Kevin Clark with Naonprecise

SUMMARY KEYWORDS

Industrial Talk, CAP Logistics, SMRP 33, asset management, reliability, maintenance, technology, RCM, data quality, AI, condition monitoring, CMMS, cultural shift, quantum computing, cybersecurity.

00:00

Hey, Industrial Talk is brought to you by CAP logistics. You want to minimize downtime, absolutely increase reliability, you bet, ensure operational profitability. Yes, you do. That means you need 24/7, 365, insights into your supply chain. Look no further cap logistics. Go to cap logistics.com or just call them. They're great people. 802 to 724, 71

00:34

you're listening to the Industrial Talk Podcast Network Scott Mackenzie. Scott is a passionate industry professional dedicated to transferring cutting edge industry focused innovations and trends while highlighting the men and women who keep the world moving. So put on your hard hat, grab your work boots, and let's go all right once again.

00:53

02:03

It's about time we get to connect again.

02:05

About time I feel like, I feel like I was at the bottom of the list.

02:09

Yeah, if you, if you're lollygagging, and because an Industrial Talk is, is quite the move here and shaker, you need to, you need to act with it

02:19

is, I just expected somebody to sign me up, and apparently I just didn't happen. Here I am.

02:24

No, here we are. Here I am because you're filling in for somebody else.

02:26

That's right. Lucky Dog, yep. Lucky person, yeah. Having a good conference.

02:32

It's been a great conference. This has been a step change from 24 step change. It's been, what do you mean? I don't understand. Well, there's a couple of statistics that I think is interesting. One is, normally we would have around 130 vendors. We've got 220

02:50

Yeah, yeah. It's a massive difference. It's huge. Yeah, and, and I, and again, I see more people. Yeah, it, it's, it's amazing. Yeah. I

03:00

think the other their statistics and this one hasn't come out yet, but I recognize it, just because I've seen this list so many years, is that there are so many practitioners here, like so many people that need to learn. There's only people that want to figure some figure out some solutions for their problems. And so while we do have more vendors here. I think the quality of the practitioners is also

03:25

stepped up. Do you think we're in a renaissance? Could be, you know, could be, there's, there's Nancy, like, right there, of course, yeah. Now, all of a sudden, she needs a seller. Now we're recording now, now we're having a conversation. Nancy, yeah, he couldn't find the floor. Couldn't find the lights.

03:50

Nancy, I, I came over here quickly and said, I need a slot before Nancy gets here.

03:56

He did. Yeah, I think we're in a renaissance, by the way, that's Nancy Reagan. She is a legend in RCM. There you go. She is. Give her she's

04:07

the big one. She is. So she she's, I think she's, I think she's helping to transform the industry. I think she might be maybe part of that Renaissance, because she's bringing a different flavor to RCM. She's bringing a different kind of excitement, yeah, making it feel like it's not an old methodology. It's, it's, it's something that that can be celebrated, and it also, the crazy thing is, it is an older methodology from back in the 60s, but, but with the further technology goes, the more RCM makes sense why it's a crazy dynamic. Because it's dependent upon data. It's dependent upon programmatic behaviors, right? And people are not good at programmatic behaviors. That's part of our reason for unsuccess in RCM. Them. But when you start bringing technologies into the RCM space, and it gets more programmatic, it gets it collects more data, it's going to get better. And I believe that's part of why RCM is starting to lift up even higher as a as a go to methodology for improving maintenance.

05:21

Why do we still have the same conversations? I mean, yeah, it sustaining the incredible gains that at, you know, result from these practitioners. And then all of a sudden it's it goes the other way, and then it just keeps evident flowing. And do you think that the advent of the technology and the advent of the the ability to collect data will help move that ball forward a little bit more productively.

05:49

Yeah, so here's, here's my perspective, Scott, is that that the data that we're dealing with today because we can say, we can often say things like, like, our data quality is low or bad, but often what we find is the data isn't necessarily bad. The data was just designed for people, right? And so when we're collecting data off of machines, there is an awful lot of that data that kind of doesn't matter for AI, because AI wants to know what it needs to know for context. The rest of it is all information for people, and that information really kind of clutters our data's landscape, true, and there's just so much of that data that we just don't need, and especially don't need it. If you're designing data to feed another system, you're going to design that data to be different than what it would be if it's at the end, if the end consumer is a human.

06:51

Do you believe the technology, the ability to collect that data, will, just in general, improve the quality of that data. I'm not concerned. You know, I've got this saying, bad, good, move on.

07:05

So you get to the point, like in a CMMS system, you get to the point where, let's say, 95% of it is all system data that's automatically generated for a work order. That's 95% of the data being created is data being created for the system, not for people, right? And so that means you've now started to limit 5% of that data that's being created for that work order is only human created or human generated, right? And so you're starting to eliminate the level of errors, and let's say not methodical or programmatic kind of data. So yeah, the by nature, the data and the results will get better once we build more and more system data into the system.

07:55

How do you? How do you? And I agree that that sounds cool. I agree with that because, yeah, you know my old approach, back when data was always crappy and it was always bad and dirty, and you had to try to figure out how to clean that data and scrub it, and nobody wanted to do that. They'd go down the road and say, Yeah, let's go, let's clean that data. And then they go down the road for a week, yeah, then they just sort of, you know, uncle, we'll just figure something else and and never really go back and never really do anything right, right? But do you think, and I in it as time goes on, these, these solutions, these ability, the ability to collect the data, just becomes better and better and more refined and more accurate, for lack of a better term, yeah, and just, just avoid the need to question the quality of the data.

08:54

Well, I guess my easy answer is yes, but, but so much of this, so much of the data that we have today is just raw data, and we consume all data. Yeah, right, yeah. And I think as we move forward, and especially as edge continues to to grow and and in present, in more applications and work is actually being done to that data, before that data is actually sent. We're going to get less data. We're going to get less data, but better data, yeah, yeah, yeah, right. And it'll be more contextual, and it will already been massaged earlier, yeah, the edge, yeah. And so, yeah, we're going to get better data, but I also believe we're going to

09:38

get less data. I agree with you. Here's, here's a story. I set something up for utility, and we were collecting data, and this was years ago, before we really had the technology to be able to handle the data. So I put this piece of equipment out on a transformer. It's pulling dissolved gas analysis, doing it real time. So it's. Good. So it's pulling, pulling, pulling, pulling, pulling. And of course, you can have a frequency associated with it. Do you want to pull it every week? Do you want to pull it every you know, hour? What is it right? But you're pulling it. And so the problem I had was that it was a tsunami of data, and not the ability to be able to say, Yeah, I even had a neural net to try to learn and do all of that stuff as I was pushing the envelope, I couldn't get rid of the data. That was just noise. Yeah, I couldn't get to the relevant data. I could, but it's right there, but it's all buried in in this stream of other data, right? It was hard, yeah, and it just sort of imploded.

10:39

12:20

again, I agree with you, and I'm becoming more and more aware of the fact that AI is great as a tool, being able to sort of look at that data, normalize the data, get rid of the noise in the data, really sort of focus in on what is important but but before you lock and load, there's that human interface that I just can't see being eliminated, to just go all in with AI being able to create the work order to schedule that technician, to say, get this and that I just don't

12:53

see it. Well, we went Ramesh and I spoke together today, and we were talking about cultural shift and and and it has a lot to do with AI, right? And that cultural shift is super sensitive. So if you've got movement and you're beginning to shift the way people see things and the way they behave and the way that they the way that they learn to trust, it doesn't take but a few false positives from AI to shatter everything you've done everything. And so I think it's one of the challenges that we have ahead of us is, how do we get people engaged? How do we get people to trust it? And I had a I had a guy come up to me after the after we did our presentation, and he said, he said, How do you use that was his question to me, how do you use AI personally, he asked that to me, no, okay, okay, and that's it. That's an interesting question. So I told him, I said, you know, I have a chat GPT account, and I use that kind of regularly. I run documents through it, and I create, you know, some messages through it, just, you know, more fun than anything else. But the big thing that I use is I use Google and I use Google AI. And he said, Yeah, that's the perfect answer. And he said, here's my point. And this was always good when the when the member of the audience comes up and coaches you, yeah, right. And he said, this is, this is how I kind of judge people on whether they're qualified to make decisions in business about AI. And he said, If you don't use AI in your personal life, then you have no authority to be making decisions on how AI is used in a business. That's

14:47

interesting. Yeah, it was a, I thought it was, I'm all in. I can see that, right? I mean

14:55

that it absolutely makes sense. And I. Play the role of detractor and I play a role of advocate, sometimes in the same sentence, yeah, and, and I think that that's important for people to understand AI in that regard, so that they can do something similar, so that they can have a fairly objective view of AI, and that's how we that's how we begin to make change. Is that people are looking at it critically, not negatively,

15:29

yeah, but there's always an element of fear. I'm all in. I use it personally. I, you know, it, it's funny, it's sort of, you know, I came to the conclusion that, from a personal perspective, do I google? How do I do something in Google, or do I just go to the AI? How do I and get, get that goes. So it's interesting. There's a shift saying, Do I really need to go to Google or just go to AI, right? And so does, is there really a need for SEO or just go to AI and let's ai do the thing.

16:03

And right? Because we're seeing that, it's interesting. We're seeing that too. When, when, when we have prospects that come to us and want to know more about the company, they generally know an awful lot. Yeah. I mean, all you got to do is get on Google AI and say, Tell me about Nanoprecise . Well, what's good about Nanoprecise ? What's not great about Nanoprecise ? Oh, yeah. And it will tell you, it'll list out. Here's what I found, and then they put it out in a story format for you. Yeah, right. Rather than if I just went and Google for Nanoprecise , then all I'm going to see is a list of things that look kind of like Nanoprecise . That's

16:37

exactly correct. And that's something, yeah, I think that there's going to be this? This is interesting. So as time goes on, you need to constantly feed that large language model so that that AI can draw from it and pull and get the information. Do, do what it needs to do to get it right. But over time, everybody's just not feeding it with information. It's not it's just it. Then it goes into this. This is just me. I have no basis. I'm not you or Ramesh, but I think there's like this information starvation, because it's stale, then it starts to pull the same stuff. There's nothing unique or new, and that's why I always think like something like this where we're talking, yeah, transcribe it. That's real. That's real information that adds to that large language model.

17:29

Yeah, absolutely prove me wrong. Yeah.

17:37

I had a I had a colleague that all he wanted to do is change his oil. And he thought for kicks and giggles, he's going to ask Google, AI, what do I need to do to change my oil? And he gave it all the specifications, exact model of the car, and what he was what he needed to change, even put in what type of oil that he normally uses. And he gave all that information to AI and AI came back wrong? Wow, yeah. And so that's

18:06

the watch out. Wow. That's that trust. That's Yes,

18:09

that's exactly right, yeah. So you've got to be very careful about AI. There are times that I'm just blown away by how right it is, or how amazing it is, or the correlation that it made to something else that I just never been able to do myself right, right, and then you hear something like that, right, like you give it everything, and it comes back wrong.

18:32

Yeah. See, I don't have an answer. I don't know why the prompt should have been adequate, yeah, right. Should have been like

18:41

something else out there skewed it, yeah, something else on the internet skewed it, right. But don't

18:47

you think eventually that'll all be worked it worked out, or is it will

18:50

publicly? I don't think, I don't know that it ever will right? I think publicly, we'll have a tough time ever getting it to the point where you could just trust it because, because if somebody can still post something out there that's completely wrong, intentionally, completely wrong. It's going to be part of the query. It's going to be part of what it looks at. Now, it may be few enough that it'll filter it out, but

19:12

you've got an interesting are there nefarious individuals that are out there doing things that are wrong?

19:19

I'm sure there are. Yeah. I mean, that's just fun, see, right? Ah, it's just fun.

19:23

I see that's cool stuff. Yeah.

19:27

And you know that the other thing about it too, is, is if, if you do want to do things like that, where you want to, you know, skew the data on the internet, yeah, there are ways you can do it massively, like volumes of data going across multiple sites that you can place out there and and it becomes part of the queue for for AI

19:48

one last because I have to ask this question. So I had a conversation with the company that was we were talking about quantum computing, cyber security, particularly. Protecting, you know, assets, protecting technology from nefarious quantum computing actors. And although we can say that we don't have quantum computing yet, I believe that somebody is going to crack that code and just like chat GPT one day we didn't have it. The next day, somebody says we got it. And I think that quantum computing, whatever that is, the race, will be on. But I just thought, Wow. I didn't think that either.

20:31

No, you know, and it's crazy too. Like you're you're talking out both sides of your mouth because, because, basically, you're encouraging everything to be online. You're encouraging everything to be available, even though you use the cyber security in that to protect it. Once everything is out there, it is susceptible. Yeah, right. As new technologies come around, cyber security has to update and keep up with those new technologies

20:59

speed too. I don't have an answer, yeah, neither do I don't I just one of those things where it's like, I don't know. Ice forward, Scott. Soldier on there. Soldier On how do people get a hold

21:14

of you? The best way, best way. That's why he's getting me at at Kate Clark, nanopresident.com, so you are on LinkedIn. I am on LinkedIn.

21:25

You filled it because I wanted to say yeah, because I just grabbed your stat car from LinkedIn, and I just slap it on my website. And there it is. I have a backlink, and I don't have to hear about

21:35

no secrets, man, yeah, secrets, all right,

21:39

22:03

You're listening to the Industrial Talk Podcast Network.

22:13

Kevin Kevin Clark, again, he just what he did. He just came in with a wheelbarrow full of insights and wisdom, and came up to the studio at SMRP and just said, Scott, I got something to tell you, and just dump that wheelbarrow full of insights at my feet and at your feet. He's incredible. If there's anybody that you need to connect with, yeah, he's got to be at the top of your list. Nanopercis is the company, and again, right in the thick of it, right in the thick of it, it just, yeah, it's data. We're talking data. Data is becoming the gold again. Well, it's been gold. It's even becoming more gold. It's platinum. How about that? Platinum? All right, I want you to hang out with Kevin Clark, because he's bold, brave and he dares greatly you do too. So hang out with them. All right, we're gonna have another great conversation coming from SMRP. Put that on your calendar for next year. So stay tuned. We will be back.