AI and the New Era of Manufacturing

Ed Ballina
Well hello fellow manufacturing folks I’m Ed Ballina.

Alvaro Cuba
Hey, Alvaro Cuba, how are you guys?

Ed Ballina
Hope you’re having a great day. Alvaro, we’ve got lots and lots going on. As you know, we were supposed to be recording this in Orlando at the conference for Society of Maintenance and Reliability Professionals. And we were thrilled because, for one thing, I was going to get to meet Alvaro for the first time in real world. But Milton, or Mr. Milton, the hurricane had different ideas. Here we are, the event was canceled. Alvaro will mention this as well, thoughts and prayers for our friends and family in Florida. Nathalie Hernandez, who was on this podcast last time, actually lives in Tampa. She works for TECO, as you can imagine, and she can’t even take a breath in the preparation for Milton. Wish you the best. Alvaro.

Alvaro Cuba
Yeah, Helene passing just last week, few days ago, and now Milton and me living in Florida, I have been on the hype of all these things. So certainly not fun. And as Ed said, our prayers for everyone that is passing terrible moments after Helene. And we hope that everyone stay safe. Now that Milton is going, it’s coming. So yeah, our prayers with all of them.

Ed Ballina
So great team let’s get on to today’s episode so welcome to the Manufacturing Meet Up podcast I got a feeling you knew where you were already, but just in case you just wandered in. This is the show where Alvaro and I and also other manufacturing buddies kick back. We talk about the shop floor and fun stuff that happens in our manufacturing plants, so Welcome to the manufacturer Meet Up podcast.

Alvaro Cuba
So as you heard, supposed to be in the SMRP conference. And when we were looking at the agenda, no surprise, lot of the topics related to digital transformation and more important, the impact into manufacturing. Not only in maintenance and reliability, but in the big picture of manufacturing. We said with that, yeah, that’s the topic. So let’s use this episode to deep dive in digital transformation and what it really means for manufacturing. Which we are all in there and we hear bits and pieces here and there. So we’ll try to put some thoughts together. So first, let’s go into framing a little bit of this. So digital transformation or 4.0, it’s because it’s the fourth biggest transformation. The first in the 18th century was mechanization and steam. And then in the first part of the 20th century, electricity and mass production. Then at the beginning of this century, it was internet and automation and electronics. Now we come to the fourth, which is about smart manufacturing, digitalization of the supply chain, synchronization, the full ecosystem and everything related to AI and internet, internet of things and specifically how is the connection between people, process and technology and how to make it really efficient and not only in specific parts, but across the entire end-to-end supply chain. So in framing this first, just a few numbers to put this a little bit in perspective. We checked some surveys and two years ago, 2022, only 26% of manufacturers had started doing something on artificial intelligence. Two years later, today, we are at 86% of manufacturers. And in two more years, it’s supposed to be 93, 95%. So everyone is going to be doing something related to digital transformation, artificial intelligence. So that’s one statistic. The other one that I was reading is market has published some numbers that are staggering. The global market for artificial intelligence today, is around $4 billion. In 10 years, it’s going to be predicted 156 billion. So that gives us a 45% CAGR, year over year. So it’s transformation and it’s accelerated transformation. And one last bit of information because there is, and we’ll talk more in the show about, okay, is this technology? What about people? But it’s not about technology or process or people, it’s all together. And just a quick statistic, in the last couple of years, 44%, almost half of the roles and responsibilities of the positions in the manufacturing arena have been changed and upgraded thanks to artificial intelligence. So this is not only impacting today 86% of manufacturing, but it’s impacting 50% of the positions of the people working in our plants. So that’s some kind of introduction on the topic and with some concrete figures, Ed…

What do you think?

Ed Ballina
So just a couple of comments. The first one Alvaro, I think we might have missed the first revolution and it actually took place many eons ago. I think many of us have seen the video of the monkey with a stick using tools, right, to get the ants out, or you know the bird that you know the other one that uses you know drops a bird picks up a clam and drops it from 100 foot and it cracks, so animals started this manufacturing thing long before we did, but yes, we have taken it to the next level. Yeah, the monkey with the stick. It’s a beautiful picture. But these numbers are staggering, right? I mean, these numbers are revolution. This is not evolution. This is S-curve changes. And with it comes some caveats, right? Because one of the things I read recently, Alvaro, is in about of  half AI/IoT implementations, they’re struggling to find the savings. And like everything else, when you’ve got an onslaught as big as this is, it’s going to take time for the organizations to be able to separate the wheat from the chaff and say, this is what really works, and this other stuff looked good, but half-baked pie type of thing. So it’s exciting stuff. And as with any revolutionary change, it takes time for the adoption rates to get to levels that we’re at, but we’ve achieved those in what, four or five years? It’s crazy. So a couple of other questions that we wanted to try and answer to dimensionalize this a little bit better. We were going to ask people, hey, where are you using this? Where are you using AI? Where’s the Internet of Things, cloud-based, et cetera? And I was willing to bet that the biggest, the number one response was, hey, I’m going to be using this to detect equipment failures. That’s almost like the entry level to a lot of this. And again, I’m speaking specifically for manufacturing production organizations. In the other business processes, there is use of AI for things like determining, doing a better job of modeling new product introductions.

As a supply chain person, I’ve experienced some of them that have completely oversold the forecast. And then we had to run around like crazy trying to get more product to some that underperformed. And we were running product off three months later and trying to get out of the trade. Applying AI to that can really have a major impact in determining what’s going to work. And all based on previous patterns of performance.

I would think that would be a starting point, if you will, or what most people would answer. But, you know, I see, we see tremendous opportunities in other areas. We’ve talked about the impact of, you know, of this predictive technology on the shop floor. And I’ll just start with something real simple. If your equipment doesn’t doesn’t blow up, right, or doesn’t run to failure, right, you usually don’t get shrapnel of bearing housings and garbage and grease all over the place. So for one thing, your plants are going to be cleaner. You don’t have to do all the cleanup before you replace the piece of equipment because you caught it way before the failure. And then that, of course, drives less human/machine interactions, which by definition is going to improve your  safety performance. Every person that know that works in a factory likes their facility to be clean and to be safe. So you’re winning two thirds of the battle there already. So Alvaro, how about you give us an idea of what are some of the challenges that folks like you and I are facing as we try to implement AI and predictive technologies?

Alvaro Cuba

Well, some of the challenges, well, first to your comment, I think we are seeing AI starting to be used in different parts of manufacturing. Here and there, the true power comes when you start combining all those and starting to think more holistic. But we’ll talk that in a minute in some tips on strategy when you go in that.

To Ed’s point, some of the main challenges is data quality. We need to make sure that it’s the right data that we put in. So sensors, and Ed also mentioned that, helps that because it takes the data directly and puts it in the system and avoids any issue. Quality of data, how to measure benefits, also put it there. Regulatory things. And one that always comes is cybersecurity. With all the hacking and all that is how by nature you need to do this in the internet and to be highly efficient, that has to be in a secure way. And that’s always a challenge. What about the benefits, Ed?

Ed Ballina

Yeah. Yeah, I think, you know, as we, you know, as we experience some of this, that cybersecurity one is a real big deal, right? I was having a conversation with a new startup that’s got a pretty cool product. I’m not going to plug them yet. I need to learn more. But they have what appears to be a very simple system of answering questions like, what is my worst performance SKU and what is its velocity and how much am I writing off? And I mean, they just go into your systems and gather our data.

And I was like, wow, I mean, the ability for me to be able to do that without having to rely on huge analytics teams and, you know, building reports. I love it. But one of the things that I shared with him is, you trying to get your hands on that information is going to be a real challenge because you’re reaching into the gut of the machine. Right. And some of this information obviously is confidential and proprietary in nature. It’s competitive information. That’s always a big one that you need to figure out how to answer that question from the beginning because you’re going to get handed it. And if you then want to turn around and do something to the equipment as it is running from the outside, good luck. It can be done. Our OEMs do it, right, with our permission.

But for you to come in from the  outside and try to actually make setting changes in the equipment is going to be hugely challenging. So with that, Alvaro, you hinted at this. Tell us about the key strategies, because we told them all the pitfalls. Now, how do we get this through the minefield?

Alvaro Cuba
Just a couple of thoughts about that. One is, and we mentioned this a little bit, is when you think about AI, the first thing that comes to mind is your phone or the technology or Google or what is on the news. But the truth of the matter is that you need to think on this one in particular, different than the previous revolutions, you need to think about the three components, people, process and technology. And exactly in that order, you first need to think in your people and the training and how to get them ready and how to get them open and excited about this versus worried or concerned. Then you need to go into the process and…AI can do a lot for processes because it gives you analytics, it gives you data, data mining, it allows you to make decisions much, much easier. Even a centerlining of the machine before is trial and error, trial and error, trial and error. Now it tells you because you did this and that happened, now you just put it here and it’s there.

And then it comes the automation of that. So that’s one tip. The second tip is the more you put it, and we were talking about this, people is putting this in maintenance, others in quality, others in safety, others in different cases, all of them are valid, but the true value comes when you start putting those together. So when you say machine health, it’s the health of the entire machine. So it’s not only the maintenance part of the machine, it’s also how it’s running and the temperatures and the product that comes in. And then you take it to the process health. And then you add quality and you add process controls and other elements. And finally, you have the human part and then you put safety and analytics. So at the time that you start combining and go first machine health, then process health, and then you go to the end to end, then the benefits start not only adding, but multiplying or skyrocketing because every one affects all the others.

Ed Ballina
They’re synergistic, right? They just pile one on top of the other. So now instead of a straight line, you know, linear, no, now you have a logarithmic increase because of how synergistic all they are to each other.

Alvaro Cuba
Exactly. And the third is, we heard this Ed mentioned and all the time is, oh my God, is these machines are going to replace us? What is this going to happen? But if you go to artificial intelligence, it basically takes the data and extrapolates what can happen with the data.

But the human component on it, things like human judgment, when you need to change, because artificial intelligence gives you better, but based on what you have been doing. When you need to do a drastic change, the human creativity and problem solving, all of these are components that today and in the future and in always is going to be put by humans. And the more you work in this interface, human artificial intelligence, the more, the highest your potential on getting the benefits. What do you think, Ed?

Ed Ballina

Yeah, no, I agree. You know, the picture that you’re painting is one of an environment where people want to work, right? So going back to my previous comment about some of these intangibles, and sometimes they’re more tangible than you would think, but a clean, reliable workplace is a place people want to go to work. A safe workplace is a place where you usually have good morale. And all of that just adds, you know, to Alvaro’s point, the people component is hugely important.

And if you don’t take it into account, you can put the fanciest technology in a facility and you’re not going to get to where you where you want to be. So there’s, there’s lots here.

Alvaro Cuba
Exactly that. So it takes us to some tips on implementations.

Ed Ballina

So one of the first things that we all run into of course is, how do we justify this thing right? And you know, the banker always wants to know how are you gonna pay me back and what interest rates you’re gonna deliver, Um, so benchmarks, right? How do you know when you’re winning? And I’ve kind of I’ve pitched the the book on on the podcast before, “The Game of Work.”

I don’t make any money on it. It’s a short little book, but essentially what it says is keep track because if you don’t keep track and score, you don’t know when you win. So the same thing with these, right? So one of the first ones that comes to mind tying into a previous answer is your mechanical uptime or some measure of efficiency to include your downtime and your waste. These are things that as you implement AI, specifically from a equipment, machine health standpoint, you should be able to start tracking. And everybody knows about that, right? Machine reliability and all that. Hey, here’s another one. We make a big deal about our sustainability gains in this day and age, and rightfully so. So what’s your energy efficiency? If you take a line that’s running at 50% efficiency and through this work, you can get 10 points of TE out of it.

That is worth a ton in terms of carbon footprint, energy usage, and waste. So don’t overlook those. Of course, there’s also some that become a little challenging because the benefit doesn’t always accrue to your P&L. So as a manufacturing person, I get kudos, and I get atta-boys and rewarded for delivering great customer service to my sales organization, right? To my GMs. And we measure it as order fulfillment, out of stock, whatever you want to call it. But essentially, if you order 100 cases, did I put 100 cases in your truck today to deliver? And when you have a line that’s running more reliably, I can do that more often than not, right? And I can probably eat my order fulfillment rates. As a manufacturing person in my P&L, that never shows up.

Now it shows up in the big P&L and that’s why I always try to encourage my supply chain folks, think like a general manager. Too many of us get blamed for being functional. I’m sorry, if you’re a supply chain person in most CPG organizations and you’re called functional, that is not a good thing. They’re telling you you’re narrow-minded and just focused on your own part of the business instead of the total P&L.

You know, you’re running million dollar enterprises, act like a GM. What’s happening to my net price? What’s happening to my volume, right? Costs I contribute to tremendously, and that yes, I should be held 100% accountable for that, but I’m also aiding the business in other areas. Many times, because it doesn’t accrue to our P&L, we forget to give ourselves credit for that in these, you know, in greater efficiency. So that’s, that was another one.

And how do you build a financial case for this? Start in the easy spot. The low-hanging fruit is going to be your line uptime. See what that does for you in terms of operating cost, out of network source product that you have to bring in because your line broke down, waste, and labor. And I think you’ve got a really good story to tell. You’re not blazing pioneers anymore. This stuff has been around for a while. There are foot treks, know, trails that have been carved through the woods for you. Believe, take it on because this stuff delivers and delivers in real terms. Alvaro, your thoughts?

Alvaro Cuba
Just couple of thoughts on the metrics and the benchmark. Don’t be afraid. The landscape is completely new. A new landscape requires a new set of KPIs. So you continue to measure these kinds of KPIs. It’s like companies or startups.

Companies measure one way, startups measure in a different way. No? They are not interested in the revenue yet. They are interested in how many customers they are getting. It’s a new landscape, requires a new set of KPIs to know if you are going. In most cases, you don’t need to reinvent the wheel. You just need to go out there and benchmark with open eyes. And sure.

Ed Ballina
Alvaro, I can add something to that, I tried that story, because I’m a startup, right? It’s going to be two years in the business in January. I tried that story with my co-owner, CEO, and HR manager, who also happens to be my wife, Heidi. She didn’t buy it, brother. She was like, that’s really nice. Where’s the money? Where’s the profit?

Alvaro Cuba
The Ballina’s business. And one last comment is about change management and organization integration. Again, this has huge potential to help break the barriers.

Ed Ballina
Yeah, sorry, I couldn’t resist it.

Alvaro Cuba
to help integrate, because everyone is going to be seeing the same data. The data is going to be available. This is going to, it will free up time instead of digging the data and putting the case and all that and collaborating and looking for better ideas and interacting with other parts of the organization. So a new set of KPIs, also comes with a new set or a new organization. So don’t be afraid or in other words, take advantage. The landscape is changing. It’s a good moment to rethink the ways you are doing and the ways you can win.

Ed Ballina
I, your comments about the data, when I was in college many, many moons ago, I remember one of our professors saying to us, when you are in school, right, and we give you a problem to solve, an engineering problem, we give you all the data, right, you got to figure out which page in the book you need to open to and what equation you need to use to solve this problem. Because that’s why you’re here, to learn that there is an answer somewhere and we’re going to teach you how to find that answer.

When you step out in industry, it’s going to flip. 80% of your time is going to be spent getting the data. And hopefully, if we taught you right, 20% of time is applying the right equations or problem solving techniques to it. And that’s the promise of all this AI, right? In fact, we’re drowning in data almost too much, right? But the promise of AI is to streamline that so that we don’t have to spend in order to amount of time gathering the data or chewing it or analyzing it. It is given to us in bite-sized pieces. So, Alvaro, I think I’m going to turn it over to you after that. Well, I think it’s time.

Alvaro Cuba
Well, I think it’s time to wrap up. Guys, we hope this quick, but a little bit comprehensive idea of what is this 4.0 and how it goes into manufacturing. So thinking holistically and don’t be afraid to go out.

We hope that this has been helpful for you and putting a little bit of context and some ideas. So with that, that brings us to the wrap up of the episode of today. We want to thank you very much for joining us. And if you like us, like us or give us a review and more important, like I always say, talk about that and invite your buddies to join us. And the more we are, the more ideas, the more questions, the more fun we all will have.

Ed Ballina
That’s terrific, Alvaro. I just learned over the weekend that my 18 year old granddaughter actually watches the podcast. It’s surprising. Every once in a while I meet somebody or a friend or whatever. Oh yeah, we watch you and Alvaro. Hey, that’s pretty cool. I was like, okay. Makes you feel good. I gave him an autographed picture, know, five bucks, you you got to make money somehow, you know, anyway.

So if you want to keep this going, email us at mmu at augury.com or find us on The Endpoint. That’s a really cool online community for manufacturing pros like you and I. And you can find that at endpoint.augury.com. And we also will have those in the links for the episode. Manufacturing buddies, stay safe and hope all of you come out of this in the best shape possible. God bless and see you next time.