Aptiv PLC (APTV) Earnings Call Transcript & Summary

Great. Good morning, everybody. I think we're ready to start our next session. I'm Itay Michaeli, Citi's U.S. auto analyst. And I'm very pleased to have Aptiv back with us at the 2021 Car of the Future Symposium for a fireside chat. From Aptiv, we're very happy to have Joe Massaro, Senior Vice President and CFO; as well as Elena Rosman, VP of Investor Relations. Joe, Elena, thanks so much for being here. We'll keep it to a fireside chat session. I've got a bunch of questions. But of course, we love to answer your questions as well. So if you have any, just feel free to e-mail me throughout the session, and I'll be happy to ask any of your questions. And with that, we can go ahead and begin. Good morning, Joe and Elena, thanks again for being here.

Thanks for hosting, Itay. Good to be here.

So let's kick things off. You just had earnings last week, but given the fluidity of the situation, Joe. Anything new to share with respect to production, semiconductor impact. We're kind of hearing it's sort of a week-by-week thing. So just curious if there are any new thoughts to share there?

No. I mean, I think it continues to be a week-by-week thing. That's really our view of how Q2 is going to play out. I think folks like GM and Volkswagen have had made similar comments around their earnings call. It's really what we're seeing today. I think just to frame it, I think we generally talk about chip constraints as sort of one topic, but there really are 3 drivers. And I think that's important to understand because it in part frames our view of how the industry sort of starts to climb out of this in the back half of the year. The first driver was the broad sort of COVID-related ship constraints that came out of sort of the back half of last year with really just the chip industry not producing enough chips to deal with a V-shaped recovery in auto. I think they had been planning -- despite the forecast they were receiving from the industry, they had maybe planned for more of a U than a V. That is getting better. That first constraint cost the industry by at least, by our estimations, 1.5 million units in Q1 and should be somewhere between 500,000 to 600,000 units in Q2. So sequential improvement there in terms of the impact. There were 2 events in the second half of the first quarter that were discrete events. One, the Texas weather, which closed down a number of large semiconductor facilities for about a week to, sort of 7 -- about 5 to 7 days. As well as a fire at a very large automotive semiconductor plant in Japan, which we refer to as the Naka fire. It's a Renesas facility. Those 2 events differ than the broad chip constraint. Those 2 events sort of create problems with work-in-process inventory. And that results in sort of downstream air pockets in supply. So in the case of the weather, in case of the fire, they have to pull work and process out, check it, test it, make sure it's good. In the case of the weather, put it back into production, but you've got a delay there of a couple of weeks because you've -- the plants were down, and then it takes a week to 10 days to get those chips back into production. And there's probably some yield loss because some of those chips aren't -- you're not the wafers, the sort of raw components aren't able to be put back in the manufacturing process. Fire, very similar, but obviously, you had to repair the facility -- pull WIP, but then you had to repair the facility, get new equipment and such. Those air pockets are going to manifest themselves in the second half of May into June based on everything we're seeing, but a material impact. That alone -- those 2 events alone could be 1.5 million to 2 million units in the second quarter in terms of not being able to produce. So the magnitude of those 2 discrete events, although they're more air pockets and you can sort of do math around when the chips will be available and the suppliers can explain when they we'll have steady flow. As an order of magnitude, that's pretty significant for the quarter, just as big, if not a little bit bigger than the Q1 impact of the COVID constraint. So obviously, dealing with a lot. And I think the week-to-week sort of discussion for Q2 is very appropriate at this point.

That's incredibly helpful, Joe. Thanks for kind of -- letting out that, that way. So I guess, given the week-to-week, do you think like by early June, like what we have seen in May, will represent most of the disruption? Or is it still -- do you think -- will the all clear sign be at some point in June and July, relatively all clear sign, I guess, or what do you think this sort of extends all the way through Q2?

I think it extends all the way through Q2, right, because you're going to have shortages. And in these types of situation, for us anyway, and I think for our customers, the all clear, you sort of wait until you've had weeks of steady supply, and you're really sure you're all clear. But I think -- so you're sort of at the end of Q2. I think there's a couple of things that -- I do think you'll hear noise around chip constraints. Maybe going back to the original COVID issues more in the back half of 2021 and possibly into early 2022, as supplies improve and more capacity comes online, but I think those will be more isolated events. They'll be maybe chip specific, a specific kind of chip or maybe a supplier specific. I think the other thing, Itay, we're starting to get, from our customers, which would be a positive, obviously, particularly around getting sort of home for the back half of the year is that to the extent there are multiple-week shutdowns, particularly in June, when the fire impact really hits, that those shutdowns are effectively the summer shutdowns. And that summer shutdown scheduled for July or August in the industry would become full production weeks. And there's communication now with the supply base. And I'm not sure the OEs have made this public, so I won't mention names, obviously. But we're in those dialogues. They're talking about supply base because they want to make sure the supply base is tied out and up and running those previously scheduled shutdown weeks in the late summer. And that obviously is very helpful in terms of creating production windows in the third quarter that weren't there on the original schedule. So I think as we work through June, into July and get clarity around chip supply from the 2 discrete events, get a good sense of what Q3 production can look like relative to are we producing those weeks of shutdown, those weeks -- that's what I think we'll -- certainly, from our perspective, that's when we'll have -- should have enough visibility to provide sort of go back to our normal forecasting process for the back half of the year.

Awesome. That's very helpful, Joe. Maybe switching gears to bookings, I think pretty strong numbers in Q1. Maybe 2 questions there. First, kind of how did the win rates for the company track relative to internal expectations? And also on the high-voltage award, hoping maybe you could go into a bit more detail on that particular award.

Yes, sure. Why don't I start with high-voltage and then Elena can go through -- just give a quick update on win rates across the businesses. So high-voltage awards continue to be very strong. We did about $900 million of awards in Q1. Compare that to $2 billion to $2.2 billion over the prior -- each of the last 2 years. So very strong start to the year. I don't think that's a run-rate number. I wouldn't multiply that number by 4 for the year. As you know, bookings, and this has nothing to do with COVID or chips. Our bookings are just lumpy, but it's a very strong start to the year. And we've really seen strength in that high-voltage portfolio, not only on the booking side, but on the revenue side. The pull from customers in Q4 and Q1 continues to be very strong. Our business -- our high-voltage product line was actually up 125% in the first quarter, which is a fantastic number. Now that's somewhat helped by -- there was a China shutdown in the prior year for a few weeks that obviously is helping drive that from a comp perspective. But speaks to the strength of demand for high-voltage vehicles, electric -- battery electric vehicles. The other thing that was interesting for us during the first quarter, as you talk about bottleneck management or constraint management around these chips, you ultimately get into an allocation discussion with your customers. You have chips, which platform do they want you to build? Because you can't build them all, right? So when you're talking to a North American customer, not surprisingly, it's all about protect the truck, protect the SUVs. Very -- you sort of know that going in, that's going to be their focus, and that's how you sort of direct supply. What was very interesting in Q1 in Europe, in particular, and to some extent in China, although China didn't have a lot of manufacturing disruption related to chips. In Q1, when there was a constraint issue though, it was very much protect the EVs, particularly in Europe. They were very focused on prioritizing -- when they had chips, prioritizing EV production over internal combustion production. And I think that just speaks to one more data point around how much the industry has really rounded the corner on the focus on electric vehicles. And we found that to be interesting. And in addition to the great performance on the bookings and revenue side, just another strong proof point of where this was heading from an electric vehicle perspective. Elena, do you -- would you mind going through win rates real quick?

Sure. Happy to. So when you look at Q1, obviously, win rates continue to track ahead of both last year and actually our 3-year average. They were just north of 60% in terms of overall win rates. That obviously translates into very strong incumbent win rates, close to about 80% and north of 50% on conquest wins. Where we've specifically focused in on high voltage, both high-voltage content on electrified vehicle platforms as well as the low-voltage content going into battery electric vehicles. We continue to see win rates there that are more, I'd say, in line with that 60% to 70% rate, that we've talked about previously.

Perfect. That is really helpful. The point, and I'll come back to bookings in a moment. But just on the point of EVs and just protecting the volume there. Maybe just remind us of what Aptiv CPV range looks like today with emerging EV automakers, and maybe you expect to be the next couple of years. Anything you can provide there in terms of an update would be helpful.

Go ahead, Elena.

Yes, happy to. So I think we talk a lot about our core value proposition with customers really helping OEMs achieve lighter, smaller much more durable electrification solutions. And we do that through our complete portfolio. So within Signal and Power Solutions that includes electrical distribution systems, connection systems, cable management. And so when we talk broadly about the content per vehicle opportunity, we start with a midsized passenger vehicle with average internal combustion engine that's got a TAM, if you will, of about $500 per vehicle. When you layer on top that secondary system that's got high-voltage wiring, high-voltage connectors, battery pack, bus bars. Some of our expansion into power distribution boxes, battery disconnect units as well as chargers. And then deduct a small amount for the wiring that would effectively have gone to the engines or the engine harness. You're well north of $1,000 for both plug-in hybrid vehicles, full battery electric vehicles. And that would -- that TAM, when you look at someone like Tesla, who we've gained considerable share of wallet with as we've grown from small levels on the initial Model S to now probably close to about 60% share of wallet on the Model 3 and Model Y applications. That sort of -- that tracks to about $600 of content per vehicle on those vehicles.

Perfect. That's really helpful. And one question we received after the quarter was on the active safety, quoting and booking, how that's progressing thus far in '21. How is the overall kind of level of industry bookings relative to expectations? And are you targeting any large opportunities out there this year, particularly as we see more automakers sort of move to these L2+ platforms that can be upgraded over time? Just kind of curious what you're seeing -- what you've seen and maybe what you expect to see there on active safety this year?

Yes. I'll start then, and Elena, feel free to jump in. So obviously, we've had strong active safety bookings the last couple of years. Q4 last year was particularly strong. Big awards, as you know, Itay, are somewhat episodic, right? So they're -- I would say, generally speaking, this year, active safety bookings are weighted more towards the back half of the year. That's just sort of the cycle. I think what we've seen, which isn't really a surprise, just given the amount of sort of full system awards we've had over past couple of years, the -- we're in full launch mode on those, right. 5 large satellite active -- satellite architecture systems that are going to grow -- within the OEs, are going to grow across all the platforms. So fewer of those now, just given the amount of awards we've had, but the ones that are there are more towards the back half of the year. We are seeing what we call a lot of extension awards where even systems that aren't the satellite architecture like Audi zFAS, Audi or the VW Group, Porsche, are pulling the zFAS into other platforms. So for instance, recently, about a year ago, a little -- probably a little more than a year ago now, had that system pulled into the Porsche Taycan. So not an original satellite architecture system, but a system that works very well. Customers are very comfortable with. So we're seeing extension awards into other platforms. So -- which, quite honestly, is good from a customer perspective because it's a known system. From our perspective, the initial engineering investment has been made. It's obviously, as you know, much less expensive to take a system and move it from platform A to platform B than to develop a complete brand-new system for platform B. So seeing a lot of those awards as well, which, again, we view as a positive because it's driving this sort of -- continues to drive this L2 penetration across multiple platforms. Elena, anything I missed there?

No, all set.

Perfect. That's really helpful. One of the -- I guess it's a pull case for Aptiv, but there's sort of thought that as you see this wave of new EV entrants that it could accelerate the speed by which your traditional customers might adopt advanced technologies, whether it's L2+ or advanced architectures like SVA, which we'll talk about in a bit as well. Are you seeing signs of that? I mean, is there kind of more activity? Maybe it's not in book yet. I mean, maybe it's more on the early stage activity, but are you seeing your core customers sort of feel the pressure, to move more quickly? And if so, does that provide Aptiv with more opportunity over time?

Yes. I think we are. I think there's some of it showing up today. There's more to come. I think Kevin made some comments specifically on the earnings call around just despite the chip issues and sort of having -- still being obviously aware of COVID in the environment, but having been through the pandemic last year. The long-term trends around technology in the vehicle, whether it's electrification, additional safety, digitalization in the cockpit, overall software content, those trends, we actually see accelerating. And I do think the pull forward of EV, which, by some measures, is -- there's sort of ramp we're seeing today in EVs, Itay. And if you talked to us 2 or 3 years ago, to us, that was sort of a 2025 and beyond ramp, right? It's really been pulled forward sort of 3 or 4 years. That's generating a couple of things. It's, one, we're now dealing with sort of this hybrid architecture, where sort of one more time the industry is trying to put a lot of technology into the existing architectures because they haven't had time for some of these first generation BEVs to go to a complete new architecture. So that's creating some opportunities for us today to help our customers make sure everything gets packaged in there and how best to use existing technologies. I referenced zFAS going into the Taycan before, right? That's a legacy -- from an architecture perspective, very capable, but a legacy active safety system. And that's what I mean the Taycan's got a bit more of a, what I call, sort of a hybrid, not a pure next-generation EV architecture. So there's opportunities to put more tech in the vehicles today. The EV regardless of model size, whether it's high-end performance or the ID4, there's an expectation. I think Tesla helps set this bar that they're technology-rich vehicles. The consumer expects them to do a lot. If you have an EV and you live with one, the connectivity is a little bit more important than with an internal combustion platform, right? Just the where are you in the charge cycle? Where is the nearest charger? You're away from the car, how fast is it charging? There's inherently a much more pertinent use of connectivity with an EV. And I think folks have realized that. They're not at -- there's certainly not where Tesla is today, but they understand they need to get there. I think the other benefit on top of that is the industry really looking itself in the mirror and say, "boy, we really got to get to SVA by 2025 and beyond, because the more and more -- we really are hitting the limits. And we knew we were hitting on this before, but there's this pull forward of meaningful electrification has really driven that home again." And I think we're seeing a lot of focus on Zonal Controller Awards. People starting to consolidate big, large numbers of domains into zones ahead of smart vehicle architecture. There's a few domains that are being added for BEVs like the battery management system. So how do you start incorporating those. And I think eventually, that leads not only to SVA, but the disconnection of some of this hardware and software where you get to really large domain controllers that handle multiple functions and run a number of different software applications. And there's an increase in demand, and we see the OEs investing in this as well to make sure they're ready for it. But you're seeing a lot of desire to start to get common applications and things like middleware into the vehicle that can handle all this increased functionality. So I -- from our perspective, and Kevin said it as well on the call, very positive macro trends in our view.

That's super helpful, Joe. And maybe kind of just to even dig a little bit more into that. This evolution of domain to zone to SVA. So is the main driver there really kind of to what you were saying around limitations as automakers want more functionality. And to the extent that, that is the case, sort of what are some of the things that would drive an OEM from a typical domain controller to zone and SVA? Like is this sort of level 4 personal vehicles. GM's talked about hands-free driving in 90-plus percent of all scenarios as their target. And kind of maybe talk about that progression and how you see that?

Yes. I actually don't think you need -- because we're saying, and Elena can go in a little more detail on zonal. We're actually seeing it now. So we're not -- you don't have to wait for sort of level 4 or level 4 functionality to get there, right? Really, what we're dealing with now is increased needs for compute power, increased needs from memory, starting to see opportunities. And I think the BlueCruise functionality on the F-150, which will be an OTA updatable, sort of turn on, turn off functionality that's going to sit on top of our active safety system. I mean that's a -- and that's growing on the F-150. That's a really big deal, right? And that's not necessary at level 4. That's just a lot more compute, a lot more memory, a lot more signal and power required in the vehicle. So you're seeing needs for more horsepower from a compute perspective. And what's been true in this industry now for the past 5, 6, 7 years is one of the ways to not only pay for but make more power in the vehicle more efficient is to start to centralize compute, right? It's what we saw with active -- the original active safety systems on the domain consolidation into multi-domain controllers. So it's a continuum along that same line. It's a precursor to SVA. I think there's also a practical element. And there's analogies to this in the enterprise over time, right, where if you're going to go from -- if your goal is to go from 100 computes to -- over the next 10 years to 5 or 10 or maybe even fewer in the vehicle. You're best to start consolidating now because flipping the switch from 100 to 10 is going to be really hard. So how do you start to wean down? How do you start to centralize? And we think zonal is really the next step in sort of that run into SVA. And it's great because it helps customers get there. It helps set the roadmap, but it's also -- it's revenue today for us, right? It's not necessarily waiting for SVA to be there in 2025, 2026. Elena, do you want to touch on zonal a little more detail, real quick?

Yes. So I think you touched on a lot of the overall benefits, including equivocal exploring time to market and lowering our total cost of ownership. We'll talk a little bit at the high-voltage teach-in about simplified manufacturing and assembly processes as well. But the path to SVA, really, we've talked about it before, starting with domain centralization. We have a number of those going on around the world, comprised of active safety domain centralization, the user experience domains as well. So that includes our Satellite Architecture programs with 5 OEMs that will be spread across more than 10 million vehicles over the next few years. The next step we've talked about is this shift to zone control, so breaking the vehicle up into zones. We have very complex wiring harnesses. And while we, as a company, sort of love having all that complexity, the reality is we talked about, we can't add any additional content, add more length to those harnesses. So breaking the vehicle up into zones and effectively extracting the input-output from those compute boxes, allows for the much simpler wiring harness design. And all that electrical function that's now in that zone is centralized into a zone controller. We call that the Power Data Center. What's important about that is we think the zone control will be about $1 billion market in 2025, growing to about a $5 billion market by 2030, where you see a lot of that up-integration of body functionality, different body ECUs into these zones. That's where Aptiv will benefit from a content aggregation perspective. We've talked about our first sale controller, our first to market, we'll launch in 2022. And then we are in very active discussions currently with a number of other customers looking at different zone control programs. So we're seeing a lot of traction there. The final step, we talk about that, that allows for further consolidation of the compute. So you have your ADAS compute, your infotainment user experience. Those ultimately then get further up-integrated into very high-performance compute modules or clusters. That could vary depending upon the vehicle between 1, 2, 3 or 4. And that's really the next wave of centralization that's going to start to take hold in 2025 and beyond.

Itay, one thing to add there because I think Elena sort of touched on it quickly. One of the advantages we have as Aptiv over the next couple of years. And this was very deliberate on our part in terms of portfolio planning action and both SPS and ASUX is, I think we're very well positioned not only to take advantage of the new content types. High voltage, zonal controllers, continued growth in active safety. But we've also positioned the business in such a way that we really have very little decremental content coming as a result of these ships. So on the high-voltage side, our only sort of SPS content really tied to internal combustion engines is right around -- it's actually a little less, but right around $200 of content for what they call engine harness, the electrical connectivity to the internal combustion engine. And all the content numbers that we talk about for high-voltage exclude our net, that loss wire harness, right? So very little decremental. And then as you move into zonal controllers, we have the advantage of -- over the last number of years, we've really pivoted the business to focus on large domain controllers. We still do some body and security controllers, but we're really not big into these -- the hundreds of small controllers that go into the vehicle. So as those get consolidated into zonal controllers. Our view is we pick up our fair share of content as the functionality of those controllers come into the large zonal controllers but we're not losing anything. There's no decremental content as those smaller controllers come away. And we've worked very hard over the last couple of years to position the business that way. And I think that's in part what you're seeing in the outgrowth, that's in part what you're seeing in the strong product line growth is there's just no real meaningful headwind that's netting against the tailwinds from some of these new products.

Yes, absolutely. That's a very important point, no, thanks for that, Joe. Maybe 2 questions kind of relating to -- one thing we've written about in our work is the idea that as the technology converges with economy and electric, the automakers will -- and of course, connectivity, automakers will start to look to realize more of the lifetime revenue of the vehicle today. They realize, we think, far less than half of lifetime vehicle revenue. And brings up kind of 2 questions for Aptiv. And one I think, Joe, you touched upon BlueCruise and these new L2+ systems that are being sold partially as subscriptions. I'm curious whether those types of subscription go-to-markets by your customers create economics for Aptiv beyond the initial content per vehicle. And then more bigger picture going back to SVAs, as we think about automakers, thinking about the second and third owner of the vehicle and being able to upgrade hardware throughout the life of the vehicle, to retain value and to offer more features. Is that part of the SVA story to enable automakers to realize more lifetime revenue opportunity as well?

Yes. Listen, I think from a revenue opportunity perspective, something like BlueCruise, at the moment, is still a vehicle sort of point-of-sale opportunity for us, right, because it's more content to go in to support that, which is good. We're yet not participating in ongoing revenue streams. But quite honestly, what we're excited about, particularly on a platform like the F-150 is that a large North American OE putting a pipe like that, a connection pipe into that vehicle is the start for how the industry creates those opportunities. And we expect over time to be able to participate in those. I think the real question around SVA, which is in terms of upgradable features, I think the first thing on upgradable hardware -- I think the first things this industry needs to work through, and I know this is hard, and we spend a lot of time with customers thinking through this, is providing ample headroom from day 1 for a reasonable sort of range of upgrades. I think -- and by that, the original compute has enough memory, has enough power for several generations of upgrades. I think that's a better place to start or we think that's a better place to start than the upgradable hardware in the future because that does mean you got to touch the vehicle. That does mean the vehicle has to come back in. And really, to us, the trade-off would be, is there enough -- can you work through generating enough incremental revenue from the upgrades of the user experience system or adding additional features on the original upsized hardware, is that a more efficient way to do it? Is that a more economical way to do it? So that, for us, at the moment, tends to be the concentration -- where the concentration is of how to justify putting in bigger hardware today and going after the revenue streams that are related to that versus what could be a fairly complex set of hardware upgrades over time. So I'm not ruling out hardware upgrades over time, but our initial focus is really put as much as you can into the first one to allow you to do -- to make money off of over-the-air upgrades, right? Because that's a far more economical model.

Yes, absolutely. That's a good point. And it kind of brings us to a question that we and I'm sure you guys often get as well around this debate about how much of the software the automakers want to do in-house. And of course, on one hand, we know the software content is increasing. We know Aptiv can be flexible in what you're offering to your automaker customers in terms of software stacks. But how pivotal is this in-source/outsource equation in achieving your kind of 6 to 8 points of GOM -- revenue GOM over time? In other words, what's sort of the base case that you're assuming in that outgrowth in terms of what the outsource kind of mix will be?

Yes, that's a really good question. That's actually a good way to ask that question. So our view -- you talk about whether it's 20 or 25 global OEs, right, or the big OEs that to us are the sort of the core customer base. We have a group of customers, as they do it today, that are going to invest, that can afford to invest in software resources. They're going to invest. And they're going to do 2 things with those resources. One, they're going to make sure they really understand the full suite of software in their vehicles that they become experts in software at the vehicle architecture level, because really, the OEs are the only ones that can see across all the software that goes into a vehicle. And two, and those will tend to be -- these OEs are going to be the maybe the higher end folks, the larger folks, the folks that are very focused on brands. And they're also going to want to develop those elements of software that they think support the brand best. So in a lot of cases -- and we had this all the way back. And this is -- I know they're talking about this now, and there's a little bit of -- there's a lot of discussion in the industry about this now. And I think some of it is sort of this endless sort of pursuit of Tesla to some extent, which I think is a good thing. I'm not being critical, but they're talking about it a lot, but we have this all the way back on zFAS and during the development of zFAS between 2012 and 2015. And zFAS has Audi developed software on it. It's around the human machine interface and sort of that -- those parts of the system behavior that Audi thought was important to their brand. And we see that from other OEs as well. Those get incorporated into our system. We're the integrator of that as well as we integrate other software applications. And I think that -- so that group of customers, that's what we're going to see going forward. And that's a business model that, in a large part, we have today, that works very well. I think you're going to have customers that try it and probably aren't as successful as others. And then you're going to have another large group of customers that just simply can't afford that level of investment, and they're going to continue to rely on the strong software Tier 1s like ourselves. And there are a couple of others that are very good at software as well. So I think -- and that's sort of a consistent landscape. If you go back to years over powertrain or other important aspects of the vehicle, there are customers that became powertrain experts. They still relied on the supply base or maybe used it in a different way. So I think that -- I think that's how this is going to shape out. I don't see a situation where you go 5 or 10 years down the road and the top 20 OEs are -- all have their own software stacks. And the top 20 OEs have all sort of gotten their own closed systems, right? There's going to be open architecture. It's just -- it's the way software works better. I think there's been a number of industries that have proved that over time. There's going to be a cost and investment dynamic to it. And I think you're going to have open systems. You're going to have some very OE-specific applications for certain OEs. But I think that's our expectation. That's really how we're thinking about it from an outgrowth perspective. And that's to date how we've seen it play out.

Awesome. That's really helpful, Joe. Thank you for all that detail. A couple of minutes to go or maybe about 5 minutes left, on Motional. Obviously, a lot of focus capital markets on AV and AV Networks. And maybe just to kick things off on Motional. What are some of the milestones? Just maybe remind us on kind of what Motional is expected to achieve this year and even into next year, ahead of the 2023 deployment?

Yes. There's a couple of big ones. The first big one for this year, they've actually already ticked the box off -- ticked the box on, which was getting permission to -- and they went through a pretty rigorous testing process for this, getting permission in the state of Nevada to drive with no one in the front seat, completely driverless on public roads. That was achieved early in Q1. Kevin and I were out there probably 6 weeks ago now. Driving in those vehicles on public roads, I'd call it dense suburban-type traffic situations. Not -- we're not on the strip yet with those vehicles. But dense suburban environment. It's the Pacifica, which is the first generation, fully autonomous, no driver, no safety engineer in the car. Next, we'll obviously continue to work on that model, but they're going to accelerate or have accelerated the planned move to the generation 2, which will be on a battery electric Hyundai vehicle. Those cars are starting to come in. We'll have -- we've got over 50 of them now in the U.S. We've left a couple in Asia as well for development there. But we'll start to get those cars on the road from a testing perspective, work through those. And that's another milestone that I'd call sort of early 2022. And then over the course of 2022, it's getting that platform ready for commercial delivery to folks like Lyft, which we expect to do in 2023. We'll pick a second city with Lyft. And we'll deploy those driverless vehicles, not only in Vegas, but in at least 1 more city with Lyft in 2023. And we'd expect all along the way here, particularly as you get towards the end of this year into next, we would expect additional commercial announcements coming out of Motional as well. They've done a nice job. They're -- the development efforts for them were -- COVID last year could have really slowed them if they let it. They did a really good job of working around COVID. When you think about developing those cars, it's a lot of people and cars together. It's a lot of people and the tech centers together working on the cars, that's not necessarily the type of development that's easy because you physically need to touch the car, it's easy to do on a remote basis. So they did a nice job of hitting their milestones, even managing through COVID.

Absolutely. That's helpful. And then as we think about -- we've done some modeling on this ourselves in the last few months, and it's really an interesting exercise and it always is. But any color on kind of the high level unit economics you'd expect with Lyft or maybe even future partners. Would Motional be paid on a per revenue mile basis, your cost of sales, just the fleet management? Any thoughts on maybe how this might work? And sort of any tips.

Yes. I won't give specifics, because they're in a number of commercial development -- commercial discussions now. And obviously, don't want to tilt that one way or the other. But as we've talked about in the past, even before, we formed a JV when it was fully part of Aptiv, we think that business is going to have multiple revenue streams. They'll have an initial point of sale. That could be the full vehicle from Hyundai. We're also -- they'll also be capable of selling the software stack itself, just the software itself, if someone wanted to put that in another vehicle. And one of the nice things Motional has done over the past couple of years is they've demonstrated their ability to work on a number of vehicle platforms, right? There was the Audis, then the BMWs, now the Pacificas, it will be the Hyundai. So they've gotten a good -- they have a good skill set at porting that system across multiple platforms. So there'll be the initial revenue sale. We would expect there to be ongoing software-related revenues, upgrades, maintenance, further development, that type of thing. And then, listen, they've certainly, particularly -- providing over 120,000 rides now with Lyft on the Las Vegas Strip, they've gotten pretty good at some of the fleet management, not the ride-hail management, not picking people up, dropping them off. But how best to position vehicles, how best to pull vehicles off the road for periodic maintenance while maintaining high levels of fleet utilization, those types of things. And they feel very strongly that those capabilities are something that are marketable to fleet operators as well.

Terrific. Any we're past our time. I'll sneak one super quick one in on this topic as well. How well represented Aptiv's content on the Motional kind of future -- the Gen 2 vehicle, the EV vehicle that you'll expect to deploy?

Yes. Listen, we obviously work very closely with them. It's, I think, the level of content that you would certainly expect us to have, things like the 4D imaging radar will be ours, the sensor fusion, those types of things. Motional took some of that with them. There's cross licensing. We continue to support them. And quite honestly, Hyundai has done a great job of coming in and really augmenting and supporting Motional's sort of vehicle-level engineering. And particularly which I think could have been a challenge in another circumstance, but just moving from ICE to working with BEV, there's obviously some nuances there. And having the technical expertise of Hyundai on the electric vehicle side has been incredibly helpful.

Awesome. Well, thank you so much. I know we're out of time. This has been really, really helpful, Joe. Thank you so much for participating. Always learn a lot from these sessions. And we thank Aptiv again for participating in the symposium, and we can go ahead and conclude the session. Thanks again, everybody.

Great. Thank you, Itay. take care.