A software-defined vehicle is a car where software controls a lot of what the car does. Instead of everything being hardwired, the car can be updated and improved with software.
Cunix is a company that makes software used inside cars. The hosts are saying it helps automakers add new features while keeping the car safe and secure.
Vector is a company that works on software that runs inside cars. The idea here is that they help make the shift to software-controlled cars less complicated.
Dr. Matthias Trubb is introduced as the President of Vector Informatic. The host positions him as one of the key experts who can explain the current state and challenges of software-defined vehicles.
Stuttgart is referenced as the location of Vector’s headquarters. The host uses the visit to Stuttgart to establish credibility and to set up a discussion of Vector’s software demos.
Alloy Core is named as a “new product” shown during the host’s visit to Vector’s headquarters. In context, it’s positioned as part of Vector’s tooling or platform for building foundational software for software-defined vehicles.
Term
case connected
This likely refers to “Connected” in the industry’s CASE idea. It means the car can talk to the internet and other services, so you can get features and updates remotely.
“Buy versus make” describes the strategic choice between sourcing technology from external suppliers versus building it in-house. In SDV programs, teams often decide which parts of the software stack to develop themselves and which to purchase to move faster and reduce complexity.
A platform approach means using one shared “base” system for software. Then different features can be added on top of that base in a more organized way.
VMs are like separate “virtual computers” running on the same physical computer. They help keep different software parts from interfering with each other.
QNX is software that helps a car’s computers run reliably. The point here is that companies like QNX help automakers turn complex software into something that actually works in real cars.
The Ford Edge is a mid-size SUV that’s built for everyday driving and family use. It can come with features that help you drive more easily and with a modern screen for controls and entertainment. The podcast mentions it because of its driver-assist and in-car technology.
IVI is the car’s infotainment system—basically the screens and apps you use while driving. The key point is that it’s software-driven, so it can gain features over time.
ADAS means “driver-assistance” tech. It’s the safety features in modern cars that help you avoid crashes or stay in your lane using sensors and computers.
A Skunk Works program is a special team set up to move faster than normal. The goal is to experiment and build new technology quickly, before it becomes a big, slow project.
A universal EV platform is a common “car foundation” that can be used for different electric models. The idea is to reuse the same base so the company can build and update many cars more efficiently.
Subscription services are features you pay for over time, like certain apps or capabilities in the car. Instead of being fully included forever, they can be turned on through the car’s software.
OTA means the car can get software updates over the internet, like your phone does. That can add new features or improve how the car works without you going to a shop.
FSD is Tesla’s software that tries to help the car drive more by itself. Even when it’s advanced, you usually still have to pay attention and be ready to take over.
Point-to-point navigation is just route guidance from where you are to where you want to go. The “point-to-point” part emphasizes start-to-finish routing rather than general guidance.
IoT just means the car is connected to the internet and other devices. It lets your phone and the car share information, like routes and charging info.
“Level three” and “level four” refer to SAE driving automation levels, which classify how much of the driving task the car handles. Level 3 generally expects the driver to be available to take over when prompted, while Level 4 is designed to handle driving without human intervention within certain conditions or geographic limits.
“New energy vehicles” is a broad label (used a lot in China) for cars that aren’t traditional gas-only vehicles. It usually includes electric cars and plug-in hybrid cars.
In software engineering, “legacy” refers to older, established codebases and interfaces that must be supported even when building new systems. The speaker’s point is that starting from scratch avoids the cost and complexity of integrating with old vehicle software stacks and their constraints.
ECUs are the car’s little computers. Each one runs a specific job (like controlling parts of the drivetrain or other systems), and they all have to work together through software.
APIs are like rules for how different software parts communicate. If you swap in new software, you still need to make sure it can “talk” to the older parts using the same communication rules.
They’re talking about how modern cars are becoming more software-heavy. The goal is to make it easier to build and connect all those software parts without delays and extra cost.
The operating system is the main software that runs the car’s computer. It controls how the computer uses its resources so other programs can work smoothly.
Time to market means how long it takes to go from “starting the project” to “the car feature is ready to ship.” Faster time to market usually means less development time and fewer integration headaches.
Company
QNICs
QNICs is mentioned as a specialized technology area that has deep know-how behind it. The point is that OEMs may not be able to build the same kind of platform expertise as quickly or easily.
The digital cockpit is the driver’s screen setup in a modern car—like the digital gauges and the main infotainment display. It’s where the car’s software experience becomes very visible to the driver.
Integration testing means checking that different parts of a system actually work together. For future cars, that’s important because software and hardware are often developed by different teams. The host is saying you need to test the combined system, not just individual pieces.
Term
application development
Application development is the work of building the “apps” or features that users interact with. The transcript is saying there’s also the platform work underneath, and problems can come from that foundation. They’re arguing for testing and readiness before stacking everything together.
A “big bang approach” in software and platform development means integrating and launching everything at once, rather than rolling out components gradually. The host argues that OEMs often use this strategy and then struggle because failures may originate in the underlying platform, not the application layer. The takeaway is that integration and platform readiness should be proven early.
OEMs stands for Original Equipment Manufacturers—the companies that build the vehicles. In the transcript, OEMs are described as struggling when they integrate everything at once, suggesting that their development process and system readiness planning matter as much as the application software itself. The term is used as an industry shorthand for the automakers responsible for the final car.
Vertical integration means a company tries to do more of the work itself instead of outsourcing. In cars, that could mean making parts like chips and software internally rather than relying on suppliers. The hosts are saying some automakers believe they should “own everything,” but that mindset is shifting.
NVIDIA is a semiconductor and AI computing company whose hardware and software ecosystem is widely used for AI acceleration. In this segment, it’s referenced as an example of a particular SoC/AI approach that can influence which AI frameworks work best.
Qualcomm is a semiconductor company known for mobile and automotive compute platforms. The segment contrasts Qualcomm’s approach with NVIDIA’s, emphasizing that AI software integration can vary by the underlying SoC.
“Physical AI” means AI that can actually control things in the real world. Instead of just recognizing images, it helps make decisions that move or operate systems.
It means the car computer is designed to behave in a predictable way. The timing is consistent, which matters when the car needs to make safe decisions every time.
Agentic AI is AI that tries to do tasks, not just answer questions. It can take steps to reach a goal, which is powerful but needs strong safety limits in cars.
“AI defined vehicles” is a way of saying the car’s main features are driven by AI. The hosts suggest it might be more of a buzzword than a fundamentally new idea.
This means the car is designed so software is the main “organizer” of how different functions work. Instead of hardwired behavior, software components can be combined and updated.
CUDA is NVIDIA’s toolkit for making computers do certain heavy calculations faster. For AI in cars, it can help the hardware run AI tasks more efficiently.
The Plymouth Cuda is a classic muscle car, meaning it was built for strong acceleration and performance. In the podcast, the name is used because it sounds like “CUDA,” which is a computer technology term. So it’s more of a reference to the word than a discussion of the car’s engineering.
Certification is the official safety approval process. It means the car’s systems have been proven to meet required safety rules.
Term
ASLD
ASLD sounds like an acronym for a set of rules or standards the car still has to meet. The hosts are saying the OEM can’t ignore it, even if they buy a compliant platform.
Middleware is software that sits between the vehicle’s operating system (like QNX) and the application software. The segment frames middleware as part of the integration chain that must be safety-certified, which is why OEMs previously had to coordinate multiple vendors and manuals.
“Safety certified” means the car’s software has been tested and proven to behave safely, even when something goes wrong. The speaker is saying that proving this takes a huge amount of effort.
A safety manual is like a set of rules for how to use a component safely. If different parts come with different rules, it’s hard to build one consistent safe system—so the segment argues for one combined manual.
Cybersecurity is about keeping hackers out of the car’s computers. The speaker compares it to safety work because you have to think through lots of different ways an attack could happen.
A cyber attack is when someone tries to break into the car’s computer systems. The question is whether certain software or platform protections can stop that kind of intrusion.
Company
Allocore
Allocore is mentioned as a system or software layer that might help protect a car from hacking. The host is basically asking whether it can block attacks at the foundation level.
A “port” here means a connection point where the car’s computers can talk to other devices. If you expose too many of them, it can give attackers more ways in—so OEMs may need to control what’s available.
CAN bus is the car’s internal messaging system. It’s how different computers in the car share information, and if it’s not protected, someone might be able to send messages they shouldn’t.
Zonal controllers mean the car’s computers are organized by areas, like front, middle, and rear. Instead of lots of separate computers doing everything, the car uses fewer, more centralized controllers per area.
HPC here means the car uses faster, more powerful computers to handle lots of tasks. The idea is that fewer, stronger computers can be easier to secure than many older small controllers.
These are new rules about how secure and resilient connected systems (including cars) have to be. The point is that automakers will need to build security into the car’s design so they can pass required compliance checks.
A flush door handle is shaped to sit smooth with the car’s body instead of sticking out. That helps the car slip through the air more easily, but new rules may force a different design so you can grab it more directly.
This means the car must include real buttons you can press, not just controls you tap on a screen. The idea is that important actions should be usable reliably while driving.
The gear selector is the control you use to put the car into the right driving mode, like Park or Drive. The point here is that regulators want it to be a physical control you can use directly.
HMI means how the car “talks” to you and how you “talk” to it—like screens, buttons, and what the car shows you. The discussion is about designing that interaction to be simpler and more consistent.
This is the difference between the main car system underneath and the “apps” or features on top. The speaker is saying the cool effects are more about the user experience layer than the deep platform work.
They’re using Volkswagen (VW Group) as an example of a big automaker that buys huge numbers of electronic parts. That kind of buying power can affect what gets used in cars.
They mean real physical parts in the car, like door handles and the switches you press. If rules change, those parts may need to be redesigned, which can affect lots of other systems too.
A product cycle is how long a car model stays in its main form before it gets a big refresh or replacement. If rules change, the effects can last for that whole time.
Tier one suppliers are big companies that provide major parts or systems directly to the car maker. They matter because the car’s final design depends on what they can deliver and update.
Silicon vendors are semiconductor manufacturers that provide the chips used in automotive electronics. The segment highlights that QNX’s platform strategy depends on tight relationships with these chip suppliers because hardware availability and capabilities affect what software can do.
A foundational platform is the main base software/system that other features rely on. The idea is that if rules or needs change, you can update the car’s behavior without starting over completely.
Changing requirements means the rules or needs for the car keep shifting over time. The hosts are saying the car’s tech should be flexible enough to adjust quickly.
“Informatics” here means formal computer-science-style training—how to design, implement, and reason about software systems. The speaker contrasts that with other engineering backgrounds, arguing that system-level work may still benefit from non-traditional paths.
“System level solutions” refers to designing vehicle software and electronics as an integrated whole, not as isolated components. It emphasizes architecture, interfaces, and timing/behavior across multiple subsystems—where different engineering backgrounds can contribute.
“EE” stands for electrical engineering. The speaker is saying fewer people are choosing the electronics track, because car engineering is becoming more software-focused.
Ottawa and Waterloo are Canadian tech/engineering centers. The speaker is mentioning them as places where automotive software talent is being recruited.
The BMW M5 is a powerful, performance-focused BMW sedan/wagon made by BMW’s performance division. The “Touring” version is the wagon body style, so it has more cargo space than a regular sedan. People talk about it because it’s meant to be both quick and practical.
The Mercedes EQS is Mercedes-Benz’s big, high-end electric car. It’s made to be efficient and comfortable, and it’s also a place where Mercedes shows off its newest EV tech.
The Ford Bronco Sasquatch package is an off-road upgrade for the Bronco. It’s meant to help the SUV handle dirt, rocks, and rough trails better than the standard setup.
LIVE
You know, we see a lot of initiatives in the market today to build a platform, and we're
already two to four years ahead of them, that this is the type of acceleration that we're
talking about.
Welcome to The Inevitable, a podcast by Motor Trend.
Hi there, and welcome to The Inevitable.
This is Motor Trend's podcast, our podcast about the future of mobility, where are we
going, how are we going to get there, probably in an SDV, that's a software-defined vehicle.
As you can see, I am rolling solo for this episode, Johnny is not here with me, and that's
because we're up here in Ottawa.
We're here in Cunix's garage, ready to talk to two experts in the field of software-defined
vehicles.
But first, I have a very important message, which is The Inevitable podcast is brought
to you by Cunix, whose high-performance foundational software powers over 275 million vehicles
on the road today.
Cunix delivers safe, secure, reliable, and scalable solutions, enabling automakers to
unlock transformative applications, drive new revenue streams, and launch innovative
business models, all without compromising safety or security.
The Inevitable podcast is also brought to you by Vector, a competent partner in embedded
software.
For over 35 years, Vector has been helping the industry simplify complexity and accelerate
the transition to software-defined vehicles.
And software-defined vehicles is what we're going to be diving into today with our two
guests, president of Cunix, John Wall, and president of Vector Informatic, Dr. Matthias
Traub.
These are two super knowledgeable folks in the field of everything to do with SDVs.
The automotive industry, they're veterans of both OEMs and suppliers deep within the
software space.
I've known John Wall a little bit longer than Dr. Matthias Traub.
John's been with Cunix for as long as I've known, and we started our relationship with
Cunix about four or five years ago when they started working with us on Motor Trend's
Software-Defined Vehicle Innovator Awards.
They've been great supporters, I've said this many times, publicly and privately, that we're
trying to look deep into the future of automotive, and we really can't do this without their
support, because this is fundamentally an area where not a lot of people are looking
at right now.
So, with the help of Cunix, their resources, their financial support, we didn't even get
to know a lot of people within the SDV space, talked to a lot of people on this podcast
about where things are going, and that includes Vector, who Cunix is a partner with, in the
software, in the foundational software space.
So, I've actually visited Vector's headquarters in Stuttgart, quite recently, a fantastic place,
saw their operation, got an hours' worth of their demos, including the new product, Alloy
Core, which we're going to talk about, so this should be a very interesting conversation with
two of the brightest minds within the space about where software-defined vehicles are going,
what the current state is, and some of the issues that are at hand within the automotive
industry when it comes to going vertical, developing stuff on your own, or working with
key partners.
So, I'll stop talking now. Let's bring the experts on, President of Cunix, John Wall,
and Dr. Matthias Trubb, President of Vector Informatic.
All righty, so Matthias, John, thank you so much for, well, thank you, John, for welcoming
me to your house, this is the nerve center for Cunix, and also Matthias, I visited your
headquarters just about three weeks ago in Stuttgart, very impressive facility, like
incredible artwork inside, I saw a great demo there, so thanks for taking the time, I know
you guys are super busy, so I want to get into it real quick, so for you guys just listening
and watching, I did this in the introduction, but we'll say it again, we have Dr. Matthias
Trubb, head of Vector Informatic, and John Wall, who is the new president of Cunix, congrats,
I think it's still in order, you know, appreciate it, very good, and I was looking at your backgrounds,
you probably saw me lurking around your LinkedIn, that's what I do, I think it's interesting
that, John, you've been here almost 30 years?
Oh, more than, more, I came with the building, yeah.
Okay, so how many years total?
I started in 93.
Holy smokes.
January of 93.
Okay, so you've been here working primarily in automotive, so particularly in the last
15 years, right?
Yeah, last 20 years.
Last 20 years solely in automotive, you've seen the industry evolve at Cunix, who's been
foundational in a lot of all different manufacturers at a very core level, contrasted to Matthias,
who's been at different OEMs, I think a little bit on the supply side, but mostly, I mean,
you've been at BMW, Volkswagen, Cariad, and now you're also fairly recently to your head
position at Vector.
How has this term, how has S.D.V. changed, like what's, and I set this up because when
we first started the relationship with Cunix and Motor Trend and doing the Software Defined
Vehicle Awards and all that, I mean, we were still struggling whether S.D.V. was the right
name, was it self-driving car, was it, what was the other term, case connected, autonomous
share, electric, or the other one, right?
And the only reason we picked S.D.V. truly was because GM had named their division S.D.V.
and Hyundai had as well, and I was like, okay, so two big OEMs are calling it S.D.V. I think
this is the right term for the future of automotive.
So what has changed, I would say, in the last, you know, Ed, even going back to 2016, Grant,
who's a member of my team, got a domain name, a Software Defined Vehicle, so we were thinking
about this along, you know, I think the ambitions within the car makers, it's been around for
a long time to add a lot of features. I think what's changed is there's been a lot of learnings
along the way, and I think a lot of car makers thought they needed to own the entire vertical
stack, ran into a lot of problems trying to do that. I think the ambitions are the same
as to be able to provide an experience with the vehicle that's branded around software.
We talked about this the last time, but I think there's an evolution in the thinking
about how to achieve that goal, and that maybe there's a balance of buy versus make, and,
you know, more of a platform approach. I think the goal hasn't changed. I think the approach
is changing.
Okay, would you agree?
Yes, fully. So I remember, therefore, I love before we started in 2014 at BMW with the
case were already already software was one big portion of it. And we started at that
time with a so-called software platform approach brought out at that time with the seven series
in 2015. And then step by step, we improved the focus on application level on one hand,
but on the other, in all my jobs, we have still a lot of challenges with the underlying
software platform. So we have a robust and more evolutionary development in place. And
nowadays, the VMs, especially on the Western side, they have to focus more and more on
what matters most for their customers, because they are so much enforced by the Chinese car
companies. And therefore, they have to rethink and say, okay, the software platform, it's
commodity now. There are companies like QNX and Vector and some others who can take care
of it. And they can push their capacities from a development perspective to the real
application and developments for their customers. I think, hey, this is great stuff. It's the
cutting edge technology on ADAS, IVI and so on.
I heard something quite recently from, Ford's got this Skunk Works program out on Long Beach,
very close to our office, and they're building their next universal EV platform. And we had
a call, and he was saying that the term SDV, now is the usual point where it's almost kind of
performative. It's a little bit of marketing speak, whereas maybe some car companies are really
good at it, but others are not, and they're just sort of mouthing the words, you're not.
I mean, SDV, it's a label. It's a label. I mean, when we started this journey at four years ago,
I remember describing SDV as something very difficult to describe. And that B, what it really
meant to me was that more and more features that were going to be important to the customer would
be based on software, whether it's from an entertainment perspective, or whether it's
from safety features within the vehicle. And the idea is that you need to make your vehicle the
most compelling vehicle, and maybe the focus is less on how the car drives, and is more on the
experience within the vehicle from a feature perspective, and from a safety perspective.
I think definitely there's some that are better than others, but I don't think, I think every car
company is trying to achieve the same goal, which is to enhance their brand, enhance the value,
provide services in the vehicle, and find other revenue streams than just the vehicle and service.
Okay. And you agree? Yeah. Yeah. Okay. So, from your perspective, and I think you just called it
out, the key definers of an SDV are going to be that in car experience, the value it brings to the
customer, whether it's an IVI in vehicle infotainment system, or advanced ADAS systems,
right? I mean, these subscription services, I feel like are part of that, but also have been a
little bit of a roller coaster, especially on the consumer side. Is there anything else? What would
you say would be for the lay people out there? What is the defining factor of the future of the
car? That's what this podcast is about. I always come back to, I think, what we talked about again
at the start of this is the ability to OTA a car and to make it sort of fundamentally different
from the last update. Is that still whole true? Is there something else on the horizon that you
think is more key in terms of all the, I mean, I've looked at BMWs, got the new fancy display
system where they're projecting things onto the screen. I looked at Audi's, got the VR, also talks
to you through the headrest. Of course, there's everybody's chasing, I think Tesla, FSD, in terms
of point-to-point navigation. What else is out there in terms of like on your very front of
mind for the team here in Ottawa and in Stuttgart? Yeah, so for me, it's the in-car experience,
but for me, at the end, it's a complete journey what we have had in mind. So really,
when I have my BMW app or Mercedes app, I have a good interconnect to my car and
can already plan a route and send it to the car. So the interaction with my IoT system,
so that the car is really integrated in that in some areas not fully solved. And then the second,
for sure, especially the level three and four steps towards the automation decrease
for self-driving vehicles, this is the key topics in the near future.
You know, one anecdote I heard from somebody in Germany just recently was the goal should be to
make the car more comfortable to drive, to reduce stress. And so you're in a strange city, you don't
do the city well, you rent a car and it has features in it that make that a more enjoyable
experience. And I think it's the interconnect, it's the autonomous or the automated driving, etc.
Yeah, it's definitely also, as you said, it kind of starts with the phone too. The phone is a high
bar for any consumer in terms of interaction with the device and how seamless it can be.
And we, and a lot of manufacturers are still not, I mean, there's still manufacturers to say
where the phone is key is very clunky, spotty just to get into the car, let alone look at the
charging data, find a charger, all that stuff. So 100% agree. I want to come back to what you
said, Matthias, about the challenge of China because, I mean, most of the successful, the big
buzzy, what they call new energy vehicle manufacturers over there, they've been startups
within the last 15 years maybe, and primarily EV, although BYD has also done a lot of
like hybrids and things. But there, how is it that they have their foundational software
sorted? Like what, because a lot of them learned through JVs from the OEM partners you guys have
worked with for decades, right? They learned how to do it there and then what, they just abandoned
and moved into their own software stack, their own, like, how did, how did they get to where
they're, to where they're at? I think that the biggest benefit what they had is that they can
start from scratch. So they don't have to deal with legacy, also thinking about, okay, I have my
car program in place and want to use some COP parts. So take over some ECUs or functions from
the form of a car line, because I don't want to reinvent the wheel. But with that step, you have
always the necessity to deal with the former APIs talking from a software level and so on.
And so they started. So they have launched their first vehicles. And now step by step,
they will having my hypothesis, more or less the same challenges, like the established OEMs,
especially they develop their software platform by themselves. So they have a lot of heterogeneity
in place. And this is the idea why John and I said, Hey, let's sit together and offer something
where our customers can focus on what is the real asset for their customers.
Yeah, yeah, it's like Xiaomi and Huawei. These are phone manufacturers who just come in and drop
their essentially their mobile software system into something now that has wheels.
Yeah. But I mean, I think to assume that the Chinese have developed a ubiquitous platform,
they haven't. But they're now looking at that because they know to be able to maintain the pace
that they basically have created a situation where there's an expectation. And so I think
they're looking now at how do we platform and to be able to reuse their assets going forward.
You kind of teased it, Matthias, the you and John sitting together to create this new product,
you debut this at CES, it's called Alloy Core, right? And I was thinking about the name. I was
like, let me just check my logic, right? Your marketing teams probably have hundreds of names
on a whiteboard somewhere, but Alloy Core, Alloy because Alloy is a metal of two or more
you know, elements, in this case, cunix and vector, right? And then core because it's
foundational. It is at the lowest, like most basic safety firsts, like all of the core elements,
and then your partners build on top. Did I nail it? You nailed it.
That the name has done its job. Okay, good. And it's a K because I'm sure Alloy Core,
CES is a little over a month ago. What's the reception been? How do you guys,
I saw, I saw like a press release come out, Mercedes is looking at it, like other OEMs,
other partners, and fully engaged. Yeah, I mean, I think, I think it really resonates. Obviously,
it wasn't new to a lot of our, our customers, we'd, we've been working together for a while and,
and we've been talking to potential customers. I think there's a recognition that this is
something that's required to ease the development of the systems within the vehicle going forward.
I think it's not a hard, hard sell from the value proposition perspective or from
the need for it. I think we're seeing very good reception. We're seeing that technically,
it certainly resonates very strongly with our customer base. And, you know, obviously, we have
a lot more customers that are looking at it and working with it than the press release.
So we, we, I think we feel very comfortable. We still believe there's work to be done on our side
to be able to sell the value and the proposition, but we, we, I think we feel pretty good.
Yeah, I come on the line that's
for me. Okay. Yeah, we chatted with Justin and Mark from your teams at CES about it and
went to just like whatever a couple of days after it and launched. But for those who are maybe new
to this, can you just, what, what would you say is the top three, top two main value proposition
of Alec or four, a major car manufacturer? For me, it's first of all, that our customers
get something pre-indecrated so they can safeguard development time, development costs because,
because we are doing the job before they start with their development of an ECU or a new platform.
This is the, the, the biggest difference compared to the time before where we are
asked by a new MHAV. We need an operating system. We need some kind of a middleware hay vector,
please offer something. And then we are starting from scratch. Now the starting point is much higher.
So you can have the development time shrink down to two years compared to four years before,
for example. The second point that is referring to the alloy, it's, for me, it's some kind of
effusion. So our technical teams are looking how we have to combine our communication middleware
closer to the operating system to, to have much more performance on one hand. And in addition,
also reduce the, the, can have a reduction of the hardware costs because the power consumption
and so on is less. Yeah. Yeah. And I think, I think to add to that, there is also, so I mean,
time to market, obviously, production and development costs as Mattias out late and
much higher performance of the parts being very, very tightly integrated and a lot of know-how
that is very specific to QNICs and to Vector that is very hard to replicate by our customers.
You know, we come from very specialized embedded software backgrounds, not a lot of companies
build operating systems, not a lot of company do the type of work that Vector does the middleware.
So we're very suited at it, but it's also, there's a commercial model around it. There's
reps and warranties around it. It's safety certified the entire, the entire stack, which,
you know, never underestimate the effort that it takes to do that. You know, we see a lot of
initiatives in the market today to build the platform and we're already two to four years
ahead of them. Right. That this is the type of, of acceleration that we're talking about.
Okay. I'm understanding the benefits Mattias mentioned, you know, hardware, the software,
you talked about the cost savings, but back to Mattias from, from your OEM background,
what does that look like from a, like a software, like a development team, like the,
like the people side of it, right? How many people would you say a major OEM
would no longer have to devote? If they decided, wait, we're going to use Alec or
we can focus now, we can take this team or this many people now and convert them to focusing on
the layers on top, the, the IVI, the, the digital cockpit experience, the ADAS. Like what,
you have any, like round figures, like what that looks like from the savings or?
Nope. I have clear, clear figures because in my former chops, the teams were in, under my
responsibility. So point taken. So I have to answer that. So my view is between 50 and 75.
Soft, like hardcore software, top talent. These are the guys you try to.
Software guys focusing on, on, on the software development. So the platform and also we are
talking about testing integration. Therefore, for me, the, the, as what I mentioned, the pre-indecrated
solution. So it's on one hand to bring the stuff together, but at the end also to test
that robot system in place. And in, in formal times, you have always in parallel the application
development and the platform development. And then I call it, it's always the, the big bang
approach. You bring everything together at one point in time. And then the, the, the OEMs are
wondering why it doesn't work. And then it's not due to the failures or errors in the application
level. It's underneath in a platform. And this is why John and I said, Hey, let's do it before.
And then we can offer something which is already robust. It's compared to, you don't buy a iOS
or a Windows, which is not still well working and put the, the, the word in the Microsoft
Office suite on top of it. All right. That makes a lot of sense. But so then I'm struggle with,
and maybe this is for John, like why would, and I sound like a total shill, but why would it,
so why wouldn't OEM like decided like, like this is, they want to do their own, like,
then we talked about vertical integration. I feel like it's really a sexy term because,
oh, Apple, Apple kind of started it, right? And they're doing their own chip and now Tesla
and now some of these other manufacturers say, we're going to do our own show all the way down.
We're building our motors, that whole thing. But there is, you make a very compelling reason to.
But you just nailed it. It's really that simple. There, there is a perception that
the big successful companies out there like Apple and, and Tesla, they did a full vertical
integration. And that that's the way you, you end up doing it. And it's, it's philosophy.
I think there's philosophy at some of the car companies that they need to own everything
up the bottom. I think that what we're seeing is that philosophy is changing.
There is, you know, even if a car company believes they have to own
most of the stack, they're starting to say, well, maybe I don't need to own this,
which so it's new to the industry to have something unified like this. I think,
you know, there's also concern about lock in, you know, there's one vendor or a couple of vendors
that are providing a platform. But we believe strongly, and it's a case we have to continue,
continuously make, is that there are so many benefits and the benefit is just, is not just
performance. It's not just technical. It's commercial. It's time to market. To your point,
we think it should be fairly obvious that this is a, an important step that the industry has to
take. And where would the phone industry be if there was no standardization on a couple of
platforms? Exactly. Okay, fair point. Well, then let me ask, and this might be a really basic and
dumb question, because a lot of OEMs now are starting to call out that they're, they're,
they're going to build their own silicon. This is, of course, agnostic to that, right? Like,
it doesn't matter. You do whatever you want on, on top of that. Okay, great. Inclusive of, and
they're nodding their heads for those of you listening. Yes. Violent agreement. What about AI?
Is that also like, I know AI is kind of going to be in everything we do, but certainly for the,
there's some really, I would think foundational things were, I mean, we're using a, well,
not we, the car companies using AI to analyze ADAS, whatever, the actual footage or simulation,
like also agnostic to whatever AI vendor and OEMs going to use?
I mean, the platform is going to be agnostic. I think there's still a challenge with AI that
frameworks are specific to some of the SOCs. For instance, NVIDIA have their approach.
System on chip. Yeah. Qualcomm have their approach. I think the way, the way we
approach AI is we are agnostic to it. Obviously, we support all the frameworks that, that work with
it. But really what we're focused on is the next wave of AI, which is physical AI. This is where
things get controlled by AI. Robots. Bingo. And that the way we're positioning our product is that
in the industry, whether it's automotive, robotics or others, you need to have a robust,
you need to have deterministic software platform, real time platform to support that physical AI
story. And that's kind of our approach to it, is that we're this foundation that has all the
properties that can make physical AI a reality. Okay. Okay. That's good. Because I think there's
a lot. I mean, it's moving so quickly. AI, I mean, what, two years ago, it was a gen AI.
And then the year after was agentic AI. Now it's physical AI. And now I'm looking at,
it's the lunar new year. I'm looking at clips of Chinese robots doing backflips and kung fu, right?
So a little, a little troubling in some, in some aspects. It, well, just staying on that for a
second. When we, we had a good chat with the folks that are in video. These of you guys know
who that is, the big system of chip manufacture. And their claim is, wow, the era of SDV is over.
It's AI defined vehicles. Is that, that seems like marketing. It's a little spike. At the end,
as we said at the beginning, SDV is a, is a buzzword. Yeah. Yeah. So for us, it's,
doesn't matter. So we can call it SDV or ADV or, right. For beer. These are software defined systems.
And this is the idea that we said, okay, the alloy core platform is not automotive focusing
only at the end. It's the foundation for our software defined systems, enabling also the AI
capabilities. So the, in the action with the dedicated stack or the specific stacks from
NVIDIA with their CUDA framework or with Qualcomm and so on. And on the other hand,
when we are talking about AI on one hand, you need the AI in the system. And on the other hand,
that we have also a good solution for, for the software developers that they can save time with
a good co-pilot in our software development framework as well for alloy core. This is
the second part when we are talking about AI. So you talk about having, you know, alloy core
being foundational and basically all the certification, the particularly safety. I
mean, just like why, how much time, energy, resources it saves that they can, an OEM can
essentially buy into the alloy core system and not have to worry about what ASLD, all of the,
all that top level compliance. So I mean, the OEM is always going to have to worry about ASLD
because the car is certified as a unit and we're providing components, but you know,
security compliance is also as big as safety now. So there are standards 21,
4, 3, 4 that are for, for safety, for security that we do as well. And that, you know, the vector
does as well. In the past, as Matthias talked about earlier, you know, the OEM would call,
we need an operating system to buy QNX. So we need, we need a middleware, we're going to buy,
we're going to call Vector and we would deliver our software all safety certified or all safety
certifiable, but we would have a manual on how to use it. They'd have a manual on how to use it.
So on its own, it's a lot of work, just safety certifying the components is a lot of work.
It's more than the development, let's say. Okay. So the testing and certification is a huge part
of the job. Now as an OEM, I'm getting all these disparate components that all have a safety manual
that by the way, they don't agree with each other. They're not asking you to do the same things
because they're done by independent companies. What we're doing is we're saying, okay,
Vector, we're going to work with you. We're going to safety certify the entire platform and we're
going to generate one safety manual. And we're going to agree that these are the conditions
by which the developers at the OEMs must follow to create a safe system. For me, it's, and the same
goes for security, very similar type approach. To me, that's one of the more invaluable parts
of the platform. There is the technical part, there is the performance, but getting that
pre-certified platform is, to me, is a very, very strong selling point of the platform.
Okay. And when you say security, specifically, like cybersecurity, like preventing all the hacking,
all the different approaches... It's the same as safety. It's thinking about all the cases. It's
very similar to doing safety analysis. Okay. So to break it down into real terms,
for again, our lay audience, occasionally we hear about, oh, there's some new entry point
into this OEMs. The kids are buying flippers off of the Amazon and they're able to record and then
go in. Is this something at a base level like Allocore can help guard against from a cyber attack
perspective? Or is, and I know I'm probably laying it on you guys a little thick here, right?
Or is it the OEMs also have a role to play in this in terms of what ports they're exposing
or what protocols they're using? Oh, yeah. I mean, look, at the end of the day,
the responsibility for security is the car makers. Okay. It's their responsibility.
What we're trying to do is we're trying to provide a platform that, at least at the level of that
platform, they don't have to worry about those topics. We've taken care of it for them. But
I can't, if an OEM puts a sensor on the car and they choose not to secure it in some way,
and it's connected straight to the CAN bus, and somebody can go and talk to it and get right
on the CAN bus, that's an architecture discussion. So the car makers are still responsible at the
end of the day for the car being safe and secure. Okay. That's the one point and the second point
what we want to offer is also that we can support them. Yes. That they have a safe and secure system.
So we know our platform and based on that knowledge, we can also help them for a bigger
integration step as well. And this is also a big benefit when our customers are working together
with us. Yeah. And you know, from when you hear those stories about some of the security entry
points, don't forget there's still a lot of legacy in vehicles. Right. There's still a lot of old
systems in vehicles that are going to evolve as we have zonal controllers. And I think
that's when what we're doing around security is going to have much more of an impact on more of
the architecture of the vehicle is HPC focused high performance compute versus all these individual
legacy ECUs. Right, right, right. Okay. So you're talking, yeah. So there's all the stuff that's
already in the car. You have to worry about a little bit of the backwards compatibility problem.
And then you also have to look ahead to changes coming to the marketplace, particularly with
like regulations, right? Like there's one, the Cyber Regulatory Act, Resilience Act, which is
coming in like next year, right? So you're going to build this into ally coordinate your other
future products. And again, it's on for the benefit of your eventual OEM partner, right?
For somebody who's, we have a fair amount of car nerds who listen. Some of them have kids.
Some of them are in school themselves. You clearly need an engineering degree,
probably this day and age, computer science or computer engineering or E or what do you think?
So in some areas, it's really necessary to have the real informatics background. But from
my learnings over the last 20 years, especially when you have topics considering system level
perspective, also mechanical engineers and especially physicians and from chemistry,
they can learn that also. So we have a special program at Vector where we said, okay, we don't hire
but only informatic guys also from electrical, mechanical, mechanical industry because they
bring up another perspective, especially when we are talking on system level solutions.
So this is what we are currently doing. So I know you're tight with all the Canadian
universities trying to build the top grads. Absolutely. And in India as well, where we
have an engineering center, I would say you're starting to see less mechanical engineers heading
up EE departments. So I think it was more traditional because I won't name the OEM,
but I remember sitting with the head of electronics at a car OEM and the plaque he had in his office
was a transmission he developed for that OEM. But he was a good administrator. He was a good
manager and he had good people below him that were much deeper into the software tech. But I
still think if you look at some of the car companies, you'll see they've brought in executives from
Apple, they've brought in executives from Tesla that have very deep software backgrounds.
Let's utilize this opportunity, this moment, this platform. We have a pretty large YouTube
audience and if you wanted to tell a junior in college in computer science or computer engineering
or whatever the favorite top candidate, why come into automotive? Because they could go work at
an Apple or what we used to call fang companies right now, it's a magnificent seven. They could
go work in the tech industry. So why come into automotive and why specifically to vector or
to QNICs? Why Stuttgart? Why Ottawa or Waterloo? Especially where there's like stone on the ground.
I'm not sure those are the selling points. I think on one hand, from my perspective,
automotive is still a very emotional stuff. So you create cars and so on. On the other hand,
talking about QNX and Vector, but please, John, we are really the top experts on software technologies
which you find also at Apple or other big tech players. So you have a nice mixture of great
products at the end, emotional products, and also the contacts to a lot of different OEMs.
So you've got a lot of different perspectives. You can learn a lot what's happening in the
industry and also we are smaller companies. So you have as a software developer more freedom to
interact, driving decisions forward in a small team and so on. And from my point of view, this is
really a really beneficial thing and I know what I'm speaking. Right, because you've been at the
Vector. No politics and stuff like that. You can really focus on the technology
and the solutions for the specific projects and not fighting for numbers and making big
PowerPoint presentations and stuff like that. That's a key. For those of you who have not been
in the corporate world, this man is speaking the truth. The bigger the company, a lot more layers,
a lot more internal selling before you can actually get stuff done. And before John goes,
I'll just also say, Vector's headquarters? Pretty nice. I was like, wow, look at this place.
No, it's huge. Super nice. I think competitive with Silicon Valley in terms of sort of like,
again, incredible artwork in the lobby. The foyer with the restaurant. Okay, John, so why?
Why? Why? Why Cunitz? I agree with everything that that Matias said. I think, you know,
when we first got into Automotive, we got bought by a company called Harman and it
married two of my favorite things, high-end audio and cars. And so very emotional connection there.
But I think beyond that, if you think about the car and the software journey of the vehicle,
it's probably the most complex consumer device in the world. Way more complex than a phone,
it is the most complex thing in the world as far as software is concerned today.
And it can kill you if you get it wrong, right? Absolutely. Like the phone breaks,
it breaks, you restart it. Yeah, it's not a big deal. And what we're seeing now in the industry,
and this ties back to the physical AI, is everything that happened in the car or is happening in the
car is going to happen in robotics, is going to happen in medical, is going to happen in oil and
gas. This is going to be the most rapidly growing area. We do not have enough highly qualified,
embedded software engineers. We have lots of application engineers that can write applications,
but this is a very specialized field. And if you're really serious about doing
hardcore software development, this is the place to be. We are very unique companies.
They're not a lot of companies that do what we do.
That's why it's president. That's excellent in terms of selling it, because I think that's the
point that people kind of forget, right? Like here we are, we spent 45 minutes now talking about
software, but the wrapper around where the software lives is sexy, right? Can be very,
very cool. And then also, but you know, a lot of people buy them, particularly in North America
and in Europe, right? Like this is the one thing we've always said our motor trend is,
you know what's great about the car business is for most Americans, it's going to be the
second most expensive purchase in their lifetime after their home, and they're going to make it
multiple times in their life, in their lifetime. And now it's a case apparently this next generation
can't even buy a home. So, you know, right? So they're going to buy a lot more cars, right?
But also, so it can be sexy. It's also at some level, people just need to get around. So whether
you love cars or you don't, you're still going to probably buy one or at least be in one, right?
And then as we've talked about this, all this cool stuff that's coming into the vehicle
that has never been seen before, like the fact that cars are driving themselves that you can
watch, you know, movies while you lay back and it takes you to your night on your journey,
all the fun stuff, which is John rightly pointed out. And that's just where it's happening
automotive. A lot of these similar experiences are going to go to other industries or just in
different ways that we haven't thought about. So, okay, well, this has been great. We're going to
close it out here pretty soon. But I do want to, because you guys both mentioned in the last question,
sort of the passion, the excitement around being in cars. So let's,
we've talked a lot about software, talked a lot about Alloicor and what you do as a company.
Let's talk about you as engineers in the car business at the top of big companies.
What do you drive? You still got the M4? Yeah, I just got an M5. Nice. Touring. Okay. What color?
Charcoal. Okay. Did you do anything fancy? Did you get any? Check all the boxes? It has everything.
It has everything. All right. It wasn't even an option. It just comes that way.
So you can't, you trade the, yeah, I actually regret trading in the M4. I would have kept it.
Okay. You're running around here in the snow. You got snow tires on? Yeah.
Yeah, we do. Of course, maybe Bavaria. I got snow tires. It's four wheel drive.
Okay. Good. Anything else? You got the sedan, you got the wagon. I You got the
wagon. Yeah. Yeah. Nice. Okay. I like that. Okay. Can you beat this? Mathias, what do you got?
I'm more serious. So Mercedes EQ. Okay. Serious. EQS. Yes. Okay. Nice. Nice. EV. I like it.
EV to the bone. Did you check out that S-Class premiere for the new S-Class for the 140th anniversary?
Very nice car here. Nice car. Yeah. I think some fun stuff in there, for sure. Any other fun
vehicles at home? Nope, no. I drive a BMW R1200RT. Okay. Which is a lot of fun. Nice.
And my wife's car, which I like just as much is Bronco, the Sasquatch package. Oh, no kidding.
Okay. Fun vehicle in a totally different way. Very serious. Okay. So this is great. So inspiring.
If you didn't know, if you work at either company Vector or Cunex, you now know your bosses are
giant car nerds as well, which is great. Yep. So this is a great conversation.
Again, if you want to check out Alloy Core, it's AlloyCoreWithAK.com. There you can find out about
their new release from CES, also linkouts to, I think, their sites. Thank you, Dr. Mathias
Trauk. Thank you for having us. Yes. And John Wall from Cunex. Again, it's fantastic to have you
on. And best of luck with all of your partners and products that you are launching. Thank you
very much. And thanks for coming to Canada in February. Yeah. And watching, well, we just
watched the Olympics in Canada once. Good job. Yes. All right. Thanks, guys. Thank you. Thanks, Ed.
About this episode
Software-defined vehicles (SDVs) are framed as the future of mobility, but the hosts question whether the label has become “performative.” They walk through how automakers struggled to own the whole stack and why a platform approach—paired with virtual machines, OTA updates, and deterministic real-time software—helps. QNX and Vector are positioned as foundational layers that simplify development, safety certification, and cybersecurity complexity, including unified manuals and reduced integration risk.
On this special episode of The InEVitable, I ventured out to Ottawa, the capital of Canada and home of the engineering side of QNX, the automotive software supplier and longtime sponsor of MotorTrend’s Software-Defined Vehicle (SDV) Innovator Awards.
Recorded inside “The Garage” at QNX, where the company’s engineers build and test all sorts of robotic and automotive software and hardware, I met with QNX President John Wall and Vector President, Dr. Matthias Traub for a conversation about the current state and future of SDVs, including what these two companies are doing to make the transition from traditional electro-mechanical automobiles to the electrified, connected, and possibly autonomous vehicles of the future.
