Simulation means running “virtual tests” in a computer model instead of immediately building and testing real hardware. It helps engineers find problems earlier and try lots of scenarios quickly.
Siemens is a big engineering-technology company. In this conversation, they’re mentioned as a provider of software tools used to simulate products and speed up development.
A digital twin is a computer “stand-in” for a real product. Instead of just looking like it in 3D, it’s used to predict how it will behave and how it will be built, so engineers can test ideas before making physical prototypes.
Multiphysics means the computer model accounts for more than one kind of physical effect at the same time. That makes the virtual results closer to what will happen in the real world.
NPI cycles are the time it takes to introduce a new product—basically from early engineering to getting it ready for production. The idea here is that better simulation helps reduce that timeline.
Altair is a tech company mentioned as part of an acquisition. The hosts say it helped Siemens’ product-development tools, including for electric-vehicle development.
Term
Realized Lab
Realized Lab is mentioned as part of Siemens’ showcase at a user event. In this context it appears to be a program or platform used to demonstrate how their tools (simulation/digital twin) translate into real engineering outcomes.
The vehicle development cycle is how long it takes to go from designing a vehicle to being ready to build it. The claim here is that better digital tools can shorten that timeline and cut down on prototype builds.
Physical prototypes are actual sample parts or cars built so engineers can test them. The point here is that simulation can catch many problems earlier, so you don’t need as many real prototypes.
Term
AI
AI here means computer “smarts” that can help explore design options faster. Instead of testing only a few versions, it can run lots of digital experiments to find better solutions sooner.
“Heavy metal AI” is a nickname for using AI to build real cars and parts. The host is contrasting that with AI that mainly generates text or answers, like a chatbot.
LLM means a large language model—an AI that’s mainly good at working with text, like writing or answering questions. The host is saying this conversation is about AI used for engineering work, not just text.
It’s a car where software does a lot of the controlling, not just the mechanical parts. Because of that, the car can sometimes get new features or improvements through software updates.
Think of a bill of materials as the car’s “shopping list” of everything needed to build it. If the list is wrong or doesn’t match the design, the build and integration can get messy.
Term
electronic bill of process
It’s like a digital “work instructions” list for how to build the car. When it’s connected to the design, it helps factories build the right thing the same way every time.
A systems model is a big digital map of how different parts of the car work together. It helps engineers make sure the mechanical parts, wiring/electronics, and software all agree with each other.
Concept
reindustrialized
Reindustrialized means increasing manufacturing in a country again, not just relying on imports. Here, it’s about building more of the supply chain locally so production can move faster.
Brownfield means using land that was used before, like an old industrial site. It often needs upgrades or cleanup before you can build something new on it.
Micro credentialing means earning smaller certificates for specific skills. Instead of waiting for a full degree, students can start building job-relevant knowledge sooner.
Concept
AI enabler
An “AI enabler” is the idea that AI isn’t just a standalone feature—it’s integrated into workflows to make engineering tasks easier or faster. Here, it’s framed as AI being built into training sessions and design tools so engineers can learn and work more effectively.
CAD is the computer software engineers use to design parts and vehicles. The idea here is that AI can help people work faster inside those design programs.
“Edge” means doing the computer processing right near the machines that are collecting the data. That way you don’t have to wait for results from a far-away computer.
A digital thread is the idea of keeping all the computer information connected from design to building the product. Instead of starting over at each step, the same data keeps flowing through the whole process.
Tokens are the small pieces of information an AI system reads and processes. If you use the AI a lot, the cost can add up because it has to process more of those pieces.
This is a strategy idea for protecting North America’s auto industry. The goal is to make it easier for car companies and suppliers to keep operating even when outside rules or disruptions happen.
A tariff is a tax a government charges on imported products. If parts or cars cross borders, tariffs can make them more expensive, so companies have to follow the rules to avoid problems.
“Tracing” here refers to tracking where materials and components come from to prove compliance with origin rules. In automotive trade policy, this can affect which parts qualify for tariff treatment and how companies structure sourcing.
Think of the supply chain as how parts get made and shipped. The value chain is everything that turns those parts into a finished product that people buy.
Stranded capital is like spending money on a factory or equipment that you can’t fully use later. If companies move production, some of that earlier investment may not pay off anymore.
More regional means car companies may build and source parts differently for different regions. Instead of one plan for the whole world, they adapt to local rules and supply networks.
This phrase means making one car design and selling it in many countries. The point here is that new rules and costs are making companies adjust cars by region instead.
Critical minerals are special materials that are hard to get reliably. They matter a lot for making batteries, so shortages can slow down EV production.
Rare earths are a set of materials used in some battery and motor technologies. Because they can be hard to source, people look for alternatives to avoid supply problems.
“Tier one” companies are the big suppliers that make key parts for carmakers. The point here is about whether many top suppliers coordinate, or whether governments set up shared groundwork first.
“Pre-competitive” means the early stage where companies cooperate on shared foundations before they compete with their own products. The idea is to avoid everyone doing the same basic work separately.
AV training data is the real-world driving information that self-driving software learns from. The better and more complete the data, the better the system can be trained.
Term
L two plus plus L three
This is talking about levels of self-driving. Higher levels mean the car can do more of the driving itself, though the exact responsibilities differ by level.
OTA means the car can get software updates over Wi‑Fi/cellular, like your phone. Instead of going to a shop, the update downloads and installs wirelessly. It’s used for fixes and new features, but it usually has to pass safety checks first.
Self healing means the car can notice something wrong and try to fix it automatically. Instead of needing a full update right away, it can adjust itself in the moment. But bigger changes still usually require an approved software update.
A driver monitoring layer is the car’s software that keeps track of what the driver is doing and how they’re behaving. It can then change how the car responds. The idea here is that it can help fix issues without needing a full software rewrite every time.
Homologation is the official safety/approval process that makes sure a software change is allowed to be used in the car. Even if the car can fix some things automatically, other changes still have to be checked and approved first. That’s one reason updates can take time.
Battery drain here refers to the power consumption that occurs while downloading and installing OTA updates. The speaker argues that modern vehicles can have enough compute demands that OTAs may be difficult to complete within short time windows, especially when the car’s power state isn’t ideal.
“Goodwill” means customers trust the company enough to feel okay waiting for improvements. If a brand has a reputation for updates, people expect new features instead of getting upset.
Tesla is mentioned as a company whose customers are more willing to wait for new features because they’ve gotten used to software updates improving the car over time.
“AI layers” means adding smarter software on top of the car’s main computer. That smarter software can learn from what the car is doing and help adjust things more safely.
Term
STV
STV appears to be an acronym for a vehicle compute/processing element in the speaker’s architecture discussion, paired with “essential compute.” Without the full expansion in the excerpt, it’s best understood as a specific in-vehicle system that supports the AI/configuration approach described.
ADS means the car’s autonomous-driving software system. Because it affects safety, updates to it usually have stricter limits than normal infotainment apps.
A configuration layer is like the car’s “settings” layer. Instead of rewriting the whole program, you change parameters so the fix is safer and less risky.
Term
learning transmissions
“Learning transmissions” refers to transmission control systems that adapt over time based on driving behavior and sensor feedback. The speaker uses it as an example of machine-learning-style adaptation that’s already existed, then suggests extending that concept to a broader software layer.
LIVE
Speaker 1: Auto Line after Hours is brought to you by Alex Partners.
For more than forty years, we have helped companies and their stakeholders around the world harness opportunity, overcome challenges, and achieve outsized outcomes Alex Partners when it really matters, and by Bridge Stone Tires solutions for your journey.
Speaker 2: He everybody, thanks for joining us on Autoline after Hours.
Gary Vasselash is out right now. He's out in Nashville.
Actually he went to go see a concert. Lost out
to Joe Jackson. But we've got Joe White and as good.
Speaker 3: Yeah.
Speaker 2: I can't play the piano, but yeah, other than that, you can play the guitar.
Speaker 3: I can play the guitar. I think probably Joe Jackson
can't do.
Speaker 4: So and Joe right now you're doing your.
Speaker 5: I have a substack high speed rodeo. And I'm also
doing a podcast with Tuley called The Wheel.
Speaker 2: Yeah, formerly Wall Street Journal, formerly Reuter A bunch of stuff, just so the audience knows. I know you already know that.
We got Steve Plumb joining us too. Steve is the
editor editor in chief. I guess I should add of
managed manufacturing, engineering and Technology. I want to say magazines, Steve,
but we don't really do magazines anymore.
Speaker 4: So there we do you do, You've got to print first.
It's still very popular with awesome. I love it.
Speaker 6: We also have a website at Advanced Manufacturing that Org knew and Joe, you do look sharp.
Speaker 4: To continue the Joe Jackson be there and I got out.
Speaker 2: Let everybody know we've got del Costi. He's the president
and managing director for Semens Digital Industries for the America, with a great automotive background or you know, history of it too, And Dell, I'm so glad that you're able to be with us. Later in the show, we're going
to have Mark Wakefield from alex Partners. But let's talk
uh digital industries, digital technologies. So I'll pick it off
and then you guys jump in. But speed to market, China.
Speed is what everybody is talking about so much in the industry these days. And one of the key ways
of achieving that, of course, is simulation. And I got
to believe this is one of the things that Siemens is deeply into.
Speaker 7: Oh gosh, it's I think it's the foundation of everything we do.
Speaker 8: So we often talk about digital twin and people have a different representation of digital twin. You'll hear a lot
of people talk about it and they think about it as a sophisticated three D CAD model, but it's just so much more than that, and it's really important to set that up before even getting into the simulation because when we represent the digital twin correctly, it's a multifaceted, multiphysics, multifunction model, and so it's going to represent your engineering domain.
Speaker 7: It's going to.
Speaker 8: Represent your float onto, your requirements, your system model thinking in this three dgometry, and it also can represent how you produce the product. And so this connected set of capabilities,
when it's there and it's there properly, what you can do with simulation is absolutely outstanding. And that's where we see,
I think, some of the key advancements to drive NPI cycles and so yeah, it's a highly important area. And
I think you guys saw we spent a couple of bucks on an acquisition that long ago of a company based in Troy, Altair, and that's been a huge, huge benefit to the overall portfolio.
Speaker 6: I don't know if the part of the Altair purchase, but I think I saw that a recent EV development program brought to market in twenty two months.
Speaker 8: Yeah, there's a we even showcase that we were talking about Realized Lab earlier our user event Detroit, and so we had over three thousand people and we have a TeXcellence Award, and so we did Luke Lucid as one of the nominees. They did some really amazing work and
compressing their vehicle development cycle and eliminating a significant amount of physical prototypes.
Speaker 7: And so that's the ability you can do.
Speaker 8: And you know, when you think about simulation, now, in the past you would do a prototype to discover issues.
Speaker 7: That's not really the case.
Speaker 8: Now if I'm going to do a prototype, I've already discovered the bulk of my issues through simulation and through the comprehensive digital twin, and so bringing all that rich data to this digital world, I can go through all these scenarios. And now when you look at AI on
top of this, it's no longer spend a lot of time building up a model, analyzing that model and going through a few iterations. With AI, I can experiment with
thousands of iterations digitally.
Speaker 5: So you've filled on something that I wanted to ask you sort of you know, position your technology the simulation technology again AI, because you hear you hear automakers and you know, I call it heavy metal AI and they're talking about we're actually going to build something physical at the end of this, not an ll M. But how
do you use your technology relate to the AI that the automakers are getting from, you know, companies like Oracle or.
Speaker 9: Like you on the name of it. It's never mind
other other AI providers. Yes, yeah, this is this is this.
We could take the whole discussion on this alone.
Speaker 8: I did a talk that reindustrialized last week as a matter of fact, and the thesis of the conversation is, we have a lot of probabilistic models and if we think about OEMs all over the world, this is what they're wrestling with. How do I incorporate Every executive is
getting so much pressure on AI, and so companies are running fast. Companies have spent a tremendous amount of money
and you guys can read all the publications, how many have really seen the value? That's exactly my question. That's
the genesis of the question, one hundred percent. But the
fact is we live in a very precision based world.
We have people in these vehicles, these vehicles interact on the roads. Like we discussed earlier, precision matters, and so
probabilistic models have a place. And what I would say
is that the thing I discussed at reindustrializes. Imagine a
design of experiment highly accelerated, so I can go through thousands of iterations.
Speaker 7: I can interact with a model.
Speaker 8: Give me this scenario, give me this scenario, go back and look at what's been done before, bring that forward.
Let's experiment with that. But I can still put simulation
at the core. Math matters. We have one of the
most robust, widely used math kernels in the industry, hands down with Parasolid, and this math engine matters and it's highly performant. You couple that with systems model thinking, AI
accelerated and validation in the loop.
Speaker 7: Now the engineer can go.
Speaker 8: Through multitude of iterations faster than ever before in the day digital world and hone in on really really good answers.
So AI doesn't have to be the holy grail. It's
a facilitator of amazing ideas. And I still have engineering
in a loop. I'm accentuating the power of the engineer
in essence.
Speaker 5: And by the way, pounteers a company I was struggling with just before more coffee please.
Speaker 4: I'm curious because I've been hearing about.
Speaker 5: Digital twins and simulation for a long time, probably not as as long as Steve has, but a long time.
But I'm wondering what's kind of the rate of adoption now?
I mean, how pervasive in mainstream is this tech? And
if it isn't, how fast do you think it will become pervasive in mainstream given the competitive issues that John.
Speaker 7: Race will be has to be.
Speaker 8: I think it's the context we have to discuss candidly in automotive. There's some great where automotive is pioneering digital
back and go back thirty plus years ago. I remember
the first engagement I had, the precursor to Siemens.
Speaker 7: Was ugs Unographics.
Speaker 8: I was working with Vistian shortly after it spun out of board and working with the manufacturing engineering.
Speaker 7: That's where it started for me.
Speaker 8: And me and this gentleman were going through and looking at.
Speaker 7: So you were tier one back in the day. What
did you have to do?
Speaker 8: You bid on work and so you were given expected volume of the particular program you're bidding on, and then you would say, okay, well, here's my bid, and then you get a task to target if you want. So
now you had to figure out how to drive costs out.
The issue was because of the way the tools were implemented.
We were digital back then, that they were disparate digital tools, not necessarily connected cial in nature. And so we would
do all this development work in the vehicle development process, we would get to a certain gate and then manufacturing got involved. Then if you looked at and we mapped
out the all the great manufacturing tools that would get involved, well just about all of them were put in after design freeze. So your ability to change change is significantly
reduced at that point after design freeze, your cost to change is very high.
Speaker 7: So we looked at the intersection of the curve.
Speaker 8: We said, if we could fun and low, we came up with the concept that we could shift everything left and bring it much earlier in the design cycle and start doing our.
Speaker 7: Manufacturing capability digitally.
Speaker 8: Then all of a sudden, we could discover these issues, we could get to the task to target cost and we could figure out how to drive cost out through the life of the vehicle program. And that was a
light bulb moment, and this was twenty five years ago.
We're still making properties today. But if you fast forward
and compare to the Palenteers of the world, a great company, a huge market cap, but it's interesting what are they doing.
They're mapping out your intology and for the viewers, and it's basically, how do these communication protocols? How are they
going to happen amongst all these data types. But you
have to figure out how you're going to do that because if you, in essence lock in your systems, how do you upgrade?
Speaker 7: I need upgrade technology and so.
Speaker 8: What we're working on and we at our customer event talked about this, the concept of Intelligence Center X and I think I think it's key again part of the Altearor and another acquisition we had done years earlier. We're
looking at this completely through a different lens. If I
can contextualize the data across the life cycle and I can link these data types and so imagine I'm getting a warranty issue. A warranty isn't an issue in isolation.
I quickly need to figure out what happened with my production runs? Is there something I need to discover out
of manufacturing? Was it a product design issue? Did I
have a requirement not met in the way I produced or designed. Connecting these systems and developing my ontology on
the fly. Now I can do pretty amazing things and
get access by AI to all those different data that are sitting out there that can help me diagnose that problem much quicker than ever before.
Speaker 7: This is the promise of what we can do now.
Speaker 8: I'm still going to be the one sitting here saying I still want deterministic models in our world. I want
to know if I asked that same question six different times, they get the same answer six different times. And that's
not often the case with Generica, it's not the case with jeneric AI models. So I think that's the line
we have to walk here in our industry.
Speaker 3: Okay, okay, yeah, yeah, you can't have hallucinations.
Speaker 2: Bill mentioned how Lucid was able to compress this. They're
doing amazing things. I happen to think that Rivian's doing
some really stuff. Tesla was the one that was blazing
the way. But they're all startups, and I know the
legacy automakers are working on all this, but they all seem to be struggling with it, especially when you get to a software defined vehicle. What are some of the
advice that you would give to legacy automakers and suppliers by the way, too, of how they can accelerate this, because the benefits you're talking about are exactly what the industry needs.
Speaker 8: John, I honestly think we're going to see some major acceleration.
And we're working with just about every one of the companies, the legacy, the startups on each continent, and so we are seeing amazing work. This is not a lack of effort.
The companies are doing incredible things. But you've got to
understand these are companies that are mid flight in a lot of programs. They're producing vehicles, they're servicing vehicles. This
is not necessarily the easiest challenge, but the progress is being made and we're seeing companies experimenting with some really exciting ideas. But we also look at it and we
look at the maturation of the bill of materials. So
we have a design bomb and electronic bill materials and manufacturing bill of materials are electronic bill of process. How
do we make sure we're connecting all those things, And we also have to look at mechanical engineering, electrical engineering, electronics software combined in the systems model. By having these
foundations connected and making sure they're connected, this is or AI can.
Speaker 7: Accelerate what we do in vehicle development.
Speaker 8: So this is all about gaining insights faster than ever before.
I would tell you it's the transition from current state we talked about reindustrialized. You're seeing all kinds of countries
do amazing things with manufacturing, and now there's all this investment in the US.
Speaker 7: What's it going to take for us to get over the hump?
Speaker 8: The fact is we have an extremely rich heritage, no more rich than right here in Detroit.
Speaker 7: I tell everyone about Freedom Forge the book. It's an
amazing book.
Speaker 8: It's amazing read, but it showed the power and impact Detroit had on us manufacturing. I think we have that
opportunity again. But the fact is we have Brownfield. So
it's the same issue in manufacturing as we see in the OEMs you're talking about. We have to take the
existing structure and start evolving it quickly.
Speaker 4: I think it.
Speaker 8: Takes to be really well thought out and how we're applying these technologies to what we have today. But yeah,
there are advancements that we talked about. How do we
link these different functional models, accelerate with AI, and then make sure we're having this simulation at the loop. This
is what's going to put this we call it comprehensive digital twin, but that's what's going to put it in flight and in motion. And these are just initiatives that
are really in flight everywhere, by the way, and so this is not a lack of effort of any of the OEMs you mentioned.
Speaker 6: Engineering and the loop workforce is a key issue through manufacturer in the auto industry. How are engineers interacting with
AI now? What kind of training is being done and
what's being done on the college level to prepare the next generation for this.
Speaker 8: There's a lot of work to be done there, there's no question about it. I think a lot of college
campuses are figuring this out right now. But candidly, AI
done right there shouldn't be a huge lift on the training portion of it. You need to interact with the
models that you have in the systems that you have.
I think the balancing act is every company like ours, we're building amazing AI capabilities.
Speaker 7: In our offerings.
Speaker 8: It's how you connect the offerings to get the insight the shop or worker A little different that is where I think we need to put a concerted effort and so we're having a lot of conversations in that area.
What do we need to do for tool facilitation or learning aid facilitation.
Speaker 7: But it starts with that foundation level. We've launched this
thing called micro credentialing that will allow students to get in there early.
Speaker 8: We want to get it as early in the curriculum as possible to understand the concept of engineering.
Speaker 7: And then how is AI enabler for that?
Speaker 8: But the AI should be incorporated right into those sessions as simple as we had the complexity of CAD environments.
A designer or an engineer working in CAD AI can study their behavior model and help recommend based on the way you work in the tool and the type of work you're doing. Here are the commands that are coming
next for you. Still want the human and the loob,
so they're controlling that session. These are the little things.
Speaker 7: We can do to embed AI and what we're doing every day, but still put the worker in some form of control. Like you still need to be involved in
that session.
Speaker 5: So can you talk a little bit about you may measure on the shop floor that people who actually need more training than some other people in some other parts of an organization. Can you go a little deeper on that?
I mean, how so what would you need? What's going
to be the difference between a production floor employees job in an environment that's got your technology fully plugged in versus today.
Speaker 8: Yeah, this is one that's it's it's kind of fun to follow what's going to happen. And you hear two
schools of camps. You hear that AI is going to
accentuate the job, and then you also hear a lot of discussion on lights out factories. The reality is, I
think the companies of the future that are going to be hyper successful are the ones that are going to be able to gain insights quickly. And this has an
implication how you produce product. And so if I can
gain insights quickly and I need to modify quickly, how do I translate that to how I'm going to build product?
Where am I going to build product? How am I
going to handle supplier resilience? And what's the implication on
the shop floor. We refer to that as adaptive I
want and you know, we go right to this idea of lights out.
Speaker 7: I go right to the idea of adaptive.
Speaker 8: How are you going to be flexible with what you produce, where you produce, how you produce, and then what's the role of the worker going to be in that? And
I think we're going to see a transition. I think
there are certain functions that are going to get more automated for work or safety, for work or productivity. We
don't want work or fatigue. So I think we need
to create a really good environment for the worker on the shop floor, the one that's flexible. And I mean
the fact is, go talk to I used to work a lot with Japanese manufacturers, and for them it's you have to go to the spot. You don't do this remote,
get to the shop floor and under stand, go to the genda exactly exactly, and so.
Speaker 7: The insights you can gain from that worker.
Speaker 8: The thing is, though before you would go, you would then go set up a team due to some discovery, go pull data that could take weeks.
Speaker 7: That doesn't take weeks anymore. I can interact with the
machine with voice.
Speaker 8: The data all comes to me on whatever device it needs to come to me, and I can solve that problem much more real time. I think those are the
opportunities that we have in manufacturing here.
Speaker 2: Don't've seen a tremendous of the evolution of the supplier industry over the last decades. And I'm more interested in
your part of the supplier industry because you mentioned earlier Siemens bought Altair, paid a good chunk of money to get it.
Speaker 4: Well.
Speaker 2: Bit also, I want to say, it's Ansis and Synopsis got together. So we're seeing engineering an industrial services company
going out and acquiring simulation companies. Talk about that evolution
that's going on. What's driving in? Where do you see
it going?
Speaker 8: That's interesting because right now the explosion of data centers.
I mean, what's the most limiting factor to AI, the ability to build data centers? What's the limiting factor to that?
Where you're going to get your energy from? So this
is all the system you know, And we talk about the Altar acquisition. The acquisition that fascinates me is and
when it happened is when we acquired Mentor Graphics. That
and there were a lot of questions, why would you guys acquire Mentor Graphics, you know with Caliber which validates the chip itself. And at that point it was so different, right,
the chip wasn't even part of the conver that's just something you purchased. Boy, Look how we've learned from that point.
So and then you see synopsis and answers. So I
feel like Siemens was onto something back then, and how that chip interacts with the performance of the vehicle and now the importance.
Speaker 7: Of that's going to play in AI.
Speaker 8: I think was a fascinating a strategic move from Semens, and it's so accentuated today because the impact of the chip, how you stack chips, the heat generated from the chips, whether it's in a product or in a data center.
Speaker 7: It's amazing the parallels in this. And we talk about
AI factory.
Speaker 8: I look at an AI factory and I look at the parallels to a modern production or manufacturing environment.
Speaker 7: The challenges are different.
Speaker 8: But I have in a liquid cool data center, I have the flow of something. I have potential power surge
issues that I got to deal with. In a factory,
I've got flow of product. And so how we solve
those things and take train digital twin models and have them at the edge so I can make a decision quickly.
And I think that's one of the great things about working for a company like Siemens is that our world isn't three D geometry by itself, it's not system modeled by itself, it's not float on or requirements.
Speaker 7: It's the whole digital thread. But it's also production.
Speaker 8: We have hundreds of plants, we have World Economic Forum, lighthouse facilities, you know, we have these things. We've applied
this technology, we've implemented it, we experiment it with it like every other company or customer of ours does so, John, I think that was a fascinating thing we did years ago, and it really opened our eyes to what it means to combine mechanical electrics, electronics, electrical electronics, and software together.
And so when you talk about a software defined vehicle, there's different levels of understanding the impact of what that system model needs to behave like, and I think those are going to give us the future advancements in working with the supply base as well.
Speaker 3: Yeah, I'm wondering.
Speaker 5: I know this is I'm kind of forcing you to oversimplify something it seems fairly complex, but I wonder if you can talk about some of the buckets of where this technology has potential to you know, cut costs or cut defects or you know, and basically speed things up, because it seems like the ultimate challenge that the Autumn Western automakers have, the Detroit automakers have is if they're going to have I think we'll talk about this a little later, if they're going to have to, you know, build an America for America American labor rates. Fine, great,
but you know that that's going to put pressure on their businesses and on their margins. So when if you
come with your technology, where's what are some of the buckets where your technology can save money? Is it in
more rapid identification of problems that are causing warranty cost?
Speaker 3: Is it?
Speaker 5: Is it in cutting time and cost out of the upfront engineering and supply chain management you can talk.
Speaker 8: I'm an answer yes to whatever scenario because if you think about what I laid out, the answer is yes to that.
Speaker 7: Where I think it gets fascinating is combined. Again. I
always look at AI from two lenses.
Speaker 8: I think there's so much potential and promise, and I think there's things we have to be very careful about and we have to be super pragmatic. And if I
if I pull the proverbial thread on that, I think an amazing equalizer for companies is how can they implement agents effectively because right now it's super expensive. You go
consume a bunch of AI tokens, you wake up you have a billion didn't anticipate at the end of each month.
I think that's going to level itself out like everything does.
And so if I deploy agents, what's my approach for deploying agents, because that's kind of I don't want to call it an infinite pool of labor, but it's a great supplement a labor force. Make sure the scenario we
talked about earlier with running all these simulation iterations where we could run a handful before, now we can run thousands and we could triangulate on really good results.
Speaker 7: That's amazing impact the time to marketing cost savings.
Speaker 8: If I have issues in the field and I have a robust, comprehensive digital twin of my product and my process and my ability to service, when those issues come up, I can explore all the reasons for that.
Speaker 7: I can use AI and agents to help me do that.
Speaker 8: An engineer in the loop, but I have that opportunity to process data like never before. My insights are off
the charts. That's the promise that can really equalize things
in terms of cost base done right, But you have to you have to have the investment now in that foundation, because if I have data sitting all over the place and it's not contextualized, how am I going to come to the right conclusion. We all know the AI output
is going to be it's probabilistic.
Speaker 7: It's going to learn off models, it's going to learn off data.
Speaker 8: If that isn't complete, what I'm going to get out of it's not going to be.
Speaker 7: So those are the opportunities.
Speaker 8: I think for global manufacturers, it's it's how far can they can they what steps are they taking right now to make.
Speaker 7: Sure they're prepared to get the benefits out of it.
Speaker 6: You talked a little bit about the difference between oms and suppliers and smaller companies and adoption rights.
Speaker 4: What about globally?
Speaker 6: Where does the US stack up in terms of adoption and implementation? Also, are there any regulatory concerns or IP.
Speaker 7: Things are going to move so fast? I don't know.
You guys are experts in the field. You talk to
all the industry leaders.
Speaker 8: The IP thing is an interesting one for me, just with how pervasive information is and how fast things move.
Speaker 7: What are you going to be what's going to be your most important IPA of the future.
Speaker 8: That's a difficult question for me, honestly, just being in the technology sector and seeing how fast things are moving, I'm curious how that goes. As far as adoption goes,
we see pretty incredible adoption all over the place, and you look at it across industries.
Speaker 7: It's automotive has been one of the pioneers and digital technologies.
I mean again, we can go back to like I brought some names, but I'm sure you remember Frank musition is factory of the future and digital factory initiative.
Speaker 2: In fact, the first CADCAM used just about in the world was General Motors in the nineteen sixties. Very basic
compared to today, right, but it was a pioneer.
Speaker 8: It contracted really the first modern PLC and that was a project done for General motors and things have been in that business in Europe in the.
Speaker 7: Fifties and sixties.
Speaker 8: So you look at that first PLC in the early seventies that was really implemented for General motors. So you
look at the innovations that really were here in the automotive industry in this town.
Speaker 7: But you look across industry sectors.
Speaker 8: I think you look at industry sectors as much as anything in terms of globally because there's advancements all over the world. There are companies that lean in all over
the world. There's startups that are doing absolutely amazing things
with technology. Back on stage at Realize Live, I mean
you saw some of the advancements happening in rockets right now for the purpose of launching satellites hand little communication, and so we're seeing them just incredible innovation based on industry need as much as we are geographically.
Speaker 4: I think NASA was one of the first to implement a digital time the late nineties or.
Speaker 8: No, yeah, you watch what what we get to a close seat with SpaceX as well, so it's it's, uh, it's fascinating to watch.
Speaker 3: That's a good point.
Speaker 5: I mean, you know, we think about the auto industry all the time, but I obviously it seems like cutting edge and manufacturing is not necessarily the auto industry. I mean,
it's the chip fab industry, right, or or maybe it's it's you said SpaceX and the space industry. I think
you mentioned you know, you know the energy industry. I mean, well,
you know they're they're they're you know, so that this is kind of proliferate, proliferating everywhere.
Speaker 8: I got to tell you, I spent Monday and Tuesday with a company UH in Texas, and I was blown away.
Speaker 7: We often get asked what does it take to deploy successfully?
Speaker 8: And the very senior executive there, they're doing a full transformation of their organization structural that they're partnering with us on a key tenet of that to enable this transformation that they want to do, and they're going to look at their business differently going forward, and they want this to be the foundation of how they're going.
Speaker 7: To carefully grow with AI.
Speaker 8: And he looked at me and he said, you guys implemented against our will. We worked with you on a
process that you told us we need to do to run this project, and I got to thank you for that.
He goes, but I got to tell you one of the things, because we were struggling for a couple of years and we were trying to get this project off the ground, he said, one of the things that you proactively did with us that I think was game changer is you educated us all the constituents, senior, mid level users of the system. He goes you educated us on
the capability that the system has.
Speaker 7: I don't want you know, I walked out of there.
Speaker 8: It's interesting you can hear something at different decades or your career and it means something completely different for me at this stage of my career. That was so profound
here because I think as leaders in folks in the industry with a lot of experience, we have to humble ourselves a bit to recognize it's too easy to make decisions and say, team, go do this. But as a leader,
if you don't invest in the details, and I think it's more important now than ever with AI, because if you're making a decision and you don't know the pros, the cons the risks of AI, and you're not participating in the leadership of that initiative, we as your partner can't save that. We can advise, we can recommend, we
can push, But trust me, I've been involved enough that where we've come in and what we're doing, we're replicating an existing process. Absent what the new capabilities will bring. Well,
that's going to accelerate more than ever before my career.
AI is going to change things very quickly. Interaction models
in these systems. If we look at the way we've
done it before, I don't know that that's the right thing.
Speaker 7: We have to look at things.
Speaker 8: It's kind of like the conversation we were having on form and function of a robot on the shop floor.
Speaker 7: Does it need to look like a human? Probably? Does
you know it needs to match the function it needs to do.
Speaker 8: I think technology is very much going to be that way as it continues to evolve. So I'm really pumped
with the next couple of years spark on how.
Speaker 4: Fast are AI progressing? We used to have more as law.
Is this an accelerated version right now?
Speaker 8: I think it's an accelerated version, But again, I think we have to be very careful. It has to fit
into what we do, and we have to constantly ensure that what we're getting out of AI is foundation to what we need out of AI. It again, if I'm
on the production floor, I can't ask a question of a system and get different answers. I need a precise answer.
I need to discover where that quality issue came from.
If I have a through put issue, I need to discover why. I need to get to root cause fast.
But it needs to be It can't be a thesis.
It's got to be something that's proven, and so I think that's going to be the key. Yes, I think
it's going to change extremely fast. I still think we
have to be smart about how we adopt and weave it into.
Speaker 7: What we're doing and dealt with that. We're going to
have to wrap up this segment of the show that went quickly.
Speaker 2: I know, I know, it really does, especially when we're so engaged in the conversation like this. But thanks so
much for coming on and you know this, this digital world is fantastic and very intrigued by what you're saying about how AI is even going to accelerate it.
Speaker 7: Thank you. This was a lot of fun, really good.
Speaker 2: We'll be back in just a moment to talk even more about what's going on in this crazy industry.
Speaker 10: We act, we grow, we transform, We protect, we rescue in moments that define the future. We are the partner
you can trust, Alex partners when it really matters.
Speaker 1: Knowing that a little rain won't slow down your day.
Speaker 10: That's what really matters, Which don't runs it by attract hires, confident control in wet conditions.
Speaker 2: We're back and now We've got Mark Wakefield and I got to read this because I can't keep this memory.
He's the Global Automotive Lead Executive, Global Automotive Lead, Executive Partner and managing director for alex Partner.
Speaker 4: Executive Partner is a new title that they gave what eight of us.
Speaker 2: That's great and Mark, you guys just presented yesterday your annual global update for the automotive industry and there's some hair raising Stone going to say.
Speaker 5: What a basket of cheer that there was some positive We're higher than ICP on ours.
Speaker 11: It's one thing that should be a rising tide. But
you know, we do it each year where we get all of the team from around the world say, okay, what'd you work on it was neat last year, sanitize and synthesize this. Let's get all on the same page internally,
and then there's some of that stuff that's able to be shared publicly, and so we do it in that process.
Speaker 2: Well, what I like is how comprehensive it is. I mean,
you go all through the world and you hit so many different topics on it, and we can't possibly talk about them all in the segment that we've got on today's show. But one thing I wanted to get into
that really stood out for me is you're advocating for what you're calling Fortress North America. Not just Fortress America,
but North America. Take it from there.
Speaker 11: Yeah, I mean, it's it starts with almost being crazy that we're doing projects to help suppliers and automakers figure out the compliance. And when I mean the compliance, I
don't just mean actually how to comply on the tariff.
Speaker 4: I mean like the administrative pieces of it. It's a
real cost.
Speaker 11: Actually, it's like one to two percent of the cost of a vehicle.
Speaker 3: Two thousand dollars a vehicle. Yeah, almost that.
Speaker 11: And so you know, if this gets worse with a US NCA two that's very focused on origin, on tracing, and it's tracing across Mexico and Canada and the US, this is going to get uglier compared to the real focus, which should be China. You know, a competitive and resilient
North America facing off against China. You have just a
natural logic to it where you've got, first of all, the supply chain and the value chain is already built that way. The market is dominated by the US, so
there's not really a fear that the US gets elbowed out of anything in this in this mixture. But you've
got things like fifteen million tons of rae oxides ready up there in Canada. You've got processing facilities in Alberta
and Saskatchewan coming online for those. You've got hydro power
and aluminium in Quebec, so they're not really at risk of becoming the next AI chip locator. But you combine
that the labor of Mexico, the infrastructure that's built up in Mexico, and you have fairly compliant then Mexican and Canadian government supporting this idea of hey, let's all get this together and work better. You then had if you
had the governments and the industry doing like a semitech style collaborations around the aight as and av chipsets that are required the kind of technology that has to go on the vehicle. You now have instead of dealing with
these sort of trade things on the continent, you're dealing with things that matter in terms of getting up to speed to prepare for the Chinese coming here or the Chinese basically leap frogging elsewhere, and this being isolated market, so to me, it makes.
Speaker 4: A ton of sense to me.
Speaker 11: And also, you've got these two thirty two tariffs that are putting billions into the Treasury's bank. Why don't we
use some of that money to invest back in making the industry actually more competitive on a carrot, not just to stick. And then we also the benefit of that
that is, over time, that becomes a lower priced car for the consumers instead of just a tariff that becomes a higher priced car because there's a lot of stranded capital that will come from moving plants hundred miles south from Canada or a hundred miles north from Mexico.
Speaker 3: No, I mean, that does seem like the problem.
Speaker 2: Man.
Speaker 5: You mentioned the compliance costs and sort of the costs of ripping up the existing kind of state of play in North America. You talked about something else though, which
seems like it's going to happen, whether this vision of what should happen in North America goes forward or dozen, which is that the automakers are becoming more regional.
Speaker 3: The idea of you.
Speaker 5: Know, build it once, sell it everywhere, which I started with and this is probably you did too.
Speaker 3: That's gone right, I mean, and.
Speaker 5: I think you talked about the automakers if they want to have any global presence at all, I think even outside out of China, there's going to have to be like a US product with US tech, US SPECK and then everywhere else. Talking about more about that, because that
seems like another drag on the industries.
Speaker 11: That's a drag, but that that does fit to a degree because you're you're trying to prevent and give a protected period for the US industry to catch up and to change. That actually can't be too long, you know,
I said, I think that's around five years is probably an appropriate time period to do it in. But that
does mean that if you're doing all of these origin and you've focused it on China to look for Chinese toier two to your three year ow materials, Chinese software and hardware, then it's really just focused there and it's an easier process than checking, oh, this thing moved from Poland to the Czech Republic, then to Germany, then to the US, down to Mexico and back up to that's not very valuable that exercise compared to did it come from China or as it involved with China as an
you know, foreign energy of concerned type of affair that still is likely needed to prevent everything just being Chinese oriented in Chinese.
Speaker 4: Storced and related.
Speaker 11: So that's sort of a necessary evil compared to the the US and Mexico orientation of trying to push the the scales there, and it distracts us from the real mission, which is getting a healthy, fit competitive industry that can take on the Chinese industry in five years.
Speaker 6: I was just at a conference and one of the keynote speakers was Hot of the Critical Minerals division of the DIE, which I don't know existed until I saw him speak, but he was talking about bringing back or finding alternatives to both rare earth and other critical minerals for batteries. How critical is that and what's being what
kind of progress is being made, how long it will take to really implement them.
Speaker 4: It's very critical.
Speaker 11: But the reason I talk about the sort of semitech thing is that doing it with twenty different players doing it between Tier one's battery companies and OEMs, versus doing it as a pre competitive base that then is able to be used by people that participated in that that are US based, makes it much easier to go through.
Speaker 4: And do you say pre.
Speaker 3: Competitive, I mean, but basically the governments should do it.
Speaker 4: Yeah, Like ceetech was this thing whereas you know, fifty to fifty industry and governments to put in they did, you know, think of AV training even beyond the hard things of chips think of like.
Speaker 11: How hard it is to get good quality AV training data. Right,
And you've got all these automakers that are putting chipsets, sensors, cameras, things on the vehicle that aren't even used to help the customer. Now they're there to get training data and
be able to do ultwo plus plus L three.
Speaker 5: And they're all duplicating and their efforts to each are subscale to what you really need to do the analysis.
Speaker 4: Yeah, and they're not.
Speaker 11: Going to come together and just use one model in one thing. But the pre competitive thing is the nature
of this data la, like of having all of this training data available for the participating groups.
Speaker 4: That's an example.
Speaker 5: Yeah, that makes I mean, it makes so much sense and yet right and yeah, it's.
Speaker 11: Not the focus is on is it fifty US content and a half?
Speaker 3: Yeah.
Speaker 2: I think you can really touch on something here that if we treated North America as is a united supply chain, we have a much better chance of competing with the Chinese.
And also, you know this whole idea of using, like you say, the natural resources and the cheap energy that Canada has and the cheap labor that that Mexico has.
Speaker 5: But you just said the magic words, so cheap labor.
I mean, I mean that's the thing. And it seems
like the sort of the elephant. There's a bunch of them,
but the elephants. One of the elephants in the room.
Here is the labor arbitrage with Mexico. Yeah, for the
United States that I imagine for the Canadian unions and Canadian workers too.
Speaker 3: That's the problem, it is.
Speaker 11: That it isn't don't think of it as a labor arbitrage of the US to Mexico. Think of it as
a labor arbitrage between Mexico and China. Okay, and then
it's a different story, right. And if you look at
the trends on employment, the productivity, the robotics, you know, putting a higher bar on things has driven more robotics to one degree. But the import data, it's Canada that's
actually through NAFTA, the one that got hurt a lot.
China's now flatlined for about the last ten years in auto imports. So the idea of the BCC is now
for LCC is now BCC and not China for best cost country sourcing. But Mexico is still going up, Canada
is still going down. So it's really the Canadian group
that's got hurt in terms of employment more so than the US.
Speaker 7: Why they're playing their China card.
Speaker 11: Yeah, but you think of the China cards already being played.
There's forty plus supplier plants in Mexico and half of them are new new plants that got put up in Mexico.
So if you had a United Front that said no, we don't want this, we want X, Y and Z, and it was a consistent industrial policy across North America, you would have a lot easier time doing the winds to fight in the next war instead of trying to sort of hold on to today.
Speaker 2: And so what you're saying instead, what we're getting in terms of compliance is a bunch more red type and bureaucracy sprinkled with tariffs and sort.
Speaker 11: Of stranded capital where you're having to move stuff for convenience.
There's a whole bunch of don't wrong, there's a whole bunch of loopholes that need to be fixed.
Speaker 4: Things like when you.
Speaker 11: Get a part that comes in seventy six percent from Canada, seventy six percent Canadian, but it's got twenty four percent Chinese and it becomes magically one hundred percent USMCA compliant.
That's the kind of loopoole that needs to get closed, and some of these other sort of transshipping types of things, And that's going to be tough for the compliance cost will be there for real. But let's focus the cost
on achieving the goals that we really want rather than.
Speaker 5: So Mark, you talked about them. So you talked about
five five year window. Yeah, and I guess I want
kind of press on that a little bit because we talked about this set in one of the conferences that we've seen each other at in the last month. I mean,
given the state of the reality of the state of supply chain, is five years enough to basically relocate the supply chain for the North American auto industry out of China.
You can't even make the dies right we need in this country without China.
Speaker 3: So how much can we do without China?
Speaker 5: How fast is it are going to take to really become regionally self sufficient.
Speaker 11: I mean I speak of that more in the sense of getting competitive on electric vehicle and on aight as and AV and on an STV. And no, I don't
think you're competitive by then, but I think if you go too much beyond then.
Speaker 4: There's not the impetus to move.
Speaker 11: And some of these things, the processing raw material processing is a ten year affair to get a new pipeline, so that's not in five years, but it has to be started and moved along. And it also feels like
if you try to hold the line too much, they'll be creative solutions around it. And in Canada essentially yeah,
in Canada, but also you know there's some other licensing things, some other different back doors that you're already seeing, the safe harbor things, break against ICTs and these other creative solutions around the laws. And yeah, new laws can be
put up. No ghosten law as clear as day is
what that's about, right, And it's but it's hard to hold that line against economic gravity, you know, economic weight of if there's a better mousetrap, not using that mousetrap.
So it's about trying to get to a competitive mousetrap.
Speaker 6: So you move away from China, what will the role of near showing fund showing areas of expertise for other countries play and the whole fortress strategy.
Speaker 11: Well, one of the things you're mentioning was this idea of this sort of fourth era of globalization. We had
a globalization window that was in the beginning, then you know, into the depression and sort of different protectionism things that took it down, and then after World War Two starting a globalization, a whole bunch of things that sort of moved it towards standard you know, even containers and station containers and not not all those laws. Right, And you
had a growing globalization dramatically through to you know, about the mid two thousands, and then it kind of flatlined from there. And you've seen more and more regionalization, difficulty
with crises that become that expose supply chains across the water, difficult with politics and difficults and different regulations and sort of non industrial logic regulations coming in that disrupt a plan that was put in place. That's going to happen
more not less. So let's you know, we now no
longer have a US hegemony. There's the acts American all
this stuff is yesterday's news, right, It's it's now at China and US not equal.
Speaker 5: But you know there's two main parties here now. Well honestly,
when it comes to the auto industry, China is sure.
Speaker 11: I'm talking about US more global and in some ways they're not even in at But the point is it's not like a US hegemony that works as a globalization.
So in that you then pick your your zone, and you can pick North America fairly straightforwardly. It's a lot
harder to do Australia or to do a Japan or a Korea connection in because there's a lot of things that could happen politically with how that moves, also in economics with how that moves. So the North America, I mean,
it's it's not going anywhere, right, and it's not an area where you're worried about a significant influence on Canada or Mexico from China.
Speaker 2: So, as I said, you guys covered a lot of topics in your in your latest report, one that truly intrigued me. Again, this is getting away from what we
were just talking about, is you know, yeah, yeah, no, no, no, no, it was really good.
Speaker 7: But man, I wish we could go on another two hours.
Speaker 2: But what you brought up yesterday is that over the year updates OTAs might become obsolete.
Speaker 11: Oh yeah, I didn't mean are obsolete in that way, because it's still with you're talking about the AI dvs, so you get self healing. So right now, there's a
lot of bug fixes that go on in in OTAs, and it's it's a frustration because you know you you want new features in the car as a customer. It's
kind of nice that something's irritating is fixed, But wouldn't it be better if the car fixed itself and you know you noticed something wrong, even if you had to press a button and say I noticed this, I didn't like it, and the car then tries to fix that thing to the degree you can't. Right, So not into
the source code, but just into like a driver monitoring layer and a driver that configures things instead of really adjusting source code. The source code of course has to
get home alligation tested, and then the OTA is still happened a lot because you still have then that vehicle saying hey, I fix this thing.
Speaker 4: Anybody else want this fixed? Kind of thing?
Speaker 11: Some things coming off a vehicle instead of a one way OTA, and then you also have you're still going to have ot as for feature updates and for other things that have to go through am alligation process and safety pieces. So it doesn't really take away from the OTAs,
but it makes them not this process of the only way to update the vehicle, and sometimes a very frustrating way, because especially with the vehicles today, there's a lot of ones that are having a very tough time, especially in the US, doing OTAs because of the amount of compute power and the amount of battery drain required.
Speaker 4: To do them.
Speaker 11: You can't even get them done sometimes overnight, and so the ability right now they're trying to do jumpstate OTA's where if you're in version one, but you need to go all the way to version five, you can just jump there. Most of them are built that way. They
have to go one, two, three, or five to get.
Speaker 4: To these updates.
Speaker 11: So it's a real bottleneck that's not really pleasing customers at the moment because it's not bringing in the new features.
Tesla probably does it the best in terms of having the goodwill of their customers. Most others don't really have
that goodwill that sort of oh good I get a new feature coming.
Speaker 4: And so having the AI pieces take care.
Speaker 11: Of that most of those most of those issues could be dealt with if you did have essentral compute and an STV, but then you had AI layers on that in configurations and in being able to learn what's happening and adjust the vehicle without needing to go off. You
still need some good edge compute, very strong compute. You
also need some cloud But so there is OTAs but it's just not in the way that it is today.
So is there anyone doing that or oh no, no, this is like five years from now. So this is
this is projecting. Yeah, I mean you're kind of seeing
a bit of it.
Speaker 4: An IBI in China, but not much.
Speaker 11: I would say, So this is more people are still working to get to stvs. They're not having the compute
on the car enough to do this yet. This is
this is probably five years in the way.
Speaker 6: Examples of the types of features and functions where this I imagine there would have to be limitations that regarding safety.
Speaker 11: And I mean ADS is an obvious one, but that if you can correct an error or a problem, then that's a benefit, right.
Speaker 3: I would think almost any kind of software glitch.
Speaker 11: Yeah, glitches and bugs like if you're fixing something that's not working and it also then doesn't work like you know, okay, it's it hasn't worked, but you can. That's why I'm saying.
It doesn't go really into the source code. It's into
a configuration layer effectively, so it's not doing something that would require that would break come alligation rules or something like that.
Speaker 4: But there's a significant amount.
Speaker 11: I mean, honestly, if you had you've had learning transmissions and other things that do a bit of.
Speaker 4: Machine learning for many years into it. This is taking
that to a bigger layer.
Speaker 11: Now that you have the connectivity to all actuators, you have data coming off the vehicle, and you have a much higher for ponderance of bugs that you need to fix, and you can then have instead of half the engineers working on bug fixes, you can have them working on true feature enhancement and do the more the more routine bug fix is handled automatically on vehicle.
Speaker 3: Yeah.
Speaker 5: JD Power was out with their initial quality survey and uh, you know, board managed to kind of get to the top of the mass market. It's good for them, but
but I was the rest of the report was interesting, is infotainment is still like the number one dissatisfire.
Speaker 3: Has been that way for years.
Speaker 5: It's kind of it's kind of like, it seems like that's like elementary school in terms of the stuff you're talking about.
Speaker 3: And yet it seems like the industry is struggling with.
Speaker 11: That, struggling with it in part because it's on old architectures electronic architectures that are just difficult to deal with that.
But yes, absolutely, And then the speed of which something gets fixed, Like if you've got to wait weeks for something to get fixed, it's not ibi, but in a you know, like a recall type element versus it if it sees it and fixes it itself. That's a dramatic difference
of a customer experience.
Speaker 4: And you know, you.
Speaker 11: Do risk this thing of these OTAs not being seen as positive and being eventually seen as a nuisance and negative.
You know, and especially when you're then going to be competing saves five years time, you've got something that does do this and something that doesn't. Now this thing, you
know it looks like you Nokia to when iPhone.
Speaker 5: Yeah, yeah, for sure, I mean and that seems like, I mean back to the first part of our conversation that seems like a big risk for some of the legacy automakers is that you'll, you'll you'll increasingly have a situation where people know that there's you know, the product that's up here and kind of doing the stuff that you're talking about, the digital tricks and not tricks but digital enhancement, all the school stuff, and then you know, it's it's you know, it's the rest of us are living in some version of Havana, and you know, with an old tech that doesn't update. It just seems like
that's a real competitive problem.
Speaker 11: It is for anyone that has a big fleet of portfolios to deal with. But this is where it's like
on the stvs and like on the electrical architectures that need a clean slate jump to be done, you're going to need basically a clean slate jump on this as well.
And so if you have the cash flows that the US automakers are having, you have the ability to do that.
And so, and it's better to be started on the next war than fighting today's war. And you know, frankly,
a company like a Tesla is very well positioned to be one of the first to do this. And then
it's in your backyard and you're seeing what it's doing, and it's it's not to say that it will obviously be a Chinese that does this first.
Speaker 3: Yeah, that's true. We shouldn't forget about Tesla.
Speaker 2: Okay, another big topic jump here because one of the other things that you guys have uncovered this now jumps to Germany with young Germans, and what I found intriguing is young Germans aspire to Chinese automotive brands, not German ones.
Speaker 11: Yeah, the avoidance. So what we find is hard in surveys,
it's hard to deduce the truth if you say do you prefer this? You prefer that. What's better is to
say what do you not want? And then you look
at how many people avoid this? Basically, so you can
look at it two ways. You can look at the
ones like you know the you can say, okay, do I have champions that are advocates the nine and tens, and yes, that's a worthwhile signal. But parsing that specifically
when you're asking about things people haven't experienced yet, is a very difficult thing to put a lot of faith in.
So what we asked is one of the many questions we asked in the surveys was what would you not buy?
What are you for sure not going to buy? What
would you avoid buying a car from which region? You know, US, China, Donda,
and you could choose more than one. And yeah, the
young Germans below thirty five and a higher avoidance on German cars than on Chinese cars, showing you know why you see excitement of b y D and others to go into Germany and to make a shot at it.
And by the way, that the avoidance wasn't very high in the US either.
Speaker 2: In fact, I think you found that only twenty five percent of Americans would not buy a Chinese car under any circumstance, which.
Speaker 4: Means seventy five percent or open to the idea. Yeah,
of the thirty four and below age Yeah, And.
Speaker 3: By the way, in nineteen seventy five, if you did that.
Speaker 5: Survey about Japanese cars, you might have found some of the same, maybe even worse for the than where are.
Speaker 11: We Yeah, I mean the older population, it's not like that, right, It's as you would expect. But in the younger population,
that's what we showed, and that boats poorly for the Germans, and frankly boats poorly for the US in terms of just an open competition all.
Speaker 6: Mostly electric vehicles on the Chinese side that we didn't ask about electric versus ice and frankly there's both being imported into Europe, and.
Speaker 11: I think it's timed to cost definitely, kind of the cost because that's one of the things when we looked at like it's not just the sort of tech and these pasts. There is an expectation that the Chinese car
is cheaper for a like to like that it's cheaper.
There was an also expectation though that it was more modern and had more technology in it. So that's also
to a degree in nature of cheapness right of putting features in and not charging for them in addition to actually having a like to like. That is there's an
expectation of that, but that expectation is also reality.
Speaker 4: The Joe's quint, have you asked this question before? To
see what the changes when we didn't.
Speaker 11: Ask this question this way we've done. The other stuff
was more oriented. The power stuff was more ented on
electric vehicles and electric vehicle readiness, and as that was kind of flatlining that we looked more broadly in part because of just how much of an existential threat it is the Chinese coming into Europe, Like everybody else is kind of shrinking and the Chinese are the only ones growing.
You know, Mark, that's fairly flat.
Speaker 4: So the.
Speaker 11: That was the impetus, the idea of you know, is this going to work and who's going to buy these?
Why are they going to buy these? And what can
both players do around it markets.
Speaker 2: You know a lot of people watch the show who work in the industry, they're probably going, hey, what else is in this report? How can they get it?
Speaker 7: Or can they get it?
Speaker 4: So it's not like a printed report.
Speaker 11: It's the repository that our team gets through. So we
all sort of know what each other know and we can then share on various topics. So you can see
on our website a list of those topics. You know,
they can reach out to us. We're happy to talk
through two or three of the things that are most important to them. And yeah, we just we don't publish,
you know, thousands and thousands of pages of things. It's
more built for us to do our work with clients.
But it does have relevance for people various topics.
Speaker 2: Yeah, we're good, but that we're going to have to wrap up the show. Thanks for coming back on again.
You know we're going to have you back. I mean
you've been for years now and it's been great.
Speaker 4: That's my pleasure. It's great fun every time. Thank you.
Speaker 2: Yeah, great, Steve Plum, thanks for joining us, and Joe always good to have you here. To be here, and
I want to thank all of you for having tuned in.
Speaker 1: Auto Line After Hours is brought to you by Alex Partners.
For more than forty years, we have helped companies and their stakeholders around the world harness opportunity, overcome challenges, and achieve outsized outcomes. Alex Partners when it really matters, and
by Bridge Stone Tires Solutions for your journey.
About this episode
The conversation connects “catching up to China speed” to engineering workflows, manufacturing execution, and trade realities. Guests explain how digital twins and simulation—accelerated by AI—can compress development cycles, reduce physical prototypes, and enable faster decisions via edge/looped validation. They also discuss shop-floor precision needs, AI agents, and lights-out/voice-driven data collection. The episode then pivots to Mark Wakefield’s “Fortress North America” approach, covering tariff compliance, USMCA loopholes, and why regional self-sufficiency timelines are hard.
TOPIC: China Speed PANEL: Mark Wakefield, AlixPartners; Del Costy, Siemens; Steve Plumb, Manufacturing Engineering and Technology; Joe White, High Speed Rodeo; John McElroy, Autoline.tv
