Podcast: Automate 2025 recap — How robotics and AI are powering smart manufacturing
Key takeaways
- AI adoption is surging, but real-world applications—not just hype—are key to proving its value on the factory floor.
- Simulation is evolving from concept to core tool, enabling better design and predictive maintenance in manufacturing.
- Integration speed and accessible automation now cater more to small- and mid-sized manufacturers, not just large enterprises.
- Open architectures and interoperability are breaking down silos, allowing seamless collaboration across industrial systems.
In this episode of Great Question: A Manufacturing Podcast, Robert Schoenberger, chief editor of IndustryWeek; Mike Bacidore, chief editor of Control Design; Rehana Begg, chief editor of Machine Design; Sharon Spielman, technical editor for Machine Design; and Linda Wilson, chief editor of Vision Systems Design, share insights from the 2025 Automate show in Detroit. The discussion centers on the widespread integration of AI in industrial automation, with a focus on practical applications and real-world use cases. Topics include the growing role of simulation, increased accessibility of automation for small- to mid-sized manufacturers, and the importance of open system architectures. The conversation also highlights evolving industry partnerships and the push for greater interoperability across platforms.
Below is an edited excerpt from the podcast:
RS: So we're all here in Detroit this week at Automate, the premiere show for automation and automation technologies. We've spent the past few days wandering around the show, looking at lots of robots and linear systems and all sorts of cool technology. What did you see that really interested you or was a surprise in any way? Let's start with you, Mike.
MB: Sure. I mean, the obvious addition of AI to everything under the sun is so prevalent here at the show. There are a few interesting bits of technology. Even Beckhoff is kind of changing what used to be TwinCAT Chat. They're rebranding that now—it's much more enhanced AI-assisted programming. I'm seeing a lot more AI-assisted programming, a lot more AI copilots that companies are designing into their software. Siemens had some interesting technology that they were showcasing for motion control.
RS: AI everywhere. In some of these cases, it looks like, oh, these are brand-new AI systems—and it's so interesting and cool. In other cases, it’s like, you just threw two more letters onto last year's product, didn’t you? But there are some nice new examples of people doing interesting things. You mentioned Siemens. I was talking to someone over there, and they mentioned that, yeah, when you start talking about doing something like a large language model on industrial software, it’s a lot harder than when you’re dealing with, you know, whatever is on the Internet. Because you don't have a lot of technical diagrams out there. You're not going to have the ability to say, “What does a CAD file that does this look like?”—because that kind of information just isn't out there.
So they're struggling—not struggling—there's a real challenge in training these AIs on real industrial data, and they're just not going to get that in the public domain like you can with all these other things.
Rehana, you mentioned you had a few concepts and things that you saw that were interesting.
RB: Yes, but before I even go there, you have to remember—Mike just said it's AI everywhere. What’s really interesting right now is that at this particular show, in 2025, NVIDIA was a keynote, which speaks volumes about how AI is being integrated. And their focus—their market focus—and the ability to bring what everyone was just teetering around with or tampering with—or, should I say, experimenting with—now really showing use cases, more so than ever before. So yes, AI everywhere, but can you prove it? Can you show me real applications?
I've seen some. Others, I don’t know. I don’t know that you're ready for prime time yet.
But what did I like? Let’s see. So, I know NVIDIA has a partnership with Teradyne. So that’s UR and MiR.
RS: Universal Robots and Mobile Industrial Robots, for those of you who aren’t familiar with all those names.
RB: So their collaboration is really tight. I’ve seen some interesting applications with what they have, including with large language models and large—I don’t know what it's called. You’ll be able to tell me, Linda—it’s the visual models that you can use. Just some interesting applications in physical environments.
AI is allowing the machines to say, 'Well, this is what I think it's going to look like,' even though we’re in an industrial manufacturing environment where I can’t really see.
- Sharon Spielman, technical editor for Machine Design
And that’s—I said concepts, buzzwords—that’s the big one for this show: physically show me how you're actually applying artificial intelligence. Whether it's through—you name the AI—you pick it and then apply it in a physical environment, in a real environment, using robotics. That’s what I’m here to learn about.
And then the other concept that I think is really important here is simulation. Simulation has been around forever in different models, different styles—you know, physics models and digital twins.
Not hearing a lot about digital twins this time, but simulation is finally taking its place, I think, and being discussed in a way that takes the “digital twin” that everyone, about a year ago or two years ago, and even five years back, wanted to elevate. I’m not using that as a fuzzy term this time.
RS: It's interesting that you mention that, because I talked to a couple of different people who were talking about digital twins in relation to simulation—or in relation to AI.
The idea that you can take an existing digital twin and say, “Well, what happens if I do this?” And you simulate the data and then train your AI on things that haven’t happened yet.
It’s all very high-minded, but if you really start talking about it, you can work out planning scenarios. Like, what happens when this part fails right now in this kind of machine? How’s that going to ripple through our operations? You get a sense of that and train the system to respond to something that hasn’t happened yet. It’s a fascinating idea. Who knows what will happen.
RB: And you can truly see how that will change the game from a machine design—from the designer’s perspective—because right off the bat, it completely opens up a whole new... I want to say a whole stream of revenue at the design phase already.
In understanding how you can use artificial intelligence to do better designs and understand, through simulation, what that is going to mean—where it’s going to work, where it’s not going to work.
And the last one—the last thing—is setup speed and speed of integration. That's a big deal as well. Compared to last year, I find that there are many more examples that are showing and are appealing more to the small- to medium-sized enterprises—which I always find difficult.
OK, this is great if you have the budget to throw at this, but what does it mean for the little guy—which is 80% of everyone in manufacturing? That’s my take.
RS: Sharon, what did you see today?
SS: My colleague just took the words out of my mouth. I was going to say—that’s what we’re seeing, especially with simulation, that it could be used with the SMEs as well.
Because before, they didn’t have the budget to do it, and now they can. And the integration of components—and being able for everything to be connected.
At Schneider Electric, I was talking with them, and they were talking about opening the architecture. That’s another big one—trying to make it accessible for everybody. I know they’re working on it. Also, Phoenix Contact is working in that same realm for that type of protocol—so that all of the things will work together instead of being proprietary. Then you don’t have to build a whole new system. I think that’s a big one.
There’s a lot of cool linear motion out there from ESTAT. They were in Schneeberger’s booth, and they have a brake that’s as thin as a piece of paper. It’s really cool—that linear motion stuff. Very cool stuff.
So that was very interesting to see.
And then also, with the AI—back to AI—if you're using cameras, you really need lighting for vision to work. But AI is allowing the machines to say, “Well, this is what I think it's going to look like,” even though we’re in an industrial manufacturing environment where I can’t really see. And even with 3D scanners, you still can’t always get the right amount of light. So having the AI software is letting them “see without seeing.”
About the Podcast
Great Question: A Manufacturing Podcast offers news and information for the people who make, store and move things and those who manage and maintain the facilities where that work gets done. Manufacturers from chemical producers to automakers to machine shops can listen for critical insights into the technologies, economic conditions and best practices that can influence how to best run facilities to reach operational excellence.
Listen to another episode and subscribe on your favorite podcast app
About the Author
Robert Schoenberger
Robert Schoenberger has been writing about manufacturing technology in one form or another since the late 1990s. He began his career in newspapers in South Texas and has worked for The Clarion-Ledger in Jackson, Mississippi; The Courier-Journal in Louisville, Kentucky; and The Plain Dealer in Cleveland where he spent more than six years as the automotive reporter. In 2013, he launched Today's Motor Vehicles, a magazine focusing on design and manufacturing topics within the automotive and commercial truck worlds. He joined IndustryWeek in late 2021.