Logo of jester cap with thought bubble.

Image source: The Motley Fool.

TuSimple Holdings Inc (TSP)
Q3 2021 Earnings Call
Nov 3, 2021, 5:00 p.m. ET

Contents:

  • Prepared Remarks
  • Questions and Answers
  • Call Participants

Prepared Remarks:

Operator

Good day, and welcome to the TuSimple Third Quarter 2021 Earnings Conference Call. [Operator Instructions] Keep in mind that this call is being recorded, and there will be a replay available at ir.tusimple.com following this call.

I would now like to turn the conference over to James Mann, Head of Investor Relations for TuSimple. Mr. Mann, please go ahead.

10 stocks we like better than TuSimple Holdings Inc.
When our award-winning analyst team has a stock tip, it can pay to listen. After all, the newsletter they have run for over a decade, Motley Fool Stock Advisor, has tripled the market.* 

They just revealed what they believe are the ten best stocks for investors to buy right now... and TuSimple Holdings Inc. wasn't one of them! That's right -- they think these 10 stocks are even better buys.

See the 10 stocks

*Stock Advisor returns as of October 20, 2021

James Mann -- Head of Investor Relations

Thank you, Brandon. Good afternoon, everyone, and welcome to our third quarter 2021 earnings call. With us today are TuSimple's President and Chief Executive Officer, Cheng Lu; and Chief Financial Officer, Pat Dillon. Cheng and Pat will review the operating and financial highlights and then we will take questions. Before I turn the call over to Cheng, let me cover a few housekeeping items. As a reminder, TuSimple's shareholder letter and a replay of this call will be available later today on the Investor Relations page of our website. This call is being recorded. If you object in any way, please disconnect now. Note that TuSimple's shareholder letter, press release and this call all contain forward-looking statements that are subject to risks and uncertainties. These forward-looking statements are only predictions and may differ materially from actual future events or results due to a variety of factors. Please refer to the risk factors detailed in our SEC filings. We will also discuss non-GAAP financial measures, which are not prepared in accordance with generally accepted accounting principles. Please refer to the safe harbor disclaimer and non-GAAP financial measures presented in our shareholder letter for more details, including a reconciliation of the non-GAAP measures to comparable GAAP measures.

I will now turn the call over to Cheng to begin. Cheng?

Cheng Lu -- President and Chief Executive Officer

Thank you, James. Hello, everyone, and welcome to our third quarter earnings call. Our mission at TuSimple is to bring the most safe, most reliable and low-cost autonomous freight capacity to the market. TuSimple is making meaningful progress developing a world-leading Level four autonomous technology, scaling our autonomous freight network and investing in our people and operations. I will share a few highlights, and I'll turn the call over to Pat to discuss our financial results. I'd like to start today's call by sharing some very exciting updates on our long-standing partnership with UPS. Over the last 2.5 years, we have accumulated over 160,000 miles of paid autonomous freight haulage for UPS's North American Air Freight division or NAF. You've likely heard a lot recently that autonomous trucking companies are trying to find ways to scale the commercial deployment of their technologies. It is not easy, and that is why we're so proud of what we have accomplished with a world-class partner in UPS. We believe UPS and TuSimple are running the world's most significant autonomous freight operations. To our knowledge, whether you measure by the years we have been working together, the cumulative autonomous mileage or the geographic reach of our UPS autonomous routes, we believe our partnership represents a significant lead in the race to commercialization. Secondly, we continue to develop new autonomous lanes for the UPS NAF, including, most recently, expanding our HD maps from Dallas to UPS terminals in Orlando and Charlotte. This represents a nearly coast-to-coast autonomous network for UPS NAF and a fantastic opportunity to expand our relationship with UPS. Lastly, we set another industry benchmark by achieving significant fuel savings during our UPS autonomous operations.

Over the last 2.5 years, for UPS autonomous miles driven, in the industry standard 55 to 68 miles per hour operating band, we were able to achieve over 13% fuel savings relative to human operators. This builds on our 2019 university study that found a 10% fuel savings from our Level four technology and provides further support for its significant environmental benefits. We believe the UPS operations represent, by far, the largest third-party validation of the significant fuel savings from autonomous technology. Fuel represents almost 25% of trucking's variable cost structure, and over the last year, the price of diesel fuel has increased over 75%. This has a significant impact on the cost of freight, including increasing fuel surcharge borne by shippers. Our technology's fuel-saving benefits can translate into billions of dollars of potential cost savings for the freight ecosystem. Our UPS autonomous operations represent a multiyear deployment of our technology in real-world conditions and provides us with a significant competitive advantage in our go-to-market and technology developments. Moving to the quarter. Overall, the third quarter represented a period of continued progress and investment in all areas of our business. I want to take a moment to recognize Michelle Sterling who we appointed to our Board in October. Michelle adds amazing perspective from her time leading Qualcomm's global human resources practice that oversaw 33,000 global employees.

Her background at a high-growth technology company like Qualcomm makes Michelle a perfect addition to TuSimple's Board as we continue to scale our operations. In terms of our people, we continue to invest in building our team of exceptional, pinnacle and operational experts. Two of our recent key hires include Dr. Jing Zhu, Senior Vice President of Software Engineering; and Dr. Ersin Yumer, our new Head of Machine Learning, who oversees the testing of our virtual driving algorithms. We added 138 new full-time R&D employees during the quarter to bring the R&D team's headcount to approximately 1,000 full-time employees. Overall, we ended the quarter with approximately 1,300 employees globally. This quarter, we continued to make progress with our global OEM partner, TRATON. In the U.S., we continued to make steady progress with Navistar on our autonomous semi-truck production program and regular joint dialogues with our fleet customers. We believe the now-completed merger with TRATON has made Navistar an even stronger partner for our AFN ecosystem. Internationally, we're working closely with TRATON and its Scania brand on near-term development initiatives, including on-road testing as well as refining our long-term go-to-market strategy. This month, TuSimple and Scania will be co-hosting an exhibit at the China International Import Expo in Shanghai, which is one of the major technology trade show events. We continue to accumulate real-world road miles with our global fleet to operate across the southern U.S. as well as our testing programs in Sweden and China.

Our fleet drove 840,000 miles in the quarter, bringing our cumulative road miles to 5.4 million. Now to provide an update on our Driver-Out pilot program. Our team is making significant progress toward removing the driver for runs over the 80-mile route between the Phoenix and Tucson areas. We expect to perform the initial Driver-Out runs before year-end and to complete the pilot program over the coming months. As a reminder, the Driver-Out pilot will consist of multiple runs performed over multiple weeks and is a major part of our ongoing technology development across many dimensions, including software, hardware and go-to-market. What makes the Driver-Out pilot program so challenging is that we're solving for both known and unknown factors that we might encounter on public roads. This includes noncompliant motorists, unplanned road construction and changing driving conditions, all of which must be continuously monitored and accounted for in real-time. Safe operations require autonomous driving technology beyond anything previously undertaken. Let me provide you with some insights into our progress thus far. We started testing along [Indecipherable] route eight months ago and so far have performed over 1,800 trips and accumulated over 150,000 autonomous road miles. We continue to see improvement in the system's hardware and software performance. I am pleased to share that the hardware quality challenges we discussed last quarter are now resolved. We expect to freeze our technology development in the coming weeks so the system can move to the validation phase.

I recently participated in one of our tests, and I can tell you that it was amazing to see our next-generation trucks embark on the terminal, terminal run. My trip, like the vast majority of our test runs, including 0 disengagements and demonstrated just how far our technology development has come. Safety, of course, comes before all else at TuSimple, and this is particularly important for the Driver-Out pilot. The Driver-Out safety case validation process has a multifaceted risk mitigation approach. It includes a significant focus on systems engineering and functional safety, which gives us the confidence to safely remove the driver on public roads. We have accumulated significant learnings from our Driver-Out safety process. These learnings will be critical to scale our technology for broader commercial deployments. In today's letter, we outlined two primary areas of our Driver-Out safety case validation. First is system safety, which, at a high level, validates that the Driver-Out trucks are safe to operate autonomously by helping each aspect of the system to be reliable, failsafe, sufficient and proven. Second is operation safety, which supports each aspect of our Driver-Out operations to be prepared and proven by creating safe processes and procedures. Operations safety validates that we have monitored and triaged every driving event that we can, assess the event's level of safety risk and assign it for resolution by engineering teams.

Our safety teams are working closely with all departments and across all levels of the organization to validate our Driver-Out safety case, consistent with our overall company safety culture. Across our organization, our primary focus is on safety, and this is embedded in our hiring, onboarding and day-to-day operations. Ultimately, responsibility of safety belongs to our Safety Committee, which is a cross-functional senior leadership group. All members of the committee, myself included, are fully focused on ensuring the safety of our Driver-Out program in all of our operations at TuSimple. Safety is one of the most significant value propositions of our autonomous technology. This quarter, we announced the initial results of our telematics study with Geotab. The primary study results indicate that our Level four technology significantly reduces harsh driving events relative to human driver. This third-party study result is significant as harsh driving events are well-established precursors to accidents. We believe we're the only autonomous driving company with empirical data to quantify the safety driving behaviors of our autonomous driving system. Having objective measurements of our autonomous driving system's performance is important, not only for our own internal development but also for our partners, including insurance providers and our shipper and carrier customers. We look forward to continue our work with Geotab and other stakeholders to accelerate our technology development. Now moving to commercial development. We remain focused on our goal of growing our current and future AFN capacity and, over time, increasing the penetration of autonomy on our network. Our high-definition map network has now expanded across the southern half of the United States.

We added 1,400 miles this quarter to bring our total unique miles to almost 10,000 miles. Further, the pace of mapping is also increasing, which allowed us to expand our AFN to UPS terminals on the East Coast ahead of schedule. This is an excellent example of our technology development translating into immediate commercial progress. We increased our revenue miles quarter-over-quarter despite a relatively flat truck fleet. We deepened our customer relationships, including adding significant logistics player as a reservation holder. And our customers continued to show tremendous enthusiasm to help us develop our technology in real-world commercial conditions. And before I hand the call over to Pat, I do want to provide a brief update on our CFIUS process. We continue to be in productive dialogue with CFIUS and we believe are moving closer to resolution. But we have now reached a definitive outcome of the process at this time. We remain committed to providing updates on the process, and we will share more at the appropriate time.

And so with that, I will now hand the call over to Pat to discuss our financial results and guidance.

Pat Dillon -- Chief Financial Officer

Thank you, Cheng. Beginning with our reservation program, we added 100 new truck reservations from a significant global logistics operator during the quarter. This brings our total reservations to 6,875. I mentioned our ongoing progress with this particular customer during our last call, and our team worked diligently to finalize the reservation in September. We plan to share more on this relationship in the coming months once we start hauling freight with them on routes in the southwest. Our team continues to engage in reservation discussions with a broad range of shippers and carriers, and we expect to continue to add new reservations in the coming quarters. We continue to focus on the quality of our reservations, working with partners who are committed to deploying resources alongside us to establish autonomous lanes in the future. We continue to be highly encouraged by the breadth of potential customers that are engaged with us on incorporating our technology into their networks. And we believe this bodes well for the rapid adoption of our AFN over the years to come. Now shifting gears to our financial results for the third quarter.

We reported $1.8 million of revenue in the quarter, which increased over threefold versus the same period last year and represents a 20% increase quarter-over-quarter. While we started to receive new semi trucks recently, our revenue fleet size was relatively flat during the quarter, and our revenue growth was primarily driven by continued improvement in asset utilization. Moving to expenses. We spent $85 million on total R&D in the quarter, including $22 million of SBC. This compares to $60 million in the same period last year. We spent $29 million on G&A during the period, including $10 million of SBC. This compares to $15 million in the same period last year. Our adjusted EBITDA was negative $81 million, which compares to negative $67 million in the same period last year. The larger EBITDA loss reflects our continued investment in technology and commercial development, including discrete items that we do not expect to repeat in the fourth quarter. We invested $6 million in capital expenditures during the quarter, driven by equipment purchases and facility costs, and we ended the quarter with a cash balance of $1.4 billion. Now to provide an update to our 2021 annual guidance. Our financial performance for the year-to-date and our forecast for the remainder of 2021 are largely trending as expected, although we do have a few refinements to share. We are refining our R&D ex SBC guidance to $210 million to $220 million, which represents the upper half of the prior guidance range. Similarly, we are refining our guidance for G&A ex SBC to $55 million to $60 million, which is also the upper half of the prior range. We are adjusting our stock-based compensation expectations to $120 million to $130 million, and there is no change to our other previously communicated guidance.

I'll now hand it back to Cheng for a few last remarks.

Cheng Lu -- President and Chief Executive Officer

Thank you, Pat. At TuSimple, we're striving toward solving one of the most complex engineering challenges that has ever been undertaken. This past quarter, all of our employees stepped up to propel us toward completing our mission. I'm incredibly proud of our team and the progress we've made toward bringing Level four technology to the market at scale. As we head into the end of the year, we'll not let up. We'll continue to lead the industry with our technology, our go-to-market strategy and our talent.

And with that, we're ready to start the Q&A session.

Questions and Answers:

Operator

[Operator Instructions] And from Morgan Stanley, we have Ravi Shanker. Please go ahead.

Ravi Shanker -- Morgan Stanley -- Analyst

Thanks, good afternoon. Maybe I can start out on the Driver-Out test. You reiterated your time line to get this done or start the test at least before the end of the year. There are only about seven weeks left in the year, so is it fair to say that you have really strong line of sight and confidence on the timing of this test?

Cheng Lu -- President and Chief Executive Officer

Good to have you. Yes, we're aiming to hit our target date. As mentioned, we have moved on -- we will be moving on to the next phase and last phase of our Driver-Out pilot program, which is a very positive development. And as we begin our validation process, we feel confident. But look, I can't guarantee that we won't discover some new issues during that phase of development. And really, that's what having a structured development process is meant to do. And so in the day, our #1 priority is to safely complete our safety case validation process.

And end of the day, this is not meant to be a demo or PR stunt that once we are done, it's over and onward. Everything we're doing directly with the Driver-Out pilot program translates to full-scale commercial deployment. So to us, if we make more improvements prior to the pilot program with Driver-Out initial runs, what we're afterwards, I mean, that all generates positive ROI for us. But again, to reiterate, that is our goal. We do believe we can hit that target at this point.

Ravi Shanker -- Morgan Stanley -- Analyst

Great. And maybe for a follow-up. I think on the last call, you mentioned that you were having some difficulty getting new trucks and getting new parts. Obviously, the world's a crazy place right now. Can you just give us an update on that? Kind of how are you looking from a technology perspective, even from a labor perspective? Obviously, you don't have the same labor issues as many other companies out there, but just I'm sure even hiring engineers is probably challenging in this environment. So if you could just give us an update there, that would be great.

Cheng Lu -- President and Chief Executive Officer

Great question. Look, from I think what we've seen and across the industry, I think the world has continued to be a crazy place. I think the supply chain issues have gotten worse rather than better over last quarter. So we do -- of course, we are having some challenges receiving new trucks, receiving new parts. We have suppliers have labor issues. But I think we can use that, and we're not going to use that as an excuse. And so we'll continue to work hard to overcome those near-term challenges.

Ravi Shanker -- Morgan Stanley -- Analyst

Very good...

Cheng Lu -- President and Chief Executive Officer

On the labor part, as you know, I think the new term that the labor market is calling after the pandemic is the Great Resignation, right? So not only are -- I think the talent pool is tighter than ever, and I think that explains for some of the additional increase in cost that Pat has alluded to. But again, those are things that as we run the company, we just have to overcome.

Ravi Shanker -- Morgan Stanley -- Analyst

There are very few jobs out there that are cool as working on autonomous trucks, so maybe you'll have an easier time. Thanks so much guys.

Operator

From JPMorgan, we have Brian Ossenbeck. Please go ahead.

Brian Ossenbeck -- JPMorgan -- Analyst

Hi guys. Afternoon. Thanks for taking the question. So Cheng, I just want to come back to supply chain, maybe looking a little further down the road. Can you just give us some sense, you mentioned moving forward with Navistar? Just a broader update there would be helpful in terms of time lines and milestones. And then similarly, in terms of the L4 Tier one supply chain, how that's progressing in terms of getting the redundancy with the quality and scale that you need to move past the Driver-Out tests and actually put more of these vehicles on the road.

Cheng Lu -- President and Chief Executive Officer

Sure, Brian. Great question. So as it relates to truck deliveries and I mean, Navistar is really collaborative. We're a priority partner for them. I think the reality is that we're starting to receive the trucks, and we believe we'll get what we ordered beginning of this year, hopefully by -- all of it by the end of this year. So in a way, these truck purchases are almost becoming sort of trucks for next year, and that's OK. I think that actually fits our plan well. In terms of Tier one supply chain, I think there's really -- there's two folds for that. One is the near-term, call it, disruption to supply chain, labor shortage and just everyone being generally very busy. And those are issues that we continue to overcome every day. And then in the longer term, as I alluded to before, one of the biggest challenges we see to the scale deployment of autonomous technology is the supply chain maturity.

And that really revolves around key Tier one components like the compute, autonomous domain control, ADC, or redundant actuation, steering and braking. And so there's a little chicken and egg that happens in this because Tier 1s don't want to commit to investments without orders. And so of course, there's a natural chicken and egg, that kind of problem that persists. And that's something that we have identified as one of the risk, and so we are taking steps to address that. We've always worked very closely with our Tier one partners, some of which are our investors from a long time ago. And so over the next coming, I think, quarters, you'll hear more announcements of us in terms of investing more heavily in the supply chain to ensure that we can meet the time line that we talked about.

Ken Hoexter -- Bank of America -- Analyst

Okay, got it. Just to clarify, I probably didn't ask it clearly enough, but on the Navistar style, I was more interested in the longer-term development of the L4 truck at scale, not necessarily the near-term orders for the ones you're retrofitting.

Cheng Lu -- President and Chief Executive Officer

Got it, got it. It's -- we're -- we have extensive collaboration. We'll continue to make steady progress. Again, I don't think we can expect -- you should expect to have a significant update every quarter. Overall, the TRATON acquisition of Navistar certainly strengthened Navistar's balance sheet and access to a global platform. And so this truly actually, I think, provides us with a global platform partner. And we're working through, together with TRATON and Navistar, other brands within TRATON on how to better realize those synergies globally. And so I feel good about the development. And again, just in terms of overall risk, I'll point to the supply chain as really the biggest risk that we see at this point.

Brian Ossenbeck -- JPMorgan -- Analyst

Okay, got it. Just one more quick follow-up, if I could. Just on the safety case, which is really helpful here in terms of the Driver-Out pilot. So it seems like this is going to be sort of the guidepost for going forward into the next phase. You got several different categories here under the two broader headings. Maybe if you could just give some broad color about which of these areas you think potentially has the most challenge ahead, just so we can kind of get a sense as to where the biggest lift is as you move into the final phase of this Driver-Out pilot.

Cheng Lu -- President and Chief Executive Officer

Sure. I wouldn't categorize a particular part of the safety case as having the biggest challenge. It's -- I think moreover, it's more about having a very structured, systematic approach to identifying and mitigating risk. At the end of the day, the technology we're developing inherently does have risk, and it's our job to mitigate and to show that we are responsible stewards of safety. And so it's a lot of collaboration and a lot of work with our teams. And there is a validation process as the last part of our Driver-Out pilot program. We believe all the pieces of the puzzle are in place, and all of us are excited to kind of go into that phase and hopefully have a very short validation process and so we can be on the road without driver shortly. Okay, thanks.

Operator

From Bank of America, we have Ken Hoexter. Please go ahead.

Ken Hoexter -- Bank of America -- Analyst

Hey, good afternoon. So TRATON, just I want to clarify, they've got -- or Navistar has new management, I guess, here in the U.S. Just I want to understand if same vision, same time frame as the old management, if that's been kind of clarified at all, just checking in on that. And then looks like the U.S. trucks went up to 58 from 50, if just taking your total truck and assuming they're all in the U.S. So still not the full 25 you targeted. Are you all set with Ghost Rider or is that fully loaded in terms of your -- the trucks you wanted? And then lastly for me, just you mentioned it's not start-stop. So once the Driver-Out test is done, do you keep running it or is it a fixed period pilot?

Cheng Lu -- President and Chief Executive Officer

I mean, with Navistar, as I mentioned with Brian's question, we believe that the acquisition from -- the merger completion with TRATON, the change in management team, which can actually happen as part of that all positive developments. There's a bigger global platform that we're leveraging. I mean, we already have existing relationship with TRATON, with Scania MAN. So I think really it's a continuation. And actually, it's refreshing that all of us are in the same room now, which we couldn't do before. So I think overall, the vision's the same. And of course, there are some changes or new people, new faces that we have to assimilate. But overall, that hasn't changed. You're right. So we haven't received the full 25 trucks that we ordered earlier this year. This has no impact to our Driver-Out. So that -- these trucks are really for what's after and continuing growing our fleet of trucks that we can provide more fleet capacity and more data collection, autonomous testing overall. So this wouldn't have an impact to Driver-Out.

And after this Driver-Out pilot program, it is -- a pilot program is meant to have a finite time. We are working with pro type components, some are A sample, B sample components. And so there is one operational cost as well as just in terms of reliability of the hardware that we -- it doesn't make sense at this point to continuously run. I think really this is to showcase the maturity of our technology operations, even working with non-auto-grade components. We will -- after this, we will give a more detailed plan of the different milestones from now to series production. So we look forward to updating you on what's next shortly.

Ken Hoexter -- Bank of America -- Analyst

Wonderful. Just another one for me, if I can. The -- you had mentioned the target for ramp-up in revenues in China was starting next year. I just want to know if that's still on target. Should we wait now, given some of the equipment issues or just timing and development? What are your thoughts on China and revenues?

Cheng Lu -- President and Chief Executive Officer

Great question. So with China revenues, a significant driver of that is the regulatory environment. We're making tremendous progress on the testing and technology side. We're working closely with the different layers of regulatory bodies to allow us to actually have paid freight. At this point, we have -- we don't have permission to do so yet. And so that's kind of where we are. We're working closely, and hopefully, we can make more progress. But I don't really have a clear update or time frame for you at this point.

Ken Hoexter -- Bank of America -- Analyst

So I guess from that, just unlikely in '22 or is that kind of your view or if you think if something develops?

Cheng Lu -- President and Chief Executive Officer

We're optimistic. We're optimistic and I'm hoping it will develop. So right now, I would assume it's still 2022, and then we'll continue to provide you with updates.

Pat Dillon -- Chief Financial Officer

And Ken, it's Pat. I would expect in the early part of next year when we report our full year results for 2021, to give some guidance on revenue, including a little bit of detail on what we're expecting from revenue from that region.

Operator

From Citi, we have Chris Wetherbee. Please go ahead.

James -- Citi -- Analyst

Hi guys. James[Phonetic] on for Chris. Wanted to touch on the order book that you had covered. You converted the 100 truck order that you had mentioned on the prior call. But just wanted to understand, should we think of that as like essentially, there's some other orders that are similarly closed in the same place in the order book is the pipeline is developing fairly continuously? And also just wanted to understand the comment around when you said you were focused on quality. Should we think about that as a certain revenue hurdle? Just wanted to get your thoughts on those 2.

Cheng Lu -- President and Chief Executive Officer

Chris, sure. The -- as we continue to talk about reservation, for us, our strategy is to focus on depth rather than the breadth at this point. And of course, announcing big numbers, of course, makes great headlines but doesn't really help developments unless we can go deep with our reservation partners. I mean, our collaboration with UPS is a very good indication. It requires significant resources from both sides. There's a lot of in-depth learning that is helpful for us, both on the operations side as well as technology development. And it also kind of pushes us to increase our levels of commercial operations, service requirements and technology requirements. So that's what we hope to see from the reservations at this point. All of our reservation partners are the bluest chip partners you can think of. Of course, we can't disclose some of their names. And so with this one for the last quarter, it's really excited to be working with this new partner and we hope to go deeper. And I think it's -- our strategy hasn't changed. I think you should expect more of a sort of a step ladder function in terms of reservations rather than, at this point, just trying to put as many as we can under the belt.

James -- Citi -- Analyst

Got it. Okay, that makes sense. Also wanted to ask about TuSimple capacity and sort of how you think about it. Is it an offering or product that you believe is sort of a strategic asset? Or is it something more along the lines of an investment in market development? Just trying to understand if, like, essentially, if there's enough momentum toward market adoption, if it's something you might pivot away from or something that you feel is an absolute necessity to build the type of business and have the sorts of moats that you were looking for.

Cheng Lu -- President and Chief Executive Officer

I would say both. Certainly, it's a critical piece of perfecting the solution to be able to sell to a third party. And so that is very critical. It is also a critical piece of our mission, which is to enable autonomous freight capacity at scale. Our job is to put -- to expand the size of the freight ecosystem, AFN to have more lanes and to put more density of trucks on the network, so that provides a better service to all of the shipper and carrier users. And so we do see this as an important piece of our strategy, and it complements the other capacity business model.

James -- Citi -- Analyst

And then one quick 1, if I could. Just wanted to touch on the guide. You're moving to the upper end. Just wanted to understand what essentially changed over the past three months. And sort of what that could be seen as an indication of sort of development speed or any sort of return that's coming back from that.

Pat Dillon -- Chief Financial Officer

Yes. So I think moving to the upper end is really just a continuation of the same trends that we talked about during our last quarter call. We've been pleased at our pace of hiring and other strategic investments that we've made that both further the technological and commercial development. So narrowing the range to the upper end is sort of what we would expect when we're actually moving at an accelerated pace. So we're pleased with where we're headed and see this is roughly in line with what was expected when we last talked to you a quarter ago.

Operator

From RBC Capital Markets, we have Joseph Spak. Please go ahead.

Joseph Spak -- RBC Capital Markets -- Analyst

Thank you. I guess just to maybe pick up on that last question. I mean, you've been hiring at a pretty impressive pace here, Pat. And I guess what I'm wondering is: one, what's sort of the marketplace like here for employees, especially on the R&D area, where it seems like maybe the -- there's more limited resources? And how long should we sort of expect this hiring pace to continue? Or does it start to slow down until you get closer to launch, if you will?

Cheng Lu -- President and Chief Executive Officer

Joe, as mentioned before, we are in a very tight labor environment. The cost of wages are going up, especially for hard-to-find positions in the computer science, artificial intelligence space. But I think overall, I mean, the value proposition that we bring to a new hire is very strong. It's a very exciting field. We're a leader in this space. There is clear direction on where we're going. And so -- and I think the mission that we're embarked on is quite significant, too. So I think the value -- when we think about hiring is really the total value proposition we can bring to new hires. And I think we feel good about our relative positioning in the market. And in terms of this is pace of hiring, kind of talked about before, when you think about what does it mean to bring autonomous freight capacity to market scale? TuSimple, we're effectively AI company, a software company, a car company and a logistics company all kind of put together in 1. And so we do continue to see that our hiring pace will continue to be similar to this year's for next year. This is a lot of work. It's really a vastly complicated system that's a lot more than just a particular software or a particular piece of hardware. So I think the capital we have, the value proposition that we offer allows us to bring on very good talent, and that itself is a competitive advantage that we have in the industry.

Joseph Spak -- RBC Capital Markets -- Analyst

Okay. And then the second question is, I thought I heard, but maybe you could confirm or clarify, that the purpose-built truck is on schedule. And I guess I clearly see the advantage to having that to do what you want to do at scale. But a lot of the other sort of competitors, if you will, don't really have that. And I guess I'm wondering, do you think that's sort of sufficient for the sort of next maybe 3- to 5-year period because we don't sort of hit scale and like an aftermarket or a retrofit sort of solution is OK? Or even for that first phase here to really scale, you're going to need that purpose-built vehicle?

Cheng Lu -- President and Chief Executive Officer

Very good question. There will be additional milestones in between series production. So I think what's important to recognize is that we have our eyes on the prize, right? We understand what it means to really bring this technology to market scale. I think we communicated that these are the key components, including series production of trucks. Now beforehand, like within any new technology adoption, we have to be practical. We have to think about what's the most important, how we demonstrate a service. And it's not going to be nationwide day 1. It's much better to start with a smaller service and grow out rather than try to go from 0 to 100 overnight. So those facts, we know. And so I think you'll see certainly different generations of our trucks. The reliability of this generation, the features will improve. And I think -- so I think it's -- I wouldn't say -- I wouldn't take this as we're focused on just that particular thing. It's -- that's important because ultimately, for a company to be successful long term, we can't lose sight of what it means to scale. But overall, there's no change to our 2024 time in terms of scale deployments. And as mentioned, after Driver-Out, we will talk about some of the other milestones that you should expect to see prior to that time frame.

Operator

From Needham & Company, we have Raji Gill. Please go ahead.

Raji Gill -- Needham & Company -- Analyst

Yes, thank you and congrats on executing on a lot of those metrics. Pat, I was wondering if you could maybe give us a flavor of how to think about 2022 revenue since we're kind of coming to an end for this year, given kind of the supply component shortages? How do you think about the drivers of revenue next year at a high level?

Pat Dillon -- Chief Financial Officer

Raji, I think while we're not in a position to give 2022 guidance yet, the drivers really will be around the size of our revenue fleet, our ability to add trucks as well as safety drivers for that fleet as well as our AFN partners. So nothing dramatically different from what we're doing today. We are focused on growing revenue, and obviously, we've been doing that quarter-over-quarter throughout this year and would like to see that trend not only continue but accelerate next year. And we're doing that very strategically. Frankly, we've been oversubscribed in the amount of blue-chip carriers and shippers that want to work with us. And they want to see our technology in action in all freight autonomously as an entry point to eventually deploying driver autonomy when it's ready at scale. So we're trying to be very strategic about how we do that. And I guess the short story is continue to see acceleration in the revenue but never losing sight of the overall strategic objective of why we're building this capacity today and keeping sight of the long-term goal. But we will give more specific targets when we meet with you again on the 2021 fourth quarter call.

Raji Gill -- Needham & Company -- Analyst

Got it. That makes a lot of sense. On the Driver-Out program and congrats on solid progress on that front, how do you view this as a validation tool in order to increase the number of reservations or increase your value proposition? In other words, the Driver-Out program, in my view, obviously, is going to be a very significant milestone not only for your company but for L4 self-driving trucking. And if you can demonstrate that and with all the safety and reliability that's required, I would assume that can be used as a significant kind of selling tool in order to increase the number of reservations and increase the order book. So how are you guys thinking about that? How closely are the customers, the shippers, the carriers are monitoring the success of the Driver-Out program? Do you feel if you do this successfully, this is kind of one of the last steps in order to get more reservations, this is like a significant proof point? Any thoughts on that would be helpful.

Cheng Lu -- President and Chief Executive Officer

Sure. I would say from my interactions, the investors are more focused on this than probably our shipper and carrier partners. Now everyone is focused on this. But I think when you come to Tucson, ride the vehicle, I think the general belief is that the future is coming. And it's a matter of when and not if, and that the rollout of economy is going to be a phased approach. It's going to be lane by lane rather than overnight nationwide. And so I think those are takeaways that our partners already have. And that also requires in-depth collaboration to embed this technology into the existing supply chain. So none of these things really changed. So I think at this point, I wouldn't say that for instance, we expect to have X amount of new reservations because of this. I think, certainly, it's a great validation point and it gives everyone more confidence. But I do say for all the folks that have come to Tucson already, I think they've already taken away with a strong positive feeling of where things are going.

Pat Dillon -- Chief Financial Officer

And Raji, one thing maybe I would add on to that is not an immediate boost by the Driver-Out translating day one into reservations. But taking the learnings from the Driver-Out and some of the things we've learned around the systems architecture to be able to expand our capacity to do A/B testing runs for commercial customers, oftentimes, that can be one of the gating items in order to get them across the line to being a reservation partner. And I can think of one in particular where being able to run certain runs for them in autonomous mode is certainly a requirement for them to make a reservation. So there's things like that, knock-on effects that getting through Driver-Out, which may not show up immediately in the reservations, but we think is a good indication of our ability to increase that reservation count over time.

Operator

From Credit Suisse, we have Dan Levy. Please go ahead.

Dan Levy -- Credit Suisse -- Analyst

Hi, good evening. Thanks for taking the questions. So I wanted to start by asking on the Driver-Out. When you do complete the program, what are the types of metrics or items we should expect you to report so we have a sense of how successful it is? I mean, I assume you'll give us some sense on the number of disengagements for miles, but what other things might we expect you to report so that we have a feel for the success and how much this is a stepping stone to other achievements?

Cheng Lu -- President and Chief Executive Officer

Yes. I mean, I think the -- I mean, generally speaking, the ability to do it. I mean, we will capture this on video. This is not going to have something where it's doctored. So I think you should be able to see that this is done fully driverless. And the fact that we passed our validation case, which we will publish in terms of the safety validation case, that's one metric or a set of milestones. And look, I think it's really just kind of taking a step back, I think the industry, I know everyone is kind of very keen on finding ways to measure progress, right? But I think the Driver-Out is -- I think what passes for the PowerPoint, the pretty PowerPoints, just a demo, we constantly are giving you more data points on our progress to give you examples this quarter and highlighting the empirical data of just the amount of miles we've driven, the performance of our Thomas system, both in terms of fuel efficiency and safety. I mean, I think that's something that is already a benchmark that TuSimple is setting the way. And so I mean, I think to kind of answer your question, I don't think there will be sort of new specific metrics that come out of the Driver-Out program. But the fact that we were able to complete it with this very stringent safety case in itself is a very big milestone.

Dan Levy -- Credit Suisse -- Analyst

Great. And then the follow-up is on talent expansion and it's a bit more specific to AI. I think we know generally, this is where you have a real arms race with Tesla doing AI Day and Cruise having something like this as well. So AI seems to be where there's a much more intense race. So maybe you can just give us a flavor of, A, what the composition of your R&D hiring has been between AI or hardware engineers or other software engineers? And then maybe you could just give us a sense of your ability to attract AI talent versus other competitors, how that's fared?

Cheng Lu -- President and Chief Executive Officer

Sure. The -- so I think generally speaking, we have -- I mean, it's hard to give an exact breakdown. Probably 30%, 40% of our software engineer hires are focused on algorithms, deep learning, machine learning areas, if you kind of just want to have a rough number, -- And obviously, right now, we feel really good about the level of talent we've been able to attract, I mean, really, really good. I think we've gotten some strong -- very strong senior hires, and that we continue to boast our ranks with junior and mid-level hires. I think the value proposition, again, that we're proposing is attractive. A lot of learning to come in TuSimple. We have a clear direction on where we're going. It's a focused strategy. And so there really -- I think, if anything, I feel really positive about our ability to attract talent compared to any point in time in our history.

Operator

From Baird, we have George Tenaris. Please go ahead.

George Tenaris -- Baird -- Analyst

Hey guys, thanks for taking my question. I had a question about partnerships in the industry. I'm sure you read the same press releases from your competitors that we do, and there seems to be sort of a chess game going on here. Can you just elaborate as to who you're seeing within these competitive bake-offs and what the other partnerships in your opinion mean relative to yours?

Cheng Lu -- President and Chief Executive Officer

In terms of these overall commercial partnerships, I think all these partnerships is positive for industry. I think we applaud more partnerships, and it means that more and more stakeholders in the freight ecosystem are getting into economy. So overall -- that will push the technology forward. I mean, I think it's not our place to comment on specific competitors. But I think it's also a timely question because for us, the depth of the partnership is so critical and that's something that we've been talking about all along. We started, I think, the trend of getting people involved in having partnerships. But now we're focused and we have always been focused on going deeper and really having action points and benchmarks to measure our progress.

And so going back to UPS, I mean, being able to have 160,000 autonomous freight models, just one partnership, to be able to [Indecipherable] how long it takes for any other competitor to catch up to that. The second part is everyone talks about fuel efficiency but the empirical data and being validated by third party is significant. I mean, that alone is a real business case for autonomy. And that was 2.5 years in the making, right? So I think we feel really good about the quality of our partners. And then more and more, it's really about the depth of the partnerships.

George Tenaris -- Baird -- Analyst

Just one more. I'm sure you guys have been reading in the press with all the supply chain challenges in the country, that part of the issue might be a shortage of drivers. Curious as to whether these current issues have accelerated any discussions you've had with potential partners or shippers.

Cheng Lu -- President and Chief Executive Officer

It certainly has. And I think the supply chain -- I think what's interesting is the supply chain issues, the driver shortage issues, I think in the industry has always been well known. It's just now, I think, is really front and center on Main Street, right? I think everyone feels that the supply chain constraints and the costs are being passed on to the end customers. The $2 per mile rate now is the $2 per mile rate and fuel surcharge is more expensive than ever. I mean, last year, fuel was $30 a barrel and now it's close to $80 a barrel. And so again, the fact that we have empirical data now to demonstrate our fuel efficiency is significant. And so I mean, certainly, I mean, I think everyone, not just people in the industry, are waking up to the extreme pressures that supply chain is facing. And it doesn't seem like it's getting better anytime soon, right?

Operator

From Piper Sandler, we have Alex Potter. Please go ahead.

Alex Potter -- Piper Sandler -- Analyst

Great, thanks a lot guys. One more question on the rollout with Navistar. Last quarter, you had mentioned that you're hoping to freeze the BOM, I guess, more or less come to a final conclusion regarding the system itself. Is that -- and I think you were targeting early next year. Is that still on the table? Or are some of these supply chain and truck delivery delays caused you to push that out slightly?

Cheng Lu -- President and Chief Executive Officer

Great question. I'll say for large -- for the majority, it is. There are some key -- there are a few key components that are -- that could be pushed out a little bit. And so we're working very closely with our partners, Navistar and TRATON to address those.

Alex Potter -- Piper Sandler -- Analyst

Okay, fair enough. And then the last question for me, just maybe a follow-up on the question that you were just discussing. I'm curious qualitatively, maybe you could include UPS in this, but also beyond UPS. Just how have your discussions changed with partners, whether it's shippers or the fleet, the discussions that you're having today versus maybe what you were having with those same people this time last year or two years ago?

Cheng Lu -- President and Chief Executive Officer

Sure. It's -- I would say, generally speaking, it's the same. And the reason is because our strategy has been very consistent. We identified that there is an imbalance between demand and supply freight capacity. And that's proven, unfortunately, because of the pandemic to be exacerbating and actually speed up the severity of the supply chain constraints. Our job is to enable new freight capacity at scale. We talked about being able to do so by developing the most advanced software, the hardware and, of course, our go-to-market strategy. So I think we definitely see more urgency from our partners. We see that, of course, I would like to have these Thomas trucks ready by next year, not a few years from now, at scale. But those things haven't really changed. And I think more and more forward-thinking partners recognize that this is something that takes time to adopt in terms of baking this into their existing infrastructure. So I think -- I mean, we're very positive by all this development. And in a way, we feel validated. We've been validated by because of this, and we feel like our strategy to really enable freight capacity at scale is -- it's definitely the right strategy and it's a very clear strategy on how we intend to do that.

Operator

From Oppenheimer, we have Colin Rusch. Please go ahead.

Colin Rusch -- Oppenheimer -- Analyst

Thanks so much guys. I wanted to get a little bit deeper into the mapping data and understand how these models that you're adding, how much of that volume is focused on a particular area? It seems to me that some density going to be really important, especially given your comments earlier about starting small and growing networks. So I just want to understand how you're tracking that and how we should think about that with where you're at right now?

Cheng Lu -- President and Chief Executive Officer

Sorry, is your question around how we're using the data or maybe if I could...

Colin Rusch -- Oppenheimer -- Analyst

No, no. It's more about like how much -- how many of those miles are concentrated on a limited number of routes? So if we think about the total mileage that you've got mapped, how much of that has really been driven over in a consistent basis where you guys are getting a real depth of experience around those roads and traffic patterns?

Cheng Lu -- President and Chief Executive Officer

Yes, got it. It's very strategic where we collect the data. And I mean, certainly, there's more concentration kind of where we operate today from I-10 from Phoenix to Tucson down to Dallas. We're starting to collect more and more data on the Texas Triangle. And while, largely speaking, highway driving across the southern part of the U.S. is consistent, but there are differences, right, difference in behavior differences and road conditions. And so absolutely, I mean, the more data we have on particular routes, the more confidence we have in order to safely traverse in autonomy. So -- and the same thing with our revenue, with our mapping. I mean, it follows the same pattern that we're not here just for sake of gathering more revenue. Of course, carriers and shipper partners want us to help them alleviate some of the freight capacity challenges they have. But ultimately, all that is centered around where we see the initial rollout of autonomy first, and then we'll expand outwards from there.

Colin Rusch -- Oppenheimer -- Analyst

Okay. And then can you give us a sense of the learning cycles? Certainly, it seems like as you go deeper in particular areas, that's going to accelerate your learning in newer geographies. So I guess, can you give us a sense of how that translates and how quickly those learning cycles are accelerating at this point?

Cheng Lu -- President and Chief Executive Officer

I think generally speaking, one of the proof points of our expansion outward and with UPS to the East Coast, the number of trucks we have and our ability to operate these trucks in autonomy, I mean, showcases our ability to scale and our ability to iterate fast, right? And that's where I think we see -- we'll see sort of us and the competition kind of diverging. I mean, we feel very good about our infrastructure platform, the way we can take data, structure it and use it to continue to train our system to be smarter and more reliable. So I don't know if there's a metric I can share with you today, but I think really all the metrics or the data points we shared about the amount of miles we've done with UPS, the empirical data that we have, the scalable operations, I mean, that is -- and the fact that we're close to being the first to have a Driver-Out pilot program, I mean, that just shows that you have to have a -- we have a very robust and scalable and fast-iterative development process. And we'll continue to improve that every day.

Operator

And from Wolfe Research, we have Scott Group. Please go ahead.

Wes Mattox -- Wolfe Research -- Analyst

It's Wes Mattox on for Scott. I had a question just about opex. With stock-based compensation, just trying to understand the trajectory of what to expect over the next few years or so. Should we think about that as a consistent percent of opex? Or is it on its own sort of separate growth path?

Pat Dillon -- Chief Financial Officer

Yes. I think in terms of what you saw this quarter, you would expect to see that grow roughly in line with opex quarter-over-quarter. The vast majority of our operating expense is around people costs and so would expect SBC to trend generally in line with our overall people costs, which make up the majority of our overall opex.

Wes Mattox -- Wolfe Research -- Analyst

Okay, thank you.

Operator

[Operator Closing Remarks] Thank you.

Duration: 67 minutes

Call participants:

James Mann -- Head of Investor Relations

Cheng Lu -- President and Chief Executive Officer

Pat Dillon -- Chief Financial Officer

Ravi Shanker -- Morgan Stanley -- Analyst

Brian Ossenbeck -- JPMorgan -- Analyst

Ken Hoexter -- Bank of America -- Analyst

James -- Citi -- Analyst

Joseph Spak -- RBC Capital Markets -- Analyst

Raji Gill -- Needham & Company -- Analyst

Dan Levy -- Credit Suisse -- Analyst

George Tenaris -- Baird -- Analyst

Alex Potter -- Piper Sandler -- Analyst

Colin Rusch -- Oppenheimer -- Analyst

Wes Mattox -- Wolfe Research -- Analyst

More TSP analysis

All earnings call transcripts

AlphaStreet Logo