Automotive News Canada Podcast Oct 10, 2025

Oct. 10, 2025 | The week’s top stories and Magna’s Todd Deaville on how the supplier uses AI

Annotations will appear as you listen

0:00

15:59

Car

Cadillac DeVille

The Cadillac DeVille is a big, fancy car that was made for a long time, from 1949 to 2005. It's known for being very comfortable and having a lot of space inside, making it a popular choice for people who wanted a nice ride. People might talk about it when discussing how luxury cars have changed over the years.

Term

luxury SUV

A luxury SUV is a fancy type of vehicle that is bigger than a regular car and has lots of nice features like leather seats and high-tech gadgets. They are made to be comfortable and stylish.

Concept

three-row SUV

A three-row SUV is a large vehicle that can fit many people, usually up to seven or eight. It's great for families because it has extra seats and space for luggage.

Car

Jeep Grand Wagoneer

The Jeep Grand Wagoneer is a big, fancy SUV from Jeep that has room for a lot of passengers. It's known for being luxurious and spacious, making it a popular choice for families.

LIVE

Hi everyone and welcome to the October 10th, 2025 episode of the Automotive News Canada podcast. I'm your host, Greg Lason, the digital and mobile editor at Automotive News Canada. Coming to you from just outside Windsor, Ontario, the Automotive Capital of Canada. Today on the show, we hear from Magna International Vice President of Advanced Manufacturing and Corporate R&D Todd DeVille. He recently spoke with John Irwin of Automotive News.

Magna is already using artificial intelligence and what comes next. But first, look at some of the top Canadian Automotive stories of the week.

Prime Minister Mark Carney was in Washington DC this week to talk tariffs in person for the second time with US President Donald Trump. Trump said after the meeting that the US and Canada can, quote, get there on a resolution to their dispute over sectoral tariffs on steel, aluminum and autos.

President described the disagreements between the countries as quote, natural conflicts. He said that's because they're competing for the same business. Says Carney, there are areas where we compete and in those areas where we have to come to an agreement that works, but there are more areas where we are stronger together and that's what we're focused on.

In supplier news, Linemar Corporation has agreed to buy an iron castings plant and lipzig Germany. It's buying the plant from Swiss industrial company Georg Fisher, the Canadian auto parts supplier says it will pay $72 million for the plant, which manufactures large ductile iron castings for forestry, construction and agricultural equipment, as well as on and off highway trucks.

Linemar CEO Jim Jarell says the purchase will fuel both revenue and income growth while broadening the company's European footprint.

And finally, on the product front, Jeep is dropping the three row wagon year SUV after the 2025 model year. The grand wagon year, however, will soldier on. It will be the brand's largest and most luxurious model.

The move comes ahead of the fourth quarter arrival of the freshened 2026 grand wagon year. The updated three row utility vehicle features a refreshing front end that bears the Jeep badge rather than the name of the vehicle, which is what previous models did when Jeep was trying to establish wagon year as a sub brand.

The wagon year is a low volume seller in Canada, however, while sales rose 14% through September, the automaker sold just 402 in the third quarter.

We now hear from Magna International Vice President of Advanced Manufacturing and Corporate R&D Todd DeVille.

Here with Todd DeVille, he's with Magna, he leads their Advanced Manufacturing, Corporate R&D, and there's a lot to talk about, obviously, in those fields.

We're here at the Center for Automotive Research's Management, Briefing Seminars, where AI has sort of been a big topic throughout the past couple of days.

I wanted to start kind of a big picture question, but for Magna, what does AI and ML, how are you guys applying that in your factories today? I mean, I feel like a lot of times we've been talking about that as maybe the future, but in many ways this is happening right now.

For sure, it's now. We think of these as theoretical and out there in advance. The truth is, use properly these AI and ML tools are embedded in everything we touch, and you don't even know if you think about some of them, you know, some of them you don't.

But think about in the factory, which you asked about, vision systems is a big one, an early one, because that sort of came out of early development in the last 10 years, especially, coming from the tech industry, using AI, neural networks to analyze images, video, et cetera, hugely powerful, relatively low cost, very accessible, something we can use and apply today.

Yes, that's in the plants now. Other areas, I would say the emerging right, well, predictive maintenance is there now. A lot of standard analytics can be used where AI comes in, is layered on top of that when you want to look at, like, really massive multi-variable scenarios with a lot of data coming from disparate places.

AI tools can kind of augment more classical data analytics tools and provide insights. So those are a couple of them. And then I think robotics, I'd mentioned right is the vision systems fit into robotics already. So those are running today in many, many places.

That is going to significantly magnify, I think, over time as we get out of just using them for vision analysis and start using AI tool sets for things like identifying objects, learning, training, programming, path planning, all of those types of topics.

Yeah, and there's so many applications. I mean, I wanted to delve into a few of these, but we're dealing right now with, I think tariffs and everything else where companies are looking deep into their supply chains to learn where their parts come from, where their materials come from.

Is magnet utilizing AI and able to really dig in on that, and how are your suppliers thinking about AI and ML and those applications, maybe others?

For sure, I mean, what you're talking about is getting deep into highly complex data sources coming from many, many different places, the formats are all different, huge issues around just data quality, cleanliness, the basics, naming conventions, these kinds of things is enormously challenging.

Yes, that's where AI and some of the ML type tool sets can come in to make that much, much easier. Imagine you have millions and millions of data points that don't quite align in the old days, right?

You had to sort of look at the each text and you're looking at the characters and this is the same as this different. Do I need to erase this space or this dash and that kind of stuff, that's where the AI tool sets can come in and really accelerate that very quickly.

So that data cleaning data handling layer, critically important, highly labor intensive, but you need it without that, you're not doing any of the analytics that comes next.

So yes, that's being used. Then you get into the analytics of, okay, what do I do with this information now, getting deep into the supply chain, what's coming from where, what's common, what's not common, how do I just use that right?

How is the adoption of AI and these ML tools impacting maybe your relationships or your collaborative efforts with your customers? How are they thinking about this and how are you guys working together?

It's an interesting question, right? Good question. I think our customers, obviously we follow our customers as we need to and should in many ways.

I think it's the same on many of the AI tools set topics. And then as our customers need more and more transparency, they're building their digital twins of, you know, well, product engineering are already deeply involved.

We use the same tools. We often co-locate, right? We have to. You've got to be right embedded with the customer using the same tools, the same function.

So at that level, already there, you get into the manufacturing side and you start to want to build for our customers to build, say, complete digital twin of their operations, they're going to need inputs from us because there will be pieces of this that only we have, only we can provide.

Similarly, when you go down to our suppliers and vendors in the supply chain and it sort of works its way down the chain, we have to do this together.

You know, led by our customers, but we got to move together.

I think a lot of times, you know, AI kind of goes hand in hand in some conversations with, you know, labor and both in terms of, you know, what does your labor force look like as you integrate more of these AI tools?

And, you know, what does, you know, your upskilling efforts look like? I know that's been a big topic for a lot of companies making sure that your, you know, labor base has the skill sets that you need as, you know, these tools are rolled out.

What does that look like for Magma? So, we've always had a sort of a strong training focus in the company, either internal, as well as networked with both colleges and universities and attic ecosystem.

You know, in the past, maybe very strong and particular engineering disciplines, trades, et cetera. We're now expanding up. We now need IoT developers. We need programmers on the university side in R&D where I work, we're reaching deep into computer science and sort of other areas that we were there before.

But now it's a much, much stronger need. So, again, it's ecosystem. How do we use what's out there? Because there's a lot of tremendously strong educational capabilities. But how do we adapt that to what we need? And in some cases, we do it in-house.

But for the most part, we're, you know, we're relying on our partners to help build that.

There's a term that's thrown out a lot. Smart manufacturing can kind of lay out, you know, from Magna what exactly that looks like in your factories. I know they might vary by factory. But, you know, what exactly does that look like?

Yeah, what does smart mean, right? It means sort of making good decisions with the data and the information. So, what does it look like on the ground?

It looks like connectivity, common systems that allow us to pull that data in a clean, manageable way so that we can use it. And when you say smart, what does smart mean?

It means actually using that data to make some decisions, make some changes, and improve things on the line. And that can happen in many ways, right? It could be specifically like scheduling different material flows, which product I build went on a particular line.

It could be integrating the camera systems we talked about for vision, for quality. It could be taking the output of that data that then feed into a different process step and say, hey, we're seeing these issues quickly and they're coming from this station.

And there's a quality thing that maybe it's an issue now, or maybe we think it might be in an hour or two hours, and then feeding that further upstream so that you can adjust on the fly and keep everything running.

So, that's really, I guess, how I would describe smart factory.

Yeah, I'm moving forward. Obviously, we're in a building right now at the Michigan Central Station that has seen a lot of history right here in the heart of the motor city.

In that time, manufacturing has evolved quite a lot for the past century or so. Looking ahead, whether it's AI or maybe something else, what's a technology or two that you think might be particularly disruptive for manufacturing, particularly transformative?

What kind of stands out do you? Yeah, okay. That's a great question. I think we have a very long way to run on what's on our plate already.

If I look at, I mean, again, I look at a lot of things historically, and I look at where these trends emerge and come from.

I mean, we're still dealing with microprocessors and chips that originated in the 1950s.

We're still dealing with the ramifications of that. We're really early in this AI wave.

I think that is going to have profound implications, but it's going to take decades to fully play out.

And right now, we're dealing with, I think, case-by-case use cases, but we're still really very much learning about what it means for the overall system.

And I don't think we know yet, but I think as we walk through that over the next, it's going to be decades for that to really play out, to transform.

But I think we'll look back at it and we'll say, yeah, it was disruptive.

Obviously, this is all, this conversation's happening at the same time as there's kind of big picture questions about the pace of,

whether it's electrification or software defined vehicles, that sort of thing.

How might smart manufacturing or the factory of the future, however you want to call it?

Does that help to maybe bring some of these technologies to market faster in any way?

I guess how do you see the relationship between manufacturing the future and the product of the future?

Yeah, okay. So, if you think about many of these technologies, they're coming out of consumer electronics.

So, we've got our consumers, you and I and everyone, we're used to this rapid pace of software defined product coming out and evolving very, very quickly.

Things change in a highly, highly dynamic.

We go to the physical world and things take a little longer. We've got some more physics involved.

Things are a little bit slower, we've got materials and sort of processes and tooling investment and all of these things.

The factory world takes a little bit longer to adjust.

On the other hand, as it adjusts and that gets build out and implemented, it's got a really long life because the impact of that lives for a very, very long time.

And so I think bringing these smart factory solutions, these digital tools and technologies into the manufacturing operations inherently makes them more flexible.

Because we can seamlessly see what's going on, we can check the data, okay, we might have to change some physical things to adapt to product.

But we now can do that relatively quickly because we can see it, we can predict it, we can do it ahead of time, which allows us to be more responsive and dynamic.

So, what does it allow? It'll allow more variety in product.

The whole industry is driven to faster model cycles, faster turns on designs.

That's coming. There is no question we're all being driven in that way.

The question is how do I do that? And the smart factory, use the smart factory tools to not have to change everything in the factory every time I'm changing all of these things and be more adaptable.

One last question for I get you out of here. This is automotive news, this 100 year.

And we're using that as kind of an excuse to look back at the long history of the auto industry during that time.

And I'm curious, you know, for you, especially in your role of manufacturing, when you look back over the history of the auto industry,

it was someone that kind of stands out as someone that made a particularly large impact. And, you know, what did they do?

It's a super broad question, but curious what your thoughts are.

Wow. Really broad question. And really, really hard to pick one.

I mean, you go back to the, it depends on time frame, right?

Obviously, Henry Ford and sort of in that early time period, absolutely amazing, that vision that drive and sort of what he and his team did was absolutely incredible.

So that early time period, I would say that approach, absolutely fascinating. And you see it now being used by others in different ways.

So definitely visionary.

More recently, I look at what's gone on with our friends at Tesla and Elon Musk and his team that sort of created and drove that.

Again, a visionary drive that right or wrong definitely has changed our industry and changed our view of many things.

And now you see the sort of the Chinese OEMs using many of those same principles and driving that forward. So just a couple.

Awesome. Well, Todd, thank you so much for joining us today. Really appreciate it. Thank you.

I'd like to thank Todd for his time and John for conducting the interview.

If you'd like to be a guest, have a suggestion or simply want to comment, email me at glason at autonews.com.

And remember, you can listen to all our previous podcasts on Spotify, iTunes, Google Play or on our website, automotivenews.ca to scroll to the podcast up in the middle of our homepage.

And don't forget you can follow automotive news Canada on x where we're at auto news Canada.

And you can find me there too under at glason a nc. Finally, look for us on LinkedIn. Just search automotive news Canada.

That does it for this episode of the automotive news Canada podcast. We hope you'll join us next time. So long everybody.