AAH #806 - AI Compute Platforms and ADAS Tech

0:00AutoLine After Hours is brought to you by Bridgestone Tires, solutions for your journey. And by BorgWarner. The automotive industry continues to evolve and so do the opportunities to define it. BorgWarner, one of the world's most admired companies, gets partners where they need to go. Let's do something big together.

0:22Gary Vasilash: Hey everybody, thanks for joining us on AutoLine After Hours. Shout out to Borg and Bridgestone for sponsoring the show. We really appreciate that. But now I've got to say, hey, Gary, how are you doing? I'm doing great, John, and you? I'm doing great now. Yeah. And I say now because this morning we were without power because this area got hit by some pretty severe storms. So I'm glad we're able to do the show today. Yeah, we've gone from no rain to a deluge. Yeah, right. Exactly right. So we've got to let everybody know we've got Chris Prediger here. He's the regional president for Bosch Mobility for Cross-Domain Computing Solutions.

1:02Gary Vasilash: I think I got that right. Yes. Did I, Chris? Excellent. And we got Nora Eckert back from Reuters. And it's great to have you here, too. Yeah, good to be here. Yeah. Thank you for having me. So when I see this, Chris, saying cross-domain computing solutions, we're talking computers and cars here. Yes. And I'm thinking we're also talking software-defined vehicles and zonal centralized computing. All that comes in with what you're doing. Yes. So this division is roughly five to six years old where Bosch started to prepare for software-defined vehicle for cross-domain. Not have all the domains separated, but really merge the domains together in one organization.

1:48As you mentioned, we do the compute to power the vehicles for ADAS, for cockpit, for infotainment, for body functions. And we do also the sensor set around it and, of course, the software.

1:59Gary Vasilash: You know, I never thought of that before, that legacy suppliers that had compute in each and every...

2:14that you would have to reorganize

2:16Gary Vasilash: how you operate within the company. Absolutely. And we have to be much more flexible now. So before, like you said, we had our classical tier one with our closed system ECUs, black box. And now we have to be very open for cross-domain, but also for different business models. For some, we supply the hardware only, for some, the software only, or some, still the closed system. So what if there's a situation where the customer chooses you for one domain and another supplier for another domain? How do you cross those domains? It happens quite frequently. So everybody has to really be open for it and collaborate together.

3:00So we even have cases where we supply a portion of the hardware and then the software is coming from a competitor. This decade is really changing or evolving over the last years.

3:14Gary Vasilash: Chris, where do we stand right now in terms of this evolution? Because certain automakers have said, okay, all you suppliers, don't worry about writing the software. We're going to write all of the software. Where does that really stand right now? And I ask you because of this. When it comes to anything that's FMVSS related, safety related. If it's a Six Sigma system, you don't want to mess with it, right? You don't want to mess with it at all. So I'm just wondering, where would you describe the industry and its path towards going away from the old individual domain more to this integrated compute platform?

3:55I would say it really depends on the OEM. We see the OEMs without legacy, the more derivians, the Tesla, they're already there. Then we have the more traditional ones who are progressing to be there with some hiccups in between. I think we saw it last year before it with a pivot in the architecture. Others have done that too. And then we see the Japanese OEMs who are a little bit more traditional. They're still a little bit more hesitant. They are fast followers, but they're coming behind. Or not behind, but after the US OEMs, for example. So really different stages. And for us as a tier one, it's very important to be flexible to support all stages of development in these architectures.

4:43Yeah, I'm curious about how your relationships with the OEMs have evolved, especially as you're facing this incredible compliance burden, which I know there's already an incredible compliance burden in autos. but just additionally with, you know, banning Chinese software and vehicles and things like that, you have to, suppliers need to be able to prove to automakers that they're not going to bring them out of compliance. Do you think you're sharing more with automakers than you used to or how is that relationship? Absolutely, absolutely. And we have to be and it's vice versa. So we have some OEMs where we collaborate where they have a software portion, we have a software portion, we are the integrator.

5:21So basically the OEM is our software supplier. So they have to be transparent to us, but we have to be open to them and very transparent. In big organizations with old processes, this is always a hurdle. You see that in the OEMs, we see that in the tier bonds as well, but that's really evolving over the last years that we have to be so transparent. At the same time, protect your IP, of course. So it's a fine line always, but it's necessary. Without collaboration and openness and transparency, it doesn't work. So let me ask you, I think when many people hear Bosch, they think about things like fuel injectors, okay?

6:00Mechanical devices. That's old guys like that, period. Or dyswhasers. Yes. Well, they probably think about dishwashers and refrigerators and things like that. But the point is that historically, the company Bosch has been very heavily involved in, let's say, mechanical engineering. Yes. Yes. How has the company transitioned to being one where there can be cross-domain computing solutions, as part of your title? The investment we do in software-defined vehicles in AI is really tremendous over the last years. We had a clear direction that we have to change from a pure mechanical hardware-driven company to hardware and software company.

6:51I don't think it's feasible as like a hardware company or is coming from that area to shift completely to software. I think software doesn't work without hardware. So for us, it's important to have the end user in mind, the software in mind, and then the hardware is there as an enabler. It's a tremendous mindset shift of the people, of our competency, what we need, of the mindset and of the entire organization. And you have to trust that even as automakers say they can do this themselves that you're still going to be needed and position yourself to where you can offer that service even if you're being told that you don't need to.

7:29It's a fine line. You don't hope that they fail. Right. But also you want to have your part of the business too. And that's why I think it's so essential to be flexible. I mentioned that before with business models, right? So we shouldn't be too proud. Tier one is like, we have only closed systems. No, we really have to change and be open, transparent, and adapt to the need of the OEM and the end customer.

7:54Gary Vasilash: Talk a bit more about that. I got to believe that for a legacy automaker or supplier, being able to make that kind of change is not easy. Not easy? That's correct. So give us some of your thoughts. I mean, you're the president. You know, what kind of things do you have to do from a leadership standpoint? I mean, you can talk to people and say, we need to do this, but there needs to be systems and places that encourage them to do it. Yes, I mean, process landscape needs to change, mindset needs to change, but it's not about educating the people, talking to the people in the organization to do this.

8:36And for me, it's always important. I always try to come from the end user, that they understand what is the end user value in the end, what is then also the OAM looking for, and how we then can approach it. You guys recently introduced something called an AI-powered cockpit. Yes. What is it, and why does Nora want one in her next cockpit? Yeah, I'm a tough customer. I feel like I need nothing. Just give me the slate. No, we want to evolve the cockpit of the future to be more agentic and really user-centric. I mean, we had attempts in the past where OEMs try to have subscription models in there for existing functions, which were there before heated seats.

9:21I think it's a very prominent example, which didn't work out. I think with our AR port cockpit solution, we really try to enhance the end user value in the vehicle with AI. at the same time make it very safe. So for example, we talked earlier before the show about parking features. We have amazing ADAS parking features, but a lot of users don't use it because they don't know how to activate it. That's where an iPod cockpit would come in place. The vehicle detects in which situation you're in based on the speed, you're looking for a parking lot and automatically would suggest, hey, would you like to use that feature?

9:59And you can take hands off and it's automatically parking for you. or take another example we are collaborating with Microsoft to get productivity into the vehicle so my morning starts pretty early usually it's take 30 when I'm leaving the house and I have my first meeting getting to the car getting to my call I have a short commute 20 minutes but usually when I arrive at the parking lot and I'm in a meeting it's like how did I get here I don't know if you can relate to that but oh yes very much of course you were driving and paying attention but it's like, oh wow, that passed so fast. Also with AI Port cockpit, we want to enable different domains, right?

10:35So you want to enable productivity in the vehicle. You want to be able to join your calls via Teams. And at the same time, you want to make it very safe because I might be distracted. So the vehicle will detect you're in a call and will automatically switch on, and keep assist, ACC, that you're safe without you having to activate it. Really intuitive and agentic. And are you hearing from customers that this is something, like what gives you the confidence that this is something customers will actually want and use? Because it's so tough when they don't know that feature's there, obviously, to justify it.

11:10But yeah, what are you seeing that's giving you confidence? The end user studies which we have done and seen is that most of the people are not aware of the great features and technology we put into the vehicles, that they need basically not trying to figure it out by themselves, but really system suggesting, helping them that it's like an easy transition and really easy to use. That will be the key. That's the feedback that we are getting, and that's about our AI-powered cockpit solution. We are really trying to make that happen for the end customer. That the technology we bring into the vehicle really adds value to end user not only is there for technology sakes.

11:50Gary Vasilash: How do you determine what your AI cockpit should do? Because I've always argued, if my phone can do it, I don't need my car to do it. Correct. But there's much more than what the phone currently can do. So if you think about a little bit further, so every vehicle by now is equipped with sensors, with cameras. But the vehicle can see and sense your phone and not. So, for example, if you drive in winter, the camera or the radars can detect there's snow on the street, right, and can automatically adjust your driving behavior or put in your snow mode what you actually do mechanically. Your phone can't do that.

12:33I like the example that you guys presented that basically you just say out loud, I feel cold, and the heat in your heated seat are adjusted. It's more personalized and context-driven. Also, your phone can't do that. And you're using everything what is equipped in the vehicle. So you have interior cameras, your voice, and the vehicle can detect your mood, what you want, get the context, and help you adjust your vehicle accordingly. So give us some examples of mood alteration. Okay. You're in a very happy mood, right? you get into your car and automatically, for example, your favorite music could come on or you come from the gym or you're going to the gym.

13:17The vehicle knows it because it's in the map program where you're going and could put in motivational music. Or for example, it sees you're sweating and then it's like, okay, turn on the AC automatically. Are you finding that any customers are sort of squeamish about that level of interaction with their vehicle or like that level of intelligence? I know it's a big divide, you know, So it depends on probably a lot of different graphics. I mean, even I have associates working on it. It's like, okay, I would never, you know, it's kind of weird. And what we have seen, and also again, with looking at the end user, that it's their cultural differences.

13:51So for example, in Europe, it's maybe more direct. You want to give commands, right? Park here. And if you're then in more Asian culture in Japan, for example, it's very gentle, it's suggestive and so on. So we really have to adapt to the end user and to the culture. in order to make it work.

14:11Gary Vasilash: How do you get that regional difference? I mean, you send out engineers into the field or how do you pick up on it? I mean, some of these things are very subtle that you're talking about. Yes. So I think as Bosch, we are a super global company and we have footprints all over the globe in Japan, in China, in Korea, in Europe, and of course in North America. So with that, we have teams with the expertise everywhere in the world who can serve the regional needs of this market. So that's really our strengths, our global footprint. And of course, we are very connected. So I have very early calls with my colleagues in Japan because for Japanese OEMs, this is a big market, right?

14:51And we want to be connected. And to serve our Japanese customers for the North American market, I need to understand the culture. And that's why we are very well connected, interconnected. I want to get back to this work you're doing with Microsoft 365 and the ability to join a Teams meeting while you're doing your morning commute. A lot of people say, I don't want anything like that. I want to have time to just drive. I don't want to be working. Absolutely. If that is possible, I would love to do that too. Reality is different now. But also if you think ahead, you talk about autonomous driving.

15:30At the moment you're busy driving. What are you doing in the future in a level three, level four, hands-free, eyes-off car. You can maximize or utilize really the time in the car for not driving, but to be productive. Microsoft is just one use case to join a Teams meeting, but you could also, like, after this, I would like to take notes, but I have to drive, right? I could utilize that with voice and dictate and basically my notes, what I want to take from this discussion. Or sleep. That's another option. Or sleep, isn't it? Yeah, I'll put that. That's a luxury. Yeah. Yeah.

16:05Gary Vasilash: Or like you say, Nora, you get a slate, you don't worry about any. Yeah, you're just. Yeah. But I want to go back to an issue that Nora raised that I think is really interesting. So the U.S. is banning all Chinese hardware and software that has anything to do with connectivity. Yes. Chris, where do you see this going? It seems to me there's going to be two platforms automotive wise in the world. one for China, definitely, one for North America, definitely. But how is that going to work out in the rest of the world? Is Europe going to be with the U.S. and with China or not? Or how do you see it going?

16:45It's a very good question. So we ask ourselves that question as well. So we definitely see the need to have local for local for China. And then we usually then focus for the rest of the world because we see that also, or we think that in Europe, some kind of protectionism, what we have with this rulemaking in the U.S., will come for Europe as well. That's why we rather focus on really local for local to serve the Chinese market, but then having solutions for the rest of the world.

17:13Gary Vasilash: But China is exporting like crazy right now, too. So that technology is going to go throughout the global market. Yes, that is correct. And I think we can protect the U.S. market for so long, But if we do it for too long, we will only have old school technology. So we also need the competition. I think the protectionism at the moment is good to catch up. But then really we have to spend that time now to really catch up and don't lean back with like we are protected in this market. Yep. We've been preaching that on their show. Very good. So one of the companies that's based here that you guys do a whole lot of work with is NVIDIA.

17:54How does that relationship work? I mean, what does NVIDIA bring to the party that Bosch doesn't have, and what does Bosch bring to the party? Yes, absolutely. NVIDIA is one of them. We collaborate with Microsoft, I mentioned before. Qualcomm, too, you're with, right? Absolutely. Qualcomm is one of our biggest partners on chip suppliers, yes. So I think it's very essential that you're not locked in with one and that you don't do it yourself, because we as Bosch, we do great high-performance computers, we do great software. But for example, should we develop our own large language model or voice assistant?

18:29They're better out there and other companies who can do that better. So that's why we partner with NVIDIA, with Microsoft, with Google, with Qualcomm, with other SOC vendors like Renaissance, MediaTek, Telechips, you name them, really to have the best suitable option for the OEM and the best, really best value for the end user. So do these companies recognize Bosch when you knock on the door, as it were? Some more than others, but usually yes. Also, not new to the automotive, but everybody has its value. We have the OEM access. They have OEM access. Together, also together with the OEM, we can bring the best value.

19:13Gary Vasilash: So your AI cockpit, that's with Microsoft? Also with Microsoft. yes we use portions of NVIDIA but we are open it really depends on the ecosystem of the OEM if you look at maps some are with TomTom some are with Google and we are open we have an open ecosystem for that and we are the orchestrator basically getting the pieces together and based on the use case then picking the right application for it Gen.ai is evolving so quickly you've gone with Microsoft take me through the decision process a little bit because there could be a breakthrough coming any day now that I'm not going to say it's going to obsolete what they got, but it's going to be better.

20:01Yes, I mean, you always have to. I mean, if you look at the past years, and especially with AI and how fast everything evolving, but you just have to be on the pulse on the market. And Microsoft is a great partner with also a long legacy like us. For us, it's actually pretty good. So we understand each other very well. Like I said, from a legacy, from a process point of view, with a similar culture, it's working quite well. I think that the fast development on AI, you can really see on an example for ADAS. You look at driving stacks of the past, they were more like rule-based. If this situation happens, the car will do this.

20:39And now within one year, two years, everything is AI-based and end-to-end. So really have to be the pulse on the market and be able to pivot and willing to pivot if you see there is a faster development. You know, it seemed to me, if we go back two or three years, that there was some question about the relevance of suppliers, that basically the OEs were going to just do everything. Um, what you're describing is, you know, intense engineering and development. And it seems to me that it's resources that perhaps the OEMs don't have. I mean, is that the case with how things have turned out? Yes, I think so.

21:26Resources, competency, and then also, So do really the OEMs want to invest in every functionality of every control unit of every functionality of the vehicle? I think so that there's value add for them. And that's why I think there is still a big role for the tier ones or for the other suppliers as well also in the future. But it looks different. So as I mentioned before, with the former classical tier one closed system with a centralized architecture, it's not possible anymore. I think the suppliers who open up and are willing to change, they will have fair share of the market also in the future and an important relevance.

22:11And maybe one more factor also from an OEM point of view. develop everything in-house, what happens if you have your failure in the field? If you have your tier one, your supplier, you can also call them to help or help, or kind of as an insurance, right? So we talked about FMBSS regulations. Do you want to be responsible for everything, what is happening in the field? Or do you want to have your supply base who is helping you out and is partnering with you to solve these situations? And share blame if you have a quality issue. That's to put it more cynically, but I have a lot of blame. Yeah, share is too charitable.

22:54Sort of like a broader step back question. Of course, they're not saying it publicly, but just contacts at the D3 have signaled that they're in the software teams have signaled their concern that, you know, with tariffs and ED write downs, there's such a cost crunching effort and a focus on the core functions of the business that software is getting sidelined. Are you noticing that at all? No, not at all. I think really not the opposite. I'm with you that before everybody like software, software, software, software, but now it's more change from software to value. But I don't see a decline in software.

23:33Gary Vasilash: That's good. I think if you talk to any part of any engineering department at a car company, they'd tell you they're not getting enough resources. You're right. It's like any company. Right. That's right. I want to go back to AI for a minute, Chris. As you put Gen AI into a car, you guys have ideas of what it would do. Are you leaving the door open at all as to what the end user might be able to do? Because the customer in the car might figure out things to do that nobody at the car companies or suppliers have figured out. Yes, it's actually when we demo our AI-powered cockpit and we actually have done a point where like, okay, you take the mic and you just try it out.

24:11and what comes out of that is like amazing. The new use cases for us. So we do that a lot actually. And I always pray like, okay, does the system work if in the use case, which we didn't try out, but it's amazing what we learned with that because we did so many brainstorming ideation sessions on use cases, what we can do. But I would say nearly almost every time we have a demonstration or have an end user study, something new comes up. So really endless. And I think the involvement from in the past, everybody was looking for this use cases like before maybe the smartphone with the killer app like now like i have this app and then i will make this money uh i really compare it with with the smartphone like you have multiple apps and that makes the smartphone itself like invaluable right so um and i see the same with ai in the vehicle there will not be this ai use case which will bring the most value really to have ai in the vehicle and the ecosystem and the variety of use cases like comparable to apps this would be the game changer so so following up on that um so if we have phones and everyone has a phone right and we have these apps and we're able to do these things and you know it's it's fairly democratized to that it's across the board when you come out with this technology say the ai cockpit is this going to start early at luxury and then work its way down or do you see it being more widespread from the get-go more widespread i think certain use cases where you need more memory more compute power in the vehicle i think that's more for the high trim um for other use cases where you can use cloud computing absolutely it can be enabled in every trim level i think it's it should be yeah i think we're going to get to the the point very

26:04Gary Vasilash: very quickly that if you do not have AI, gen AI in your car that people in the car can use, you're not going to be competitive. Yes, correct. I really think that will be one of the competitive advantage who figures that out from the OEM in the future and differentiator. We're going to look back in the not too distant past and laugh that we used to worry about cup holders and you know, Sweated over all of that, you know, because the industry's moved way beyond it. I haven't. Cup holders are key. Oh, I know. I agree. They're still key. But, you know, we in the media would say, oh, my gosh, they have seven cup holders in this minivan or nine or whatever it was.

26:48Gary Vasilash: There was, in fact, I think the Pacifica had at one point, maybe it still does, more cup holders than there are seats in the van, you know, just for the convenience. That's America. Yeah. but when Chris is on one of these team meetings he may need several beverages to keep him going depending on the content of the meeting you never have to leave exactly Chris we're going to have to wrap it up at this point but thanks so much for coming on the show very interesting we're getting a little snapshot of where the interior cockpit is headed yes and if you want to stop it if you want to try it out please stop by and we can show you what the in-vehicle cockpit of the future looks like.

27:33Great.

27:33Gary Vasilash: Thank you. Thank you so much. Thanks for having me. Great. We're going to take a quick commercial break. Thanks to our sponsors. And we'll be coming back to talk all about the latest news in this crazy industry.

27:46Gary Vasilash: Making a life full of memories. One road trip at a time. That's what really matters. Bridgestone Weather Peak Tires with a 70 ,000 mile limited warranty.

28:00the automotive industry continues to evolve and so do the opportunities to define it borg warner one of the world's most admired companies gets its partners where they need to go let's do something big together

28:15Gary Vasilash: all right we're back talking all about the news or gary do you have a are you going to stump us here try to stop well i'll try to stump you okay okay so you have your choice this is either something that happened in 1967 or something that happened in 2012 on September 3rd. Okay, well. Well, I was alive for both, so I'm ready to ask me. I'll go back to that day, 67, okay. Which, which, okay, so, so both of these happened in Sweden, okay, so that, that, that. Oh, okay. Okay, so, so. I know the 67 answer already. Well, okay, then we won't use that one. Yeah. We'll move on to 2012 then. You know what's at 67?

28:56No, I don't.

28:57Gary Vasilash: When they switched from left hand to right hand side of the road. Oh, really? It was H-Day. Yeah. Hoger Tropic Omlegaden Jen. Exactly. Just like H-Day in Sweden. Yeah, that was my state. But yeah, they, you know, they used to drive on the left side of the road, like the UK and other countries. And overnight they switched to go to the other side. That's so cool you do that. Which is like crazy when you think about it. Yeah. there's been other countries who have done it but much much earlier sweden was fairly well developed obviously in 1967 so yeah just imagine that like you wake up the next morning and yeah suddenly everything's changed all right so let's move on to the 2012 one okay so this this has to do with business okay september 3rd of 2012 correct oh you want me to just say what happened So you're multiple choice.

29:49In Sweden on that day. Do we get a hint, Gary? Is it about something with EV sales? No. EV sales? Sweden's always been ahead. I don't know. Oh, Norway. What was the year again? 2012? 2012. This has to do with an automotive company.

30:07Gary Vasilash: Yeah, so. So this will make it like. So I'm going to guess. Obvious to John. That's when the Tesla Model S went on sale. Oh, that would be. Or in production. That would be an EV sales answer, to be fair.

30:22Gary Vasilash: Oh, Sean is the end of Saab, right. So is that Saab's got closed down? No. Oh, Gary, I've got nothing. Come on, I need help. I need a phone to find over there. So National Electric Vehicle Sweden bought the assets of Saab in 2012. They had thought that they were going to start making EVs, but it didn't work out. And a couple of years later, it was like, it didn't happen. So the legendary sob. Yeah. I don't stand a chance against you guys in auto trivia. The humiliation ritual. I didn't know that one. So you're okay. You got that H day one much too, much too fast. I know, that was so impressive.

31:04Gary Vasilash: So Gary said you're working on a Volkswagen story. Oh, yeah. Can you talk about that? Oh, it actually, it was just a short story that came out today. Yeah. So we were following after all of the personnel changes they announced among - This is Volkswagen of America that you're talking about. Yeah, yeah, yeah. And we need to talk about those personnel changes too, but continue with your story. Yes. Obviously, there's this broader global restructuring happening with Volkswagen. I mean, it's just a downsizing really at this point. And Friday, there's a big meeting that my colleagues in Germany are covering, but we were keeping an eye on some product announcements out of the U.S.

31:39and Volkswagen today, it was pretty vague, but they tease that they are exploring new SUVs and pickups for the U.S. market. But they said they were class B body on frame, which I think we might have some different definitions about what that might look like, because we're like, I don't, I don't know what you consider it.

31:56Gary Vasilash: A B-class car, small vehicle. Right. And they were saying it was a B-class body on frame SUV pickup for potential. I'm like, B-class body on frame pickup. So we're puzzling over that but they made that announcement yeah and besides don't they have scout coming right pickup yeah that was that was our question that body obviously different company just yeah don't be confused it's a completely different yeah so not the hugest news but we're following that it was a bit more vague than we were expecting but um yeah i mean it's just interesting thinking about all the competitors in that space especially if we're looking at smaller trucks coming out there's there's a lot so we'll be following that no details or timing or anything yeah so have you been following what's going on in Germany with Volkswagen?

32:38I have, but not as closely as my colleagues, but it seems to be a bit of a mess. Yeah.

32:43Gary Vasilash: Yeah. So we have the board of supervisors, which is the big board that includes labor and politicians. And then you have the board of management and they seem to be at each other's throats to where this company is going to go. You know, the board of management has said, we got to get rid of a hundred thousand people in Germany. Actually, the last number I saw, if I got this right, is 115 ,000. Oh, is it up to 115? Wow. Close four plants. It's in flux. But what I'm hearing, or what Reuters is reporting, or colleagues are reporting, is that what the Board of Management may do is call a special shareholder meeting.

33:28Gary Vasilash: Right. Because they're not going to get anywhere with the Board of Supervisors. Like I said, half of it, I think it's more than half at the moment. Half of it's labor, half of it's politicians. They don't want to see any jobs go away. And so maybe the management takes it to the shareholders, which is largely the Pieck and Porsche families, and then they'll approve it. Yeah, and I think that that measure has never been taken before we reported. It's an extraordinary measure. That's my understanding. Yeah, it's fascinating. So we'll know maybe more later tomorrow. We'll see what happens. Because tomorrow is the big meeting.

34:07Yeah, but it makes things at some U.S. companies look a little better. Looking at that, it's just been fascinating to follow, especially the labor side of that in Germany and the securities that these workers thought they had. And they're not being taken away. It's like everything that they took for granted in the global autos. Yeah, which I guess is the biggest story of the entire industry. I think it's interesting, you know, the changing of the head of Volkswagen of America. Yes. Jill Gruner after 22 months is going to be exiting the scene and replaced by Marco Schubert. Yeah, who was a long timer, I think.

34:49And Gruner came from Rivian, if I remember correctly. So it's just been interesting to follow that transition. And of course, with that unionization effort at that plant and how that affected the leadership changes. But yeah, the Volkswagen release was very short on Gruner exiting. He just said he's pursuing other opportunities. And yeah, there's more like institutional knowledge is winning out. I mean, do you get a sense that being the head of Volkswagen America now is basically a thankless job? Yeah. I mean, I think it's like what automaker is not a thankless job? I feel like, I don't know. But especially so, I think their market share right now, just check the motor intelligence data, I think 3 % in the U.S., which is roughly what they've been at for the last decade, maybe even a little higher.

35:37But yeah, you just wonder where they can make headway in the U.S. market. And they've talked about prioritizing hybrids and maybe leaning into these segments that they say are popular with SUVs and body-on-frame pickups. But yeah, it's a really tough task for sure.

35:54Gary Vasilash: Right, but even when you announce that today, you're looking at three years before it gets in the market. Yeah, yeah.

36:03We have a tornado coming in.

36:06Gary Vasilash: Tornado warning? Oh, do you think? Maybe. Oh, good. Everybody, we hope this show gets to the end before the electricity goes off. I want to go out on AutoLine. That's the perfect thing. That's fine. If that's my last. Here's a tip, folks. So you turn your thing off on your iPhone and yes, lo and behold, National Weather Service tornado warning in this area until 415. Well, I'll have you know that this building is tornado proof. We're in a very safe building right now because our out of line studio was originally a fire station for the city of Farmington Hills. And it was a tornado safe ground.

36:44Yeah, this thing's built like a mountain.

36:48Gary Vasilash: The show will go on. The show will go on as long as we have power. Well, hold on tight here. Speaking of small pickup trucks, the word came out that Ford plans on selling a lot of fathoms. Yes, 100 ,000 in the first year. It's interesting because when I was hearing about this, just through my sources, the number struck me immediately as... Unobtainable? Well, yeah, there's some of that. But I also thought about like, think of the T3 truck that they're playing in Tennessee. They were expecting annual output there of 150 ,000. And then the Lightning, they capped at 200 ,000 reservations. And we know reservations don't equal sales.

37:31But if anything, I was immediately struck by how it's, you know, it's lower than what they've projected. But still, I mean, that would, as my competitor, Ryan Felton reported, that would put them in Tesla territory, which would, yeah, which would be a new feat for Ford. And it does seem to indicate that they feel they can take from gas and hybrid drivers a lot more than just EVs. There's a confidence there.

37:55Gary Vasilash: And remember, a slate launching this year. And that's kind of in the same segment. I think they'll appeal to different buyers. But yeah, what do you think, Gary? I thought that was a mighty aggressive number. All right. So General Motors in 2025 sold 169 ,764 electric vehicles across the board. Yeah. Okay. So that's the Equinox, the Blazer, the Lyric, the Hummer, the Optique, the Silverado EV, the Vistik, the Sierra EV, the Escalade EV, and the Bright Drop vehicles. Okay. So word's coming out. We're going to sell 100 ,000 of these things? Yeah. Do you take these projections seriously anymore? Like you two following the industry?

38:36Because it started, I've only, you know, been following autos for the last five years, but even I am starting to feel like, okay, I've heard enough of these in the last five years that you just kind of immediately, I don't think they hold much weight.

38:46Gary Vasilash: They must have debated this. Yeah. Because they would know that all of us would look at that and go, what, what are they smoking? This is crazy. And wouldn't you come out and say, oh, geez, man, man, if we could only sell 40 ,000 of these, it would be so good. And then if you hit 100 ,000, everybody goes, these people are geniuses. Yeah. So they must have debated this very heavily. And they knew that they got burned on their projections for Lightning and for Mach-E. And so they must have something there that gives them the confidence to say, yeah, we'll sell 100 ,000 of these in the first 12 months.

39:23Gary Vasilash: Now, the follow-up question is, what's year two going to be? Right. Yeah. Yeah. And to be fair, they haven't formally announced it. So we were just I was just hearing it from sources. I think that's what the journal reported as well. So maybe that's the new strategy is to, like, you know, quietly tell people and maybe let it leak instead of having it be projected, you know, from the rooftop. But that's I didn't realize they had not. No, they didn't confirm it and they didn't announce it. And I know that there there has been like a strategic shift there after they were burned, like you said, that they don't want to go out on these numbers.

39:54But then again, if you if you're telling people that figure,

39:57Gary Vasilash: you know it's going to well you know i'd love to know where it leaked from because you know suppliers get production orders yeah and uh i wouldn't be surprised at all if this leaked out of the supplier community yeah you know somebody is saying yeah you know we've got uh orders for 100 000 for the first year but i saw uh how tai tang who used to be at ford post recently on LinkedIn, he says, look, every supplier takes the OEM's numbers and they shave automatically 15 % right off the top because automakers will go to suppliers and say, oh, we're going to sell 100 ,000 of these. So now you got to amortize all the investment and give us this piece cost based on 100 ,000.

40:42Gary Vasilash: And suppliers look at it and go, you're never going to sell that level. And so that's where all the negotiations start to come in on what's reality. Given that the Mach-E sales so far this year are down 55%. Yeah. 55. Yeah. Yeah. Well, remember, at one point it was outselling the gasoline Mustang two to one, but it came out in 2019, 2020. I mean, it's six years in the market. Every model starts the nosedive unless you refresh it and do things that make it look different. make the public think, oh, here's the latest and greatest. And they haven't done that. So that doesn't surprise me at all that it's dropped off that much.

41:31Gary Vasilash: I would think the lightning is down even worse. Well, okay, so the same year that the Mach-E came out, the Bronco came out.

41:43And Bronco sales are up 3%, which is actually good because the overall market is down. Right. So the point being that the Bronco hasn't been changed significantly since its introduction. Correct. But it's EV versus. Yeah. Right. But then, yeah, exactly. Exactly. Yeah.

42:08Gary Vasilash: Well, we'll know soon enough. Here's another crazy one. Today, Tesla doing the big media launch of the CyberTab. What do you guys think? Because my understanding is NHTSA has still not approved these things to be sold because under NHTSA's regulations, you have to have a steering wheel. You have to have pedals. And Tesla says they're not going to include those on the cyber cam. Right. So they got to be counting on NHTSA saying, OK, yeah, no, we see what you don't. We'll give you a carve out. Go ahead and start selling them. What I find interesting is that its competitor, Waymo, has said on more than one occasion that they don't think that a one-sensor vehicle can be as autonomous as it needs to be.

43:09and therefore you know in the cyber cab camera only so depending on the camera cameras plural well cameras and camera technology the um ai system as well as the massive number of miles that tesla vehicles have accumulated over the years versus waymo saying you know you need lidar you need cameras you need other sensors in order for this to work now who has more cars out on the road right now that are making money, you know, or at least getting revenue. Tesla. Yeah. Right. Well, no, I mean, in terms of robo cars, robo taxis, Waymo. Oh, yeah. Just for robo. Yeah. For robo cars. Yeah. It's just like kicking Tesla's ass.

43:57I mean. Yeah. So I don't know how successful this is going to be. Although Wall Street will just, you know, whatever he says, you know, he says tonight, they'll just, let's give them more money. Do you think that his leash is getting shorter, though? I mean, it seems like the stock hasn't been quite as great as it was. And I feel like it doesn't really move on good or bad news anymore. Do you think that's right? I need to monitor more closely.

44:26Gary Vasilash: I think there's a wait-and-see attitude. Yeah. You know, because there's a lot of people who question whether Tesla can pull it off. And then there's the true believers who think this is going to be the thing that kicks the stock to a new level. Kick it up, that is. What I find so interesting, I saw, we have seen multiple sightings of cyber cabs around the Outerline studio here. Oh, you have? Oh, yes. I saw one two nights ago. Okay. And, which I find interesting because I've seen people post online. Wow, that's so stupid. Tesla can only operate them in Austin in their geo-fenced area. Well, then why are they driving around Farmington, Michigan?

45:11Gary Vasilash: And I've seen other people post stuff online from around the country, pictures of cyber cabs in their areas. So I think Tesla is planning on going big with these things, just like Elon's always talked about. because if they can run around Hermington, Michigan, they can run around a whole lot of places in the country. Yeah, and if they roll them out in numbers without NHTSA approval, what's the recourse there? Oh, I don't think they can. They can't. I doubt that they could legally sell cars without NHTSA giving them some sort of dispensation for the regulations. Yeah. Yeah. Well, if anyone's going to test that.

45:57Gary Vasilash: And then who's going to insure it? I mean, is Tesla going to insure these cars? Because I could see a company like the AAA going, we need some field experience to figure out how to win. You know, what do we charge you for this? Yeah, that's such an interesting question. I know that came up a lot in our reporting on Level 3 systems too. Yeah, the whole liability question and the insurance one. And then I thought CyberCab was going to be a robo-taxi service just operated by Tesla. Yeah, but it doesn't seem... I think I got that wrong. I think the plan is, no, we're going to go retail with this thing.

46:32Yeah. So in effect, you buy one of these vehicles and then you use it as a service to make money for yourself when you're not working. Yeah. So you would use it as like, well, then I'm thinking about the Uber cuts too with all this, like just how, yeah, all these services that... Yeah, Uber laying off 10 % of its global workforce. Yeah. And I thought that's more Waymo than CyberCab, I think. Definitely. Yeah. Not to change topic, but I, yeah, I was just thinking about that more because like that whole value proposition that Uber provides and Waymo is basically taking it away and their only hope is to partner in some way with Waymo.

47:14It's just fascinating. Yeah. Yeah. I haven't been following that as closely as I should, but yeah, you do wonder for Uber and Lyft, if they can only survive for the next few years.

47:27Gary Vasilash: Well, you know, they're partnering with companies that have the autonomous technology. So it'll be really interesting to see, is Waymo going? They had a deal with Uber. They were partnering. Now that's starting to drift away, I think as they both recognize that they're competitors to each other. But, you know, Gary, you triggered a great thought. But can you do autonomous vehicles camera only? So I was talking to the CTO of an unnamed large tier one supplier recently. And he said, yeah, Tesla's argument is that you, we human beings, we have our eyes and ears and, you know, we see. And so Tesla's AV system is going to operate like human eyes do, except as the CTO pointed out, we don't see very good at night.

48:24Gary Vasilash: And we don't see very well through the fog and the rain and all this other stuff. So maybe you can use a thermal camera. But when you start getting into fog and rain and snow and things like that, you want LIDAR. And you want radar. Yeah. And you want ultrasonics. The whole suite. That's right. But I mean, camera only. I mean, this guy convinced me camera only is probably not going to work or your car will just say, sorry, I can't drive in these situations. And how how would it not work? Would it just be unsafe? Are you saying or it would just not operate? Not operate. Okay. So think about that.

49:07You have that in your driveway and you, you know, cut off your baby finger and you go, oh, I got to go to the hospital. But it's a foggy night. Sorry. Yeah. Sorry. That was very specific.

49:21Gary Vasilash: We're saying this now and Tesla is doing its big rollout today. Maybe we'll have more answers to this tomorrow. So, I mean, we're just speculating here to a degree, but I thought that CTO's point was pretty good. Yeah. And most, well, I'm thinking more of like the systems we have with level three and automakers, but most of them are banking on that. Yeah. Banking on that premise that you need LiDAR, even though it's a lot more expensive and not to mention the whole connected Chinese software and hardware. When you get into LiDAR, that gets really, yeah, really big. So, John, you mentioned AAA.

49:57AAA, as you know, tracks gasoline prices, fuel prices. And so today,$4.14 per gallon nationwide. Last year on this day, it was$3.19 a gallon. So it's up 30%. But the really astonishing thing is that the average diesel price today is$5.78. cents per gallon. Highest it ever was, was$5 and 81 cents. Okay. And that's average. So this is the average.

50:32Gary Vasilash: And I'm just interjecting here because I just got an email yesterday from one of our viewers out in California, diesel there, seven to eight bucks a gallon. Right. Yeah. So, so basically it's three cents from the highest ever. And that was in June of 22. And that was caused by the war in Ukraine when they started attacking the Russian oil fields and refineries and so on. So what does this do? I mean, so the diesel part of this is going to affect everything that we basically buy from... Everything that moves by truck and train is going to be affected. Right. That's right. um so this will go way up and then you know you think about these gasoline prices i mean what will this do to car sales well do you think this is any different of a story than what we've been tracking over the last few months like really things haven't changed that much and ev sales haven't except until right now so i'm just oh are you seeing some movement for last month okay car sales in the united states down 6.3 percent and do you think that's related to gas I think there's a confluence of things happening.

51:48Gary Vasilash: Yes, I think gas is part of it. I think interest rates are part of it too. You know, you go to get a car loan right now. If you've got prime credit rating, you're paying close to 7%. If you're subprime, you're paying 13 % to 14 % interest rate. That's crazy significant. So I think it's a confluence of things. And there's uncertainty in the world, but, you know, an almost seven percent and other automakers like Ford was down, I think, almost 10 percent. Yeah. Audi. Now, Audi, I think it's got more to do with the tariffs and everything. They took a big nosedive. Yeah. But yeah, I you're right. Car sales have not been bad up until right now.

52:35Gary Vasilash: And now we're you know, we saw a sharp drop in August.