you know, revered as a machine learning and AI expert. But I'll give it a shot, 'cause you asked. It does start with our mission, right? Our mission: to deliver the benefits of self-driving technology safely, quickly, and broadly. So that means it starts with safety. And we deploy a verifiable AI approach to our software. And you know, what does that mean? It means we're taking the best of modern AI techniques, which we've been doing and talking about since its inception. There are things that AI does that are remarkable to deliver human-like driving behavior, right?
Things that you just simply can't code, that allow us to be able to operate, and so we use AI heavily on that. But we also encode it with some simple follow the rules of the road. So what does that mean? You don't need to teach a system to stop at a stoplight, a stop sign. You can just tell it. And interestingly, you know, if it was a learned behavior, I'm not sure it would learn the right thing, because only 11% of the people actually stop at stop signs.
Hmm.
So you guys need to really stop a little bit more. But, I think for us, the other thing is it's the introspectability of our approach, right? The ability to understand what's going on in the system. It allows us to have the necessary transparency with our partners, with our regulators, and to build trust with the public. So for us, we think that this is just different than this end-to-end approach, where you don't exactly know what's gonna happen. And for us, we wanna make sure that the system operates and behaves exactly as it's intended. Again, we're not coding how to drive these complex situations, but we do wanna know what's going on when we don't think the system will operate as intended.
If we focus on, right, like, three true giants, you know, from, you know, working on different parts of the autonomous stack over the last, like, ten to fifteen years in the founders, right? You settled on a go-to-market to start with a limited ODD of long haul. Can you talk a little bit about how we think about moving from long haul to side streets, to eventually maybe looking at passenger vehicle? Just why the focus on long haul to start and sort of, you know, taking bite-sized chunks of the elephant, so to speak, to start?
Yeah, sure. I think there's a couple of things to, to point out, right? Why trucking? First, you know, we've always ensured and developed a technological approach that would allow us to capitalize on all markets. We chose trucking as the first market when we acquired our now proprietary FirstLight lidar technology in Blackmore. That gave us the confidence to be able to operate a tractor-trailer at 80,000 pounds at highway speeds. Before that time, frankly, we couldn't operate in trucking where we felt confident and to be safe. So the first thing was, we needed a technology unlock on the trucking side. The trucking side is massive, right? It's $1 trillion. It's got great unit economics. In comparison to ride-hailing, the drivers make about 3x.
We have the ability to immediately create value in the trucking space by addressing industry pain points, and those are different than in the ride-hailing space, right? So we think that there's a tremendous opportunity to start in trucking. And everything that we do in the trucking space is actually transferable to the ride-hailing space. And just also to confirm, we do operate surface streets, right? Like, it's not like we're just driving between two highway endpoints. We have to operate on surface streets. We drive to our terminals. We drive right by future endpoints of our customers. So, but we think trucking is just the right first market.
I think once we've been able to demonstrate the technology, create value, get to scale, we will have then driven down the cost substantially on the hardware and even really on our operational support, to enable us to be more successful in ride hailing. Because ride hailing has got a different cost structure. You need to actually drive down the cost in ride hailing to grow the business, whereas in trucking, that's not really a requirement. There's a ton of market opportunity out there without actually growing the business, so that's why we're doing it this way.
Now we really get to sort of your wheelhouse. We talk about go-to-market plan and making money. So you addressed OEM partners, you talked about scale, you talked about Conti. One of the things that we have appreciated, particularly from the public markets, is that each quarter, you've sort of laid out that plan towards getting to year-end launch. Sort of the updates we received each quarter have been, largely, we're on track. You've also talked about a meaningful portion of your 2025 capacity is already contracted. So the question that I think would be, and also helpful for people in the room who are less familiar is: What are some of the commercial goals for the rest of 2024, and how should we think about what this looks like in its very nascent state in 2025?
That's fair. So I can't tell you. You know, we have earnings coming up-
Yeah
... in a little bit, so I won't give too much away on some of the progress. But what I can tell you is, for 2024, the remainder of 2024, our focus is on working with our pilot partners, right? Continuing to deliver and increase the number of loads that we operate for them. Integrating our systems so that it's seamless. Continuing to work on operational procedures, on how we want the service to work, not only today, but in the future. So those are some of the key elements in 2024 that we continue to work on.
You know, recently, we had a partner summit, and we went and had a deep dive with all of our existing customers, as well as you know, maybe half of the crowd was new customers that are interested in it. And we talked a little bit more about our approach to safety, how we're progressing, but also, they got to hear from regulators, law enforcement, including Customs and Border Protection. So, a lot of what we do is about transparency and making sure that we answer as many questions as we can and really just being the benchmark relative to everybody else in the industry in terms of our transparent approach. In 2025, we're gonna deploy a deliberate, kind of a, call it a this is a safety-critical industry.
Mm-hmm.
Throwing a ton of volume in there just to start is just the wrong approach, so it's gonna be a deliberate crawl, walk, run. So I would say when we first start, our focus is gonna be on having a great product, demonstrating the value proposition that we said, and starting to integrate and operate from, you know, having a single truck, to, you know, maybe ten trucks, over the course of, time. I think in the first part, it's all about demonstrating the value in the product. The second half of 2025 is gonna be about expanding our capacity and expanding our customer base. So, we'll be able to share more later, but, for now, that's what I can tell you is kind of the main elements.
And then I guess the last question from me, then I'll pass it over to Doug, is: How much do you think, in terms of the launch, is now in your own control, versus you rely on getting also the green light from your partners like Volvo and PACCAR? Meaning, is it on you to validate the last, you know, six months of software, or are there other factors that could be out of your control that could, you know, theoretically delay or slow down the ramp?
Yeah, that's fair. So I think a couple months ago, we talked about our path to commercial launch and focusing on kind of four elements, right? Are the customers ready? Are the regulators ready? Are we ready? And are our partners ready? We've made tremendous progress on all those fronts. I think we've been pretty clear and actually already talked about the regulatory front. When we're ready to go, Texas is ready to go. We've had customers who have said, "When you're ready to go, we're ready to grow," and we actually have contracts in place to support that. I think the focus right now is really it lies with us, right?
It is continuing to work on finalizing and completing the safety case, as well as ensuring that the vehicle operates or the truck operates as intended on our anticipated launch platform. So for us, we still have some work to do on closing the safety case. We still want the system to behave great, and for us, we're more focused in on ensuring that we're gonna have an outstanding launch that we can all consider safe than we are gonna rush to conclusion. So I think it's mostly in our hands. I would say the great part about this, and you had also mentioned this, the AV industry has had a tremendous amount of foot faults. We're our own worst enemy sometimes, and we used to measure risk in quantum of years.
Now we measure it-
Yeah.
And we're confidently measuring it in months. So we are very excited about the progress, and we'll have more to share in our next update.
Great. David, let's talk a little bit about BOM costs here on the Aurora Driver, how your partnership with Conti is going to progress. You know, Conti, another AV forum participant who has a pre-record live online right now. Shameless plug here. But, you know, how are they gonna help with scaling and driving down some of the significant costs associated with just the sensing and everything associated from a hardware perspective with the Aurora Driver?
It's a great question, and I don't mind the plug 'cause they're a great partner. When we think about BOM costs, let's start with where we are today. Today, we have a designed, engineered, and manufactured, commercially ready hardware that we actually deploy on our vehicles and will deploy on our early launch vehicles that's done completely by Aurora. It's great technology. It is not scalable, right? We are not a manufacturer at scale. So we're gonna be able to take the units we have and deploy them. Before we get to Continental, we actually add in Fabrinet. Okay, so Fabrinet is a contract manufacturer based in Thailand. They're very well-respected. So, later in twenty-five, we're gonna shift to our next generation of hardware.
Again, designed and engineered by Aurora, but in this case, manufactured by Fabrinet. That will allow us to actually get a step change reduction in cost-
Mm-hmm
... as well as being able to build a more reliable product at higher volume. So we're going from kind of tens of builds to hundreds of builds on a kind of monthly basis. And then we get to Continental, and Continental is yet another step change in terms of reduction in cost, right? And in this case, we're co-developing the system with Continental. So we're taking the best of what Continental has, including some of their high-quality off-the-shelf parts and a lot of their engineering activities. As a matter of fact, they're spending a substantial amount of money helping us engineer and manufacture this product, you know, north of $300 million that they actually are investing in the business, and they will be repaid on a price-per-mile basis. So for us, it's great.
It allows us not to have to spend so much in terms of the development, to find a partner that goes from hundreds of trucks to tens of thousands of trucks.
Mm-hmm.
It's a great progress. They're a great partner. They have a lot of new technologies. One of the things that they're gonna bring, which is gonna be different, is they would like to bring in a secondary or a backup computer, which we call an MRM. But they're able to bring in a secondary compute, which is of lower cost, and it's kind of a heterogeneous design, as opposed to today, where we have two homogeneous computes that operate separately in the case of a failure. In this particular case, we'll have a main compute, and then they'll do a second compute, which is gonna be lower cost.
It will still be capable of doing everything we need to do 'cause the second system we call the fallback system essentially is designed to get to a safe haven, right, in the event of you know a critical failure.
Why don't we do two long-term financials? And then we'll open it up to Q and A. So 2028 math that you put out earlier in the year, $1.3 billion targeted rev. It's something like 70% software gross margins. It works out to 10,000 trucks, 180-200,000 miles, something, $0.70 plus. Rough math. That's not my question. The question is, is the miles... Clearly, we need to get there over time. How much of the actual contractual relationship with some of these customers have been negotiated, or do you have to sort of do that? Meaning, like, the structure of pricing and things like that, have those conversations started on, you know, signing commercial deals?
All right, so let me clarify one thing, just so we get this straight, so we actually think that the vehicles will probably operate 200-250 thousand miles, right, so it’s a little bit higher. In terms of your question, absolutely, we have conversation with customers. That is our long-term driver-as-a-service business model, right? And that is essentially when we have our generation of hardware that is being line-side installed at the OEMs. We expect the hardware will come from Continental. Some of it could come from Fabrinet actually early. But we expect it to be line-side installed.
At that point in time, a customer would purchase a truck equipped with the Aurora Driver hardware and software, and they would sign a subscription service, a contract, a per-mile contract with us. All of our customers. We've validated the appropriateness of this model with our customers. They wanna own the trucks. They would rather do this driver-as-a-service business model. It matches what they do in the rest of their business. So for them, it's very consistent. They've helped us shape and form how we do this and when we do this. So, they've been along. I, we do not contract, and the industry doesn't contract, like, in 2027, 2028. They contract earlier. So our focus on contracting is 2025 and 2026, when we're more of a Transportation-as-a-Service provider.
But we have every intent to be able to roll that into a driver-as-a-service business model.
Well, I have one more, but I wanna open it up to any Q&A first, just so we have time. I need to check because people have shot. There's got to be some. I see familiar names. Anyone? All right, then let me go into the funding question. So you raised $483 million about six weeks back. You talked about from a liquidity position that basically on that 2028 plan, it probably that raise puts you through about half of the capital that you need through 2028. Talk a little bit about what the roadmap for, you know, sort of funding looks like. Did I phrase it and sort of, you know, give the summary correctly?
Yeah, I think that's a great summary. We had said earlier in the year, well, even back to our prior fundraise, the quantum of money that we expected to want to raise until we achieved a point where we were positive free cash flow and self-funding. We've consistently been opportunistic about our fundraising efforts. This last one was another opportunistic raise. We actually started out thinking about three fifty-
Mm-hmm.
It immediately upsized to strong demand to $420, and then with the overallotment, the Green shoe, we ended up at $483. So really excited. We have an amazing investor base. Really folks that believe in the market, they're long-term in focus, and they're some of the most astute investors out there, some of them being in the room here today. Prior to the raise, we had said we need about $850 or so, so we've taken about half of that down. We think that's still a good estimate. You know, things happen.
Yep.
We might raise a little bit more in the future if an opportunity arises or if we have risks come in. But right now, we think we've taken half of it down, and we're gonna approach future fundraising the same way we've done it in the past. We're gonna be opportunistic. We believe we have tremendous catalysts coming in the future that consistently demonstrate the proof points, the value proposition, and the confidence in Aurora. And we need all of those to be true, right? We need to have these external proof points. We need to build the confidence. And so for us, we'll probably be opportunistic again. We're excited about the future, and we're excited about our potential.
Maybe we could follow up on one of the... Oh, is there a question in the back? Yeah, go for it.
Can you talk a little bit about your cost structure, like both CapEx per vehicle today, and where you hope that to get you over time? And what are the drivers of potential cost reductions over time? Is it scale? Is it high articulation, what have you? Then also the OpEx per vehicle, how do you think about that? Like, I don't know, mile hours per mile or perspective.
Yes. Well, I can kind of talk about it. So we don't publish kind of like the per truck cost the same way, but let me take a step back and see if I can answer your question. So today, we're, as a pre-revenue company, we focus on controlling our operating expenses, our OpEx, and ensuring that we're deploying that for the sole focus of delivering the mission on hand. And so for us, we look at the development, and we're very much focusing on having the right people. So we have reasonably high personnel costs, as well as making sure that we are working on projects that support our future vision. So not just launching a truck today, but also being able to scale a business in the future.
On a capital front, we have a pretty capital-light efficient, capital-efficient business model. So largely, capital today is centered around two things: building, buying trucks. Well, three things: buying trucks, building the hardware, and where we operate our vehicles. In terms of all three of those proof points, the first one, buying trucks, we're moving to a driver-as-a-service business model. So, our customers will actually be buying the trucks. We will not be buying them. So we expect that to be, super valuable for us and very capital efficient. In terms of the hardware, our Continental arrangement, part of the benefits of the Continental arrangement is not only are they co-developing and, engineering and manufacturing, but they're also doing sales, service, and financing.
So Continental will be financing the hardware kits in the future, and we will be paying on a per-mile basis. So we will not need a lot of capital to carry that inventory. And then the third thing is our terminals, where we operate today. So we operate. We have four terminals that are officially operational today, although we did announce the anticipated add of Fort Worth, Phoenix, coming up in 2025. But today we operate our Dallas to Houston lane, which is in South Dallas, going to Houston, and then Fort Worth, going to El Paso. In the future, we will not own and operate terminals. We are going to go to predetermined points.
They could be owned by third parties, they may be customer endpoints, et cetera, but we will not own and operate a broad terminal network, so that's how we're being capital efficient. In terms of the cost downs, we measure cost. Let's put it this way: We have a plan to get to gross profit positive in 2026 to achieve this gross profit positive in 2026. We're really focused on the BOM costs. As I mentioned before, we have iteration step change reductions of roughly 50% each time. First, when we go to Fabrinet, and then when we go to Continental, and then the two other cost drivers, the big ones, are the ratio between the remote assistance to number of trucks operating on the road. We don't publish those stats today.
We say that we're gonna start with one-to-few and go to one-to-many. We're highly confident the one-to-many is gonna be achievable in some point. I won't tell you exactly when yet, but we'll more to come on that. Then in terms of rescue, that's. All trucks break down, right? You have engine failures, flat tires, et cetera. Our goal is to make sure that our trucks needing on-site support are as low as possible, and we actually expect in the long term with how we inspect the tractors and trailers before they go out on the roads today, we actually may be able to improve upon that from the industry benchmark. So those are the main drivers, but we don't publish any specific stats on that that I can reference and give you numbers on.
Yeah. Okay.
Just gonna talk a little about the competitive environment. Who do you view as your biggest competitors, Kodiak, anybody else, and how do you differentiate?
In fairness, we don't spend a lot of time worrying about competitors, and it's not that we don't respect them. Every accomplishment in that space we think is great. But if we spend a lot of time on that, we'd just be spinning our wheels a lot. Like, last year, the three most talked about competitors were TuSimple, Embark and Waymo in trucking, and all three of them aren't in trucking right now. We think Embark has maybe a slightly different approach in terms of how they're thinking about commercializing the technology, and, you know, we wish them the best. There's a lot of new competitors that are coming into the market.
I don't know that we totally agree with their approach in terms of how they're thinking about deploying a safe and reliable system, but you know, we don't spend a ton of time on that, just in all honesty. Our focus remains on delivering our mission, and we've got enough challenges as it is to make sure that we can do that.
And Dave, I think we have to end there just for time. I know we could go on. For anyone who wants to learn more, please reach out to the team. Also, happy to put you in touch with Aurora. Round of applause for Dave and Aurora.