AMA: GenAI & How to Avoid Machine Health Mistakes

Ed Ballina (00:06.106)

Well hello there, I’m Ed Ballina.

Alvaro Cuba (00:09.134)

Hey, good afternoon guys. I’m Alvaro Cuba.

Ed Ballina (00:13.306)

And I think he’s still overseas because it’s like barely noon here. Actually it’s nine. So good to see you traveling in Finland.

Alvaro Cuba (00:20.398)

Actually, I continue in Finland and I’m enjoying the weather, which is not too hot, not too cold. And also the system that I mentioned in the previous episode, it’s something special to live it. It’s not only what they say, but how you live this and the experience. How about you Ed?

Ed Ballina (00:40.986)

You’re experiencing it.

Well, I stepped outside for a second and I had a delivery of cardboard boxes for our business. Okay, how exciting is that? Except that when the truck pulled out, there was a picture of Dunder Mifflin Paper Company. Any of you that have watched The Office, which you know is filmed in Scranton, I was blown away to see this truck pull up with Dunder Mifflin on the side and it’s PA paper, but they play that off very well.

So that was, that was my little a-ha from about 15 minutes ago, because I am a rabid fan of The Office, I think it’s one of the funniest shows out there. So, but hey, enough of that. Welcome to Manufacturing Meet Up podcast. The show where we kick back. we talk about our downtime in the plants and, efficiency on the plant floor, so, with that, let’s kick this off. Ask me a question.

Alvaro Cuba (01:52.013)

Okay, so today we have two questions coming from our listeners. So the first is, how are you seeing generative AI being used in the plant?

Ed Ballina (01:52.602)

Okay, so.

So it is propping up in so many different places, right, in manufacturing. Everything from digitized training material, but we talk a lot about using predictive technologies with the AI to keep our equipment from failing. And that’s one that’s starting to be pretty common. But I personally have seen folks using

Google Glass, for example, right, to talk to a technician in Italy from Atlanta, Georgia. And it’s just amazing how you can really get almost 80% of an on-site person helping an operator or a mechanic adjust a piece of equipment when they can actually see and direct the work. Also, there’s more and more development taking place in using AI to help us predict the demand on our products, right? Everybody’s got different algorithms to figure out how to produce what the consumer requires at that point in time, but we’re getting much better at that with AI. How about you? What are you seeing out there, Alvaro?

Alvaro Cuba (03:23.213)

Couples that I have seen, and you mentioned one, is the training material. It’s impressive how it is tailored to the operator and the machine. Even starts talking the same language, like you were talking to a pal that is beside you. And then the other is, remember Ed, in our times, the amount of time that we were using to do operating procedures, SOPs, or design of experiments. Well, it’s not necessary anymore that you do it and you sit down and write it down, no? So now just generative AI is doing it for us. But the really simple and cool, and we have repeated several times, is the mobile.

Ed Ballina (03:53.754)

Yes.

Alvaro Cuba (04:23.213)

Now that they are putting generative AI in iPhone or in Google or everything, it becomes all in one. Now you have in your hands that and you can use it and you can connect to other plants. You can show the video, you can talk to the people, you can even show the machine and they can tell you, okay, that’s that.

And improve quality or many other things. So that’s something that I’m seeing and as you said, more and more frequent.

Ed Ballina (05:00.442)

It’s very cool. And people think because this is kind of new and emerging technology that it has to be daunting and you need to have a degree in IT and you really don’t. Part of the magic of this stuff is they’re making all this information imminently available to you in consumable fashion. That just cuts so much of the time, right? To be able to generate that. One of the… my God.

Alvaro Cuba (05:23.053)

Chat GPT with Apple.

Ed Ballina (05:27.258)

I was going to say, you know, I’m on LinkedIn and I post probably once a week and ChatGPT, you know, you can choose, Hey, can you clean up? Some of this stuff is really good. Some of it I’m like, I better double check this because it took my conversation a different way. But one of the coolest uses of AI that I have seen is there, there are some sugar mills in central America that actually are using AI to determine the exact time to pick sugarcane. When it will yield, it’s provided highest yield, the time best flavor, right, high sucrose levels and all that. And they literally will pick that load of sugarcane and get it to their mill. And as it goes across their gate, it’s tagged as a batch, and they will modify their process to the characteristics inherent in that. And it’s amazing and maximizes yields and reduces waste. So, lots of really, really cool things going on in the AI field.

Alvaro Cuba (06:27.309)

Let’s go to the second question. What three tips you can give listeners to avoid machine health mistakes?

Ed Ballina (06:41.882)

So that could be an episode on all into itself. But I’m going to pick two. I’m going to say properly trained operators and maintenance people. In a lot of locations, we still have the buddy system for training, where you learn from another operator, experienced operator. But unfortunately, that at times perpetuates errors or the wrong way to go about it. So properly trained operators, I think, and maintenance folks is really, really critical. And the last one that I’ll throw in there, I know you have several others, is what I call not closing the last yard. So what I mean by that is we have all these great systems, right, that provide us with terrific information in terms of causes and actions to take to correct issues. But if you don’t close the last yard, nothing happens. What I mean by that is if you don’t have somebody, a mechanic or an operator that will take that AI information and actually do something with the machine to fix it or to correct the problem, it doesn’t matter how great your technology is. You got to close the last yard. Alvaro.

Alvaro Cuba (07:58.724)

Couple from my side, one is avoid the “I know better.” What I mean by that is standard settings in the machine. That’s what it should be. Everyone should respect those. Not the new operator coming and say, I have my way. And by the way, it’s better than anyone. And then the next one that comes have to redo everything and every shift you are going up and down. The other is, it’s one very used in operational excellence, autonomous maintenance is CIL: is clean, inspect and lubricate. No, you clean to be able to inspect, you inspect to be able to detect earlier the things that you need to adjust and then you lubricate your machine so it runs smooth all the time.

And I’ll throw one bonus there, the quality, it has to be checked end to end. No, it’s not only, it’s raw and packed and work in progress and everything. So you can have a repeatable results and the quality that we were discussing previously.

Ed Ballina (09:19.674)

Great point. That’s why they call it a supply chain, right? A chain is only as strong as its weakest link. So you could have a phenomenal product quality performance, but if you don’t deal with packaging quality, right, the consumer never gets to what’s inside the container or what’s inside the package if all of a sudden it breaks or there’s an issue. So, I think there’s a couple little nuggets for our team here to take and run with.

Alvaro Cuba (09:52.9)

Yeah, so, and with that, let’s close it.

Ed Ballina (09:59.226)

I think, you know, we tried to keep this one short and sweet, give us some feedback. So that wraps up our AMA. It is not the American Medical Association. You didn’t come on the wrong podcast. Although Alvaro and I talk a lot about safety. But hey, Thank you so much for tuning in today to our Ask Me Anything episode.

We’re only as good as the questions you get. So we want to make sure that we are listening to our Meet Up community, our Meet Up buddies, so that we’re tackling those topics that are important to you. I mean, at the end of the day, we’re here to provide value and hopefully get more people to engage with us. So you can email us at mmu@augury.com. You can find us on The Endpoint, which is an online community for manufacturing pros like you and you and you and me, and that’s at endpoint.augury.com and we’ll throw those addresses on the show notes, and look forward to more questions from you. Like we’ve said before, Alvaro and I will always have an answer. Not necessarily the right one, but you know, what we think is the best. We’ll have an answer for you. Alvaro.

Alvaro Cuba (11:06.66)

But we’ll have the answer. Yes, And we’ll share it. Well, we are doing this, ask me anything, because there is a lot of questions out there and the couple that we discussed today is an example. So keep sending us those. We’ll keep giving us our point of view and more important, exchange ideas.

Don’t need to be questions, could be comments, could be your own experiences on how you are doing things and working. And we promise we’ll share those with the rest of manufacturing buddies So if you enjoy this episode and please follow us, subscribe. If you watch YouTube or iTunes, give us a review or like us.

And it’s always a pleasure to spend this time with you and all the best buddies.