Aptiv PLC (APTV) Earnings Call Transcript & Summary

Alright. Hello, everybody. Delighted to continue on with the 8th Annual Morgan Stanley Auto 2.0 Summit. We have a fascinating panel here discussing the future of mobility, where we have Glen De Vos, Senior Vice President and CTO and President of Advanced Safety and User Experience at Aptiv; and we have Chris Urmson who is CEO and Founder of Aurora. I'm going to give a bit of bio here for both of these very accomplished and important technologists at this historic time for the industry for color, and then I'll read a few disclosures. But Glen De Vos has had this role as the CTO and President since 2017, as I mentioned. In this role, Glen is responsible for leading the company's innovation strategies on development of advanced technologies. Glen also leads the global engineering organization at Aptiv, which includes more than 18,000 technologists. So Chris, you might be approaching -- careful, Glen and Chris might need to be approaching some of those as he's growing the business as well. Those 18,000 people located in 15 major technical centers around the globe. And he's also President of the Mobility and Services Group. There's a lot here. I'm going to cut it off. But Glen has been a really a friend of Morgan Stanley, and has not just been a business leader, but also, I think importantly, an educator to Wall Street for so many advanced technologies that are still changing. So Glen, we still rely on you. And then this is the first time I've had a chance to meet someone I've been following for quite a while. Ravi, you've met Chris Urmson, Co-Founder of Aurora, and has been CEO and member of the Board of Directors since its formation in 2017. But before that, Urmson, as many of you know, helped build Google's self-driving program from 2009 to 2016 and served as a Chief Technology Officer of the Group. He had over 15 years of experience leading automotive vehicle programs. Director of Technology, Carnegie Mellon, DARPA, Grand and Urban challenge teams. And it's a long -- we need to save time to talk about real s*** here.

I'm deeply offended. No, that's okay.

I think you'll be all right.

So listen, with that, I want to give you each opportunity -- and opportunity. Maybe Glen, starting with you on your key messages for investors, and how you plan on spending time, and what excites you most for 2022. Then I'll move to Chris. Thanks Glen.

Sure. And greetings from CES, by the way. And really, I'll talk about -- maybe talk about Aptiv first, and thanks for the introduction, Jonas. It's always, it's a pleasure to be here with your audience. For Aptiv, as we think about 2022, it's really a continuation of the themes that we talk about [ truly safe, green, and connected ] and really focusing in on a couple of areas of that. One is high voltage at a great year with bookings and high voltage, about $3.5 billion there. The other is around just the whole transition of vehicle architecture to more software-defined, software-enabled architectures, what we call SVA. And then it's the continued advancement of advanced safety systems, ADAS systems and being able to expand performance at lower cost to the end consumer, the democratization of ADAS. Those are really the key themes that we're going to continue to build on. 2021 was a great year for us in terms of bookings, almost $24 billion. So obviously, going to manage that. It's a record year of managing that and really a kind of accelerating on that going into '22. For Motional, our JV with Hyundai, a little bit different story in terms of -- it's really about preparing for the commercial launch in '23. So here in Vegas, what we doing the commercial driverless launch in '23. That's really what we've been building up to. So it's working through the technical preparations for that here in Vegas, expanding the site, getting ready for that. So a lot of work being done there. And then as well, expanding into other markets. We had the announcement with Uber relative to Santa Monica and L.A. and those types of things. So really getting prepared for that commercialization phase, which is really exciting after all these years of prep.

Chris?

Yes. Well, thanks for the invitation to be here. Really appreciate the chance. I follow you as well. So really nice to have a chance to meet in person. '21 was a pretty incredible year for Aurora. We kicked off the year with acquiring Uber self-driving car business and bringing those teams together. We were able to build amazing partnerships with PACCAR, Volvo Trucks, Toyota, Uber, FedEx through the year. And of course, we took our company public as well. And so really -- and along with that, all an incredibly exciting technical advancements as well. We released our first safety case framework that is the first of the industry sharing. How we're going to convince ourselves and ultimately, regulators in the public that the Aurora drivers are not creating unreasonable risk on the road. '22 is going to be another stellar year, I expect. We're building towards what we've shared guidance of trying to launch a product at the end of '23 in trucking. And so we're working towards that this year. You'll see us demonstrate the transferability of the Aurora driver from truck to [ pass through ] vehicle. This is something that several industry have concerns. Can you actually have the same hardware and software drive big trucks as cars, and we're going to showcase that happening. We look forward to increasing the -- or demonstrating, sorry, the ability for the vehicle to handle significant faults. So it's nice that the truck works when everything is up and running, but what happens when something breaks, and we want to be able to showcase that, and how we're able to handle those situations in [ sweatshop ]. And then we'll continue to build towards scale that as we think about launching a product without a driver onboard. It's really -- it's not kind of a big bang moment. It's a continuum of progress and increased performance and maturity. And so we look to share the progress we'll make on that over the course of the year. So exciting time for us.

It's been a busy year, busy year for you both. Before I turn it over to Ravi, I just want to ask one question, kind of in the form of you're at a bar at CES tonight and some [ slashed ] exhibitor comes up to you and says, I'll start with you, Chris. Chris, when are we really going to start seeing the decline of steering wheels, like in a significant way, not in a significant way. I realize there's no -- this is a moving target, but I wanted to get each of your answers on that to the drunk exhibitor at the bar tonight, starting with you Chris.

Yes. So for us, we're going to be delivering...

The person shouldn't be driving that I'm referring to.

So absolutely not. They should get a safe right home, hopefully, with an Uber, our partner. So we -- first, we're launching our product in Trucking first. And there, we'll begin with vehicles that are being adapted from existing platforms today. And that just makes sense economically. It allows us to plug into their programs. It allows us to use proven technology. And then over the coming decade, I would expect you will start to see the transition from using adapted vehicles for people to custom vehicles for operating without a human driver.

Glen?

Yes. So for us, and our focus really remains on the kind of the urban mobility piece, moving people, moving logistics. For us, it starts in '23. And we would expect, as you think about our first vehicle, the IONIQ 5 with Hyundai, that's still -- that's a fully integrated platform. So when it launches in '23, that's fully integrated into that vehicle. That vehicle gets built like any other vehicle. You think about the generations to removing those steering wheels and really getting to that step, I would expect that to be closer to the end to 2030, when that starts happening. So you're going to get mixed usage cases between now and then, where you still have the -- potentially the steering wheel in place. But eventually, those go away, I expect that closer to the end of the decade.

Ravi, over to you.

Chris, maybe I'll kick off with you. obviously, you are a legend in autonomous driving industry going back to the DARPA and Google days. Yes. Aurora is 1 of only 2 entities that's working on both truck autonomy and [ PACCAR ] autonomy at the same time and kind of given your background in your history kind of coming from the Google, sort of things. How did you get to this point, right? So at what point did you figure out, hey, it's truck first and then [ PACCAR ] and kind of why is that the case?

Yes. So we were -- when we founded the company, we were very intentional about the idea. We wanted the Aurora driver to work for light vehicles and big Class 8 trucks, right? We -- as you think about the driving task, fundamentally, it's the same in both, right? We have young -- when you go to become a commercial driver, you begin by getting your normal passenger car license, you have to have that as a prerequisite. It's a foundational skill and move on to -- onto the truck. So it's a common capability. And if you're thoughtful upfront about sensor selection, about the architecture used for representing the world about the perception of motion planning systems, you can build a capability that operates across both. And that's the way we've built our system. We began work with light vehicles, and we did that for a variety of reasons. One is they're just easier to deal with, right? It's -- they're easier to maintain. They're easier to operate. They're smaller. When we've started the company, we also didn't believe anyone had the sensor technology necessary to see far enough to drive safely and robustly at speed. And that was why we spend as much time as we did over the first couple of years trying to find the right LiDAR technology to complement the work we do in camera, and the work that the industry does in Radar today. We ultimately found that with the company we acquired, brought in-house with -- and have now turned that into the FirstLight LiDAR, which is a very long range but also has instantaneous [ stop or ] measure. And so -- which allows us to more rapidly respond to other actors in the environment. So that put us in the right place to be able to go tackle trucking. And then as we understood from day 1, the economics in trucking and the demand was there, right, a truck driver, we value that dramatically more than we value a local goods delivery driver or a ride-hailing driver. We pay much more for that capability. And so from an economic perspective as a business that makes it a really interesting place to start. And then there's an immense demand. So drivers in trucking today -- truck fleets today have something like 90% turnover year-on-year. So trying to find drivers is hard. There's a shortage of truck drivers. It's about 80,000 short in the U.S. today. We expect to be 160,000 short by the end of the decade, right? So there's a real need. And it's not about replacing drivers. As we talk to customers, it's about -- we just need more of them to serve our business. And so we expect to come in alongside the human drivers and support routes that run [ undesirable ]. These long routes that have people staying away from home. And so it was being in a place with an architecture that supported both capabilities, having the technology that was necessary to be able to see for enough. And then a deep understanding of the business, and the economics that led us to say, okay, trucking is the right place to scale initially, but let's not seed the ability to operate light vehicles because that business will be as large as trucking in the long term. It will just take longer to get there.

Got it. Maybe a question to both of you. I mean, autonomous driving, obviously, is an incredibly difficult problem. Glen, I think you've been coming to our CES on it since the first one, what, 7 or 8 years ago. At the time, we were saying, "Oh, we'll have these fleets of robotaxis everywhere by 2020." I'm sure you're like, yes, we'll see. Obviously, you were right. But clearly, it's a really difficult problem. How do you guys -- maybe you can have Chris -- sorry, Glen first and then Chris. How do you guys collectively solve the problem between tech startup versus Tier 1 supplier versus OEM versus the regulators? I mean how do you guys come together and make sure that this is already a frontline?

Yes. It's a great point. And if you think about 7 years ago, a kind of the projections, there were some pretty aggressive projections for us, we've been operating the fleet here in Vegas now. I guess it's going on about 3 years -- 4 years or so. And that's been a tremendous educational backdrop for us. And it's up to solve that problem that you were just talking about. And that really was the genesis of the JV with Hyundai. Because as we looked at it, we're -- at Aptiv, we're a systems integrator. We're a technology provider into the OEMs. Motional, really develops that core technology maybe driving. We can help them industrialize it. And then we can help put it into the production with Hyundai, an OEM that's developing a vehicle ready to receive it. So you know, we have all 3 of those pieces working together. And we think 1 of the key enablers for commercial success and scaling of autonomy is cost. And this is the way we believe you get to a cost-optimized product, automated vehicles. And so that's how we've looked at it. And then partnering with the Lyfts, with the [ Veos ], with Ubers to then have that market access, to be able to deploy those vehicles into operation. And so that's the approach we've taken to date and through the experience we had here in Vegas, really, it's -- we've seen that model we think works. And the regulators TUV, SUD, they brought them along at each step of the way, the local municipalities is keeping them involved has also been very helpful. But we think that clears the pathway to deployment. Ultimately, though, you have to have an ODD that's commercially viable. So a service area that works, and you have to have a cost-optimized product that gives you the, basically, the monetization and the benefits that you need from the investment.

And I'd echo a lot of what Glen said that when we founded Aurora, this is 1 of the founding principles was to build a company that works in partnerships with others. As we -- and we think kind of maybe this is a simplistic view of it. There's people that build the driver, people who build the vehicle, and then the people who use that vehicle. And we recognize that it turns out building vehicles is really hard, and the people that are really good at it. So let's work with those folks. And then building Uber or building FedEx. These are very difficult complicated businesses. Let's work with them as well. And so that spirit has been there from day 1, and that's how we've -- if you look at the constellation of partners we have, that's what led to that with the regulators. Again, this is one of the places where even from back in my days at Google, recognize the importance of education and engagement, and not surprising them, frankly, right? People get in the government to help to do good things, right? And they can be a great partner. And so over the last decade, that's where I've spent a chunk of time is making sure that we don't leave whether it's policymakers or regulators surprised. And so they can make informed decisions, they can play their role in this ecosystem. And that's both at the Fed or all of the federal state and local levels.

Got it. Maybe just one more for me before I pass it back to Adam. There's been a lot of focus, obviously, for yours on passenger robotaxi opportunity. There's been a lot of focus in recent years on the commercial trucking opportunity. I think if you put both those circles together, I think the 1 of the joint slight to the Venn diagrams probably last mile delivery, which has the commonalities between both those use cases. Obviously, both of you have partnerships with Uber, obviously, for Aurora, they're a key stakeholder and a key partner. Would love both your views on where you think autonomy for last mile delivery kind of fits into the picture here and what's that like?

Maybe -- because I think just -- I'll build on what Chris has said earlier, which I think is spot on. It's -- the technology applies to both. And for us, our primary focus is on urban mobility solutions last-mile, that area because those use cases tend to be the most similar, quite frankly. And the vehicle types tend to be more similar, but it scales across. So it's not a question of do you do heavy-duty trucking? Or it's a question of when can you do both -- both are good markets? Can you do both and when? So for us, our initial focus has been really on the mobility in the urban areas, the last-mile type of applications, people and product. I think there's a direct overlap. And what's been interesting, if you see the trend like the massive growth in Uber Eats and the use of apps for food delivery, for product delivery, it's been stunning. So I think that represents a really nice adjacency to robotaxis per se, and one that we're -- we see as being very viable, a very viable stepping stone.

And I think that I look at local goods delivery, and ride-hailing is very, very similar applications from a technology perspective. From a business perspective, it seems that the local goods delivery seems like a more challenging space to operate in than either of the other 2, just given a kind of the value we ascribe to the driving vehicles for each of these applications. I think the other challenge that we see with last-mile delivery is the kind of final 50-foot problem, and that's less of a problem with prepared food. But as you think about package delivery, it's -- it becomes nuanced and complicated and -- but again, we've got enough to chew on already in the commercial freight space and the people in this space. But we expect to serve that for very similar reasons to decline ultimately.

Adam, over to you.

Guys, one of the most common questions I get and I know Ravi gets it to is who's in the lead? Is it Cruise, Waymo, Aurora, or Motional, too simple. And my response, not just because I'm intellectually lazy is, I don't know. I don't really -- a lot of this is self-reported. There's different definitions. I mean, what's a disengagement really. I mean, it's not -- it's -- and then even if there are disengagements, the routes are different. I mean do you really adjust for that? A lot of -- it might not be possible. So my impossible question for you to possibly answer is, starting with you, Chris, is how do you measure who's in the lead? What can you -- what kind of KPIs or outputs from these various systems do you value when trying to get some semblance of leadership?

Yes. We mostly don't play that game because it's really just...

Make me feel better.

Yes. Again, I don't mean to be intellectually lazy either, but -- so I think Waymo has vehicles operating today with no driver in the mono road, right? Limited scale, but they have that, and that's compelling and exciting and meaningful. Cruise, to a smaller degree. So that's an interesting milestone that it's actually sustained operation without an operator onboard. And so that's an interesting point. Beyond that, it's really hard. We have a high degree of conviction of the path we're on. We look at it as a continuum. We're operating in commercial ways today. We happen to do it with vehicle operators onboard. As we roll the clock forward, the role of that vehicle operator will diminish over time to a point where it is now able to operate continuously without an operator onboard and then off to the races and build the business. And so we've really focused on how do we make sure we've got the right team, how do we make sure we've got the right technology, and how do we have, make sure we're on the right space for the business to build and execute.

Glen, anything to add or?

Yes. I think that question will sort itself out over time in terms of who ultimately gets to the market with a commercially viable product or service. And I think time will play that out. The key is, do you have a robust plan to get to market? How are you deploying? And do you have the resources and the capability to get you there? And that's where you just got to focus on those things. And that's where promotional, it's the launch next year in Vegas with fully automated driving. It's just those stepping stones. Making sure that you're hitting the ODD milestones that you've laid out that support that time line that you have the ability to get there, and then once they're you have the ability to scale. And I think that's how we think about it. That's how we approach it. We like where we are. It's great to see progress being made broadly speaking, though, I would say that. I think that's always good because it tends to validate and reinforce the direction that all of us are heading. So it's nice to see things happening in the market and with the others. But at the end of the day, it's -- time will ultimately play out who wins and who doesn't.

Ravi, I'll know about you. But I mean this session is one of those where it's just too goddam short. I wish we had 5 hours with these guys. I think they would disagree, but you and I would agree. So I have a question on Uber. I have a question on Apple, and I have a question on LiDAR, and I want to try to get back to Ravi, too, before we're out of time. We're going to do that in 10 minutes, okay? So both of you, in some ways, especially either commercially or strategically, et cetera, relationship with Uber, right? And if I recall, Travis Kalanick, back in the day, said like, this is existential for us. If Uber doesn't own the driver, we don't have a business. Now that's 1 view. But in -- it does seem there's some logic to me at least is if you guys have the localization, the mapping, the edge compute, the cloud and all the things to have in FAA, I'm putting that in quote, kind of certified kind of terrestrial system with a commercial airline style safety, then what is Uber's value add? The brand? The customer? That seems like really easy stuff compared to what you're doing. So where am I wrong? And obviously, they could be listening to this webcast. Chris, I'll start with you.

Yes. I think you kind of underestimate their business and the operational complexity of it, the value of having that customer base the understanding they have of demand and market forces, and I think we would much rather work with them, right? Why do I want to build all of that? And so we'll provide a driving capability. We'll have a partner in Toyota that provides a vehicle. And together, that will go and serve their customers. And I think we can help them over time, grow their business and we'll grow ours. So yes, we -- I think that's just a -- the pie is large enough. The opportunity in front of it is large enough, like let's not get greedy. Let's be part -- let's each have a slice of this pie that gets bigger over time.

Glen, what about you, I mean, like you get rid of the driver and the car is the driver. It is your car. You already have the systems to manage the safety efficacy and utilization of this thing, what's left?

Yes, it's interesting because I'd have to add to what Chris was saying, one of the things we learned dealing with our engagement with Lyft, our partnership there, was we had a very upclose and deep understanding of what it means to stand up, operate, and maintain a network. And all the way into the customer experience in the vehicle in terms of the HMI, the user engagement, how that all work, the transactions, how that all works. And even though we would provide the car, we provide the tech, there still is a lot of room for the network operator to play in terms of that customer experience and how that operates. And so I think there is tremendous effort and value still there. And we recognize standing up a network is a massive capital investment. And we said, well in form of some tech on the technology. So I think there is still a tremendous degree for differentiation on the network side, the customer experience there. The key for us is as we serve multiple network operators, how do we enable that? How do we enable them to differentiate or have a different experience in 1 of their vehicles, 1 of their network operated vehicles versus somebody's.

Okay. A quick one on Apple. Let's make this rapid fire with 6 minutes left here. An Apple car, is that a threat to your business, a validation or both? Glen?

I can say, it still ends an opportunity and a validation. And the reason I say the validation is I think it reflects this rapid transition towards much more fully connected, advanced architect, and software-defined vehicles. It just again reinforces, our industry is headed in that direction.

Let me ask it differently from -- to Chris. And if an Apple car were to not have a steering wheel, that's just my belief is I think a steering wheel would be beneath Apple. That's just my own opinion. Why would they ever have in something where you're like you're expected that would just be so -- I mean so old. It might mean that Apple is very, very late to the stuff that you're doing possibly. But if you have a car without a steering wheel, do you own that car? Is that a car that's owned by an individual?

So I think it could be but I think it makes much more sense to be, for it to be part of the service. Get better utilization, economics are better for everyone. I think it's -- I think -- There's a lot of things that push it in the direction of vehicles as a service as opposed to personally owned vehicles.

And that kind of stretches the kind of concept of if Apple were to get into this business or for companies that have a great platform and a balance sheet and capability, and want to get into this, pushing how they engage their consumers in a totally different way through a service. Okay, quick one on LiDAR. Elon Musk has been very, on record, if you will. Maybe , we do -- we actually have 10 more minutes. I'm sorry, that's good. Thank you. Okay. Thank you. Sorry, Chris. So, Elon, I'm going to paraphrase. You said something like in a real serious tone. Anyone doing autonomy that depends on LiDAR is doomed. It's passive optical, baby, whatever. Is this -- Chris, this is just Elon being Elon, and it's outside of his control or kind of why is he wrong or possibly right?

Yes. So we think the right answer is to use the tools that you have available to solve the problem. And that you're going to be able to build a more robust system by using a combination of camera LiDAR and Radar. I think if you asked any ML person, could I have -- do you want 1 camera or do you want a camera and some other sensor data? Any ML person is going to say, give me more data, I will get a better result from it. I think part of the meme around, you just use cameras is, hey, people drive with 2 eyes. Therefore, that's the right answer. I would postulate that horses run with 4 legs, and yet we put 4 wheels on cars because it's an engineering option that we have available to us.

Yes, I would add, Adam, that when you think about whether it's ADAS or automated driving, the fundamental limitation is the perception system, what you -- how you perceive the world around the vehicle and your ability to interpret that. And so can you drive with just vision? You can drive to a certain degree with just 1 sensing modality. We've done those things. But it doesn't open up the ODD to where you really want it to be. You can really expand your operational design domain through the use of multimodal sensing. It gives you the richest environmental model. To Chris' point, if you're developing algorithms for self-driving, you want the richest environmental model.

Okay. So let's -- before we leave that because I could just easily go back to Ravi and I will. Elon knows LiDAR from SpaceX as well. And he has that tool at his disposal. So why do you -- why could you -- not that you're trying to get in his head, but what could be some plausible reasons for they're not being LiDAR on Tesla's vehicles. And now they're also solving a different problem. They are not making. Is that the answer? They're just not like let them do a Level 4, Level 5 vehicle first, and then we'll talk. Is that like just a question?

There's a couple of reasons why he may eliminate one or another sensing modality, because ultimately, you have to fuse it all together. So if you're at a point where fusing it together actually degrades your system performance, then that's the wrong thing to do until you're ready to do that. I mean, again, I'm just speaking from my perspective, where if you -- if the focus is on vision, maximizing vision, and not on the other sensing modalities and bringing them in, infusing them effectively. Yes, you could actually degrade system performance depending on how you do it. So there's some philosophy behind it. There's system architecture behind it. But ultimately, where we've landed is, I want to use every tool that's available to me that I can.

Chris, anything to add?

Yes, just 2 parts. One is I think if you're, if your business model is, we're going to give it away, and then we'll charge an upcharge for it later, then you can't -- you have to keep the piece part price low, and so that just creates a business constraint that may ultimately limit the bounds of what you can do with it. I think the other is just kind of the mathematical argument here, and this is maybe a little more math than you'd like. But if you think about like, let's say, for a moment, you could build three-nines system, a system with a reliability that's three-nines. Let's call that 1 chunk of work. To move that to four-nines is probably 10x is hard. So it's 10x as much work to five-nines, 100x hard. 6 times 1000x hard. In contrast, if you take three-nine systems and put them together and our approximately orthogonal, right? These are not things where it's going to feel in exactly the same way, then you've now done 3x as much work instead of 1,000x as much work and you have a system that is comparably reliable, right? And that's just -- again, that's kind of a handwavy way to think about it, but the way that you can think about the failure math working is something like that?

I can grasp that math. I thank you for doing that. Ravi, you got a couple of -- 5 minutes here?

I do have a bunch, like you said, I loved for this. I kind of wish this was an in-person dinner at Vegas for a while. But yes, we do have a bunch of questions to the audience. So maybe I'll ask Chris and Glen those questions. Maybe some on with you, Glen, how is the OEM push to move software and engineering tasks in-house impacting Aptiv? First of all, would you agree with that characterization and be kind of what are some of the opportunities and risks for Aptiv?

Yes. Yes. And I'd start by saying that's not new. It's been going on for 30 years, actually, ever since the first computer-enabled software-enabled system, engine management systems were brought into the vehicles. So this is an ongoing process. It's not really new, the scale may be a little bit different in the software-defined architectures. The way that impacts us is on one hand, it means that how we work with our OEM customers changes. We still do the systems integration. They provide more of the code and content maybe than they have in the past. But as we kind of was alluded to earlier, the amount of software that's being put into vehicles now continues to grow significantly. So we're still -- our software content growth continues, their software content work continues, third parties as well. So for us, where the big change needs to happen and is happening is in how do you actually bring all of that together? Through continuous integration, continuous deployment methodologies. Automotive is behind other verticals in this regard, to be very transparent. That's where the big change is occurring, and that's where we're focused on how do we enable that to happen much more efficiently.

Got it. Chris, next one is for you. What are some of the main opportunities and challenges facing Aurora in the next 12 months. So maybe it's a good opportunity for folks who are really new to the story, they're going to give us a very quick overview of kind of where you are right now. And what's the path to commercialization of the time line now?

Yes. So last year was incredibly exciting, as I mentioned, big grew, the team dramatically through acquisition, put in place amazing partnerships, took the company public. This year is really about -- last year was make that team work well together and do all of this setup work. This year, it's about executing with this incredibly powerful team and capability going forward. And we look forward to showcasing this year. The fact that our technology does work on both trucks and cars and being able to demonstrate that. We look forward to showing the fault handling capability, and how do we deal with the off-nominal situations and how the -- if something breaks, we can still be safe. And then we'll look forward to continuing to share more of the events on our safety work and our technical advantages as we increase the amount of time the vehicle operates without a driver having to engage.

Adam, over to you to wrap.

Yes. I mean, unless you guys want to add something else, we'll give you a little time back here. And I do hope next year in Las Vegas or anywhere else, we could get the same group together over at dinner or something to really, really, really jam out on some of these issues, and to be able to measure the space from home from this year to next because perhaps, non-linearly and unpredictably, we're going to see some unbelievable stuff happening here. So thanks for making our job so exciting. Thanks for making -- thanks for the things you're doing to improve the world. I don't think -- I think we're just scratching the surface here, but I've got a bunch of ideas coming in my head right now. I need to kind of -- I don't smoke, but if I did, I'd go out for a cigarette right now. All right. Well, with that, we're going to end this panel. I want to thank Chris and Glen for your time. And Ravi, thanks for your great questions as well. Thank you, everybody. This concludes the webcast.

Thanks so much. Take care.

Bye.