Episode 442: AI in Manufacturing: Smart Factories and the Future of Production

Hosts Kris and Brad sit down with Bryan DeBois, Director of Industrial AI at RoviSys. Bryan is manufacturing veteran pushing the boundaries of autonomous agents and machines on the plant floor. In this episode, we discuss how agentic AI is reshaping production lines. Here are just some of the ways this is happening: AI is capturing vanishing tribal knowledge as seasoned operators retire; it's reducing material waste; and it's enabling systems that anticipate failures before they disrupt flow. Bryan shares frontline examples of AI bridging skill gaps, driving predictive uptime, and providing operators with decision-making tools that adapt in real time. Tune in for an optimistic talk on why autonomous industrial AI isn't tomorrow's promise, but today's competitive edge.

Chapters

• 0:00 Cold Open And Holiday Errand Tale
• 2:33 Tesla Update Surprises And Autopilot Chat
• 9:56 Guest Intro: Industrial AI In OT
• 14:28 Adoption Shifts And ROI Reality
• 21:45 Workforce Change And The Expertise Gap
• 25:39 High Mix Low Volume And Changeovers
• 29:19 How AI Speeds Changeovers And Cuts Scrap
• 33:04 Flat Glass Case: Digital Changeover Plans
• 40:23 Safety, Labor Shortages, And Automation
• 47:16 Will AI Plateau And US Competitiveness
• 55:45 Closing, Resources, And Sign-Off

Transcript

SPEAKER_02: 0:01

Welcome everyone to another episode of Dynamics Corner. I am always curious how things are made and the stories they share. I'm your co-host, Chris.

SPEAKER_03: 0:11

And this is Brad. This episode was recorded on December 22nd, 2025. Chris, Chris, Chris. How it's made. How things are made. I'm fascinated by that. And you'll hear in this episode. And I'm also fascinated with how AI is working in the manufacturing industry. With us today, we had the opportunity to speak with Brian Du Bois. Brian, good afternoon. How are you doing? Hey guys, how you doing? You want me to lie or tell you the truth? Doing well. Yeah.

SPEAKER_01: 0:57

Oh, yeah. Tell us a story, dude. Tell us a story, Brad. What happened? A fun one. I think it's a funny story.

SPEAKER_03: 1:10

So I just it it's close to Christmas. So I just went to the supermarket market store, whatever anybody calls it. I don't know. It depends on where you're listening. And I wanted to get in before the Christmas rush to buy groceries so that I could make Christmas dinner on Christmas Day. And I didn't want to go too far out because the food would potentially spoil. You know, you do meats and fruits and vegetables. Okay. And you don't want to do it too close because one, it will be insane. And two, we're in the middle of tourist season and snowbird season. And if you ever visited snowbird season, snowbirds don't move too fast. And they and they flock to the stores. And not only do they flock to the stores, but they're the types that will push their cart in the parking lot at like an angle. So the cars can't go up and down. So they walk slowly from one side to the other. It's almost like when individuals park, you know, walk in crosswalks. They don't walk straight across the street, they like almost take the longest angle possible. So I went to the store, and I was like, oh, this is great. You know, I'll go a little early, beat the lunchtime rush, um, and be back for lunchtime for this. And wouldn't you know? I got outside, my car was doing an update. No. Yes. And I'm like, what is going on? So I don't know how it got into doing an update. Like, because I had been letting the update sit. And it literally, this update has been sitting there for a couple weeks.

SPEAKER_01: 2:49

It did a force update on you.

SPEAKER_03: 2:51

It did an update on me while I was in the parking lot. So I put all my groceries in the trunk of the car. Now, mind you, it's not that it's not like back home in the Northeast, where you could, you know, during this time of year, I I use I leave food outside anyway this time of year because it's cold enough that it's like a refrigerator. Right. Uh and you get a little further in the season, you could actually leave like ice cream and ice and stuff outside. Uh back when I used to drink beer, by the way, I used to dig and we used to have a lot more snow when I was younger. We used to actually dig dig snow banks. You know, they own the snowbanks. Used to pile up the snow, and I used to box out in the snowbank a room for beer. Yeah. You just leave it outside, and then the same type of thing. Around the holidays, if you haven't meats, you just kind of throw it in there. It's almost like an ice box. It is an icebox, actually, because it's snow, it's uh snow. So anyway, so it did the update, and I'm thinking, um one, I'm not going to get back. Two, it uh when is this thing ever going to be done? It was like watching paint dry. It took 50 minutes.

SPEAKER_01: 3:54

No, yes. It did it does warn you that says it could be up to 55 minutes, so it's pretty close there. You got 50 minutes.

SPEAKER_03: 4:03

Usually though, when it's outside, I don't care. I don't know what triggered it to do it. Right. But it did it. Were you stuck in the car?

SPEAKER_01: 4:11

You can't get out, right? I could get out.

SPEAKER_03: 4:14

Thankfully. I mean, there was a couple points where it it starts like uh clicking. So I I knew it was doing something. I was able to put the windows down because I think, ah, just be a couple of minutes. I'll just sit here because for a while the AC stayed on. But then once the AC died for like 40 minutes, I'm like, oh, this is over. I mean, I was sitting there so still that the birds were landing on the car. So I could hear these like these little things, and then I started hearing these clickings. I'm like, is somebody throwing something at the car? So anyway, I rushed back uh for our lunchtime fun.

SPEAKER_02: 4:47

You made it on time.

SPEAKER_03: 4:49

I made it on time and uh almost choked eating lunch, uh, but that's okay.

SPEAKER_00: 4:55

Uh Brad, was this a Tesla?

SPEAKER_03: 4:57

Yes.

SPEAKER_00: 4:58

Okay. Huh.

SPEAKER_02: 4:59

That's so weird it forced update you because yeah, it does it does have to you have to tell it to say yes. I acknowledge that I could be sitting here for 55 minutes. It's wild that it just did it for you.

SPEAKER_03: 5:13

I do like the update because the because the map looks better, and then it also has the I didn't have the chance to enable it, but it has the Sand is Toys feature now, too. I want to see that. What is the dash cam? There's an update for the dash cam thing, too. Like are the dashboards.

SPEAKER_02: 5:27

Um so it wouldn't when it records when you're doing a dash cam. So used to be very limited in view. Now it actually it actually uses the other pillar, camera pillars as well. So now you have a full 360 for your dash cam. So it's pretty neat. That's because everyone complains about that, right? Like, oh, it's getting all the other angles, except it's not using the pillar um cameras. So they've enabled that. So now you have a full 360.

SPEAKER_03: 5:55

Got it. I can't wait to check out the update. Can't wait to enable the Santas toys and see what some of the other things are there.

SPEAKER_02: 6:04

You what you need to do, all right. So I did this in my in my uh in my Tesla. You can change the chime when you walk away and it locks, right, and makes those weird noises. You can change that chime. So what we did, my son and I uh used uh the movie from Elf, Buddy the Elf, when that wall, that uh walrus or whatever uh character says, hey bye, buddy. I hope you find your dad. So we have that turned on. So every time we walked away, it would chime that way. You should do something like that.

SPEAKER_03: 6:34

We'll talk afterwards, I'll enable that because I want to start having fun with this. I and mind you, listen, I love the vehicle. I'm not complaining about it at all. I'm just complaining that I had to sit in the sun because I don't even know how to drive anymore. Not that I ever drove well, but now forget it, because I put that thing on, and now that it goes point to point, or and it it's you know, I w I went out for dinner the other night and it found a parking spot for me.

SPEAKER_01: 7:01

Heck yeah.

SPEAKER_03: 7:06

I can't wait to the point where we just get into a vehicle and you can do things.

SPEAKER_00: 7:14

Yeah, right. And yeah, because you still have to have your hands on the wheel, right? In a Tesla. No, not necessarily attention.

SPEAKER_03: 7:22

Your eyes test. Uh it does, it does have hands-free. I don't understand in the latest update, on or off, because it can detect now if you're using your cell phone. So it doesn't it warns you if you're using your cell phone, like in your distracted driving. Yes, it the cameras catch you. And also, you don't have to necessarily hold on, but I position my hands, like I pull my legs up and I have keep my hands on my knees, so it look may look like I'm near the steering wheel, and every now and then you may have you have to put some pressure on the steering wheel. Which is fine. I don't mind uh any of those safety features to make sure you're paying attention, but to this day, the vehicle sees things far before me.

SPEAKER_00: 8:06

Although before me, so I'm not I haven't put the trigger on the whole Tesla thing yet. Um, it's it's it's on my radar. I I was really kind of waiting until like that point-to-point. That was really so the last time cycle I went looking for a new car was 2019 and it wasn't there yet. So I was like, well, I'll kind of wait and see until we get to that point-to-point. And now that that's kind of there, now I've I'm thinking it might be time.

SPEAKER_03: 8:30

Once you go for a drive, I pretty can guarantee that you'll pretty much want one because it handles extremely well. Uh now you have the different modes you can put chill mode, snail mode. I think it's snail, chill, standard, and hurry.

SPEAKER_02: 8:49

Oh no, Mad Max.

SPEAKER_03: 8:50

They added Mad Max. Okay. I haven't gotten that. I've put it, I put it in hurry mode, and I thought I was going to die. No, Mad Max is worse, dude. That thing standard is good for me. It goes, it goes with the flow of traffic. Uh, so yeah, it's in it's anytime a minute, I feel like it's in control. Don't I'm not trying to say anything ill about it because I love the vehicle. Hurry mode's a little too fast for me because it starts to swerve. I'm one of those, I drive like a grandpa for real. Uh Brad, you gotta check.

SPEAKER_02: 9:21

I think you have Mad Max in yours. I'm pretty sure the update, there's Mad Max in the city.

SPEAKER_03: 9:26

I'm not looking at it. My daughter even says that the car drives better than me. So I am one of those. I'm far right, speed limits 50. I'm going 50, and I'm just going along. So the vehicle drives better than me. It will pass, it will do things. And anytime someone's in the vehicle with me, they're like, wow, this is good. We're actually getting somewhere. So anyway, that was my story. So I was a little, I am uh a little hurried, but uh, but we made it. And speaking of AI and such, uh, would you mind telling us a little bit about yourself?

SPEAKER_00: 9:58

Yeah, for sure. So um, Brian Du Bois, and and uh, you know, this is a kind of our second time together. We did our our previous recording earlier this year. So um, for those of you your listeners who maybe didn't get a chance to listen to that one, Brian Du Bois, I'm the director of industrial AI for a company called Rovasys. Now, they are we are a company that focuses on system integration specifically for manufacturing and industrial customers. And we only work in what we call the OT space. So that is that's operational technology, that's what that stands for, OT. And that is different from IT. So this is the world where assets and equipment make a physical change to the real world, right? So IT is about moving data and information. OT is about machines that actually do things and make things happen in the real world. And so that is what we typically refer to ourselves as is an OTSI, an OT system integrator. Um, the company's been around for 36 years. I've actually been at Rovasys for 25 years now. Um started out, I my background is I've got a computer science degree from the University of Akron. Um and uh yeah, and I've been working in this um in this OT space my entire career. Um, been in a lot of factories, and for the last six and a half years now, I've been focused on bringing industrial AI to the manufacturing customers that we service. And so I'm sure that's what we'll be talking about today.

SPEAKER_03: 11:22

Absolutely. When you say OT space, you know what I think. What's it? Outer space. Oh it's outer space. You work in outer space. Uh no, thank you for joining us again. And uh within the world of AI, I with uh several guests that work with AI in different industries or different areas industries, as you had mentioned, it's not always just software uh technology. There's, as you had mentioned, machinery or asset technology. We're again talking about the Tesla, which is another it's a vehicle technology. I'd like to do some follow-ups because AI is moving so rapidly that if you blink, you may miss an update or miss what's going on. Uh so it's been several months since we had spoken with you last about how AI is enhancing uh manufacturing industry within machinery. I wanted to jump back in with you to see what other changes that uh you've witnessed, uh experienced, or how you've seen businesses adopt AI with machinery or AI automation uh or OTSI or outer space.

SPEAKER_00: 12:35

Yeah, for sure. Um, well, I mean, let's jump into it. So I think that um even in the time since we spoke last, um, there's there's been some interesting advances. So um one of those is that we are seeing more and more of an uptick of folks who are starting to um look into adopting AI on the plant floor. This has always been, you know, I've been at it for six and a half years, but it's always been a little bit of an uphill battle because there are folks who are like, we're not ready yet, we're not ready yet. Um and so now finally we're starting to see folks who are who are ready and they're ready to have those initial conversations. Now, whether or not their data infrastructure is ready, and we can talk about that, that's kind of a separate thing. But they're finally ready to start embarking on this. The other interesting change that I've seen in the last six months is um when in fact, I think you guys asked me this question last time, like, where do these use cases come from? Where do the ideas come from in terms of adoption of AI on the plant floor? And we talked about how we do these AI workshops and we try to prime the pump and get them thinking about AI, and so then they can help generate ideas. Now, one of the changes I'm seeing is when I sit down with them and have that initial conversation, they are typically coming to the table with three to five pretty solid typically AI use cases that they've generated on their own. And then it that actually makes my job a lot easier because then I just have to vet those ideas, those use cases. I can say, hey, you know, out of these five, let's focus on these two, these seem to be the lowest hanging fruit. And a big part of that then is how do we connect that to the ROI that is going to be required to make these projects a success. Um, one of the things I've been at this long enough now that one of the primary conversations I tend to have with customers is look, these projects don't typically succeed or fail based on technical problems. Like we can get over all the technical issues. They they they succeed or fail typically based on internal political types of issues. 100% Yeah. Big part of that is did did we do a good enough job and did the customer do a good enough job of connecting that project, that AI project, whatever the technology is, to an outcome, to an ROI. And and can they go to the bean counters and the folks who are above them in the C-suite and stuff and say, hey, we got 20% improvement, we got 15% improvement, we got millions of dollars saved every year, or increase our throughput by X number of percentage. If they can't do that, then you know, I typically put the brakes on very quickly and I go, hey guys, it I know this looks like a really like cool use case, it's super sexy, like it would be really awesome, but I don't know that the ROI is there. And because the other aspect of this is in an industry like industrial, like manufacturing that is just now starting to adopt AI, you kind of get one swing at this. You get one shot to do your first project, and it better be a home run, right? Because otherwise you're not gonna get any more at bats. So we got to make sure, I gotta make sure that I'm vetting those AI project ideas that we're getting, you know, the the the lowest hanging fruit and stuff that's really meaty right out of the gate.

SPEAKER_03: 15:49

You hit on several points for this uh for me, and uh I wanted to see in your experience where some of these points fit. Uh spit them out there, hopefully we can remember to get to them. One, you talked about adoption, we're at a different rate of adoption. Yeah. Is the rate of adoption changing because of comfort? Is the resistance less? And what is the usually the biggest cause of resistance? Is it people being fearful of the technology, fearful of the technology because they're afraid it's going to replace them, feel they're full of the technology because it's not going to work. And then the other part, another part, there's three points I wanted to remember, but again, with the one that I'm having, we'll see what happens. The other point is that you also spoke about the ROI. And I see this in technology everywhere that oftentimes projects fail because sometimes people like to do things because it's the cool thing to do. I want to do this because it's cool. So, what are some exercises that somebody can do to go through and validate that they will have a proper return on the investment again? Because you may not get the return on the investment that you think of by implementing something. Again, if you build a system that will pick something up and move it across the room, but you only need to do that one time, and it takes you six months to build that system, you're really not getting the return on the investment, as for example. So and then I'll get back to the third point on that. But I want to go back to the the first point of you had mentioned you have witnessed an uptick in adoption or at least curiosity to try to explore and implement uh the use of AI. Where do we sit with that and the resistance to it as well?

SPEAKER_00: 17:41

Yeah, so um, so kind of two parts there. Why, why we're seeing an uptick, I think part of that is that I think our industry is getting more sophisticated. So I think that, and I, you know, I I this is I think you guys are my 33 or 34th podcast appearance on various manufacturing, industrial, business podcasts, right? So I'm getting the word out there. I'm hopefully helping in this in this effort to make our industry more sophisticated in their in their thinking about this. Um, but you know, there's other there's other um thought leadership that's happening around this that are getting people to realize that, you know, part of it, we talked last time about generative AI has been a blessing and a curse. It's been a blessing in that it's getting people to think and talk about AI. It's been a curse in that that's not typically the AI that we're implementing on the plant floor. So so they're starting now to get sophisticated enough to know that there's different types of AI, that they have, that it has, that they have different strengths and weaknesses and limitations and and where you should use the different types of AI. So now when I'm getting into those calls, they have a level of sophistication that actually enables us to start having productive conversations right out of the gate. Where did that resistance come from? I think it came from a lot of different areas. I still I just talked to a steel customer uh a couple weeks ago, and they said our CEO just reiterated his policy, no AI here. Won't won't do it, can't do it, whatever. Now, you know, there's probably a generational aspect there. I think their CEO is is is kind of fairly old, maybe baby boomer-ish. And so, you know, I think that there's some aspect to that. You know, I'll tell you, so I'm a I'm a Gen Xer, proud Gen Xer. Um, there's very few of us, by the way. I don't know if you've ever looked into the generational uh, but Gen X is actually the smallest generation. That's why you don't run into many Gen X.

SPEAKER_03: 19:28

Well, one, I was going to ask if you had witnessed or if your increase in inquisitiveness about AI is due to a generational change in leadership. Number one, number two, Chris, can you look up the I don't want to make clicking here?

SPEAKER_02: 19:46

The years, because I don't know what I am sixty five to eighty, nineteen sixty-five to nineteen eighty is yeah, and if your birthday and then you get that little gap, I think they said between nineteen eighty and eighty-three or eighty-four, where they have the Xenyals and Where it's just like it they still had that combination of like Gen X and sort of millennials. Different, by the way, different kinds of people. I've met them. And then the actual millennials were, you know, maybe closer to like 86 to 90s, is when they correct AOL, you know, things like that.

SPEAKER_03: 20:17

65 to 80. So I fall into the Gen X, and I don't want to digress. Now you're saying it's the smallest generation.

SPEAKER_00: 20:24

Yes. But I just in terms of hit count.

SPEAKER_03: 20:26

I also think with the generation that has gone through the most, and I'll I'll cover that in a moment, but how do they come up with the years? Wouldn't you think that they would all be equal? Like if we're talking generation here.

SPEAKER_00: 20:41

I think that there's certain things where there's specific cultural things that happen, like the internet, that just create this natural like division between different cultures and and stuff like that. But anyway, where I was going with that, yes, I think the fact that baby boomers are are retiring in droves, that we're losing that the that expertise out of the plant. Um, but they're also taking with them some of their preconceived notions. And so I think that the leadership that is remaining um are younger, they're uh primarily millennials, and that they are much more willing to look into this. Um now I I reached a milestone of my career though, guys, uh, a couple weeks ago. I I called my wife after I left a meeting and I said, Well, Tiff, I said it finally happened, I was the oldest person in the meeting. Everyone on the other side, everyone on the other side of the table on the customer side, they were all millennials. Uh, all the people from Rovasis that were there, my company, uh, they were all millennials, and I was the oldest guy in the room.

SPEAKER_03: 21:37

But how does that make you feel? Because I've I've had conversations talking about references and things, and now I'm getting I wasn't even born then. Oh, yeah. Well, that's fine. It's starting to shift now. Now we've become the other side when we started our careers. We're the other side where now we're the old guys.

SPEAKER_00: 21:59

We are the old guys. So, okay, so to get to that, so what is what is that like? So, what I saw across the table is they were overwhelmed. The the this I was working with um the uh the plant manager, and he's a millennial. And you know, again, like he's got high turnover, he's got he's lost all the SMEs, all the leadership in that plant that knew how to keep that plant running, and he just looked overwhelmed. And so, again, and we we touched on this a little bit last time, AI then can be that bridge to help. So for him, he's like, yes, like whatever we can do here, I have a serious problem. What are the technologies we can bring in? So AI is not the only answer, but it's part of the answer to be able to get to um to be able to fill that gap, that expertise gap, and be able to get his novice operators, these operators that are not gonna stick around for very long. You know, most people are seeing an entry-level manufacturing job as, you know, like a six-month type of thing, a three, you know, maybe a three-month, maybe a couple week thing, and then they lose interest and they're gonna go move on to the next the next job. So um, how does this guy, this plant manager, how does he continue to get high levels of performance, not have quality issues? This happened to be a food and beverage customer, not have serious quality issues where you got to issue uh expensive recalls and and things like how do you avoid all that and still be able to put out good product while you've got a workforce that has changed so drastically? So for him, he was welcoming. He was like, Yes, give me your ideas, I need it all. Yeah. So it is.

SPEAKER_03: 23:30

I was wondering how much is that generation because uh I see a shift in leadership, and I think that generation grew up with technology, so they're less resistant to technology, more trusting of technology. But it's still to me, it comes back down to the proper application of technology. Granted, some things are invented as byproducts of multiple technologies, and it's people are doing some experimentation to see it, but I I think not having the fear of it and then finding a place to implement it at the proper points is is helpful. And as you had mentioned, with I think a reduction in workforce, there are some functions that are monotonous, mundane, and that uh basically are repeatable. And it's those repeatable functions that you can try to automate. It's it's um yes. So I'm glad to see there's a shift, there's an uptake in the adoption of AI. So what have you seen with the adoption of it?

SPEAKER_00: 24:30

Um where where we're starting to see, so you asked about kind of ROI. So where are we seeing that adoption and and what is driving that ROI? So um again, like in a lot of cases, we're seeing it in things like um uh changeover. So this is the there's a long um arc that's been happening um for decades now in manufacturing where we we've gone from when I started my career in the early 2000s, um, it was uh high volume, low mix, right? So you were making the same thing and just cranking out the same product over and over again. We have seen over the decades, um, and you guys are a little higher on more logistics supply chain like ERP site, so you're probably seeing this in terms of like configuring these bombs and things like that. Um, we're seeing now high mix, low volume, right? Everyone wants their own configuration of that product. We want customized this and you know, and and that's and I'm seeing it across all the different industries that we service, right? It doesn't matter if you're servicing automotive, you know, downstream or or you're servicing, you know, you're in CPG. Um, everyone wants their own customized thing. We need we need 10 versions of this product, not just the one. And so um, you know, in a lot of cases, that changeover, okay, so that's the term that we use when you're changing between different products on a line. That changeover starts to become really expensive. It was one of those things that in the past, it was just kind of the cost of doing business and it wasn't, it didn't become a serious uh, you know, kind of cost driver. But now that changeover becomes a big, a big problem, especially when you have operators that don't know how to optimize for changeover, right? So, like, oh, we're running a whole nother product, man. I gotta, you know, I gotta start all over again. And so changeovers then go from two hours to six hours to eight hours between changing over between different products. So one of the primary ROIs that I'm seeing over and over again brought up by customers is let's see if we can um optimize changeover. Let's see if we can get to a level of changeover that's as fast as our experts were at being able to, you know, let's see if we can get back to a two-hour changeover instead of a four-hour changeover. And look, even if we can shave an hour off and get to three hours, we're in much better shape than we are now. And I'm in I've had multiple customers that have said that's a million-dollar problem. Annual, annual ROI on that problem. If we can shave an hour or two off of changeovers, that's that's a huge thing. Now, there's a lot that goes into that because um, so in any kind of changeover, you typically have sequencing issues where I can't make this product after this product uh maybe without a wash down. So we did a project for um a major paint manufacturer that you guys have all heard of. And you know, they've got very specific rules, as you can imagine, around tints. So, like I can make this product after this product, so I can go from a lighter tint to a darker tint of paint without a wash without a washout in between. But if I'm gonna go from a darker shade to a lighter shade, I gotta make sure that I schedule a wash uh washout of the equipment. And every industry has something like that, where I can make this product after this product, but not vice versa. So knowing that and being able to incorporate that into the AI, but then part of it too is that every changeover requires a certain amount of scrap. So you're just gonna run for a while, and until you get to what they call steady state, until you get all the parameters kind of tuned in and you're making good product in that starting time. Now you're just making scrap. And that's where, again, we can easily optimize that and shave off that time where you're making garbage and get to making steady state product much, much faster. And so now, boom, that's an instant ROI. So we can talk about some of these other things where people think you're gonna get ROI, um, like predictive maintenance, and we can talk about that, and we can talk about um uh uh energy optimization, and those are all great use cases, but honestly, changeover has for whatever reason seems to have dominated the conversation um within the last uh six months.

SPEAKER_03: 28:28

And how has AI helped with that changeover? Uh so if we are getting a return on investment for the changeover, and it's great examples, and it's it's working with different businesses of different industries. I'm always fascinated by what it takes to make things. Sometimes you don't even think about what goes into producing a product, even your reference was something as as what we would call as simple as paint, but usually is a little more complex. I one time worked with someone who manufactured soap, and I had no idea that soaps and shampoos, with the recipes, with all of this, was such an involved industry. And just listening to you talk about that uh brings back memories. Again, uh that's another sign of aging because now I think back of when I was younger, I got stories to tell. So I have lots of stories to tell. But in the case of changeover, in this case, you're talking how AI can assist an organization to decrease the time of changeover to help them have more production runs because they're having less downtime for washout, as you had referenced, or changing over other uh pieces of machinery so that they can produce a product. How is AI helping with that task?

SPEAKER_00: 29:46

Yeah. So there's a couple couple different ways. Sometimes it's it's just a matter of um reducing the amount of that scrap, that startup scrap, that is typical of any kind of changeover. So if I can get to making on spec product faster, then I can reduce the amount of that scrap. Um, and so what does that look like? That can be as simple as I'm giving the or not me, the AI is giving better set points. Um, and you know, so the set points are the things that control, again, for your listeners, the the set points are the things that control that process, right? It could be what's the temperature, what's what what's the um uh the pressure at this point in the process? You know, what what's the heat if I'm heating up a piece of steel or something like that what am what's the the target temperature that I'm going to? That's what's called your set point, right? On thermostat, whatever you set the thermostat to, that's the set point. It's not the current temperature, that's a different thing, right? But it's that's where you're trying to drive it to, and your thermostat drives to that. In the same thing on the plant floor, we have all these really smart thermostats, effectively, and we have all the you can set all these different set points and it will drive to that set point. That's what the control system does. So, what are better set points than what I have now? And how can I optimize the sequence of set points? Because oftentimes with startup uh and changeover, startup's similar. Startup would be like, I have to bring that down the line for some reason and I need to get back going again. Um, it's a similar kind of problem. Like just to get back, get that flywheel going, and get back to making good product, it takes a certain amount of time. So if I can optimize those set points, and expert operators knew how to do this. Like they'd say, hey, so I tend to crank this for five minutes and then I turn the pressure down, and then, but here's the trick. I do this thing, you know, I make this change or I turn up the RPMs here just in time, and then that's how I'm able to get to making good product faster. Novice operators don't know how to do any of that, right? So that's where the AI can step in. Um, another thing, so we've got um a uh uh so this is an interesting use case specifically around changeovers. So we've got a company that makes flat glass. So flat glass is like um think windows, doors, you know, big panes of glass. Now, the state of the art, the way they've been making flat glass forever, is um you have a bed of molten tin, molten metal tin, and they extrude the molten glass on top of it. The beauty of that is that gravity does all the work. So the glass will flatten out as much as physically possible just because of gravity. The glass actually floats on top of the molten tin and it'll flatten itself. Then you have these arms that come out and they got wheels on the end of them. They're called ADS or assisted direct stretch arms. There's about eight of them on each side, so 16 total. And an operator then hooks into the edge of that glass, and as that glass is solidifying, that operator can then um push and pull that glass like taffy. And that's how you make different thicknesses and widths of glass, right? Now, when we first sat down with that customer, they said, look, we um we know how to do this really well, but only our expert operators do we allow them to do changeovers. Why is that? Because it's very easy to lose control of that sheet of glass, and if that happens, it's moving at a pretty good clip. If you lose control of that sheet of glass, it can go crashing into the side of the the tin bath, and a lot of bad things happen then. So, so they only had two operators that they trusted to do changeovers. So they'll bring in, they got plenty of operators to handle it when you're making the same, just you know, putting out the same width, same um thickness of glass. But for those changeovers, they would literally do what they call crushing glass. So they would just make the glass and then they would crush it and immediately melt it down, just waste it until that next operator shift started so that they can come in.

SPEAKER_03: 33:28

I'm paying attention to this like a kid because I I figure like, what was that show, how it's made? Yeah, I really am trying to visualize this now. Like, I'm like, this would be a great episode to see how glass is made. That's that's that's sort of right.

SPEAKER_00: 33:40

I I love how it's made, by the way, just because you mentioned that. Before I so I knew nothing about manufacturing before I came into this industry 25 years ago, but I loved how it's made. And so now since then I've been in probably a hundred glance and seeing how things are acting. I love my job.

SPEAKER_03: 33:56

Fun fact for you as we get into that. I did an implementation once, many moons ago, with an organization that I developed uh an ex you know, a basic, I guess you call it an extension, but they were using the vision at the time that I created in a whole additional add-on. Chris, this is the one we were at the conference, and somebody talked about the saying, here it is 20 years later, and we made a business around it. But they were on how it's made. It was it was great because I actually I had been in that plant, I had known those individuals, and to see that it was fascinating. You know, it's almost one of those things that uh it was very interesting. Not to interrupt your story, it's a very interesting story, but you had mentioned that there's only two operators that they trust. Right. And I think I think that is something that's becoming more uh more common. And with that, do you think it's becoming more common because generationally individuals don't stay at an organization long enough to gain the experience? Do you think it can be attributed uh to maybe those individuals that have that expertise not uh sharing that expertise and holding on to that expertise because not others can do that? Or do you think it's a lack of for better terms, interest uh generationally, where it's uh I don't I'm just here to stretch glass and oh well.

SPEAKER_00: 35:27

Yeah. Yeah, well, I'm I'm just here to get a paycheck and and kind of clock out at the end of the day. And and again, like I'm not you know, I never get into the whole like this generation's better than I think every generation had their strengths, weaknesses, different approaches to problem solving. So I I don't get into any of that, but it's just a fact that the the incoming generation, yeah, they're not interested in spending the next 10, 15 years of their life becoming an expert on how to stretch glass. Like that's just not of interest to them. They have different interests. And and I mean, again, like there, I I manage millennials and Gen Z's, right? So uh Rovasys hires 90% of our hires are right out of college. So we're constantly facing the generation gap. We've been doing the generation gap thing for decades, and and so I I'm of the mind that there's a lot that you can learn from different generations, and I'll tell you, like I've learned from millennials and Gen Z's about this work-life balance. Gen X had no work-life balance, that wasn't a thing, right? But but they've taught me a lot about how to achieve that. So, whatever. It doesn't really matter what the drivers are. The big the point is that they're not going to spend that time to learn how to do that. Well, the plant still has to make product, right? So, so how do we close that gap? And so that's where, like I said, that's where they're looking to AI. So, what was so what was the solution then that we built? So, we built when their expert operators do a changeover, they lay it out, as you can imagine, again, because because failure leads to big problems. So they they they go through a whole process where they write down on a piece of paper, there's a template, and they go through and they say, Okay, we're gonna, it's and it's stuff like you can't make big changes with the glass. So it's stuff like, okay, I'm gonna change the angle of these for three minutes, and then I'm gonna, and then I'm gonna push in for two minutes, and that's that's the process. So they write it all down. You almost think like a poll 13, like when they were going through their different like checklists, like that's the kind of thing, and and then they've they go and they execute the plan. Now, one thing though that was interesting is we asked them, so do you ever deviate from the plan? No, no, no, we don't deviate from the plan. This is the plan, right? And then, of course, as you can imagine, we sit down with them, we we our team sat and we watched them do a bunch of these changeovers, and they deviate from the plan all the time.

SPEAKER_03: 37:41

That's part of the plan.

SPEAKER_00: 37:43

That's part of the plan. And so they'll go, you know, it said, I don't know, it said stay at this angle for three minutes, and they kind of look at it and they're like, I'm not liking what I'm seeing. And so they stay for two more minutes before they make the next move on the on the plan. And then they do typically document that that they made those those deviations. So so we built for them an AI-driven um changeover plan. Now, it's digital. So one thing is we built in the deviations right from the beginning. So we we built it like Google Maps. So on Google Maps, you can imagine, you know, it it's got a route, it's got a plan, but then if you take a wrong turn or there's a uh detour or whatever, then it reroutes. It says, okay, rerouting you, and it finds a new optimal plan. That's what our system does. So it's a digital changeover plan that finds new new ways of doing things if the real world doesn't cooperate the way that we want it to. And so that we were actually able to put that in the hands of novice operators, and they're able now to execute changeovers as well as their expert operators. And in some cases, one of the secondary goals was to, and remember last time we talked about autonomous AI, this is autonomous AI. Um, so it's that super smart AI, it's the based on deep reinforcement learning. Um, one of the secondary goals it pursues then is try to minimize the total time of the changeover. So don't just get it done, get it done as fast as possible. Because we talked about changeover, you're not making good product during changeover. That's all scrap that you make during changeover. So let's minimize the amount of time. And so that was a secondary goal of this system was try to minimize the amount of time. Wow.

SPEAKER_03: 39:23

Yeah.

SPEAKER_02: 39:25

That's that is that is amazing.

SPEAKER_03: 39:27

So now the operator follows the digital plan or monitors the digital plan. So then you have a higher rate of consistency with product and decrease in change time, which also minimizes waste.

SPEAKER_00: 39:40

Right. And this is a human in the loop, by the way, process. So we were not directly controlling the process. This was put the plan up on the screen and and then monitor the yeah, the to make sure that the operator is doing what they're they need to be doing. Yeah.

SPEAKER_03: 39:52

Human in the loop. It's I I hear human in the loop, but now you talk about AI when you uh I refer to it almost now as Person. I feel like I'm talking with someone. I do a lot with AI from the development point of view. And going through and prompting and seeing the responses and seeing the answers, I find myself saying, Thank you. You did a good job. And just to see some of the response that comes back, it really is like having a bunch of eager individuals developing for you that are at a certain level.

SPEAKER_02: 40:23

I think that example is a perfect human in the loop because they still have that control. Um, they're still part of the process. And I think a lot of people are afraid of that, right? With AI being brought into the organization that they're going to be replaced to some degree or replaced entirely, um, which is not necessarily the case all the time. It's just enhancing your production, enhancing the business process so that you can be more efficient and then be able to do other things. Or in this case, they could become an expert, right? Because they're learning more. There's a uh, you know, something's helping you out to get there.

unknown: 40:58

Right.

SPEAKER_03: 40:59

That makes me think of two things. And you know, again, I go with what what pops to mind. One, it's it's it's I say it often, AI is not going to replace you somebody using AI will because they can gain those efficiencies. And then also individuals are talking about being replaced. I saw something, I don't know, maybe it was in a vision, or I don't know if I was reading. I do a lot of reading, but I forget the way that it was phrased, but it was basically AI is AI, if you're the tool, you're going to be replaced. And this has been happening for generations. It's been happening since the beginning of the Bronze Age. The tools get replaced, the people don't. So if you are the tool, you have a high likelihood of being replaced. If you look at back, we had the horse and buggy. The horse and buggy were the tool to move things, to move people. Automobile replaced the horse and buggy. You still moved things, you still moved people, you just use different tools. Same thing in the construction industry. You had hammers, air guns, then power tools. You still have a human, you know, to use the the phrase human in the loop somewhere. So that's uh that's one way to kind of like it opened my eyes to look at it. So you have to look at the functions that you're doing and adapt to use those functions and not be afraid of technology because technology has been adding to our lives since all before uh long before any of us were born. Uh like I said, you go back to the Bronze Age, at the beginning of the Bronze Age, you know, we made a uh uh a knife and a sword and and such. Uh the rate of change is far uh greater now, but uh we're still there. And then um with that, it's uh see, I lost my train of thought. I was getting so excited. I'm not afraid to admit it.

SPEAKER_00: 42:48

Part of what you were talking about there, Brad, while you're thinking, so and it's it's even more poignant in the manufacturing and industrial sector, because the last number I saw was something like we've got a half million person shortfall in manufacturing. So they can't find people to take these jobs. It's not a question, they wouldn't fire people, even if somehow I was magically able to, you know, to bring in AI and automate huge, they wouldn't fire those people anyway. They're like, we need these people, like we'll put them into other positions, we'll put them into higher value tasks, but we're not gonna fire anyone. We can't afford to fire anyone. We're doing everything we can. Like, and you know, again, better, worse, indifferent. One of the things that I'm seeing at my clients is a new, newfound respect for culture and a newfound respect for what's what's this work environment like? And are we creating a work? Why? Because they're having to compete for talent. That's not a terrible thing, right? So, so again, there's positive outcomes that happen. So, no, they're not the I've not done any of these projects that have that that have resulted in a in a reduction in workforce. It just doesn't happen.

SPEAKER_02: 43:51

I think that's a perfect higher. I think that's perfect where AI fits in because there are short, you know, there are um shortages on some of this uh industries where you can't find people because like you like you point out, Brian, some generations don't want to do those things. So, but you still have to do those things, and you're relying on people that's been there for a long time, and eventually they're gonna move on and retire, but you still have to fill that role. Well, you're trying to fill that role, nobody maybe wants to do those things, and so that's a perfect opportunity where AI can help uh mitigate some of those uh shortcomings as well. That's uh lots of opportunity.

SPEAKER_03: 44:31

That's the hidden, I don't want to say hidden, but everyone's so I think it becomes professionally or um uh industry specific. And I say industry specific and function specific because there are certain tasks where everyone's fearful of it. And I'm glad that you brought out that there are shortfalls in some areas because there are tasks that can be automated. There are tasks that are uh not to mention dangerous that if you can have AI or have automation assist with that, it's less dangerous for someone to work on. You know, you look at some of the even some of the, I'm sure you see with some of the manufacturing facilities you're in. Some of these, some of these uh products that individuals are making can be dangerous, whether it's mixing chemicals, we talked about uh some uh implementations that I'd worked through, or even I can imagine even working with glass, you have some extreme temperatures and you have molten that could hurt someone. So if you can automate this, and uh and what's close to my world is I'm more into the software application point or software design. So it's a little bit different per it's a different perspective on the use and the implementation of AI. And that's what I was thinking of when you were talking about where you had the shortfall of individuals, and the individuals aren't available to do those functions, which is why we may have a deficit in uh certain industries and certain products. It's not because of lack of need, it's a lack of function because we don't have those to do it. So AI can also help make for a safer workplace, just like my vehicle making me drive, uh driving me around makes the roads a lot safer. Uh so uh it it is it is we should welcome that.

SPEAKER_00: 46:08

I mean, it both in the United States and globally, like you know, I'm in these plants, and you know, and I would never call out any specific companies or whatever, but I'm in these plants, I'm like, boy, that doesn't look great, like what I'm seeing over there. Like that that looks like not a great situation to have a human that close to the process or having to interact with the machinery at that point in the in the in time. And and it's just kind of scary. And so, yeah, we should welcome that. We we want humans to be as far away from these processes as possible. Um, let's put robots in there, let's put AI, let's put automation, and let's let's get people out of those situations. I I I um, you know, I was talking to a co-worker of mine, he's like, Man, I just got out of a, I just got back from a plant, and he's like, that was one of the scariest things I've seen. He said it, and there was a press, and and a guy, and this is in America, like this is not, you know, and someone is literally like putting stuff in and then pulling it out like in time with the press. I'm like, this was like this. This is an hour up the road. This was not in, you know, you we see these videos from other countries and stuff where you know, we don't have OSHA and they don't have those type of type of programs, but um, yeah, we don't want that. So let's let's leverage the technologies we have to make people safer. And um, and again, like I think it's a as this new workforce is like, I'm not doing that. Exactly. Exactly. No, let's let's do it.

SPEAKER_03: 47:36

I I think it's because of maybe exposure or understanding as well as why someone's not doing it. And I do respect what you had mentioned about generationally, because there are different types in every generation. Uh, I think to me, I look at generation as more of how you grew up, because sometimes how you grew up, what existed, if even if you look at the boomers or any that had gone through the depression, they save everything, right? They they re- they'll use every scrap of every piece of everything. Why? Because they lived through the depression and they realized that you had to ration things and you couldn't waste things. So I think that's where it comes into play. But I do think that having the mindset of maybe having at some point some resistance to say, hey, that's not safe, that's not uh something that I want to do because uh I don't even want to say you could get hurt, but also if you have some individuals doing some of these monotonous tasks that can be automated, repeated, you don't have to worry about someone being tired, fatigued, right, and which also could affect your output, and it also could be a hazard for that individual or others near them as well. So with AI and all of these uh uh inventions or enhancements that we have to an organization, whether it be from the software within the manufacturing plant, and our ability to solve problems. Do you ever seeing a plateau? And what I mean by plateau is it's almost like the AI paradigm where AI knows what AI learns based on what it knows. It doesn't create new things, right? It takes a collection of known, all of the world's knowledge that has been uh created by individuals, and you can use it in manufacturing, as you had mentioned, for uh assessing the proper temperatures or proper uh pressure, or you had mentioned moving the width of glass and and the stretch on the glass. Do you ever think there would get to be a point where we plateau? I know me personally, I when I was growing up, yes, I did use an abacus, yes, I did use the slide rule. Uh abacus was a little more for fun. Uh the calculator was invented. I looked it up when I was born, but we didn't use them until later. I used to do math in a moment. And now sometimes I joke when I say to someone, like, what's two times a hundred? or you know, give me what's 125 divided by three. Yeah. A lot of times now individuals think about it, or even I had a conversation with someone this morning. Do you remember when you used to remember everyone's phone number? Yeah. Now, if I lost my phone, I wouldn't call anybody because I have no way of getting in touch with someone. So do you think as we rely more on AI, whether it be in the manufacturing industry, uh AI automation and such as uh the great things that your organization's doing, or in the software world or any other place where you implement AI, that we will reach a plateau?

SPEAKER_00: 50:35

Yeah. So so it's a really good question. And um the here's the good news, and I know we're coming up on time here, so I'll I'll leave you guys with this point. So um the uh the good news is is that this industry is ripe for AI. We won't hit any kind of plateau for a long, long time. So here's the here's the analogy I typically talk about. So we've got this renewed focus on onshoring and US-based manufacturing, right? And and that's great. And I'm I'm all for it. I'm a U.S. citizen. I was born and raised in Ohio and raising my kids in Ohio, and I love this country, so great. So okay, so we want we want US-based manufacturing. Here's the reality is that most people who aren't in this industry, who aren't in plants every single day, they think that manufacturing factories that you've got all these robots flying around, and you've got, you know, it's super high tech, and it's and there's you know, bare barely any people out there. Maybe there's one person in a glass control room making everything happen, and there's just like a room full of robots. But that does not describe US-based manufacturing in the slightest. Okay. There is still so much of manufacturing that is an art, not a science, where the outcome, the the quality of that product or the speed that you can make that product is materially tied to how good of an operator you have, right? That's that's an art not a science at that point. That's not just I I just need someone to press a button and it'll make product. That's someone going out in the line and saying, well, it kind of looks a little thick, so I'm gonna run it for a couple more minutes and see if I can thin this down, right? So that's a problem because when you have a workforce then, so the incoming workforce, the way that they're approaching these types of factory jobs is like McDonald's. Okay, so let's look at McDonald's. When McDonald's goes out into the market, they don't go out looking for cooks like a diner hires a cook, right? They go out looking for people and then they put them in front of a machine. And we've all seen the videos, like that person then loads a bunch of frozen pucks into a machine and they press a button and the machine does all the work and out comes burgers. And then they load the burger, you know, they pull the burgers out and then they load the pucks again. Like, so McDonald's has automated, and especially now where they've got the screens that you can type out, they have automated virtually as much as you can automate, right? Short of like now you get into diminishing returns. So you need some people to do some certain things, but for the most part, they've automated. That's what people think factories work like. They don't. But that's where we have to get to because if we've got a workforce that is gonna quit in two weeks, and you've got to be able to train someone in a couple hours and get them making product and make being productive in a in a couple hours of training, you've got to have machines that are much smarter than they are today, that have more automation, more robotics, more AI to be able to overcome that gap. And so that's where we have to get to. So to answer your question, Brad, like the sky's the limit. There's so much work to be done here in the United States. And in a lot of ways, for whatever reasons, again, globalization, free trade, you can whatever, it doesn't matter. All of these things have led to, in a lot of ways, the US manufacturing um kind of lagging in a lot of cases. And um, and in a lot of ways, it was because we were we were going up against countries where you can just throw bodies at the problem. You can just have 15 people working on that part of the line, and you don't have OGS safety standards either. So, like all that stuff that we've been talking about all plays into it. And and you could so if we're going to outcompete, we have to do it smarter. We've got to make the machine smarter, we've got to be able to close that expertise gap, and we need to be able to get it to the point where that person on the line really doesn't have to think. They don't really have to make judgment calls. Some the guy at McDonald's is not making a judgment call as to whether or not the burger's done. The machine does all that. So that's where we have to get to if we want to, if we want, if we truly want to be as competitive as as I think everyone wants to with the rest of the world.

SPEAKER_02: 54:38

I love the comparison between McDonald's. You're right. I think a lot of people have that mis misconception of what the manufacturing space looks like. And because of that, it it kind of ruins people want to get into that space, you know, and and you you hit it right on. It's like you we have to find a way to to get us to be competitive here in this US market in terms of manufacturing.

SPEAKER_03: 55:02

Well, I guess to make it uh to make it sexy in a sense, right? And I think you hit it, and I start to think is is the hysteria around AI fear-based because it's new and it's masking because you have the perception, as you had mentioned, that we have all these technologies. Here we are in the United States, all these technologies exist, but it doesn't mean it's being used universally. Right. Exactly. Just like I have a Tesla we talked about at the beginning of this episode, you don't have one yet. Right? And it's for whatever reason you don't have one. But again, you would think based on conversation, oh, everybody has cars that drive them around because those cars, those vehicles exist. So it it is uh it is nice to hear, and I think it it sheds some light that AI and technology and advances have been occurring for generations and many generations. It definitely isn't going to go away because uh what what's it? What is the old saying? You know, lazy men is the father, like what is it, like the father of inventions, laziness or something like that. Oh, oh, yeah. Well, I think people come up with these tools because they don't want to do these tasks. And I'm not saying that's a laziness thing because sometimes I say productive laziness. Well, it's because you create things to do things, so you don't have to do something, but the fact of or the act of creating that thing is the genius sometimes, right? So it's it makes it that you uh simplify your life. So it is good to know, and I think for anyone listening, it's uh it's important to realize that the technology is here, the technology is not going to shut down the entire world where we're all going to sit around on the couch and have robots waiting on us.

SPEAKER_02: 56:48

Well, maybe we will, but I I do want to make a quick comment about the manufacturing space. I I do feel that here we're are a nation of cons consumers, we consume a lot compared to other places in the world, and so having the idea of like things being manufactured, it's kind of like, oh, it's being done elsewhere, I don't really care. Uh I just want to consume it. And so I I think that's maybe one of the reasons why uh we don't think about the manufacturing here and no one's interested in it. I there's so many variables, but that's I do think about that from time to time is that we we tend to consume, we're gonna we just order it, it shows up, but don't realize what it took to build something like that. Yep, yep, for sure.

SPEAKER_03: 57:35

I think it's great. Well, Brian, thank you very much for taking the time to speak with us again. We appreciate it. It's always a fun conversation. We'll have to follow up uh uh on uh interested hearing a few more stories you have and see what some advances in AI technology within the manufacturing space uh uh bring us over the next year. Very cool stories. I know within listen, uh I know within our industry the world is changing daily. I'm sure in yours it's changing just as rapidly. Maybe it's a little bit uh sometimes uh difficult to adopt because you have to change out machinery, change out processes, but it's still, I think, moving at a rapid rate. But I'm interested in hearing it. And if anything, I get to hear some how it's made stories, which I think all those are on Amazon now, by the way. So I've been watching them and it's I could put that on. It's almost like a Christmas story. You know, Christmas story is my favorite, it's not my favorite, but it's one of those movies that I've seen like probably a thousand times, but I've never sat and watched it end-to-end because they always play like a 24 hours of Christmas story, and you catch it at a certain point. How it's made to me is the same type of thing as I could put that on and just let it run for days in the Twilight Zone. Uh but thanks very much. We appreciate you taking the time to speak with us today. Um, time truly is the currency of life. Once uh you spend it, you can't get it back. And anytime someone spends speaking with us, we greatly appreciate and appreciate all the great things you're doing with AI. If anyone would like to learn a little bit more about AI in the manufacturing industry or what uh Robesys does, how is the uh what is the best way to get in contact with you?

SPEAKER_00: 59:00

I would say the easiest way is go to Robasys R-O-V-I-S, R-O-V-I-S-Y-S.com, Robesys.com, and uh or look me up on LinkedIn, Brian De Boya on on LinkedIn. So great to talk to anyone.

SPEAKER_03: 59:12

Great, thank you very much. Uh have a pleasant holiday season and look forward to speaking with you soon. Thanks, guys. Appreciate it.

SPEAKER_02: 59:17

Happy holidays.

SPEAKER_03: 59:19

Thank you, Chris, for your time for another episode of In the Dynamics Corner Chair. And thank you to our guests for participating.

SPEAKER_02: 59:27

Thank you, Brad, for your time. It is a wonderful episode of Dynamics Corner Chair. I would also like to thank our guests for join joining us. Thank you for all of our listeners tuning in as well. You can find Brad at developerlife.com. That is D V L P R L I F E dot com. And you can interact with them via Twitter, D V L P R L I F E. You can also find me at mattalino.il, M A T A L I. I-N O dot Io and my Twitter handle is Mattalino16. And see you can see those links down below in the show notes. Again, thank you everyone. Thank you and take care.