This episode of elliTek's Industrial Automation podcast is about Predictive Maintenance. We are talking with a special guest, Keary Donovan. Keary is the owner of Pathways 7.
Pathways 7 is a consulting firm helping SMEs (Small and Medium-sized Enterprises) capitalize on up-to-date information systems so they can become more competitive.
Keary recently published a white paper about Predictive Maintenance titled "3 KPI Considerations for Maintenance and OEE."
During this podcast, we interview Keary. You will hear more acronyms, as well as references to "Tennessee Talk."
You will need to listen to the episode to learn what "Tennessee Talk" is, but here is a glossary for the acronyms in the order in which they were discussed during the podcast.
Stayed tuned throughout the entire episode to hear, Keary's insights into the interdepartmental conflicts and how those conflicts can be resolved.
Brandon Ellis 0:24
Hello, everybody, this is Brandon Ellis, the host of Industrial Automation - It Doesn't Have To... Welcome. Thanks for joining us. I'm here with Beth Elliott, Beth, how are you?
Beth Elliott 0:33
I'm doing great today, Brandon. It's December. We're getting into the Christmas and New Year's holidays. But there's no time to Yap. So let's get right into it.
Brandon Ellis 0:43
That's right. So tell us what we got today. It's a special day today.
Beth Elliott 0:46
It is. Today's podcast is named Industrial Automation - It Doesn't Have To... Be Unpredictable. We'll be talking about predictive maintenance. And we have a special guest. It's quite fitting, because this is episode number seven. We have with us today, Keary Donovan, and Keary is the owner of Pathways 7. Keary, can you tell us a little bit about Pathways 7?
Keary Donovan 1:10
Sure. Hi everyone. Pathways 7 is an Information Systems consulting and contract firm that we started to serve SMEs, small and mid-sized enterprises. And our specialty is seeking practical solutions for not just manufacturers but the supply and sales channels that support them.
Beth Elliott 1:32
So what prompted you to get started in this business, Keary?
Keary Donovan 1:35
Well, I was I was raised in San Diego, but it was by a bunch of Midwesterners who grew up in southern Chicago, in Oak Lawn. Shout out to the Weidners and Donovans. So I grew up idolizing some of the elders in my family for being builders and being in manufacturing, whether it was an industrial sales or otherwise. And so that's why I kind of sought it out after I studied business and Information Systems. When I was in college. After about 10 years in motion control. I started my own web store to help manufacturers replace obsolete parts.
Beth Elliott 2:11
What's the name of that web store? Keary,
Keary Donovan 2:13
That is about to change, but it's MotaDistribution.com.
Beth Elliott 2:17
Keary Donovan 2:18
And so during that time, one of the product lines made a claim that they could prepare data for databases with one of the modules they had for what's called a PAC or a P-A-C for Programmable Automation Controllers. And that didn't, that didn't sit well with me and I have, I have a terrible personality flaw that once I hear something can't be done, or I get curious about something that people don't have an answer for, I must do it. And that, to make long story short, is about when I met Brandon, in his horse trailer.
Beth Elliott 2:59
His horse trailer?
Brandon Ellis 3:02
It's a racing race, a car trailer. Apparently, you had never been around the race circuit, but you're not a Southern guy so...
Keary Donovan 3:11
Well, for the guy who grew up in San Diego, it was a horse trailer. So and the only way I really ever have a describing my first impression of Brandon is, you know, walking into that trailer after nobody could answer that claim. You know, we were at this distribution conference for this manufacturer. And they all said Brandon Ellis when I asked who knows about this module in this technology, and he of course wasn't in the meeting, he was outside actually tinkering. And, and only, the only way to paint the impression from that first impression is, you know, imagining opening the curtain on the Wizard of Oz with everything moving around him with the actuators and, and that's when Brandon and I started talking about how PLC data can be prepared and transported and exchanged with databases. And that led me down a road to expanding my web store for just replacement parts to creating a kit that will help people replace the parts before they fail using this data exchange.
Brandon Ellis 4:22
You know, let me expand on that a little bit. It wasn't a horse trailer. It was a racing trailer. What what it was was we for that for that conference, there was it was a demonstration trailer. So we had a lot of things that we were demonstrating in that trailer, but one of the things that was featured was an attempt to do data transfer and Keary, if I recall, that was probably 2011 - 2012 or something that you and I met.
Keary Donovan 4:50
Early. Yeah. 11 almost sounds late, a little bit
Brandon Ellis 4:52
Yeah, I think that's right. Oh, yeah, maybe maybe it was earlier than that. So that's when I had first started getting into. You know in 2009 when elliTek started, I was writing stuff, but it was all on the operational technology, which is the plant floor side, between equipment there, that was still desperate systems. But again, like a CNC machine or a robot that wouldn't wouldn't was not able to talk other than I/O to a PLC or a PC or something that's on the. But still, on that solid side, we've talked about in other podcasts, the operational technology side, or the plant floor. And so I was just getting into, by the time Keary and I met, how to get that information into the IT side. Kind of crossing that gap, or that chasm, if you will. And so that's when Keary and I met and so, yeah, he kind of the reason I was out there tinkering is because I was supposed to present this and there was a problem, and I couldn't figure it out. And I was trying to figure it out. And so he found me, doing my Wizard of Oz stuff. But anyway. Okay.
Keary Donovan 5:56
Well, that product, of course, had a terrible flaw in it, and that it was
Brandon Ellis 6:02
Keary Donovan 6:03
It was PC-based. And so, as Brandon has heard, you know, I don't have Windows PCs in my house. And I always been baffled about why anybody would put it next to their machine or the production, much less on the machine control. And so that module was had a fatal flaw in it. And you know, that instance, developed into what Brandon and elliTek have developed, and what we now use for our kit to take in sensor information and communicate it with computerized maintenance management systems.
Brandon Ellis 6:38
And to be sure that module is not the IIoTA or the Data Commander, that's what he's saying came out of that. Because we were, we were still trying to do a lot of stuff then and still that was in the days of SCADA systems and all the things you've heard me go on about that, that were just a point of frustration, and Keary was just as frustrated. And that's so when we met we instantly connected because, you know,
Beth Elliott 7:01
You both shared the same frustration.
Brandon Ellis 7:03
Misery likes company. And so we were both pretty frustrated, because he was getting asked, you know, out in the western United States, he was getting asked for the same kind of things. And of course, you know, had folks that told him, we just can't get there from here. And that's the determination that he has. And then I had the same thing, because, but my determination was more about we should be able to get there. I don't understand why this is such a hard thing. So that's where we kind of are.
Keary Donovan 7:30
That's really what it was, right? That's what it was, it was such a tease, because for me, it started with, well, you shouldn't be able to do that someone's gonna have to actually show that to me, because I don't believe it. And then Brandon actually showed it to me, and then, you know, and so you get really excited about it, but it was, it was such an almost project - product that it was, it was awful.
Brandon Ellis 7:56
Keary Donovan 7:57
Because, because it was there somebody, somebody had it, but did it wrong, they implemented it wrong. And and so, you know, it's, it was one of those things where, like, somebody will just do it, right, we, you know, we'd really have something and
Brandon Ellis 8:12
We did it.
Keary Donovan 8:13
Yeah, and that's, you know, and then we're big on actually seeing it. And that's kind of what my white paper is that we're that I wrote is we you know, we just don't talk about the theory of things. We're gonna talk about how to actually get it done and how the data actually moves.
Brandon Ellis 8:29
So yeah, that's so the white paper. And that's a good segue. So the reason that they first of all, let me say that Keary Donovan is a longtime friend, and actually former business development manager for elliTek. But Keary has had developed or developed Pathways 7, and now is one of our premier partners, not just with this kit that we're going to be talking about, but also he's our distribution arm for a large portion of the western United States. And so it's certainly a pleasure to have you with us. And you've been with me, like you just said since the beginning, as far as this being a common goal for us. And so this white paper, I'm privileged and honored to say that I mentioned in it, because I think you based it based upon one of our past podcasts that that you had listened to, that was called Industrial Automation - It Doesn't Have To... Be Myopic. And Beth came up with that word, not me.
Beth Elliott 9:28
What's the name of the white paper?
Brandon Ellis 9:29
"3 KPI Considerations for Maintenance and OEE" And so in that podcast, we talked a bit Beth and I discussed a lot about KPIs and OEE and OOE and TEEP and availability and quality and all these different things. And Keary in this paper has really expanded upon that nicely and in my opinion, more even more effectively because of your, Keary, your focus on the maintenance - specifically the maintenance side. We've talked about Operational Technology. We've talked about IT or the Enterprise side, the Information Technology side. But there's this other maintenance side. And that's what I think that this paper really likes to, or really, really points out. Do you agree with that, Keary?
Keary Donovan 10:19
Yes, yeah, with our customer base with the parts, we do, we do a lot surrounding the subject of maintenance concerns, beyond just replacing parts.
Brandon Ellis 10:33
Because maintenance is a very vital, and an additional part of what I would call the Enterprise Resource Planning or the ERP system. So there are a lot of management or I'm sorry, maintenance software's programs, things of that nature that are and typically I think, Keary, this is certainly if you got a comment jump in. But my experience is most of those are PC based, running in a PC based or which is more it based environment. But they need information that's coming from OT. So again, we've talked more in the past, we've talked about getting information into for IT, and it's in the database and or the ERP system, and then we kind of stop our discussion. Keary has carried it a step further to say there is a another software platform. And it's true for accounting, and the others, too. They have software platforms that the ERP system feeds data to. But how do you get that to the maintenance software, when a maintenance software will not necessarily work directly with the ERP software? And so that's, that's the gap he's bridging with this kit. And so Keary, what I'd like to do is walk through - you've got you kind of got this broken down in your paper very well into into three kind of categories, I guess, are three sections. I thought we have Beth, take us through each section.
Beth Elliott 12:02
So the first one is OEE, OOE, and TEEP rule operations. Do you want to expand on those, Keary?
Brandon Ellis 12:11
What do those mean? What's OEE
Beth Elliott 12:15
Keary Donovan 12:17
This is a quiz. So Overall Equipment Effectiveness
Beth Elliott 12:21
Keary Donovan 12:21
And then you have Overall Operational Effectiveness. And then you have Total Equipment Effectiveness. And I always forget what the P is. Program?
Brandon Ellis 12:36
What is it process? I can't remember the P now
Keary Donovan 12:40
In this white paper, so okay, let me just tell you a little feature about our blog post, we include a glossary. So in our white paper, it's Total Effective Equipment Performance.
Brandon Ellis 12:53
Performance. That's right. Yeah. Yes, yeah.
Keary Donovan 12:55
Well, we like to do, what I like to do in my papers in my blog post is we include a glossary for all our short speak and acronyms. So people can translate what we're talking about.
Beth Elliott 13:08
Brandon Ellis 13:09
We have no shortage of acronyms.
Keary Donovan 13:12
Brandon Ellis 13:15
So so in that section, I that's that's where I got mentioned in your paper. So thank you for that, I think. But you, you were actually referring to a metaphor that I made during that podcast. And you refer to I guess, I had referred to it before to you as my Tennessee Talk. But my metaphors and because so I use a lot of analogies and metaphors. And sometimes they're good, and sometimes they're not. But usually they're interesting. And so in this case, I had a metaphor about a refrigerator. What do you think what got you on the refrigerator, Keary?
Keary Donovan 13:49
Well, what caught me about it first is I've been witness to many of your metaphors taking wide subjects, and for all of the considerations of production, and narrowing it down. And that refrigerator, one caught my ear because of what's happening in terms of data exchange that that I've seen. So because industrial IoT has been so difficult to implement, you know, it's people have have started feeling the realities after the hype. And now that they've embarked on their projects, and what I've seen is many requests from the IoT hype, get narrowed down to just - Can you tell me the machine is on or off or not?
Brandon Ellis 14:39
Right? Yeah, you're exactly right.
Keary Donovan 14:42
Now, in the, in the maintenance world maintenance, has, you know, kind of a has been disparaged, right? They're only responsible for keeping the machine on or off. You know, and of course, there we probably go into that later, but there's kind of a built in conflict organizationally between production and maintenance, kind of around around that subject, like just keep the machine on. But what I liked about the metaphor was how just on or off, doesn't really tell you anything, relative to even say, When do we want it on or off? Much less What is the performance or the quality relative to overall production. And I have been at the time reading some articles, as I like to do, that included some critiques of the mantra that what gets measured, gets managed. And this is one of those mantras that marketers and middle manager type, you know, self improvement books glom on to. And so right when you said that, right when you said that metaphor, and we're making a critique of your own, about how just conceptually data needs to be looked at. I was in the middle of reading an academic paper, criticizing not only the fact that you have to be careful about what you measure, but that the mantra that it came from, wasn't even quoted correctly. The full quote was what gets measured gets managed, hyphen, even when it's pointless to measure and manage it, and even if it harms the purpose of the organization to do so.
Brandon Ellis 16:31
That's right. That's huge.
Keary Donovan 16:34
Yeah, and so it was very timely, you caught my ear. So yes, I used your "Tennessee Talk" in my very academic portrayal of data measurement, and especially relative to what what is difficult and, and some barriers that maintenance departments that we hear from, have had to struggle with to prevent present themselves in, you know, in a good light for the organization while they're doing what they're meant to do.
Brandon Ellis 17:07
Well, the the analogy, and so again, that was from the "Industrial Automation - It Doesn't Have To... Be Myopic" session. And so I want to encourage anyone listening that if you didn't, haven't picked up on that one to check it out. But the refrigerator metaphor was essentially - and Keary, you've got it broken down in your paper, and so I'm just going to stick with that break down. But the, we were talking about a refrigerator and kind of dividing that up between, if you're just monitoring off and on, which is what Keary's making reference to is - is the machine running or not? So and Keary, you pointed out that's availability, right? That's the availability of equipment, and so is the refrigerator running. And then the point I made was, that doesn't mean that there's food inside the refrigerator, it just means that it's not stopped. And whether or not there's food inside would be, as Keary has, has pointed out performance. And then as I said, it doesn't necessarily mean the food is good, which would be quality. And then finally, it just means that the refrigerator is running. So we could probably fairly assess that whatever is inside is cold, whether it's food or air. So it may be perfectly cold, cold, but the food may be bad. And so you can't just derive that from the fact that it's just running or not running. You have too many assumptions for it to be quality what I call quality data. And you've heard me say, Keary many, many times, and Beth, I'm sure you as well, if you've got 10,000 data points in one data point is suspect. How much of the data is suspect?
Beth Elliott 18:37
All of it. Isn't it?
Brandon Ellis 18:38
All of it. And there are some processes out there granted that you don't have to be that much
Beth Elliott 18:46
Brandon Ellis 18:46
Yeah. But, and again, that comes down to what we were discussing in that podcast. If it really is subjective, it comes down to this specific customer and what they're making and how they're going about it and that kind of stuff. But Keary, you point out that you just pointed out that from a maintenance standpoint, you can't just necessarily look at one thing. You have to look at the other data and so and what those indicators are, and so I think that kind of leads us into the second section of your paper.
Beth Elliott 19:18
How do maintenance KPIs, how do they contribute to operational objectives, Keary?
Keary Donovan 19:25
Right, so not only does one piece of data may, it may corrupt the rest of your data, but the assumption is that we're acting on that data.
Brandon Ellis 19:37
Beth Elliott 19:37
Oh, acting on the corrupted data or all?
Keary Donovan 19:40
Any of the data.
Beth Elliott 19:41
Keary Donovan 19:43
So and in terms of in terms of OEE, which started your podcast, right, is that? Well, I'll wait for that. So the point is, is that that data is going to get acted on. And in the first section in that academic paper, it lines out what you have to be careful about acting on and acting on the data. And whether it's if, if you miss aim at just one piece of data, or you use multiple pieces of data without weight, you will, you will have contradictory goals between departments. And even if you have all the multiple measurements, if you don't have an overriding performance goal, and the means to support that overriding performance goal, you will still even end up with - kind of sabotaging the performance for their own gain. And so what a KPP is, are the parameters starting from the maintenance level for that department, which are the guidelines for the KPIs that are indicators, that means we'll use to measure their performance that will feed up to the overall organizational goals and their KPIs.
Brandon Ellis 21:08
So the parameters Okay, so the KPIs the key performance indicators are the individual data streams we've got coming in, and we're gonna assume we have more than just off and on. Is that fair? In your analogy.
Keary Donovan 21:20
Brandon Ellis 21:20
The key performance parameters are taken into account, of course, when you if you do it, I think, I think if I'm, if, if you and I are on the same page, and from my line of thinking, the key performance parameters should be considered initially when determining what KPIs need to be - need to exist - what data do we need to measure - what indicators do we need to stream in, but then once those parameters once those KPIs are in place, then the parameters decide how much you would rely upon, you know, how much how much weight in you said, weight, assume you mean weight? How much
Keary Donovan 22:03
Brandon Ellis 22:03
And not delay, but how much weight or how much of a priority you place, and, and how so on each of those KPIs. So I had an analogy about flying an airplane, and how in airplanes that aren't super digital nowadays, the six pack of instruments. And each instrument gives you viable data. But you have to look at that instrument for what it is. If you try to look at it for what it's not, then you crash to Earth. The other thing is, if you only have one, sometimes if you only have two or three, you're gonna die, you're gonna you're gonna wreck, you know, if you're, if you're flying in the clouds, you can't see up from down or left to right or whatever. You're flying blind, you know. Probably three can be okay, you'll survive. I don't think you'll land but unless you come out of clouds, but nevertheless, you. The point being you need multiples of those KPIs. But then you have to look at each one for what it's meant for, and emphasize that single thing. And and don't try to infer what another instrument is telling you from that instrument. Keary, do you agree with that?
Keary Donovan 23:16
Yes. Yeah, the reference point is essential for the indicators to mean anything. And
Brandon Ellis 23:24
And that's the KPP.
Keary Donovan 23:26
Correct. Yeah. And we go into the reasons why. We have examples from old, old studies of that were done of the Soviet productions and why they went wrong. You know, whether it was from anything from ratcheting down performance. So they didn't rewrite your performance goals. To like you're saying, if you just look at one, the plane might go down. So in terms of production, that would mean, I want to be a hero, plant manager, and I'm going to meet my production quantity. But I've taken all the work orders, and that are nice and short and sweet and chewed through and ignored the profitable ones. And so the overall performance of the company is subpar, or maybe even fatal to the company's success. And what what's interesting is, so what we provide and where I, where I got the idea from about the parameters and the guidelines that maintenance departments can share about what they're measuring, is from this Maintenance Audit -Maintenance Audit Handbook. And what I really liked about it was, it's a really extensive look at a very simple idea, which is, ask your workers their opinions and keep track of the surveys to then base your numerical indicators and your weights on your numbers for your composite analysis. It's just simple, you know, keep track of the surveys. And then this book kind of lays out how to go about it, which ones to measure and all those things by with the end goal of here is a complete reference point from not just from the math, or for business intelligence, but from the workers themselves. And what they're telling you is important tickets to consider when you're measuring any data.
Brandon Ellis 25:22
Mm hmm. And so let me ask you this, in that you mentioned, you kind of summarized the Maintenance Audits Handbook, who wrote that? Kumar
Keary Donovan 25:34
It's a group. Yeah, it was a group of engineers that took a lot of course materials and research and studies, then put them all together, basically, for maintenance departments.
Brandon Ellis 25:45
So you mentioned that the book provides, and I liked what you said this in your, in your, in your paper, both indicators and reference measures, identifying the user or the owner of each and, and setting the basis for external or internal audits. And so you know, when you hear audit, you think IRS sounds scary. But audits in this case, can be quality audits, they can be performance audits, they can be maintenance audits, you know, just uptime, downtime, kind of audits, those kind of things are really - they provide the checks and balances that are needed to measure success. And so we never stop measuring, you can't stop measuring. And so I really, like - and Keary, your point about the KPP, the key performance parameters, because you have to decide, and this is this, this is what I was referring to when I said everything subjective. Every company has to first setback if they're going to do this, right. And if you're making decisions off of this, you need to do it right. And as I've said in the past, Beth, there's no point in doing an IoT system if you're not going to make decisions based upon it.
Beth Elliott 26:47
Good business decisions.
Brandon Ellis 26:48
Yeah, well educated anyway. I guess still, you can still look at the data and be in denial. But But nevertheless, so Keary, as we're kind of talking about these KPPs, as far as real time predictive maintenance, how does how do these KPPs and what KPIs come into play? If you were, if you were talking to folks about how they're going to measure this stuff, how they're going to get to KPIs, that kind of thing? What what kind of thing should they be looking for?
Keary Donovan 27:19
Right. So it starts with a simple concept of a leading indicator versus a lagging indicator. So a leading indicator would be identifying the work that needs to be done, or planning and scheduling the work. And then how long does it take to execute that work?
Brandon Ellis 27:38
Now we're talking about, I'm sorry, to interject, but we're talking about. You said the work you're talking about the maintenance, whatever the maintenance is
Keary Donovan 27:46
Yeah, I'm coming from a maintenance perspective.
Brandon Ellis 27:49
Okay, I'm with you.
Keary Donovan 27:51
Alright. And then you have the lagging indicators, which are, you know, include the equipment effectiveness, the cost of what you did, and how effective that is, and how you're doing in terms of the overall safety and environment. So you have things like, meantime between failure and OEE and how are we doing while we're doing the work we identified and planned.
Brandon Ellis 28:17
So did the this was the work we needed to do? And this what we thought it would take? And then after the fact the lagging is, Did it work? Basically.
Keary Donovan 28:24
Brandon Ellis 28:26
Beth Elliott 28:27
Do you want to go into the third part?
Brandon Ellis 28:29
Yep. Let's take it in.
Beth Elliott 28:30
So we're going to talk about the symbiotic maintenance and production data for OEE. Do you want to go into that, Keary?
Keary Donovan 28:37
Yes, yeah. So with all of that said, ultimately, we're talking about maintenance operating in the overall production picture. Of course, OEE started this conversation back in your podcast. And so I use that as a reference point. And OEE is defined as availability by performance by quality. And in terms of the organization, you could think of that it commonly would be thought of as maintenance responsibility for availability by production responsibility for performance by production responsibility for quality. Of course, whenever you have interrelated responsibilities where you're being measured, or your performance is being measured, there's a tendency to point fingers.
Beth Elliott 29:29
Ah, who's responsible for what. Yeah.
Keary Donovan 29:33
And there's a tendency, you know, and then of course, there's also a tendency or their efforts to conflict with each other. And so what I did in the paper was expand the idea of the, the theoretical refrigerator machine and ask the simple questions like, between maintenance and production, who's actually in charge of making sure the temperature is correctly measured? So you just take the simple like, Is it cold? Right, you just take a basic statement that Brandon said, and then who determines what cold is? And who's responsible for what it's supposed to be? The second, the second question I asked is, who determines something simple like the light going on and off in the machine? When the door opens, does it affect safety? Is it a critical maintenance issue? Of course, maintenance wants everything to work, because that's how they get measured. And production wants the things that are critical to work so production gets done. And then you take something like the door being open and letting the cold air out, and consuming energy, who's responsible for making sure the door closes? And that and these kinds of things could go on and on and on. And so what I describe in terms of laying the groundwork for people to understand how to overcome this is an analogy that I read in a Reliable Plant article. It was they shared the idea that maintenance, the relationship between maintenance and production is like the military and their civilian oversight, where the military knows perfectly well how to do their job, but they're completely restrained by the rules of engagement of their civilian bosses. And that's, that's kind of what maintenance - that's the maintenance world.
Beth Elliott 31:23
So maintenance is like the military. No, I'm just asking.
Keary Donovan 31:28
Well, you know, they get pigeon holed into only good for one thing, which isn't necessarily true. But they also have performance measures that they're not necessarily in control of which can be defeating. Right? So what we've headed down the road of developing in terms of empowering maintenance to do their job, but then also to contribute to the overall production picture, is our predictive maintenance kit.
Beth Elliott 32:00
Oh, can you tell us a little bit more about that predictive maintenance? Well, can you describe what predictive maintenance is first? And then that way we can lay the groundwork?
Keary Donovan 32:10
Sure. So the idea behind predictive maintenance is that not only are you planning or servicing the equipment to kind of prevent failures, you are taking readings and leading indicators to get warning signals before anything fails.
Brandon Ellis 32:31
You're tying to the check engine light.
Keary Donovan 32:33
Right. And the simple analogy for to, to share it with any with any crowd with any background is everybody has it in their car. You get the check engine light before there's a catastrophic failure.
Brandon Ellis 32:47
Keary Donovan 32:53
And, you know, just like your car, that check engine light could be a faulty sensor may not even be something. So that you know, or it could be you saving your machine and or your car. And so it's not predictive maintenance can be overstated, I guess is my point. Predictive Maintenance gives you the warning lights, it doesn't necessarily tell you what's wrong. Now, our Predictive Maintenance kit is measuring vibration of moving parts, the temperature of moving parts, and the current of the moving parts. And, and the reason for the all those three is that when any of those go out of line with their thresholds, usually relates to the mechanism or the bearing of that moving part. Not operating efficiently. So it's a warning.
Beth Elliott 33:51
So the vibration, temperature and current are the KPIs? Am I understanding that correctly?
Brandon Ellis 33:58
Yes, they are. They would be, because if you just look at temperature, it doesn't mean the bearings getting going bad, it could mean that there's a heater on the machine that's gone awry that's sitting next to, you know, the point of measurement. If you just look at vibration, it could mean that the machine, you know, the machine, this this gearbox or bearing or something like that is is ready to, you know, fall apart, or they set a new 20 ton press 10 feet away, that came in yesterday and just struck its first lick, and it's shaking all the way to the front offices. And then current, of course, the current would have to do with motors and how much they're trying to, that's how much effort the motor is putting into. So each of those is a KPI. Each of those is an instrument so to speak, and if you just look at vibration, or you just look at current. I mean the current, you could see a current spike in a motor, but it might be that a cable has worn and frayed. And it's it's shorting up against the machine frame, and really have nothing to do to say that the motor is about to go bad, or there's something going on in the gearbox.
Keary Donovan 35:14
See that's why I would have said I, the reason I hesitated is because I would have called that a measurement that feeds KPIs.
Beth Elliott 35:24
Brandon Ellis 35:25
And okay, and that's fair. That's a fair point. But that's a that's a big point. So I want to elaborate on that a little bit. So KPIs are traditionally from a marketing standpoint, and I'm making that emphasis because I think and Keary, you got a marketing background as does Beth so so don't get upset, but I think that
Beth Elliott 35:42
Do we ever, Keary
Brandon Ellis 35:43
Key performance indicators - when you do all the math, as you said, you defined OEE (Overall Equipment Effectiveness) as being a an equation. And I'm an engineer and Keary is not and so he says "by" I say "times", so the availability times the performance times quality, that's an equation, equals OEE. And so once
Keary Donovan 36:09
I remember but I think in my math class,
Brandon Ellis 36:13
By, you're right.
Keary Donovan 36:14
Referred as by
Brandon Ellis 36:15
Not not Tennessee Talk. It's Tennessee Talk.
Keary Donovan 36:22
It's like you can use the X or the Asterix when you're trying to multiply.
Brandon Ellis 36:26
If you're Jethro from the Beverly Hillbillies, it's a gozintas. But anyway, availability gozintas performance gozintas quality and that gives you. Each one of those is a measure is a KPI is a key performance indicator. But then I'm dividing it down even further, in order to get availability to get performance to get to quality, it comes into these the sensor measurements, and so it's a data string. And Keary, you've heard me say this a million times, it doesn't matter what it is I view everything is a data source. It doesn't matter if it's a weigh scale, or a current current transducer or a PLC with, you know, with with information inside one of the registers, it's just a data source. And so I refer to each of those as an indicator on a very, granted it's a very micro level, I guess it can be
Beth Elliott 37:19
Granular, yeah, okay.
Brandon Ellis 37:20
But to me, a vibration is a is a is something that's measurable, it's something that I can have on a dashboard, a visualized dashboard, or in a report to where I can take a vibration measurement on average, probably with vibration, we would do an average, average temperatures, that kind of stuff, but also I can create with temperature, I can create, and vibration, I can create a trendline. So I can, a trendline would be trending meaning a more historical view. So an average temperature can mean that whatever the data source is, is averaging, and then giving you the answer, or it's just giving you all the raw data, and you're taking that and creating an average. And so whether or not the device is giving you the temperature sensor, or the vibration sensor, what not, is giving you an already averaged value. And then you may average that further. You're, you're beginning to become historical. And so to do an average, you have to have history. You have to take so many and add them together and divide by the number of samples you have. That's how you do an add, compute an average. And so you have to have history. Just taking a snapshot right now real time look at what the vibration is, well, it's gonna be all over the place, because vibration is vibration. And so that signal is usually pretty erratic. Temperature is usually if you look at it right now, and you look at it a half second later. Heat transfer takes, I mean, physics, heat transfer is going to take place as fast as it takes place. And you're probably not going to see a unless you're looking at a very, very miniscule, which means you have a very high resolution thermocouple or something like that, to measure the temperature, you're probably not going to see a huge temperature change. Because temperature change, unless maybe it's a nuclear explosion or something, temperature change doesn't often take place very aggressively. It's a it's a it's a slower deal. Preheating your oven. You can't make your oven preheat any faster, it's going to take the time it's going to take it may take longer if the ambience or whatever, but the fastest that it's going to do is the fastest it's going to do. And so, but those are data raw data streams, so I call those indicators. They may not be key performance indicators, but they're indicators in my mind. Keary, tell me I'm wrong.
Keary Donovan 39:50
So Beth. Yeah, Brandon just described why we created our kit.
Beth Elliott 39:54
Keary Donovan 39:55
Cuz if I'm in charge of the maintenance machine of the machine. I just rolled - my eyes just rolled back into my head with everything he just said. Even though I know that everything he said is really important. All I heard was words, words, words, numbers, numbers, numbers, words.
Brandon Ellis 40:18
Thank you. Thank you that that was my, that was my goal. That's exactly what I set out to do.
Keary Donovan 40:27
So, I now what I mean by that is not that maintenance people aren't capable of being technical enough to understand what he says and is instructing or even that it was part of their education, training or anything like that. It's just they're not being measured by that. They're met, they're being measured by how many work orders had to go out? How much time did it take to repair it? And if I can reduce those things, I don't need the kind of calculations that are being used for return on assets of the machine. So I'm generalizing. And I'm overstating it a bit. But all of the trend lines, and all of the baselines that Brandon are talking about is really important for me to know, because I'm gonna, I'm going to replace their part before it fails. And because I have an indicator, I can create my information system so that they don't have to carry as much spare stock for that machine. Because I know when I know before it's going to fail, I need to send out that motor from the manufacturer of the components stock to them. So I'm leapfrogging supply chain issues. And I'm leapfrogging all kinds of things by using the information system attached to the indicators that Brandon's talking about. But maintenance wants to reduce their number of work orders and the time it takes to complete those so that it feeds productions thing, performance indicator saying maintenance isn't interfering with production. They're not going to show the baseline of the vibration to prove that. They're going to show the lagging and leading indicators of what I talked about in our last section.
Brandon Ellis 42:11
So what I heard Keary say is Brandon, you were absolutely right in what you were saying.
Keary Donovan 42:16
But that's how production. That's how production walks out of the room.
Brandon Ellis 42:21
Keary Donovan 42:22
They're agreeing with everything I just said.
Brandon Ellis 42:23
The thing, the thing that Keary's kit is doing, and that's the reason I'm excited to hear about it and to see what he's doing. Is he essentially, in our IIoTA product is the we could call it the iPhone of MES. And so getting the data point click across making it easy. We're taking care have a lot of stuff in the background, so that you don't quite have to understand how it's done. You don't really want to get bogged down in those details. And Keary's kit does just that. So with what I'm talking about is that's the how it's done. Now Keary's kit does that for you.
Beth Elliott 43:00
Brandon Ellis 43:01
And so he's managing all of that stuff with within the kit. Because at the end of the day, and Keary, you've heard me say this before in many of the panels that sometimes you and I have both have been on or, or been a part of. At the end of the day, it really comes down to a specific goal. And the goal is get product out the door, as as quickly as possible at the highest level of quality, safely, and for the lowest amount of cost. And, and that includes when you're going through Keary you were going who you were talking about who's in charge of making sure the temperature in the refrigerator metaphor is measured correctly, who determines if the lights going off and who's responsible if somebody leaves the door open, and how maintenance is responsible for availability, production should be responsible for performance, and and also production should be responsible for the quality aspects. And I'm sure that most any any of our production engineers that are listening to the podcast are agreeing with the performance and the quality. Or at least they're going to say, well, it's our job to know whether it's right or not. But then they're saying quality, they'll get quality involved or, or something, but they're probably saying but we need that, you know, we still rely on maintenance for all that. And mainly the maintenance guys, the maintenance guys that are listening are saying no kidding, they come to us for everything, for all of that. Somehow that's all of our and so that is a bit of that philosophical conflict. But Keary's kit helps minimize that to a point because he's helping, it's helping, it's doing all of that behind the scenes stuff with the current, the vibration, the temperature, those kind of things, the trending and all the stuff that I went on about. And it's it's packaging that up and ultimately, it's just giving the answer or the indication to say this is something that needs to be done. And then he's able to actually springboard it a step further and say say, you know, here's the replacement part, you know
Beth Elliott 45:02
Brandon Ellis 45:03
That kind of thing. Is that right, Keary?
Keary Donovan 45:05
Right. And so I just want to add, you don't have to just take my word for it. Because we're using an IIoTA that will break down any data used down to the most granular level you can think of.
Brandon Ellis 45:19
Keary Donovan 45:20
So if a work order, so what we do is we have these sensors, they take the baselines and the thresholds over time, they will send an alarm, whether there's a catastrophic, or there's a long term degradation. And then the IIoTA prepares the data and transports it to what's called a CMMS for maintenance. This is maintenance main software application. It's called Computerized Maintenance Management Systems. And it's where they do all their work orders, and they track all this stuff. And then it talks to the rest of the enterprise.
Brandon Ellis 45:55
That's the that's the maintenance software that I was referring to earlier. That needs this data. And doesn't necessarily when you say it interfaces with the ERP system, but sometimes it doesn't even do have, it's according to the ERP system, right?
Keary Donovan 46:09
Sure. Yeah. It's, it's based on the overall information system that the company has set up. The, what I'm saying is it can, and we make it easier to do that, because the data is in database format, because of the IIoTA.
Brandon Ellis 46:24
Amen. All right, I like that. Like, like, like,
Keary Donovan 46:31
And therefore and therefore. Yeah. And so what you know, you you can see how, when you have these gray areas, like availability, or I'm sure you would have maintenance, say, how would you get a quality product out without me. And so we spend our time educating people about how these gray areas can be solved by having the right data for anyone who needs to access that. So if you're questioning, why is maintenance doing this, or that, you can break it down to a very granular level. If you want a report about OEE, you know, and how you're contributing to the production reform performance, well, then it's in your software. And it's very empowering for not just maintenance departments, and not just production departments to have all this data. I was for lack of a better phrase swimming together. But it's important, the importance of having the data together, is that you can, it I guess, it's just the insight of it. To put it shortly. I'm about to just go on too long.
Brandon Ellis 47:45
So we're kind of we're kind of we kind of need to wrap up. And so I want you to first of all, I want to thank you for your insights of your hard work on this. And certainly for your your dedication using the IIoTA. I'm glad it's working well for you. Of course, you and I've been working together on this as you've developed this kit. And so I'm proud of you, I think you've done a great job. But how do our listeners if they want to learn more about it? How tell them tell them the best way to do that?
Keary Donovan 48:12
Well, the simple way, the best way is to reach us through Pathways 7. That's Pathway with an S and the number seven.com [www.pathways7.com]. You'll be able to reach us there.
Brandon Ellis 48:22
Beth Elliott 48:23
Thank you for your time, Keary. And I hope all as well out in Boise with you and your family.
Keary Donovan 48:29
Thanks for including me.
Brandon Ellis 48:31
Thanks, Keary. Yeah, I really do appreciate Keary being on with us today. Beth,
Beth Elliott 48:36
Brandon Ellis 48:36
Episode number seven.
Beth Elliott 48:37
Brandon Ellis 48:38
Is in the books. Yeah?
Beth Elliott 48:40
Brandon Ellis 48:40
All right. Perfect. Listen, guys, as we get ready for Christmas, stay positive. And wish everybody you meet a Merry Christmas, and wonderful New Year because when the New Year comes, it won't be 2020 anymore. And we'll just keep looking forward and staying positive. So thanks for listening. As far as the podcast, "Industrial Automation - It Doesn't Have To..." walk us through how people can get the word out.
Beth Elliott 49:06
Rate, review and subscribe
Brandon Ellis 49:08
Beth Elliott 49:09
Ratin' and reviewin'
Brandon Ellis 49:10
Yeah, that's right. And that's some good Tennessee Talk for ya. Guys, thank you very much. Keary, thank you very much for being on with us. Beth, thank you as always. And we'll see you on the next one. Talk to us in the comments, www elliTek.com is how you can get in touch with us or give us a call at 865-409-1555. Beth, have a great day.
Beth Elliott 49:32
Same to ya.
Brandon Ellis 49:33
Transcribed by https://otter.ai