AI For Manufacturing, Made Practical

We sit down with Danny Norman, Advanced Manufacturing Consultant with the Center for Industrial Services at the University of Tennessee Institute for Public Service, to talk about what it really takes to put artificial intelligence to work on the factory floor without getting lost in hype.

We dig into the AI for manufacturing training delivered with the Tennessee Chamber of Commerce and why the first win is often basic but powerful: learning how to prompt well. From there, we explore a practical use case many plants feel every day, shift planning. 

Transcript

Why AI Matters For Industry

SPEAKER_01 0:05

As industrial companies navigate a complex and changing landscape, technology advancements are creating new opportunities to improve speed, productivity, and agility. Artificial intelligence is identified as a potential solution to labor shortages, skills gaps, quality control, and managing external pressures. Hi, and welcome to In Touch with Tennessee, the official podcast of the University of Tennessee Institute for Public Service. Joining us today to talk about AI and manufacturing is Danny Norman, Advanced Manufacturing Consultant with the Center for Industrial Services. Thanks for joining us today, Danny.

SPEAKER_00 0:49

Thank you, Susan, and thanks to everyone at IPS. This is a real treat. I'm excited to be here.

SPEAKER_01 0:56

So let's start out. I know that the Center for Industrial Services recently provided AI for manufacturing training through its partnership with the

Training That Builds AI Foundations

SPEAKER_01 1:05

Tennessee Chamber of Commerce. Can you tell us about that training, what it included, and how the attendance was for that?

SPEAKER_00 1:13

Yeah, so I know, Susan, I know you're a big baseball fan, so I'd like to use that as an analogy.

SPEAKER_01 1:20

Okay.

SPEAKER_00 1:21

And it's a perfect one because baseball looks very simple. But once you really start to understand the game, you realize how complex it really is. And AI is a lot like that right now.

SPEAKER_01 1:36

Wow. Okay.

SPEAKER_00 1:37

It's it's easy to jump on chat GBT, right? But the deeper you go, the more sophisticated it becomes. And so our help our training helps manufacturers build a foundation so they can confidently manage that complexity and actually put AI to work. It can just be something as simple as being able to prompt well. That is basic the basics of our our training is around prompting and then how to use it to apply prompt the prompting skills to real world manufacturing prop problems. And we've seen strong attendance. And this is Oak Ridge, Knoxville, Nashville, Memphis, Chattanooga. And that makes me so proud. It tells me that Tennessee manufacturers are innovating, adapting, embracing this new technology instead of waiting on the sidelines to see what happens. Right. And just to understand why that matters, just a little bit, if it's okay, Susan, to talk just a little bit about Center for Industrial Services.

What CIS Does For Manufacturers

SPEAKER_01 2:48

Sure.

SPEAKER_00 2:48

Yeah, so we we are ourselves a trusted partner for Tennessee Manufacturers. We have an internal team of experts with decades of manufacturing experience, many of them with 20 to 30 years of experience in automotive and advanced manufacturing. And we don't do that alone. We connect manufacturers with professors, students, researchers, experts from the private sector, and from other mission-driven organizations like ourselves. And we're all trying to solve real-world problems and create practical innovation. And we've been doing this since the 60s, believe it or not, not in over the years, we've consistently supported workforce development, quality systems, operational excellence. And we just live, breathe lean practices, applying research and technology where we can. And we're trying to help manufacturers stay globally competitive. And so that's why we're just so excited about the training and seeing the attendance that we are. It's an exciting time, and it it shows that that Tennessee manufacturers are going to lead the way.

SPEAKER_01 4:02

Great. So you know, going back to AI, I mean, how can that help manufacturers? Just a few of the areas where they could use that.

AI For Shift Planning And The Three Cs

SPEAKER_00 4:12

Yeah. So so back to our baseball analogy. Something like shift planning sounds really simple, right?

SPEAKER_01 4:19

Uh-huh.

SPEAKER_00 4:20

Kind of like just running the bases. But when you get into real-world details, it becomes incredibly complex. You may have 25, 50, or even 200 employees and all with different schedules, certifications, and daily life constraints that are going to make the scheduling difficult. And traditional software is it's helpful, but eventually that just it becomes almost too much for the software to work with and the human as well. And that's where one of the just one application where AI becomes powerful. And so you can ask AI to plan a difficult production day, tell it what workers are are good at certain machines, what their constraints are, and and it can it can do a really good job. And you might say, well, that's pretty cool. If you try it, I do invite the listeners to try that out and see how good it is. And you probably have some sense, but when you when you think through all of the ramifications of that one one scenario, you start to realize that there's sort of a cascade effect, what I call the the three C's of AI: cost, capacity, and capability. So this cap this cost is reduced. They don't have to buy, the manufacturer doesn't have to buy new software uh to add this new feature. And there's so many areas where manufacturers are are holding on, they're they're they're watching their budgets, they're trying to be cost conscious and financially secure, and they want to buy new software and do new things, but they maybe maybe don't know how to do it now with AI. They're going to be able to do a lot of those things that they couldn't before. And when you have this the right people in the right place with this just the first application that we're talking about, it increases your capacity to get product out the door. And that cascade effect again leads to more capability because you're not just focused on getting product out the door. You can actually focus on the things that are going to help your business grow, help you train new new people and do all the things that you would love to do, but maybe before didn't have time to. So AI is is just a it's a great tool, and that's just one simple application. There are so many more, and we're we're excited to be helping. What our training does is it doesn't give you it gives you this the foundational skills that you need to move forward and and to pursue any application in AI. So we have a really good a good solution for for our manufacturers.

Tennessee AI Success Stories

SPEAKER_01 7:19

Okay. Well, what is the state of AI and manufacturing in Tennessee? And of course, in comparison to the rest of the country.

SPEAKER_00 7:27

So this I am so excited and so again so proud to report that Tennessee manufacturers are leading the way. We are constantly Center for Industrial Services, myself and others, we're interacting with other MEP centers across the country right now, especially. We're on AI engineering committees, uh we're very connected with uh NIS researchers, and then you know, we're just just trying to get information and that can help our manufacturers, and we're we're we're telling uh all the great things that we're hearing. And I again I'm proud to report that Tennessee manufacturers are creating some of the best AI success success stories I've heard. And just a couple of those real quick. Susan, a company in Nashville doing heat treating, they have reduced their trial and error for new orders by almost 90%. So they're able to tell the machine or tell the the AI, hey, I've got this new alloy. I've generally on an alloy that was similar, I was able to run at this temperature and this time, and and then you know how and then based on that, how can we what's the likelihood of what we need to change? And that's that's helping them. And then another manufacturer's saving five thousand dollars a month just by using AI to troubleshoot a an decades-long problem.

SPEAKER_01 9:06

Oh wow.

SPEAKER_00 9:06

That they just didn't have time, didn't have time to research. So AI is is helping just pull in information and and then do basically help us humans think through the logic that we know how to do it, but it just helps us and assists us in doing that. And Tennessee manufacturers are off to the races.

SPEAKER_01 9:29

Great.

SPEAKER_00 9:29

Yeah.

Barriers: Cybersecurity, Sensors, Software

SPEAKER_01 9:31

Well, what are some challenges that might be keeping manufacturers from implementing AI?

SPEAKER_00 9:36

Yeah, so I I still think we're in the same basic set that I've had for a while to think through cybersecurity sensors and software systems. So for cybersecurity, everywhere we taught, everywhere we go, every leader is reassessing how their data is being secured and on the very basic level. Like how who do we know that has access to these files? And all of that has to be secured. And I'm gonna talk a little bit more about why that's so important at the end. Uh, but you can get a general sense for that. Now, on the sensor side, AI feeds on data. And so we are still in our infancy when it comes to digital transformation and deploying sensors for specific purposes. Right now, manufacturers buy a machine, it comes with a set of sensors, and we do what we can with the data that we have. But the data that we have is not necessarily the data that we need to solve a problem that we have. So we've got to get better at deploying sensors. And also, Susan, this is something else that our is kind of baked into our training. Humans are the most sophisticated sensors that manufacturers have.

SPEAKER_01 10:57

Right.

SPEAKER_00 10:57

And if we use the the humans and interact, you know, engage, engage our workforce the way we should be, they're gonna be able to help collect the data that we need to get better. And then on the software side, did you know you probably didn't, there's over 70 different types of software that are needed for a manufacturer to be, especially if they're in automotive or some other they've got they've got to manage all that. Wow, and so we've got, and again, we're we're addressing all these cybersecurity sensors and software. Dr. Enrique Macias on our team has just gone through the SESME road mapping certification, and that's helping us plan out alongside our manufacturers how best to navigate all the software out that's out there and then plan for the future so we have less less issues going forward and we can leverage all this data that's moving around already.

SPEAKER_01 12:01

Okay. And where do you see AI and manufacturing, you know, five or or ten years from now?

The Next Decade And Data Value

SPEAKER_00 12:10

Yeah, so I think all of the the buildings in Tennessee will be painted orange inside and out.

SPEAKER_01 12:19

That would make us happy, wouldn't it?

SPEAKER_00 12:21

Yeah, that would be. So seriously, generative AI makes most what you know what we're seeing with Chat GPT and things like that. That's just the beginning. We're we're going to be seeing an entirely new level of interaction with data, with our scientific community here at UT engineering, and we're gonna be able to connect manufacturers in in new ways. And it's just gonna be a great time. The future of manufacturing is still unwritten, and I think Tennessee manufacturers are gonna hit some home runs, Susan, over the next decade or so. And I'll just leave you with one final vision that I that I promised to talk about, and that is the importance of data security. There you're starting to read a little bit more. There was actually just an article yesterday that came out and Ubers talking about how they're really abandoning their AI model development, and they're actually repositioning themselves as a data aggregation, data collection company. So they're using their the positions of their cars as the the data. They're realizing that that's the value that they can bring to the table. And so what I what I'm hoping to communicate to our manufacturers is that the data that they have is extremely valuable. And in the future, in 10 years from now, the companies that realize that now, they're going to be able to say, hey, we have an ecosystem of AI and data that's specific to our customers, our suppliers, our workforce, and it's creating a competitive advantage that really nobody else can really compete with. And so that's where we're headed. I think we're off to to a great start. And again, there's gonna be some some home runs over the next decade for CIS and for our Tennessee manufacturers.

Closing Thoughts And Where To Listen

SPEAKER_01 14:29

Great. Well, that'll be that'll be exciting to watch for.

SPEAKER_00 14:32

Yeah.

SPEAKER_01 14:35

Well, uh Danny, I appreciate you joining us today.

SPEAKER_00 14:39

Yes, thank you so much. This has been fun.

SPEAKER_01 14:41

Well, good, and thank you, listeners. Be sure to look for In Touch with Tennessee, where you find your favorite podcasts.