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

Okay. We are back here for our last and final presentation on the transport track sticking with the freight tech side of things. Very happy to have Aurora here with us, CEO and Founder, Chris Urmson whom we spoke with for quite some time, one of the leaders in this space, as you all know. And so I'm just going to turn it over to Chris who's one of the members of the team here as well, Richard and Stacy. But in terms of -- I think it would be good to have an update, you've got some slides here. Clearly, there's been a lot going on in the broader industry, but the company still seems to be more or less just keep into what you said you were going to do and following that time line. So look forward to putting a lot together. If you have questions in the room, raise your hand, we will get you a mic. But with that, let me turn it over to Chris.

Great. Well, thank you, Brian. I appreciate the chance to be here. Thanks for everyone joining us today. Yes, it is -- I think the industry is dealing with some challenges. But as a team, we feel like we've been able to keep our heads down and execute and do what we said we're going to do. So honestly, I'm more optimistic than ever about our position. So good morning, everyone. This is our -- before we jumped in, the lawyers, of course, asked me to put up our safe harbor statement. So we're going to be giving you a little bit of an update and answering some Q&A. Hopefully, the lawyers are happy. Okay. Okay. It doesn't seem like clicker is working. Here we go. So we're going to talk to you a little bit about the business. But before we jump into that, just about the promise. So this is our company's mission is to deliver the benefits of self-drive technology safely, quickly and broadly. And we see the opportunity to fundamentally transform transportation, both on the logistics side and in the longer term on the personal mobility and local goods delivery side. And that means -- this leads to an increase in safety, increased access, easier to move goods through the world and the opportunity to transform logistics. What we're building at the company is the Aurora driver. It's the capability for all kinds of vehicles to move safely through the world. It's a combination of software, hardware and of course, the offboard onboard data services that enable those vehicles to drive, and it will drive the same software and hardware drives big trucks, drives light pass-through vehicles and ultimately, the block drive delivery vehicles as well. The market we're going after first is trucking. It is a gigantic opportunity. In the U.S., the truck market is somewhere between $7 billion and $800 billion and globally, it's about $4 trillion. The technology, as I mentioned, we're developing also works for light vehicles, and so we ultimately will expect to operate in that space as well where personal mobility can be a $1 trillion market in the U.S. and local goods delivery driven, of course, by e-commerce is also a major opportunity. As we enter the trucking market first, it's great that we have some cool technology, but let's make sure that it actually does something useful. And as we talk to our partners, these are the key pain points that we see our technology addressing for them. They have a massive shortage of drivers. This is not just a recent thing. This has been a trend that's been building for decades. And so we can provide them a scalable, stable source of drivers. Trucks today are limited in their turn ability to generate revenue by how long the driver can be behind the wheel, which is 11 hours in the United States. We have an opportunity to double a core metric of their business, which is revenue per truck because the ore driver won't be limited by 11 hours a day. We can reduce fuel costs, which is about 30% of the cost structure of moving goods to the world. We expect to be able to reduce that by about 10% and of course, insurance is one of the rising costs in the industry. We both expect the trucks to be safer because the driver is always attentive, he's got a 360-degree vision. And then in addition to that, we'll have better data to adjudicate these events to help manage claims through that process. Well, today, we are primarily focused on getting the technology to market. We have worked closely with partners to springle the company to be able to deploy and scale. Today, we work with PACCAR and Volvo trucks. Together, they make up about half of the trucks sold in the U.S. market. And of course, we work with Toyota, the world's leading car manufacturer. We have a number of partners on the trucking freight. And today, notably, we are hauling loads today with our trucks operating autonomously for these partners. We currently still have an operator on board those trucks, but we're actually driving the trucks autonomously all and afraid for them. We work with Uber, a major shareholder and partner, obviously, the world's largest ride-hailing network. And then we've recently added the partnership with Ryder to allow us to provide fleet services to our customers and to our vehicles. The technology that we're building uses a multi-mobile sensor approach. That means we don't just rely on cameras, for example. We have a combination of our proprietary LiDAR and other LiDARs, Camera and RADAR data to be able to see the world robustly around the vehicle so that we can operate safely. And this is an example, just to hammer home the value of this technology we've been developing in-house. It's called frequency modulated continuous wave LiDAR. And the advantages of it are that we can see dramatically further than others can. And we also can measure the speed of the thing we're seeing instantaneously. So what you're going to see here is on the left, data from one of our trucks showing what we get from our conventional LiDAR, state-of-the-art LiDAR. And on the right, you're going to see what we get from the Aurora first light LiDAR and how that allows us to react earlier. And the vehicle is going to be driving along in the right lane of a freeway in Texas and were to come across a work vehicle on the side of the road. We're going to hit play here. And as we start moving in a moment now, we get 4 little dots you can see on the bottom right from the first light LiDAR. That's enough for us to know that there is something parked on the side of the road and has a good road user, we're going to slide to the left a little bit, create some space and make sure we're not overly kind of creating danger for those folks on the side of the road. Now we'll play the video forward here. And if you will wait -- and as you can see, the Aurora driver is already reacting and making a lane change to get over, and it takes about 9 seconds before the conventional LiDAR is able to get anything from that target. In that time, we have already moved over and gotten out of harm's way and reduce the risk of the road users. This is the difference, right? That 9 seconds is something like approaching 3 football fields of length that we're able to move or detect things beforehand, right? That 9 seconds is the difference between a demo that works most of the time and a real safe, compelling actual product in market, right? This is what you need to actually be able to ship something in this space and this is a technology that is ours, and we believe we're the only folks who can do it. Some examples of the Aurora driver operating and these are sped up just so we can kind of keep things moving here. Here, we're operating on surface Street pulling out of a depot, we're going to make this Texas U-turn, which is this lovely bit of road geometry where they allow you to make -- change directions on the surface streets without actually having to stop. And so the truck has to navigate the mergers and then, of course, gets onto the freeway here and deals with traffic. So this is the kind of thing you need to do to be able to deliver a product that actually is useful for a customer. We go to the next slide here. Obviously, the world is constantly under improvement. And so here, we're navigating various construction scenarios, dealing with land closures and shifting over again, our ability to perceive things of range allows us to react to these in a safe and predictable way. And then, of course, when people putting down cones, they're not really careful about making sure they're exactly in the right place. They shift over, and so the driver has to be able to adapt to that. This is the kind of thing that you would all do intuitively, but it turns out, it takes a little bit of effort to do well as a vehicle -- as an automated driving vehicle. And then finally, the road, of course, are not pristine. There's debris in the road. So here as we're driving along, we're able to detect in a moment here, you'll see it pop up a little bit of debris on the road. This is a tire arcus, I think we nudge around it, saving the wheels, saving the underbelly of the vehicle from whatever damage. And then for added degree of difficulty here. It turns out there's a vehicle stopped on the side of the road. We again see that early enough, we make a lane change over. You can see us detecting all the little warning cones behind that vehicle as well. So these type of driving behaviors are necessary for the vehicle to be able to operate robustly again in the real world. So about a year and a quarter ago, maybe about a year ago, we shared our first road map for how we were going to get from where we were then through to commercially launching the product. And I'm really proud of the way the team has been executing. We've just been delivering. We've said what we're going to do. We've gone out and done it and done it on time and on schedule. In Q1, we have a major milestone for coming up. We expect to accomplish being feature complete. At that point, the Aurora driver is basically doing everything we think it needs to do to be a commercial product. It doesn't yet know it well enough that we're confident in the safety of it to release it, but everything is in place. And now we're at like the really fun part, right, that engineers love that you get into just kind of burning down the list of things, right? Once you get to a discrete list of things, it's just execution and go. From there, by the end of the year, we intend to get to Aurora driver ready. At that point, we'll have validated and refined the system to the point where we would be confident to operate on the road without a person on board. And then in '24 is when we expect to commercially launch the system where we'll be able to marry that driver up with a vehicle from one of our OEM partners to start hauling loads with nobody on board for our customers. So the technical development program is going about as well as we could hope. On the vehicle -- on the operations side of it again, the business is going very well. We had hoped to get to about 30 loads per week in Q4, and we exceeded that. Then Q1, we had hoped to get to 40 loads a week, and we exceeded that in the first month of the year. So again, seeing reasonably good demand for the product. We're in pilot phase. We're learning. Our customers are learning, that's moving very well and we continue to engage closely with our truck OEM partners, folks that we're engaging with on a daily basis, bringing those platforms into place, so they'll be ready to marry up with the Aurora driver and actually deliver customer value. For us, the launch bar is really having a close safety case, and we've been kind of consistent about this over the years. We're not about doing a demo here or demo there. We're about actually having a product that's useful and that's safe to be on the road. The way we do that is by closing what we call a safety case. This is a description of kind of a structured argument that explains why you could trust it. And the top level on this is that it's proficient that the Aurora driver is capable of operating in the road in a safe way, that it's fail-safe. So if something breaks, we can detect that, react to it, mitigate the risk from it, that we're continuously improving. So we're learning from what isn't working. We're making that better, and we're taking the things that are working and making them better as well. We were resilient, so we can deal with things like cyberattacks and defend the vehicle from that. And then trustworthy is that we have a culture at the company that basically fosters safety that we have adjust culture so people can raise concerns, we can recognize that and there's something curative that's done there. When we think about deploying the technology, of course, one of the questions that's top of mind from folks is the regulatory environment. The great news is that in the vast majority of the United States, if we had a truck that we were confident in the safety of, we could put it on the road today. As you can see in this illustration here, there's just a couple of states in the Northeast where we wouldn't be able to operate the vehicles autonomously and California for us is actually one of the primary focuses. There we're allowed to operate light vehicles, but not heavy vehicles. There's progress now with the Department of Motor Vehicles moving to make -- to enable heavy trucks because back in 2011, the legislature approved both light and heavy duty trucks operating in autonomy. We had a major regulatory win over the last 6 months where Pennsylvania previously had been a state that allowed testing, but not deployment without operators. And the Pennsylvania legislature passed regulation passed a law that enabled us to operate without people on board. So really good progress there as well. One more slide. So finally, we continue to see Aurora positioned to succeed, that we've been steady and executed consistently against the road map we've said. We look at the landscape around us, and we see the field clearing in trucking. Some of our competitors are dealing with inner turmoil. Some are shutting down, some are backing away from trucking. And when we look at this incredible landscape opportunity, this $700 billion market space, having the technology of the team, the partnerships to go and take it. We're incredibly excited about that. We continue to execute. We expect to have feature complete meet that milestone at the end of the quarter and then continue to be, we believe, on track for Aurora driver ready by the end of the year. I can say there's this I was talking with somebody earlier this morning about the Gartner hype cycle and feeling like we're kind of in the trough of disillusionment. And as with many things, the media and the narrative is a little bit out of phase with reality. And from where I said, I've never been more optimistic about our prospects and the opportunity to have a transformational impact in the world. It's just an incredibly exciting time for us and I think for the industry broadly. So with that, I think we're happy to jump to Q&A.

[Operator Instructions] Maybe just to start where you -- one of the spots you ended with is the field consolidating and certainly investor interest in pre-revenue companies has come down as interest rates have gone up, of course. But when we do see however this consolidation unfolds, are there opportunities to look at that technology can fit on board? I mean, of course, the people are huge assets that's probably a given. But is there -- because the models have been different for a while now, which one is going to be the one you choose, whether it's the fast car and the freight together and LiDAR versus camera. So I don't know if there's really a horizontal consolidation here or is it just like picking up more people and talent.

Yes. So we've been pretty consistent for -- about the model, right? The model is we're going to deliver trucking first. We're going to build a technology base that applies a platform that works for both, but trucking is our focus. We've also been clear that our business model is one where we have no intention of competing with our customers, that channel conflict just seems like a fundamentally bad idea. Why go and figure out how to become FedEx or Werner, they do that pretty darn well themselves. Let's help them build their business, and that will help us accelerate into market and scale and grow as a company. So we feel very good about that. In terms of consolidation, at this point, we went through a major consolidation as a company. We acquired Uber, self-driving car assets, and that's turned out wonderfully for us. We're well through the pain of that at this point. The team is just executing in a way that I couldn't be prouder of. We feel like we have the team. We've got the tech. At this point, it would mostly be a distraction to acquire something else. And so we'll, of course, take every situation as it comes up and assess it. But right now, we just need to execute, and we're doing that.

And one of the other things I think is interesting in this dynamic now is some of the supply chains and the time lines get pushed out, there's a push to maybe try to monetize something different or pivot a little bit differently. We've seen that in a couple of different areas. But I'm just thinking of the Aurora offering and the technology, is there anything that would sort of fit that for -- in your mind, clearly, you've got the road map you just saw in your marching along toward it, but something, I guess, like first light by mark, could that be monetized in some way? Could you do licensing of it? Could it be maybe not sold off, but anything else in the portfolio, I think you kind of add on as you continue on the road map, but clearly, supply chains haven't been cooperative. So maybe you can expand a little bit differently.

Yes. And so for us, the supply chain has not been a major challenge, right, that things are working along. We certainly manage that. There is a broader impact. But today, the scale of what we do is small enough that we have been able to be dynamic and find paths and access to what we need. We certainly broadly, it would be wonderful if the world were in an incrementally better place, but that's not hindering us. There's certainly additional opportunities to monetize the technology we have. We look at our perception system. We look at -- as you point out, First Life, we see long term an opportunity to see that technology brought to the ADAS marketplace. But we've been very clear about our road map. We're executing that road map, and we see that as the right opportunity trying to compete with a Tier 1 who has the -- all of the inventory management, all the supply chain processes, it just doesn't feel like a game that we're well equipped to go win at. So let's again find the right partners, but CPaaS perhaps for that technology to come in. But we look at the landscape. We see how well our system is executing and how well our team is executing. Let's play to win, and that's what we're doing.

Let me get a microphone over to Chris [indiscernible].

I actually had a couple of questions on the proprietary LiDAR. Follow-up questions on the LIDAR. It kind of struck me as I saw the video maybe why Tesla vehicles tend to slam in the parked cars on the shoulder vehicles...

Well, it's like looking at the video, you can see that it's hard to tell from far away what lane that vehicle is in and if it kind of locks in on it is, oh, I'm following that vehicle and then it's too late when it realizes. But anyway, so the questions are, I guess, first, is there also some software aspect to that and mapping aspect to be able to perform the maneuver that you showed where it knows that it's a curve because the curve thing is tricky, right? Like when you're approaching a toll booth, you don't know I'm I on lane 10 or 11, right? But -- so how does it know that that's on the shoulder. Is that also part of being able to see further?

Yes, absolutely. So thank you. We're -- so one, we've talked in the past, we didn't highlight it today about our mapping technology. So we do build high-resolution maps. We would have loved to buy them from other people. We were not able to find once that met the requirements of the space, and so we deploy and build them. The point you make about the necessity of software here is really on target, right? So people often talk about the urban driving problem as being more challenging than the freeway one. And I consistently told them it's different, right? We don't think it scales as easily as the freeway driving, but it turns out that if you're trying to locate something 450 meters, 500 meters plus that down the road, that's a lot harder than doing it 150 meters. And 150 meters, you can just kind of like, oh, that's where it says it is. That's good enough, but 450, you got a little bit of error, you're a lane or 2 over and game is over. And so I think we just shared a blog post yesterday about some of our approach here with automatic recalibration on the freeway and not calibration tuning. But yes, there's some very interesting machine learning work that we've done to take that raw superpower we have with First Light really be able to deploy it well on the vehicle. And this is again, one of the messages we've been saying for a while, bringing that sensor technology in-house allows us to co-mingle right, the smart people that we have on the software side of the house doing perception with these brilliant physicists type people that are building the LiDAR like, hey, what do we need to make this really go? And it's a lot of furnace.

And on the sensor itself, can you give an idea of kind of the price point? I'm just thinking about to Brian's question of the broader application for other vehicles. I mean, LiDAR has traditionally been really expensive, which is why people like Elon really haven't used them.

Yes, right. And Elon, it's a business model problem, right, rather than anything else as far as I can tell, it's that they want to put the sensors in the car, so they can harvest the data from it, and it turns out that cameras are basically free and LiDAR is not. So one, there's nothing fundamentally expensive about a LiDAR right, that it's all stuff that when it will go down the industrialization scale curve or a price curve because of scale, it's just going to happen, right? There's no reason this technology can't be sub-thousand dollars and ultimately less than that. For us, we are investing in integrated photonics, silicon photonics. So today, the inside of the first light LiDAR has got a bunch of discrete optical components that makes it relatively expensive. The team we're developing is these basically little optical chips, right, that we put together that will dramatically like order of magnitude, reduce the cost of the sensor, right? Like anything else, if you take discrete electric components and compare the cost of assembly and the reliability of [indiscernible] versus electric is on a chip, it's, again, order of magnitude difference. And so we're seeing really good progress with that. Again, we've shared some of the progress there, I think, in our last business review update. We've done an early integration of some of the early optical chips, and we initially expected the performance of the first light LiDAR with them to be degraded, and we'd have to do some work to bring it up and turned out in a number of ways that performed better than the discrete components. So we don't see that intercepting in the next -- in the very near term, but that is a way that we are set up to bring the cost out of that.

Are there any moving parts in it or is it all solid state?

There are moving parts. So if you -- in a LiDAR, you can think of there's 2 core -- there's a variety of core technologies for this moment we'll talk about the -- basically, the range measurement engine and then where you put the light, right, that steering. And so for us, there's a moving -- there's actually 2 moving parts today on the beam steering side. Ultimately, we expect that will come down to 1. But it turns out, there's lots of moving parts in the truck in a car. And so in particular with us, we use a rotating Polygon. And so just spinning a thing in a circle, I think we've got that. So thank you.

So just on the road map going back to the feature complete, and we've seen the different capabilities, including the one we're just talking about with the finding 9 seconds before, which is pretty substantial.

I think when I was here a year ago, I think it said -- I think it was a smaller number, I think it's 7. So we've pushed that out, which is exciting, too.

So -- but in those terms, like so what's next? And they don't have to drive in snow and all this other stuff because that's not the ODD, but what else are you looking to specifically get to test? And I know you said you're not -- when you get to that point, you're not like ready to go, it's just at that point where you know you can do it and now you just have to prove that you can do it and prove to others perhaps. But what's left from a technical capability standpoint? Because it looks like it's getting filled in pretty well, but I'm sure there's always something else that you want to get full confidence on.

For sure. And so that's what feature complete is about is to get to the point where we do all the things we need to, to launch on the route in Texas, right, and then begin to build out. And so for a long time in the future, we're going to be adding features and capabilities and improving things, right, much like Henry Ford rolled the first [indiscernible] off the line and what was it, 1912. And it turns out we didn't fire all the engineers the next day, but the engineering and automotive has increased over time. So I expect we will continue to do work there for the next century. But between now and feature complete, which I think is about 2 weeks away, we're working towards, right, things that we expect to come online are an example. Despite our best efforts, our vehicles will be in collisions. We need to understand that, that happened and then take some kind of action in response to it. Our vehicles won't initially want to operate in very inclement weather. So we need to be able to have a methodology to understand that the weather is worse than we expected or worse than we allow the vehicle operate in and then take a response to that, whether that's slow down or pull side of the road or just not launch, right? So there's a variety of these kind of what we call of nominal situations because it's not fair to say that -- so again, maybe we'll get a little geeky for a moment here. So you said ODD, maybe not everybody knows what that means. That means operational design domain. So that's what did you design the system for. Well, it turns out if you're in the real world, you can't say that we're not going to have heavy rain in Texas, even if most of the time, it doesn't, rain like 98%. But your truck might be on the road and weather forecasting is still not perfect, and it might really rain heavily. So we can't just say we're not going to deal with that because that would be unsafe. But we don't intend the vehicle to operate in it. And so this is what we would call an off-nominal situation. And so we want to detect the rain is heavier than we expect to drive safely in, let's now slow down, pull aside on the road. And so there's a variety of those things that the team has been putting in place. And the stuff that doesn't show up most days, some of them very rarely, for example, the collision situation, but that you actually have to do if you're going to launch a product. It's the not sexy, not glamorous part of what we do, right?

But you don't have to solve everything all at once, just what you're going to?

Right.

So for you to tackle on that...

But even in Texas, it snows, right? And so I was -- I actually have to be there when it was going. And so we have to have some way to mitigate that risk, right? And it turns out that we've been very thoughtful about this, where people talk about the long tail of problems that you have to drive in self-driving. Well, it turns out, if you think about it, there's a lot of them that kind of coalesce the same mitigations, right? And so we've been thoughtful about that and have been building and doing that analysis for some time. And now we're going to the point where we're implementing those mitigations. And again, we're feeling very good about our prospects of meeting future complete.

Well, it seems like the -- even the freight markets are softer and trucks aren't as expensive to hire. It seems like there's still a lot of interest in the product. So maybe you can just talk through the partners, the willingness to look long term, which I assume most of them still are because they got into that to begin with. But are there -- it seems like also this -- the partnership ecosystem is pretty well filled out as well. So maybe you can just give us a quick update in terms of if there's anything else to add along those lines. And if there's still a lot of interest even if there may not -- even if we are in that disillusion part of the hype curve.

Certainly. And everyone reads the press, right? And so it's -- I think, much like our pretty else, our partners are incrementally less enthusiastic than they were 2 years ago, right, or a year ago. At the same time, the premise of why we're doing this has not changed at all. There continues to be a shortage of drivers. And there continues to be -- I look at an industry where we have the opportunity to double the core metric of their business. right? Like that doesn't come along very often. That just on its face is profound. And then when you add to that, the safety benefits, when you add to that, all the rest of what we can do for them, it's very clear that they're enthusiastic about it. And it's -- they want much like everybody else, they want to see it work for real, and we look forward to demonstrating that.

So one of the things I would like to see work for real would be California move after what 2011 you said. So it seems like every time we talk it's still like moving a little bit forward or maybe not supposed to move forward, but it sounds even being too optimistic, but it sounds like maybe you actually are moving a little bit closer to it. So maybe you can get a quick update on the regulatory front, not that you're launching there to begin with, but certainly, you open up the full [indiscernible] quarter, that would be pretty meaningful...

It would be... We'd be very excited about it, right? And the wheels of government turned slowly. But we're -- we found an engaged audience for us as a company and as a resident of California, it's embarrassing to me that we have -- we're a major footprint there. The technology certainly feels like it was birthed in California. The fact that the state is not taking advantage of that, the economic growth that will come from it is sad. As a company, we see a lot of opportunity. Obviously, in California, we see that as an important market in the long term. But if California doesn't play, there's a lot of other trucking that happens in the U.S. And while we may have to get up to the California border and stop there, it turns that all the way to Atlanta to Florida, to the East Coast, right, there's -- it will get itself figured out. At some point, again, it is obvious, this is the right thing to do from a societal point of view, from an economic point of view. I have confidence in the long term we figure that stuff out and it works itself out.

On that point from all the safety aspects. I know you guys have had the safety case and the framework out there for a while now. The question always comes up, like, when do you know that it's safe and how many people do you have to prove to that I'd be interested in hearing a little bit more on that? And then also with one of your competitors having an accident that got a lot of press and just went through the FMCSA review. How are the -- how are your interactions with the regulators post that accident? And then also since then as you continue to progress on this time line?

Yes. So a bunch of different questions in there. I'll try and piece them back together for so. The safety case is really this explanation for why you can trust the Aurora driver on the road. And what we have shared is that we expect, after meeting feature complete, we'll begin to share with the market the autonomy performance indicator -- the autonomy redness measure. So if we have 2, I'll get better at that. The autonomy spreader measure, which is how complete that safety case is, right? And that's the -- effectively think of that as how much work did we have to do between feature complete and Aurora driver ready. And so we look forward to sharing that and you'll be able to kind of see the progress there concretely. We actually -- because of the regulatory environment in the United States, we don't actually have to prove to regulators that the vehicle is safe. We have to have confidence in. They have a number of sticks they can use if they don't have confidence that we have confidence, right? And we demonstrate that we're not meeting those, their safety expectations, but we're able to deploy the product. We don't have to go through a regulatory review process. At the same time, we have taken the stance and the approach that the regulatory folks our partners in this, right? We are trying to do something transformational for this industry. It would be naive to think that you can do that without actually engaging with the regulators. And so for -- I was one of the very first people back when I was at Google to go and spend time with the Department of -- Federal Department Transportation with the California Department of Transportation and help them understand what the opportunities were, what the challenges were. And we've been able to continue to build that relationship and that trust over the last good lord, how long has it been? A decade plus. And so I think that the -- we have a very good relationship with them. We've had senior members of Department of Transportation down in Texas, in the trucks, and we continue to engage with them. I think there is a complexity of understanding that not everyone in the industry is the same and that I think we have done a good job of building trust and demonstrating why that trust -- why we're worthy of that trust through our actions and communication with them over time.

And any good more question. Any other learnings feedback or -- from other incidents that have come out to the rest of the industry in terms of your interaction, do they feel more comfortable? Do they worry when that happens? Or are they not as focused industry-wide, it's just more company specific?

I think they obviously have questions, right? They kind of -- that happened to them, could it happen to you? No. Right? Like that -- at least not that I don't -- like we will have a bad day, right, that our trucks are on the road, something will happen. We're doing everything we can to minimize -- to kind of keep reasonable risk here. And we've had a team of humans, right? And we're at some point, there's going to be some kind of thing that is an issue, but we'll be able to address them and it will benefit the whole fleet when we fix it. So we feel very good about that. When we talk with regulators, they seem to appreciate our approach, they appreciate our transparency. We've been a leader, both publicly and with the regulatory folks in being transparent in how we approach and think about these things.

Maybe the tougher question or the more important question is not how do you convince the regulators it's ready? How do you convince the customers and shippers that it's ready because they're kind of ultimately taking the risk of deployment, right? So what's...

Absolutely.

How do you think about that?

Yes. And it's really about engagement, right? And I say this a lot, and I know some people use it as a soundbite, but we think of our customers and our OEM partners as partners. And just yesterday, actually, Sterling, our Chief Product Officer, was on stage of South by Southwest with one of our customers and one of our OEM partners and one of them unprompted said we're the gold standard in safety, all right? And so I think they get through our education with why the approach we're taking. They understand how seriously we take it. And we keep them updated on a lot of the kind of performance statistics that they're looking for on that.

Do you get the sense that there's some sort of metric or a bogey that FedEx or USX is thinking like, okay, I need to see them do x before I'm ready to put this on the road with my -- on my truck.

I think one of the biggest things is really having the opportunity to experience it with their experienced drivers, right? And one of the things we look forward to doing in the coming months and years is doing -- right now, we do pilots for them where it's our operators in the trucks hauling loads. We look forward to doing kind of field operational trials with them where their operators get to see the behavior in the vehicles.

And then I wanted to come back to one other thing. You showed before the TAM of the $800 billion. Actually, this is something Brian and I talked about a week or so ago. Like what do you think is sort of the real addressable market because there's a lot of short haul stuff in there that wouldn't make a lot of sense for someone to deploy a complicated system like this. So if you sort of distill down to routes over 500 miles or whatever you think they're focused on in lanes that are kind of straightforward and better parts of the country? Like what's sort of the subset of that that's sort of really appropriate?

And so I'm sorry, I don't have that number off the top of my head, but we can certainly find that out and get you a concrete answer. In the long term, I expect we will do all of it, right? I think it will take time to get there. But the strategy we have is to build out the interstate network first, which is, I want to say, the majority of the miles. But again, we'll get you a concrete number there and then move into the more capillary type movements. But the opportunity is pretty profound either way. Yes, and I'm sorry, I don't have the number off the top of my head.

So just one last quick one, excuse me, following up on just the safety component. Another stakeholder would be public, do you feel that it's still -- this is a super human driver. This is Optimus Prime, whatever you want to call it, they have to perform perfectly to be -- to get that license to operate? Or I mean, we spend a lot of time, obviously, with the statistics and to you. So we get the idea as the regulators would. But what about the -- what do you think the public is in that?

I think particularly in trucking, most people won't notice, right? There's a big truck driving down the road. It's driving well and safely, and it will just kind of -- it will be there, right? And of course, you look for it, you'll see that it's got extra stuff on it than a normal truck doesn't. But right, what you'll see is a more courteous in many cases, driver that's out in the road taking care of other folks. What we've -- what my impression is that there's been an OEM that's had a product in market that we think is not truly what we're doing, but they got a market like that. And they've had a variety of incidents, and that really hasn't dampened enthusiasm really surprised, right? So I think that -- I think it's -- as you say, we live in a world statistics. We always try to recognize that each of those statistics is a person, right? And as I think about bringing this to market, I want to make sure that if my kids are on the road, my parents are on the road, that I would feel confident that they could be around it and we certainly will.

Understood. Okay. Well, with that, we are out of time. But thank you very much, Chris, for coming and giving us an update, and we look forward to seeing what's next.

Thank you. Appreciate it. Thanks.