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

Greetings. Welcome to the Aurora Fourth Quarter and Full year 2021 Business Review Call. [Operator Instructions] Please note, this conference is being recorded. I will now turn the conference over to your host, Stacy Feit, Vice President of Investor Relations. Thank you. You may begin.

Thank you. Good afternoon, everyone, and welcome to our Fourth Quarter and Full year 2021 Business Review Call. [Operator Instructions] As a reminder, this conference call is being recorded. We announced our results earlier this afternoon. Our Shareholder Letter and a Presentation to accompany this call are available on our Investor Relations website at ir.aurora.tech, and were furnished with our Form 8-K filed today with the SEC. On the call today are Chris Urmson, Co-Founder and CEO; and Richard Tame, CFO. Chris will provide an update on the progress we have made across the key pillars of our business, and Richard will recap our 2021 financial results. We will then open the call to Q&A. A recording of this conference call will be available on our Investor Relations website at ir.aurora.tech shortly after the call has ended. I'd like to take this opportunity to remind you that during the call, we will be making forward-looking statements. This includes statements relating to the expected performance of our business, future financial results and guidance, strategy, long-term growth and overall future prospects. These statements are subject to known and unknown risks and uncertainties that could cause actual results to differ materially from those projected or implied during this call. In particular, those described in our risk factors included in our Form S-1 filed by us with the SEC on November 12, 2021, and our Form 10-K for the full year 2021 that will be filed by March 31, 2022 as well as the current uncertainty and unpredictability in our business, the markets and economy. You should not rely on our forward-looking statements as predictions of future events. All forward-looking statements that we make on this call are based on assumptions and beliefs as of the date hereof, and Aurora disclaims any obligation to update any forward-looking statements, except as required by law. Our discussion today will 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. Information regarding our non-GAAP financial results, including a reconciliation of our historical GAAP to non-GAAP results may be found in our Shareholder Letter, which was furnished with our Form 8-K filed today with the SEC, and may also be found on our Investor Relations website. With that, I will now turn the call over to Chris.

Thank you, Stacy, and welcome everyone to our first Business Review Call. Today, we are pleased to share our fourth quarter and full year results for the first time as a publicly-traded company. 2021 was an incredible and transformative year for Aurora. We acquired Uber's self-driving vehicle business, which brought an incredible team of talented and important technology to Aurora. We solidified a number of powerful partnerships and pilots. We partnered with 2 of the top 3 North American truck OEMs with PACCAR and Volvo, with the number one carrier in the US with FedEx, with Uber Freight, a transformative logistics platform, with the number one ride-hailing platform with Uber and with the number one auto OEM with Toyota. We continue to advance our cutting-edge technology. We're making progress on our sensor suite, which includes our proprietary, FirstLight Lidar, our Aurora Driver software, and with our on-road and virtual testing. We introduced Aurora Horizon, our trucking product, and Aurora Connect, our ride-hailing product, both of which are underpinned by the Aurora Driver. And of course, we took our company public. Already in 2022, we have continued to make powerful progress through our development and in our partnerships and pilots as we work towards the launch of our autonomous trucking business in late 2023 and our ride-hailing business in late 2024. Before we talk more about our progress, I'd like to take a moment to share why we founded Aurora. Together, Drew, Sterling, and I founded Aurora 5 years ago because we saw an incredible opportunity to deliver the benefits of self-driving technology safely, quickly and broadly. We believe in the meaningful impacts off-road vehicles will deliver to our society and we saw a path to bringing self-driving technology to market, focus first on safety, then speed, then breadth of impact. To deliver on our mission, we architected our company, our technology, our path to market and our partnerships to uniquely facilitate this emphasis on safety, speed and breadth. We began by developing on light vehicles. This allowed us to move more quickly, while we built out our foundational capabilities. We understood that to operate safely, we needed a multi-modal sensor suite that could see long ranges and enable more rapid reaction than the sensors that were available in the market. We designed the Aurora Driver with a common core of technology that is vehicle agnostic. This allowed us to focus on a single architecture, while also enabling Aurora's technology to integrate with multiple-vehicle types from a Class 8 truck to a passenger sedan. We expect this approach will unlock a powerful market opportunity and allow us to move quickly in both spaces. Along with our approach of technical development, we also focused early on to establish deep collaborative partnerships, which we believe will accelerate our path to commercialization. When we think about what we need to accomplish and how we're executing against our plan, we measure our progress to our technical development, the incremental validation of our safety case, our lessons learned through our pilots and progress with our partners. I'd like to start with an update on our pilots. Why they're important and how they're going? Pilots are an integral step toward building our commercial business. They're one of the ways we gain valuable insight as we continue to refine the Aurora Driver, Aurora Horizon and Aurora Connect as we prepare for commercial launch. Through our pilots, we gain a deep understanding of our customers' needs, including operational requirements. We build trust with our partners as we regularly autonomously move goods under the supervision of vehicle operators while maintaining on-time performance and striving for operational excellence. We measure progress by what we are learning and how we're improving. We look at product refinement, network integration, and the operational learnings we achieve. We use these experiences to build a safe, reliable and robust product. We are running a multi-phase commercial pilot with FedEx and PACCAR that launched during the third quarter of last year. Our pilot with FedEx uses Aurora-powered Peterbilt trucks to autonomously haul FedEx loads between Dallas and Houston, which is a 500-mile round trip. This is an industry-first collaboration between an autonomous technology developer, a logistics provider and a truck manufacturer. During the fourth quarter, we launched our second multi-phase commercial pilot with Uber Freight. We're autonomously hauling loads for Uber Freight's customers and we are also working with them to explore integrating access to their digital freight network with Aurora Horizon, our autonomous trucking product. We see this is an exciting opportunity to provide carriers with an additional set of tools to maximize the utilization of their Aurora-powered trucks, broaden opportunities to haul goods and streamline supply chain operations. During the fourth quarter, we began pulling loads for our partners 5 days a week. In January, we started nighttime hauls, demonstrating the Aurora Driver's capability of operating day and night and further increasing our opportunity to learn. Texas is an importing testing ground for us. It is the largest trucking market in the country with more goods hauled by trucks than in any other state and the freight moved on Texas highways is expected to nearly double in the next 25 years. Today, we are announcing that trucks powered by the Aurora Driver are now regularly driving between Fort Worth and El Paso in preparation for an anticipated new upcoming pilot launch. This lane represents the middle leg of one of the busiest commercial thoroughfares in the US trucking industry, Atlanta to Los Angeles. The Fort Worth to El Paso lane is over 600 miles and takes approximately 9 hours to complete. Its long-distance and monotony contribute to this lane's reputation for being undesirable for truck drivers. By deploying the Aurora Driver on such demanding hauls, we are building toward a future in which autonomous trucks can handle long routes, while human drivers can handle more convenient hauls that are conducive to more desirable lifestyles. In addition to our pilots, earlier this week, we announced a collaboration with US Xpress. Through our collaboration, we intend to leverage the intelligence of Variant, US Xpress' digitally-enabled fleet to identify where our autonomous technology can have the greatest impact in terms of both addressing unmet demand and improving operational efficiency and productivity. Together, we have also committed to exploring how autonomous technology can create a positive impact on the labor market by investing in programs that provide opportunities for new jobs. To recap, we are currently engaged with the largest LTL freight carrier in the United States, a significant freight broker, and now a [large in] technology forward truckload carrier, gaining distinct and important knowledge from each. As we continue to engage with these partners and other carriers, we have begun the first phase of the Aurora Horizon subscription reservation process. We have received initial non-binding indications of interest that exceed our anticipated launch fleet capacity through 2025, underscoring the strong underlying demand for autonomous trucking solutions. We believe our collaborative and pragmatic process will provide important insights into customer demand, inform our trucking lane expansion plan and ultimately support the optimization of our final subscription allocations. As we continue to learn, increase our operational capabilities, expand into new operational design domains and formulate our trucking launch plan, we are also making exciting progress developing Aurora-powered vehicles with our vehicle partners. In the North American freight space, we've partnered with PACCAR and Volvo Trucks, 2 of the top 3 trucking OEMs, who together make up nearly half of the market. They're both premium brands and they're just incredible partners to work with. These are again amazing companies, and we've made tremendous strides with both. Through our work with them, we have continued to invest in our autonomous fleet, which now comprises 18 trucks, more than double the size of our operational fleet we had in the fourth quarter. Expanding our fleet allows us to test and develop our technical and operational capabilities at large scales, while we continue to form the vast majority of our testing using virtual development tools. We're also thrilled to work with Toyota on their vision for Mobility for All, bringing self-driving vehicles to market. Through our long-term collaboration, we are laying the groundwork for the mass production launch and support of autonomous passenger vehicles and ride-hailing networks, including Ubers. We're able to work towards the launch of both our trucking and ride-hailing businesses with our partners because of our common core technology in the Aurora Driver. When I say common core, I mean that the Aurora Driver's hardware, software, infrastructure and development tools are designed to work across vehicle types. This commonality ensures that lessons learned, development efforts, hardware improvements and cost reductions made to the Aurora Driver benefit every vehicle it powers. As we make progress on the trucking front, we're also forging ahead with both our development and commercial milestones on Aurora Connect, our planned product to serve the ride-hailing space. At the end of the first quarter, we expect to launch the Aurora Driver Beta 2.0, not only on the next generation of Aurora Driver-powered trucks, but also on the next generation of our passenger vehicle fleet, the Aurora Driver-powered Toyota Siennas that will ultimately support the launch of our product for ride-hailing. Achievement of this important milestone will demonstrate one of our key competitive advantages, the transferability of our common core technology. Notably, the Aurora Driver Beta 2.0-powered truck and Sienna fleets are targeted to launch with the same expense and maturity. Aurora's product roadmap is designed to deliver the Aurora Driver at a large commercial scale. With each quarterly product release, Aurora plans to deliver an Aurora Driver that's incrementally advancing on 2 axes: expanse and maturity. Expanse represents the breadth of capabilities and domains in which the Aurora Driver can operate. For example, training the Aurora Driver to handle different types of construction on highways, dense suburban settings, and ultimately, dense urban environments are increasing the driver's expanse. Maturity represents the degree to which capabilities are ready for commercial deployment and will evolve from development to validation to completion. Maturity will be exemplified through longer durations of commercially-representative autonomous operation with a safety case that supports it. A critical milestone we expect to achieve in the third quarter of this year is the Aurora Driver's ability to safely pull over to the shoulder without the help of a vehicle operator. When we look at the key capabilities necessary to safely operate a vehicle at highway speeds, it is critical that the vehicle can safely handle possible system failures. We believe this milestone will demonstrate the maturity of the Aurora Driver and our progress towards satisfying the fail-safe core claim of our safety case framework. In the slide deck, we've included a video of our initial development and virtual testing of this capability. You'll see that we are able to bring the vehicle safely to the shoulder of the road and stop. We're now expanding our development and testing to mature the capability to be able to handle more complex scenarios like avoiding other vehicles stuck on the shoulder. We are working to ensure the Aurora Driver's capabilities address varied road environments like highway, suburban and urban roads, changing lighting conditions like day, night, dawn and dusk, a myriad of weather conditions, including clear skies, fog, rain and snow and different types of construction and responding to system failures. A great example that we include in our slide deck is a video of the Aurora Driver operating autonomously in a monsoon rain storm in Texas. This is an incredibly challenging scene. There's a lot of rain in the camera lens. There's water being thrown up on the road and there's heavy rain in the air. Despite all of this, the Aurora Driver is able to reliably see other vehicles on the road and drive safely even as people are pulling off to stop. As you can imagine, making our technology work reliably is incredibly challenging, not only as Aurora's CEO but also as a robotics engineer, it's incredibly exciting to see the strides we're making with our technical progress. As I mentioned, when we started Aurora, we believe that one of the most critical steps for building technology to transform the industry like transportation was to invest in the foundational elements to both accelerate development and enable scalability. One example of where we've done this is with our proprietary, FirstLight Lidar, with Frequency Modulated Continuous Wave technology known as FMCW. This novel measurement technology allows us to see further than conventional pulsed lidar technologies and instantly measure the speed of things moving through the world. Our FirstLight Lidar has unlocked our ability to deploy our products at highway speeds, where even fractions of a second can affect the vehicle's ability to react to unexpected obstacles. When it comes to sensors, there has been some debate in the industry around the right approach. At Aurora, we feel strongly there is only one right approach, a multi-modal sensor system that enables our vehicles to get the richest set of data to perceive the world around them. With that, our world-class hardware team has designed and built a multi-modal sensing suite comprised of FirstLight Lidar and other Lidars as well as powerful cameras, radars that optimizes range, field of view, data quality and data processing efficiency. In the spirit of continuous improvement, we are always considering ways to further enhance our hardware stack. We have hypothesized that our FMCW technology may also deliver advantages in midrange distances. We have recently tested these hypothesis and observed that FirstLight demonstrates superior probability of detection of dark targets relative to commercially available mid-range lidar. With FMCW technology, we also get the additional benefit of simultaneously returning velocity measurements at these midrange distances that traditional lidar can't provide. This analysis is ongoing and will inform the architecture of the lidar component of our future hardware generations. In terms of testing our hardware and software stack, there is a limit to how much meaningful data can be efficiently gathered from on-road driving. So we've also invested heavily in the development of proprietary high-accuracy and scalable virtual testing technology. In our Virtual Testing Suite, we've built proprietary technology that allows us to accurately model how energy and light moves through the world. This allows us to build physically accurate sensor simulations for cameras, conventional lidar and of course, for our proprietary, FirstLight Lidar. We've also built a procedural generation framework that allows us to create a basic scene and then create complex variations on that scene. This becomes particularly important when we are testing scenarios that are too rare or dangerous to test in the physical world. We continue to accelerate our development and validation through simulation in virtual environments, testing at scale and costs we couldn't achieve in the physical world. We believe our virtual testing is the quickest and safest way to train, test and ready our self-driving technology for deployment at commercial scale. In the slide deck, we included a video of a new capability, nudging for debris in the roadway that we have been developing and testing in simulation. While vehicles can run over small obstacles in the road, for obvious reasons, they should avoid larger pieces of debris. This video shows the Aurora Driver was capable to detect an object at a far enough distance to gracefully move out of lane to avoid it. We're incredibly proud of the technical progress we continue to make and look forward to sharing more on this front. Another important part of bringing this technology to market is educating our various stakeholders on Aurora's approach to development and deployment as well as the far-reaching benefits self-driving technology can provide. Earlier this month, Aurora had the honor to be the only autonomous vehicle company asked by Congress to present at the House Committee on Transportation and Infrastructure hearing on automated vehicles. This was an incredible opportunity to show our leadership on safety and a great example of how we work collaboratively with legislators at the federal, state and local levels as well as with regulators, safety advocacy groups and other self-driving companies to bring safe and innovative technology solutions to market. We're committed to being transparent with our work and our progress to help bring the government and public along on the self-driving journey. As we wrap up our first business update as a public company, I'd like to reflect for a moment. Building a great company and technology to transform transportation means not only making tremendous progress technologically, but also demonstrating the societal and economic benefits self-driving technology can provide. On that note, I want to reference the principles for transportation innovation released recently by the Department of Transportation and Secretary, Pete Buttigieg. As he mentioned, innovative technology is not an end to itself, but is meant to serve the public and society. We couldn't agree more. We're building the Aurora Driver to save lives, increase access to transportation and support and improve supply chains. As a public company, we are proud to have our investors on this journey as we work towards building and delivering a commercial self-driving product at scale. I will now pass it over to our CFO, Richard Tame, who will share more about our financial results.

Thank you, Chris. 2021 was a capstone year financially for Aurora. As Chris mentioned, to kick off the year, we acquired ATG, Uber's self-driving vehicle business, which further bolstered our technology stack and gave us the depth and breadth of talent to better tackle the complex challenge of developing a scalable self-driving product. As part of this transaction, we received $400 million to capitalize the combined company. This transaction also had a one-time impact on our reported stock-based compensation that I will review in a moment. During 2021, we had $83 million in collaboration revenue for development work on the Sienna platform associated with our agreement with Toyota. This work will ultimately support the launch of our product for the ride-hailing space. As a reminder, this collaboration revenue is recognized using the input measure of hours expanded as a percentage of total estimated hours to complete the project. Operating expenses, including stock-based compensation, totaled $813 million. Excluding both ongoing and acquisition-related stock-based compensation of $92 million and $128 million, respectively, operating expenses totaled $593 million. Within operating expenses, R&D expenses, excluding $206 million in stock-based compensation, totaled $491 million, primarily comprised of personnel costs as we continue to invest in our industry-leading autonomy work. SG&A expenses, excluding $13 million in stock-based compensation, were $102 million, which included costs associated with building our public company infrastructure. There were 2 items I want to flag on stock-based compensation, which are reflected in the reconciliations that we provided in the non-GAAP financials in the Shareholder Letter that we have posted to our Investor Relations website and furnished with our Form 8-K filed today with the SEC. First, on a full year basis, $128 million of the $220 million stock-based compensation presented is one-time in nature, related to our acquisition of ATG in January 2021. Prior to the acquisition, employees of ATG received RSU grants in Uber, the former parent company of ATG, who, based on their investment in Aurora is considered a related party following the closing of the transaction. While these are not Aurora awards, given the related-party characterization, they are accounted for as a stock-based compensation on our books. Second, when looking at the fourth quarter of 2021 specifically, in addition to $30 million for related-party awards, you will also see a $47 million catch-up in stock-based compensation expense for Aurora awards that had previously satisfied time-based vesting requirements, but also had a liquidity event performance condition for full vesting. This performance condition was met with the closing of the business combination that took Aurora public in November. Another accounting nuance related to our business combination to be aware of is the treatment of the public and private placement warrants as well as the sponsor earnout shares. These financial instruments are classified as derivative liabilities measured at fair value, with changes in fair value reported each period. We recorded $20 million in non-cash expense during the fourth quarter to reflect the associated change in fair value. Turning to cash flow. In 2021, we had $563 million in operating cash spend and $48 million in capital expenditures. The latter primarily reflecting investments in our facilities, IT and other infrastructure. We closed our business combination, taken Aurora public in November with $1.8 billion in gross proceeds and cash on hand, which we expect to fund Aurora through the commercial launch of our trucking product, Aurora Horizon, and into 2024. We ended 2021 with a very strong balance sheet, including $1.6 billion in cash, enabling us to continue to execute our mission of delivering the benefits of self-driving technology safely, quickly, and broadly. Going forward, our intention is to provide a fulsome financial review like this one on an annual basis on our fourth quarter and full year business review calls as we encourage the investment community to consider our results on an annual basis, consistent with the long-term approach we take to executing on our mission and working towards delivering a commercial product at scale. With that, we'll now open the call to Q&A.

[Operator Instructions] Our first question comes from the line of Steven Fox with Fox Advisors.

Congratulations on all you accomplished this year. Chris, I was wondering if you can maybe talk a little bit about digitally-enabled fleets. You had the announcement this past week. And at the Uber meeting last week, Uber Freight highlighted some of the hopes for autonomy. How do those opportunities do you think differ in time-wise and revenue-wise for Aurora versus, say, traditional methods of trucking? And then I had a follow-up.

No. Thank you, Steve. Appreciate the question. So as we think about this, what we're trying to do over the next -- over this pilot period and exploration period with our partners is understand the unique needs of each of the different segments of the freight space. And so we're really excited to have the partnerships with U.S. Xpress, with Uber Freight, with FedEx, and we look forward to learning from them as we move forward. We're excited about the technologies that U.S. Xpress is bringing to bear through Variant. And we look forward to offering their customers and other customers the benefit of the understanding we get through that.

Great. I appreciate that. And then in terms of just some maybe checks we should be looking for in terms of progress this year. You did mention demoing the fail-safe on roads later, I guess, sometime this year. Can you just sort of talk about what you have to do to get from where you currently are with doing that in the cloud and online to doing it on the open road?

Yes. We -- so we're excited about the things we have in front of us this year. So at the end of Q1, we anticipate showcasing the transferability. This is one of the core kind of architectural choices we made with the Aurora Driver, enabling it to operate both trucks and light vehicles. In Q3, we anticipate showcasing one of these very important elements of bringing the vehicle to a safe stop. We've shared our Safety Case Framework and one element of that is fail safety. And part of that is being able to understand when to trigger and then actually having the capability on the vehicle to bring it to a stop. And this is a significant milestone in the development of it. Today, we're running this in simulation. We're doing early work on vehicles in closed test courses. And between now and the time where we're going to showcase this on public roads, we'll be -- continue to advance that capability, enhancing the fault detection capability we have on board, the vehicle to handle a more large, broad set of problems that might occur at some point in the future so that we can operate safely.

Our next question comes from the line of Tom White with D.A. Davidson.

Congrats on your first public company earnings call. I guess we'll have to wait 12 months for the next one of these, it sounds like, though. 2, if I may. Chris, you talked about some of the different kind of milestones and kind of markers of progress here over the next several months. I guess I'd be curious to hear your updated thoughts on the usefulness of metrics for investors, things like real-world miles driven or maybe miles driven without a human intervention. Do you think those are useful things for investors to try and track? Or do you guys anticipate maybe at any point in the future, kind of publishing sort of a regular metric? And then could you just double click on the comments around the non-binding indications of interest? I'm just curious, has the pace of that really picked up in recent months? Or has it been kind of healthy all along? I'm curious whether it might cause you guys to kind of increase your pace of investment or the urgency of investment to make sure you don't -- that you take advantage of all that interest?

Sure. And maybe I can just speak. So on the timing of calls, we do anticipate providing more regular, roughly quarterly updates on the business. I think the depth of the financial disclosures is where we intend to do that on an annual basis. So I have no fear. We look forward to engaging more often than once a year. We think we're making exciting progress, and we look forward to the opportunity to share that with anyone who joins us. On the metrics front, so there's a lot to unpack there. As we think about metrics, we continue to think that the disengagement report work that we put out there is, as many people, not just us believe it's just not a super useful indication. As we've shared repeatedly, we think there's an immense amount of value in simulation. We leverage that dramatically, and we do orders and orders of magnitude, more testing in simulation than we would do on physical road because it allows us to do much more focused testing. It allows us to do more efficient testing. And it allows us to test things that we just wouldn't encounter by just operating trucks out on the public roads. And so that can give us more confidence in the ultimate safety of our vehicle. We do -- we have been trying to -- we've heard the investment community around the need to provide progress and metrics of progress, and we're continuing to try to find a good way to express that. We think there's something around the way we talk about the maturity of the product and look for us to try and share more of that in the near future. So stay tuned, I guess, on that one. On the non-binding intent, so we've tried to take a very collaborative approach to engaging with partners and customers in this space. And instead of going out to them and kind of trying to beat them to kind of shake them -- sorry, to shake loose their checkbooks, it's really about what can we do together to learn so that we can deliver a product that's of an immense value to them and ultimately to their customers. And so the approach we've taken is a 3-phase process, where we began engaging with partners and potential customers around indications of interest, and we're in that phase of the process right now. We expect later to shift to a conversation around initial allocations, and we'll do that collaboratively with our partners. And that will start to be more firm. And then, ultimately, we'll get to final allocations. What we've heard from our partners is they really appreciate the transparency of the engagement we've had with them around this. And so we're very excited for that. And as I indicated, we have, at this point, initial indications of interest exceeding our ability to deliver through '25. So we're really proud of that. And I think it speaks a lot to the work our team is doing in both developing the technology, but also building the relationships with these key customers in different segments of the freight space. Does it mean that we will want to accelerate investment to kind of keep pace with the demand? We're still assessing that. And of course, we'll share more as we move forward.

[Operator Instructions] Our next question comes from the line of David Vernon with Bernstein.

Chris, understand or get your updated thoughts on your expectations for when in a particular sort of use case like, for example, FedEx, where you're running between El Paso and Dallas. When you might be able to be taking some of the support capabilities of the safety driver and safety engineer, that kind of stuff out of the truck? I'm just trying to think from a big picture standpoint, how is your thinking on the timing of when you may be able to validate the technology without a driver in the cab changing right now?

Yes. So we continue to work towards a commercial launch of the trucking product at the end of '23, and we're excited about the progress towards that. Our philosophy on this really is one of continuous advancement. And that at any given moment, there's kind of a duration of which the vehicle could operate without oversight. But until we get to the point where it's continuously viable and doable in a commercially relevant way, it feels like it just doesn't make a lot of sense, right? If we have to operate our trucks with 3, 5 other vehicles surrounding them to be able to do it safely, it's just kind of distracting from the core objective of addressing these increasingly rare events that hold up operation until we're ready to do that and meet that acceptably safer. It feels like the most practical, efficient and safe way to do the development and testing is to have an operator in the vehicle instead of kind of 5 of them strewing about the highway around the vehicle.

Okay. So it sounds like the 2023 is still what you're looking for there. And then maybe as a follow-up. Richard, could you help us understand kind of the clean cash burn rate? I'm not sure if the stock-based comp charges that you had for the Uber stuff would have affected kind of the underlying cash burn rate in 4Q. And can you give us a sense for where you intend to be exiting fiscal '22 at? I'm just trying to get a sense for what we should be expecting in terms of cash burn for 2022.

Sure. So I think, to the second point first. We don't think as a pre-revenue company that it's meaningful to provide the forward-looking guidance. So we won't speak to that. The cash flow is relatively clean from the Uber RSU perspective, obviously shows up in the non-GAAP reconciliation for stock-based compensation there. But I think we had $563 million of operating cash spend and $48 million on CapEx. That's basically a clean cash flow for full year 2021. And then in quarter 4, I think that's what we had. And we can follow up.

I appreciate the statement that you don't want to give too much guidance right here. But I guess, if you were to think about just directionally, are there investments in R&D and a step-up in OpEx that we should be expecting that is going to make that number get worse? Is it going to get better? Just anything directionally you can give us on that front would be helpful.

Yes. I think the directional thing is that we have $1.6 billion of cash on the balance sheet at the end of the year. And as we work towards launching the product at the end of 2023 trucking, we feel like we've got enough cash. We believe we can get through the commercialization and then into 2024.

Thank you. Ladies and gentlemen, we have reached the end of the question-and-answer session. This concludes today's conference, and you may disconnect your lines at this time. Thank you for your participation. And have a wonderful day.