skip to Main Content

The Garage Podcast : S4 EP6

How are AI and SDVs Reshaping the Supply Chain?

with Derek De Bono of Valeo

Recorded live at CES 2026, John Heinlein and Valeo’s Derek De Bono discuss how Software Defined Vehicles (SDVs) are revolutionizing the industry through hardware-software decoupling and continuous AI-driven updates. Their conversation highlights the shift toward partnership ecosystems and predicts that SDVs will redefine vehicle longevity, business models, and the path to Level 3 autonomy before 2030.

Listen to audio only version:

Episode Transcript | How are AI and SDVs Reshaping the Supply Chain?

0:00 Introduction to Software Defined Vehicles

Today in The Garage in an episode recorded live at CES 2026, our guest is Derek De Bono, who’s Vice President of Software Defined Vehicles at Valeo. Valeo is a famous tier one supplier in the industry, and they’ve been doing incredible innovation in software and evolving their systems and vehicles to, Software Defined Vehicles. In today’s discussion, we have a wide ranging exploration of how Software Defined Vehicles are evolving, what are the opportunities and challenges, and we look forward to the future in how vehicles are evolving into the future. It’s a very exciting conversation.

Let’s go!

0:44 Introduction to Derek De Bono

Welcome to The Garage. I’m John Heinlein, Chief Marketing Officer with Sonatus. We’re here at CES 2026 at the Sonatus booth in our podcast studio.

And we’re so excited today to welcome Derek from Valeo. Derek, welcome to The Garage.

Thank you. And it’s great to be here.

Please start by telling us about yourself.

I guess, when I think about myself, I’m a engineer by training. So I started off…I don’t know how far back you want me to go?. What kind of engineer what kind of engineer were you? I’m a mechanical engineer. Okay. I grew up, and I lived my, my whole life in Montreal, until, my, I guess, early thirties.

Then I branched out, joined Ford Motor Company, spent several years working at Ford in Canada and the US, and then made the big move to Paris, moved to France, and joined Valeo.

First in a product marketing role, I’ve been and I still am in charge of product marketing for the group.

But as we started seeing this new thing emerge we have called Software-Defined, we started getting more and more interested in what’s happening with software and how Valeo should transform and what should transform within Valeo to take advantage of this new opportunity.

I became in charge in charge of that, what we call the software defined vehicle transformation at Valeo.

That’s great. I wanna get into that in more detail. You gotta start with us a fun fact. We like to get to know our guests. Tell us a fun fact about you.

Yeah. I guess a fun fact is always it makes you think a bit what’s different in your life and what’s unusual. I am Canadian. I grew up in Montreal, moved to Paris, created a family in Paris.

But we have brought a bit of Canada and a bit of Montreal to Paris. We have three kids, and all of them are active hockey players. So maybe when you think about Paris, you think about croissants, so you think about the Eiffel Tower. But now you think you think more and more about hockey, and it’s great with Montreal this year. There’s even a French hockey player playing for the Montreal Canadiens, so it’s kind of fun seeing that little link in there.

That’s exciting. And do your kids play different positions?

They’re all defense. They’re all they they’re all the first, the eldest, was a defender, and they all folded in his tracks, and they all became defenders. They referee as well. So, it’s been fun to see, you know, learning how to play and then learning, how difficult it is sometimes to to judge what is a what you’re able to do in the sport.

I’ve pushed them as well to become, to become refs and and to learn how you have different elements of the game and you have to keep your control.

That’s great. I always say some fun fact back. My daughter this last past summer took up pole vaulting. Oh, wow. Which is really interesting, very unique sport.

And we’ll see if she sticks with it. It’s so interesting and she was able to do quite well. But so there you go. It’s fun sports, I’m glad you’re able to bring that.

Tell us about, Valeo. Tell us about the company for those people who may not know. Valeo is a famous tier one supplier in the industry, but you do a lot more. I know you’re looking to expand the role.

3:48 Overview of Valeo

So tell us first about the company and a little bit of background of the company.

Valeo has been around for many years. We had, our 90th anniversary a few years ago. Great. So, we started off as a really…mechanical systems, within the vehicle.

Today, I think what’s driving Valeo forward is really, working on the major trends in automotive. So everything to do with electrification from 48v mild hybrid systems all the way up to 800V full-electric vehicle systems. Everything around lighting systems to see and be seen and make sure that we’re doing everything we can in regards to safe systems in the vehicle. And then I say for last, the part about ADAS and AD and automation, the transformation that we’re seeing there when the responsibility in the vehicle is going from the driver to the vehicle itself to make sure that we’re not only protecting the protecting the person in the vehicle by avoiding accidents, but we’re anticipating other road users and and making sure that, the experience, and mobility is as safe as it can be in the future.

That’s fantastic.

5:01 Transition to Software Defined Vehicles

So let’s let’s shift over to SDV then. You’re responsible for SDV strategy and and and implementation at at Valeo. Tell us about how you’re considering, that train that change and and what are some of your strategic focuses in that area?

When we first started detecting, the separation of hardware and software, we we took a look at the role of Valeo within..we’ve been supplying embedded software systems into components for years, whether it’s a parking system, a lighting system, or even different powertrain-related systems. What we’re seeing is as we split, is software defined really a transformation of software or is it about hardware? It’s neither, it’s both, it’s all software defined.

It’s really about the vehicle and how we approach the vehicle in the future with the emphasis always being on the user. The user experience, we should be taking that to a new level. And how we take that to a new level is by always being able to adapt the vehicle to the driver, to the consumer in the vehicle. And we can do that through software.

So I think that is we can’t lose focus on, you know, why are we doing this, and we’re doing this to make sure that we have the best possible experience within the car. Software defined is allowing us to add the functionality, but we can’t do anything with software if we don’t have the hardware to go with it. So it gives a value of enormous opportunities, of course, on the hardware side, but also on the software side, and really the system understanding on how you put that together from our years as embedded software system suppliers. We differentiate ourselves by understanding how the vehicle goes together and, what it means to deliver a system to the end consumer.

I think the shift to SDV…First thing, I like everything you’ve said that it’s it’s not just cram more software in. It’s, I think, it’s a more holistic way of thinking about it from a system thinking, thinking about decoupling hardware and software in different ways than would have been done before. Because historically, the hardware and the software… software is always there, but the hardware the software was tightly coupled. Oftentimes, it was never changed. And so the architecture choices you made, things you cared about were different than in a world where you know that maybe software is going to coexist with other software, software is going to change over time, capabilities will change over time. I think it causes different thinking.

7:44 The Role of AI in SDVs

Exactly. Exactly. I think the big change there is AI. When we think about the software changing over time, the more and more AI systems, whether it’s end-to-end or however you want to call it, the vision language action models that we hear a lot about at CES this year.

These are learning models. So as you get more data, you see more edge cases. You don’t want to say, “Oops, the vehicle is already in production, we’re not gonna touch it.”

No. You want your your driver, your consumer to get continually updated versions of the software to make sure that they’re living the best experience in the car, that you’re capable of giving them. And we’re able to do this because of software defined, because the software is now centralized, into either a domain controller or a central compute unit. It’s connected to the outside world.

The the vehicle is always connected. It’s a connected device, and the, the car manufacturer is able to, over the air update, that algorithm to make sure that they’re getting the best experience possible.

Yeah. Just before we started recording, we were out in the in the demo floor, and we were talking about just that, an example where by monitoring the way the vehicle operates, by monitoring environmental situations, you can make improvements, you can improve safety, improve capabilities. We were just talking about that before we started, so it’s very, very topical.

9:03 Collaboration in the Automotive Ecosystem

How do you think SDVs, the shift to SDVs is changing the way automakers and suppliers, writ large, you know, can have to interact?

Yeah. I think when we think about, how we’re acting today, we’re interacting much more as partners.

Over years, there was also a hierarchy in automotive, a pyramid of sorts between the car manufacturer and the different levels of suppliers. We’re seeing less of that to deliver the real state of the art vehicle systems, the SDVs of today. Everyone has to work together, and there’s no one’s more important than the other. Everyone’s delivering key elements of that vehicle.

So it’s created much more of a ecosystem, a partnership environment, and it’s refreshing. It’s nice. It’s nice to see that it it works. And when it works and when you see it really working fluidly in between all the elements, it’s incredible what our industry can deliver.

I mean, historically, in older systems where the software is more deeply embedded and and deeply, tightly coupled and deeply embedded, OEMs cared a lot less, about what the specific software did as long as it achieved the mission, I guess, I would say. Do you see that changing where they’re more interested in, for example, understanding the architectures that are inside, understanding and having access to the software. How do you see that their level of abstraction, if you will? Do you see that evolving?

It’s evolving, but the vehicle is complex. There’s a there’s a limit to, being able to see everything and understand everything. And that’s why I chose the word partner because you really have to rely on your ecosystem and the expertise of each and what they’re bringing to the table.

You need to have a understanding, but not a complete understanding of everything, I would say. I think the vehicle is too complex today to understand everything.

And if you have a partner bringing maybe the VLA model, the the vision-language-action model, I think you have to delegate that to them. And they’re the masters of of that really core piece of of software.

There’s no way a car manufacturer could become an expert of that at the same time as as they are. So you need to understand, you need to be able to to put it all together and deliver a vehicle to your customer that you’re proud of and and responsible for.

But I don’t think you have to be the expert of every different component.

I totally agree. And I think there’s no doubt there’s still gonna be the right, I would say, a smart division of labor between the parties. In a world you talked about lot about software updates and and over the air updates.

How much do you see the pace of OTA updates changing? We chatted about that a little bit beforehand. Do you do you because historically, the OTA updates were relatively infrequent. And even in vehicles now where you’re starting to see OTA updates, it might be every six months or every three months or something. Do you see that that’s changing?

Yeah, definitely.

I think, more the incumbents, the traditional, car manufacturers as we, as we know them, they tend to be a little slower in the updates. They’ll group many small updates into, major, updates. I think it’s more the way we used to do things in the past.

You look at some of the new players, and when I think about new players, either on the West Coast in North America or in China, the iteration and the number of updates is probably two, three hundred a year.

Right.

They go a lot faster. They’re not waiting for a big one to to add a little small ones. I think that will probably be the trend more in the future.

But I understand why we group because we need to test and validate. We have digital twins running in the background to make sure that the interaction is working with everyone.

But by doing many small ones, I I think if we see an issue, we can always pull back. I think that’s a bit of the approach to some of the some of the companies as well.

13:18 Opportunities for OEMs and Consumers

And then, you know, we were talking about the the process on the front side, I think. But then on on the flip side, what do you see are the opportunities, you know, for the OEMs, for drivers, for value creation, for, you know, loyalty and so on that SDV can bring perhaps compared to what would have been possible before?

Yeah. I think SDV is is a huge opportunity for the end consumer.

The the the pressure of specifying all the options you want on the vehicle when you buy the car, it could technically disappear in the future. Because if maybe when you buy a car, you’re a young college graduate, you know, you have a certain budget, so you spend everything you can on your car to get the best of what you want, but you can’t choose everything. So as you start working, car can evolve with you. And there’s this notion that it’s a purchase that might last three, five, ten years.

It doesn’t mean your life is not gonna change over that three, five, ten years, but your your car can now change with you. You can add more horsepower. We’ve seen that in in some vehicles from Mercedes or from Volkswagen. We can change lighting systems, another one that we’ve seen with Audi.

You can definitely make the car physically change, if you have the hardware in the vehicle, and then you can also add new functions. And, we see that a lot, with Tesla as well when they’re they’re continually offering new surprise and delight features in the vehicle. So I think for the end consumer, there’s a clear way the vehicle could age with you. For the consumer, it’s a win.

For the OEM, we don’t talk about it much, but, you have this big spaghetti wiring harness in the vehicle today. With SDV, you’re simplifying it a lot. You’re simplifying the number of controllers you have in the vehicle. You’re going from about a hundred to maybe a handful.

Right.

There’s less copper in the wiring harness as well. So there’s more of a cost save, and that cost save could hopefully lead to further further features in the vehicle in the future.

Further features or improved margins for OEMs or reduced cost or some combination thereof Exactly.

Could benefit could benefit all.

And even if you look at the overall system cost and the ability to to do over the air fixes as opposed to recalls if you have issues with the vehicle. There there’s lots of different angles to it. We haven’t seen all of them yet, so there’s have been a lot of, maybe apprehension on will SDVs deliver what they’re supposed to deliver for the, the car manufacturers? Will end consumers really buy the features and functions that we’re proposing? I think that we’re still early, on the path there, and we’ll we will, we will see benefit for both.

16:11 Revenue Models for Software Updates

Well, I’m curious about your opinion of this topic. And if you think it’s too controversial, okay. But there’s a there’s a question of whether these incremental features that get provided are something that’s revenue bearing, you know, that you charge for features, or whether they’re provided almost as a free upgrade to make your car continuously better. There’s an argument for both. And and I think you’re seeing different models out, you know, a significant number of Teslas, for example, Tesla’s features that come are generally free. They have full self driving, a few they charge for, but broadly free.

And then you’ve seen other makers, you know, charge for some features and give other features free. Do you see the benefit as more refreshing, evergreen the car, or do you see it for ongoing revenue, or is it perhaps a mixture?

I think it’s definitely a mixture. I think you’re not really giving it for free.

You’re giving it to increase loyalty to the brand That’s exactly right, that’s a very smart way to say it because it’s a long term investment.

When a customer says, “My car keeps getting better. I love this car! My next car is going be this one.” Or maybe I don’t need to change cars because this car gets I can stay with it longer. Exactly.

And that’s such a different way of thinking about vehicles, right?

It wasn’t the way it’s been for such a long time.

No, exactly. I think that’s why it’s a loyalty…they will understand what works with different consumers and which will generate new functions and features in the future, which is exactly what we want. You know, we want to improve this consumer experience between the vehicle and the driver or the passengers of that vehicle.

That being said, there will always be some that you sell.

You know, I talked about powertrain before, and the whole idea of feature on demand, I think, is is enormously powerful in automotive.

When you when you make an electric vehicle, for example, whether you choose to make it with a 300 kilowatt or a 200 kilowatt electric motor, the the real difference is the amount of copper and and steel sheets you put in in the vehicle, for your engine, for your electric motor. It’s it’s it’s literally a question of physics. Yeah. Physics and cost.

If you choose to sell it at two hundred and then allow the the driver to opt in and more extra panel actually.

Salary might increase or as they realize they might like that. I think that’s a really good point because it’s controversial. People say, “Wait a minute, you’re giving me this car, but I don’t have a full performance?” Well, that’s a very glass is half empty way to look at it. Glass is half full. Way to look at it is you cut the car at a much better price. And if you’d like to upgrade, you have some headroom that you don’t have to buy a new car.

Well, I always joke about it, and I say, it’s like a four cylinder and a six cylinder. Yeah. In the past, depending on your budget or even if you if you really could afford it, you could go with a V8. But you were limited on what you can spend when you went to the dealership.

And depending on your spend, if you were like me, you spent it on on the on the power of the motor, and the rest of the options in the car would follow. With SDV, don’t have to make that choice anymore because the power of the motor could increase over the year as a feature update if the OEM anticipated in the future. So don’t think of it as I bought a car that has this capacity and I’m not using it. Think of it as a V6 and a and a four cylinder. You afforded the four cylinder, and you can upgrade it to a six, something you can never do before in combustion engines. Absolutely.

And and I think there’s a similar point about resale value as well because if if you have a car that you know is fundamentally limited, you know, sort of hitting its head on the proverbial ceiling of what’s possible, then the resale to the next buyer, when perhaps you’re ready to trade up, you want to sell your car.

If you know you have headroom, you know the car’s upgradable, I think the next buyer is going to be more confident possibly to know that they can get the car at at whatever level they want as well. So I think there’s downstream benefits to to not be ignored.

Yeah. Definitely. Even for the OEM, for the car manufacturer, at least in the car, they don’t have to, then resell it with the options you chose. They can tune it to, what the market is looking for in terms of options.

Absolutely. Well, look, you talk to a lot of folks around the industry. You’re in a quite a strategic position. You know, here at the show or just in general, are there other trends or other things that are on your mind these days?

The the big one definitely is, a higher level of autonomy. So, we’re talking about navigation on autopilot, urban highway. When is that gonna be introduced? How is that gonna be introduced?

I think with the the vision-language-action models and a lot of the work being done on these models today, we’re seeing a huge acceleration. Think we’ll see definitely pushing to level three before 2030, which I think is a great choice. I say 2030 because I’m trying to be, honest with the market, but I think we’ll see before. I think in 2027/2028, we’ll start to see a higher volume of launches giving you full level three autonomy.

So and that for the first time, you know, hands off, eyes off, much safer, much more relaxed driving, especially on long distances. So I think that is a huge one, really enabled by these new models, which then drives this race to have more powerful processors and compute, and how we’re gonna manage that, and how the how do we ensure that the the vehicle in the future has the compute we need to run the models that we’re working on? I think that is a bit of the challenge in SDV today.

Yeah. We we don’t tend to focus on autonomous driving here in the podcast, although in fact, that was my I spent many years focusing on in my previous role. But it is it is exciting. I drive a software defined vehicle.

When I when I bought my car, I specifically chose one that was — I’m not gonna name the name — but I specifically bought a software defined vehicle because I felt like I had to walk the walk that I’ve been saying. And I really love it. I really love the the capabilities. And just the other day, I got a big upgrade of the driving capabilities, autonomous driving capabilities on, in this case, is on highways and some sort off highway roads.

And it’s fantastic. And while there’s a lot of hesitance about autonomous driving, when I’m in autonomous driving mode, it’s probably safer than when I’m driving because it has so many cameras looking six different ways. Any of us can sort of miss a blind spot or something, but they don’t it doesn’t miss the blind spot. So the capabilities are really impressive.

And so those people who haven’t tried out a car in one of these levels, you know, full autonomy is years away, and we understand that. But in certain scenarios, certainly highway for sure and some other sort of highway-like environments, it can be very compelling and very impressive what it can do. At our office in Sunnyvale, Waymo is just down the street. So there’s a big deployment of Waymo in Sunnyvale and in the Bay Area, and you can now get those and drive those.

And the numbers are compelling that driving, –those are full autonomous driving, of course, with LiDAR and incredible capabilities — they’re often safer than human drivers and the numbers are very clear. So it’s exciting to see the industry evolving. It’ll be a journey.

Definitely.

But it’s exciting to see.

No. I think it’s funny you relate autonomous driving to driving like a human.

I think it’s one of the main points that Valeo has been trying to promote over the years.

We shouldn’t hold autonomous cars to be as good as a driver. We have to hold them to be better than a driver. Yeah. And that’s why we always push and promote really triple redundancy: Cameras, radars, LiDARs. Yeah. Because we don’t want it to drive as well as a human. And humans make, you know, many accidents over the course of a year.

We don’t want autonomous vehicles to make those accidents. So I think, Waymo has definitely been a huge pioneer in in the US and we see other models coming up and other competitors coming up and now we start to see it branch out. I think, Waymos moved to Japan. They’re moving to London.

It’s fantastic to see. I think, we see Tesla with their model. We see, you know, definitely other other competitors come on. I think it’s a wonderful wonderful time to be in automotive.

24:52 Conclusion and Closing Remarks

I think that’s a perfect way to wrap up. It is a wonderful time to be in automotive, and it’s been wonderful to have you here with us today. Derek, thanks for joining us.

Thank you very much.

It’s been great!

If you like what you’re seeing in this episode, please like and subscribe to see more like it, both from CES and from our other events around the world. We look forward to seeing you in another episode of The Garage very soon.

Back To Top