Austin Anderson with Fluke Reliability

SUMMARY KEYWORDS

Industrial Talk, Scott MacKenzie, Austin Anderson, Azima, Fluke Reliability, diagnostic system, vibration data, reliability engineers, user interface, AI accuracy, work orders, data storage, oil analysis, thermography.

00:00

Hey, industrial professionals, before we get into the show, I want you to be aware of a free ebook that we have out on Industrial Talk. It takes and expands upon the five. These are common five elements that make companies successful. This is it. One, they educate. Two, they collaborate. Three, they innovate. Four, they invest in the culture of the organization. And then five are able to, in an effective way, communicate to the masses of what they do ineffective way. Those are the five elements. That's what the book expands upon. Go out to Industrial Talk, download it, and while you're at it, get the free workbook too, as well. Now on to the show.

00:58

Welcome to the Industrial Talk podcast with 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 all right

01:15

02:09

02:35

You know, I'm sorry you already did your your your background, so we all know that you got the street credit. You just had to chime in and say diagnostic. I mean, you got a gazillion terabytes of information and data that that exists within a Zima, and I'm not exaggerating. We got years of it. And how do you, how do you ensure that that information is still accurate. I mean, what are you doing?

03:06

Well, it is a good question, because, I mean, you've got years of data, like we said, it's stored, and, you know, we've acquired it all the way back into years and years ago. It's amazing. And to be to be honest, it's like, you know, at some point that data does get a little bit stale. Little bit stale, you know, yeah, still relevant. I mean, it's historical, but the way that we've maintained it, we've always the way we capture data has been standardized. So we've taken high quality data. I mean, the way we take data now, in terms of vibration data and what we look for when we're taking it, that's been standardized for the last several years. So, I mean, really, data 15 years ago, 20 years ago, it's, it may be old, but it's still relevant today, because it's still usable by our system. You know, you have a standardized way of how we collect it, what we look for in our data, and so, you know, it really, we have a data solution that really helps enable us to get the most out of it?

04:00

Are you vibration tech? Are you gonna get certified?

04:05

I am a category three vibration analyst, but I, I don't do it day to day. You know. I work very, very closely with our team. I do work on the software side, but my background is in vibration. So I, you know, prior to coming to Z mind, I did work at a work for the department the Navy, and I did vibrate, that's where, that's where Zima came from, right? Yes, yes, back in the 60s, way back when.

04:26

Yeah, see, yep. So you were in the Navy. I

04:29

was, I worked for the Navy. I was a civilian, but, you know, I did work. They did vibration related research and development for, you know, their their torpedo systems, basically from the submarines. And so that's right, so I cut my teeth in the world of vibration, and that's kind of what pointed me this way. Was so cool.

04:47

I get all giddy on that one. Okay, so people talking about Azima, people are talking about what's, what's taking place, what's on, where, where's the Zima going? What's, what's that strategic path? Look like. Well,

05:01

what we're looking to do, we're looking at multiple time horizons here, and what we're trying to accomplish. And you know, the name of the game right now is building something that's going to be useful for top level executives. They want to understand what the reliable you know, what they're getting out of their reliability program. What are you doing? You know, you're spending all this money. What's it doing for you? And so we need to communicate that to the the executive leaders, the people making the decisions. At the same time, we need to communicate what you're doing for the workers doing the work at the bottom, you know, in the chain,

05:30

you know, ones that are in the trenches do exactly the real work, you know.

05:33

And essentially, at the center of it all is, you know, we're trying to make Reliability Engineers, you know, superstars, so to speak. You know, that's cool. One of the platform it says, you know, if I'm already I've got all the tools at my fingertips to make sure I'm well connected with the state of my plant and, you know, easily understand what's my highest area of risk and what are things that that are doing fine and I don't even need to pay attention to. And so really, it's all about curating where that attention needs to go.

06:02

So Fluke reliability acquired azema. I don't remember what the name of it was at that time, but let's say it was a Zima back there, and then it's up to Fluke to sort of take that that platform, and to make it into a user friendly, value added solution. So is that, is that an accurate statement? Is that where you're taking it? We acquired it, we see the value now we've got to, you know, sort of make it a fluke, reliability,

06:38

absolutely, and that is where we're going. And so we've demonstrated that here at this conference, you know. And just kind of the way that we've laid it out is, you know, we're taking all this data. We've got the hardware solutions, you know, between our proof technic, you know, they develop world class hardware, yeah, capturing all this great data. Our systems built to, you know, the azimuth solution is built to take all that data, get insights from it, generate, you know, actions out of that. And then, of course, that's where email comes in, where you can actually manage that workflow. And then, of course, we have our alignment and balancing services teams and all the the additional technology that comes with that. So really, we're talking about end to end, connected reliability. So any solution you need is going to be, you know, Fluke is going to be there for you. Essentially, alignment,

07:22

how does that fit within Azima?

07:25

Because I it does not. We will find it. We'll say you're misaligned. You know, you need to align this unit, but yeah, because I

07:34

had that conversation about Azima, it's always vibration again, a gazillion years of vibration data, and then I'm going to ping it up against some other so with that said, so the the user experience, the user interface, is being modified in such a way that it's like because Here's the reality with fluke and Fluke reliability is that many of the tools that you have are are designed in such a way that it makes it very user friendly. And if a tool is not that way, then nobody uses it.

08:18

That is exactly it. And we are rebuilding our user interfaces, you know, we're sort of taking the opportunity with this acquisition to to make a more user friendly application in order to interact with this data. And one key element of that is the simplicity we're trying to make it, yeah, easily understandable. So as an example, as a user, you don't want to have to click five different times somewhere. You want to see where that data is, and you want to get there. Want to get there easily. That's that key element of what we're trying to design in our next generation portal. I

08:48

hear that you're testing the two questions, so I don't forget if I and I'll get to it if I send if I use the proof technic, solution, vibration, whatever it might be, vibration detection, and I hit it into a Zima, how quickly can that sort of there's a result that spits out real quick. From the time

09:15

that data arrives to results being generated is a matter of moments really see that's so cool. Yes, of course. The caveat there being that oftentimes it'll be reviewed by human, you know, in some time period following that, but you can get results very, very quickly, you know, out of our diagnostic

09:29

system. How, how accurate has that been?

09:33

I think we tossed around some numbers. And I think if you were to do, you know, forget the human element, if you were to just focus solely on the AI side of it, and assuming it's set up properly, that probably gets you about 85% of the way there, 85 I would say, you know. And that again, that's, it's hard to verify that number due to swag, the multitude complexity, you know, but, and then that added element of that, that remaining 15% really gets covered by the analysts, you know, human humans in the loop here. So it. You know, we believe that AI and humans working together, yeah, provides the most optimal outcome for our end users.

10:06

Yeah, no, I agree with you 100% on that. I do like that solution where you have that the marriage between the technology and the human element and being able to come up with real, real sound. You know, tactical solutions, absolutely. How does that work with E mate.

10:34

The E mate side of things, that's sort of a recent development pillar that we're focusing on is sort of taking what we're looking to do, in a nutshell, is to translate the predictions the element we've identified a problem, something needs to be done about it. And E mate comes in is like, let's actually do something about it. And so the whole purpose of this connection between the Azima platform is to insert, it translates to work orders, basically. So when we identify a problem on the Azima side our, you know, through our AI and analysts, you know, solution, they're going to find a problem, and then that's going to immediately translate into a work order that you can actually take action on in your your system of record and email. So it's all about work orders.

11:19

When will that take place where you're just, it's, it's, it's connected, it's, it's all a part of the service offering that exists for for that connected reliability,

11:31

exactly. We are in the very early stages of it, and we're still doing well. I would argue, a solution is never perfect, and what we're going to deliver those soon, those core functionality in terms of that event driven architecture. So what I'm what I'm talking about is when our analyst, as soon as our analyst reviews a test and says there's a problem, you need to you need to take action. You need to replace this bearing. Now, as an example, the second they click save, and if you are connected in x5 that's going to be a work order right and right inside x5

12:02

and the information is just auto populated, so I don't have to have this human intervention of saying, Well, no, this is the bearing, x, y, z, and so on and so forth. It just creates the work orders with the supporting documentation and gets it in the queue,

12:20

Exactly, yep. And if it will, once it's in the x5 system, a record, you know that that's it's then on the customer. Do you know the add additional supporting documentation. But you know, we're gonna, we're gonna populate what fields we can based on the evidence that, as much as possible, as much as possible, you know, give you that information. Give you that context. You know about what it is that we're seeing and what you need to take action on. But the

12:42

same thing can exist. Today with your proof technic solutions. Are you? Because I know that there's the focus of everything being connected in some way, shape or form, so that that data flows effectively and efficiently and accurately. The proof technic solutions. I don't need a Zima. I can still use that as a standalone vibration

13:08

you can tool, yeah, yeah. If you, if you know, they proof technic builds world class hardware. I mean, the expert two has been in service for years as a, you know, very, you know, in depth analysis tool, and that can be used with, you know, without, without a Zima, really, you know, you don't, it's an addition to, but you still need a human element, still an element you just, you rely on, on that human expertise quite a bit more with that, you know, in terms of interpreting that data. But again, from a data quality perspective, you know that hardware is top notch, really. It can do so many things, do anything you can

13:41

think of. And as you continue to populate a Zima, as you continue to fill it with data, how does it? How does that structure look like? If it's just you're constantly feeling it with with with vibration data. Is it? Is it specific to assets, yeah? So, here's a motor. It's an XYZ motor. Here's the vibration, here and so on and so forth.

14:06

Absolutely, yeah. So, yeah. I mean, the machine is kind of like a like a person, so to speak, and you have a medical record with all your, all your data you're collecting over the years. Yeah, it's no, it's no different than us. Your machine, machine training might be a motor pump kind of combination. Yeah, that's a machine. You know, how we consider it? And, yeah, that data is going to live with you. It belongs to you as a machine. And so it's going to we'll have that history. So if you've been taking data for 30 years, we'll have 30 years worth of data to look back on, including everything in between. You know, all the all the different motor replacements, overhauls, all that stuff. It all stays with the machine and No, no, no, no,

14:44

go, do the end. I was

14:45

gonna talk about, you know, the terms of data storage. I was gonna get a little bit, maybe a little bit nerdier here, you know, all that, all that data does really take up a lot of space. And so, you know, I think that's what's really one of the things that's been exciting about, you know, from, from a data. Perspective is the new modern database technologies that the way we can compress and store that data and make it available on demand without taking up excess space, you know. So, you know, from a data management perspective, it is a challenge because, you know, we pay for that. You know, we have to, we have to pay for that. And so, you know, it's in a new way of storing that data, that time series data, you know, there's, there's specific architectures out there that really help enable us to store that data, right, right, right, right, load it on demand when you need it. Yeah, wow. So it's kind of exciting,

15:27

you know, does it? Does it also sort of clean the data is like, there's a lot of noise. Do I need that really? Or do you, you? Do? You just save all of the data. We will save

15:38

it all. I mean, you can delete it, of course, after the fact, if you want, but we will store it. Um, store it and then, but there'll be, you know, additional metadata attached to it. So as an example, immediate, it's a fairly common use case where you'll take data on a machine with, you know, with a wireless sensor, an online system, you know. So, yeah, you know, the machine may not have been running, you know. And we do try to catch that on the edge and try to only take data when that machine is running, right, right? But it slips through, you know, it does. It's and so. But our system will identify that, you know, by the time it reaches our diagnostic system, we'll be able to say this machine wasn't running, you know, this, this is flat line data. Let's, let's categorize it as such, and we sort of set it aside.

16:15

Yeah, that'll get resolved too, and it'll all be taken care of. Yep. Where do you see it going? I mean, where's that future look like to put on your future hat. What's what gets you excited for tomorrow?

16:28

What's getting me really excited? Well, there's many things, but if I had to focus on, you know, maybe one key element of this, of this conference in general, is vibration is a major pillar and reliability. But it's not the only one. You know, there's other ones. Oil analysis is another key component, yeah, so that's one that's keeps me up at night thinking about and if I were to, you know, if I were in charge, and I would focus on developing a platform that can suit vibration needs the way that we do now, and enhance upon that, but it would also encompass, you know, more, more modalities in terms of reliability. That'd be thermography, oil analysis. I mean, one stop, one solution, you name

17:13

it, can you can you take, to a certain extent, the the existing architecture of Azima, and just populated with different

17:23

data. Absolutely, yeah, it's extensible, and we'll, obviously, we got to develop it. There it

17:27

is. You're the man. So if I wanted to get a hold of you and we'll find out more about a Zima, how would I do that?

17:33

You can reach out to me on LinkedIn. So Austin Anderson, you can find me at flicker liability, and I'd love to talk to you. There you

17:39

go. Now, there's got to be more. Austin Anderson's out there on LinkedIn. Oh,

17:45

there's a million of them, sure, but yeah, yeah. Flip reliability, blue girl,

17:49

18:16

You're listening to the Industrial Talk Podcast Network.

18:26