Episode 12: Ask the Experts!
Rob Kalwarowsky brings years of unique experience with a background in Mechanical Engineering from MIT, as well as experience in economics and maintenance. After working as an Economist, Rob transitioned to working in Reliability Engineering. Today, Rob is an Asset Management Specialist and is breaking new ground on the grassroots maintenance movement. Rob is the creator of Rob’s Reliability Project, a maintenance content website that produces video, audio, and visual information to spread the word of reliability.
Joe Anderson is the COO at ReliabilityX. Joe has more than 20 years of experience in maintenance and management excellence in various different industries and plants throughout the US. He’s published articles, lectured throughout the US, and been recognized as one of the top 50 leaders in the country including the US Congress.
Listen to this week’s episode online here or on Apple podcasts here! You can also watch this episode on YouTube here.
Join the Masterminds in Maintenance Podcast!
Are you an industry leader in the fields of maintenance and reliability? We want to hear from you! If you would like to be featured as a guest on our podcast, please sign up here.
Stay tuned for more inspiring guests to come in future episodes!
0:00:01 Ryan: Oh my gosh, I know. Alright, so should I go ahead and kick us off and basically just do a quick round of introductions? How does that sound for everyone?
0:00:11 George Williams: Sounds great.
0:00:11 Joe Anderson: Good.
0:00:13 Ryan: Alright, well, it seems like we’ve got a few guests here. Welcome to our very first Ask The Experts, live Q&A panel, thank you all for joining. Amongst our guests, we’ve got three decades worth of maintenance and reliability expertise, represented across, I think two, three, two different countries, three different time zones. Sound about right? [chuckle]
0:00:37 GW: Well right now we’re in the same country except for you.
0:00:41 Ryan: Oh fine, that’s fair. [chuckle.
0:00:45 GW: We’re still in your reliability channel right?
0:00:48 Ryan: Oh man. Well, let me quickly introduce myself, and then we’ll do a quick round of introductions for you guys and then we’ll jump into all the questions that we’ve received over the last couple of weeks, so a little bit about myself. My name is Ryan, I’m a CEO and founder of a company called Upkeep. We build software for maintenance and reliability teams. I got my first start into this space coming out of Cal Berkeley. My first job was working in a manufacturing plant and my role there was thinking about how do we improve our process and make our line speed go from 20 feet per minute to 21 feet per minute? But then as our entire business realized the dire need for better process, we shifted the entire thinking for our business to shift it from just thinking about improving line speed to improving reliability because if we did that, that would have the same outcome and effect. Saw the software that was out there, said, “Hey we’ve gotta build something much, much better. And I started a company called the Upkeep, and that’s a little bit on me and my company. Let’s kick it off to George.
0:01:53 GW: Thanks Ryan, my name’s George Williams I’m a founder and CEO of a company called Reliability X. And we focus on implementing holistic reliability, so working with the plant floor, all way up the senior leadership to ensure everyone understands their impact to overall reliability goals and truly implementing sustainable change.
0:02:16 JA: And I’m Joe Anderson, COO of Reliability X. And what he said.
0:02:23 Ryan: You make it pretty easy. Alright. And Rob?
0:02:28 Rob Kalwarowsky: Yeah, so I’m Rob Kalwarowsky. So I started at Reliability, would have been about 9 or 10 years ago now, and worked in mining for a while, and now I’ve moved over into oil and gas so moving up in the world, I guess.
0:02:45 Ryan: You did a lot of different things. Well, great, we have a bunch of different guests from so many different industries, so many decades of experience, I’m super excited to help lead the session today, just to give everyone a quick heads up over the last few weeks, we’ve been fielding a bunch of questions from the community, we’ve compiled the top 8-9-10 questions that we’ve received and the goal for today is just to have a pretty informal conversation about some of the most impactful questions that our community gave to us. So if you guys are ready, give me a thumbs up and we’ll jump right into the very first question.
0:03:28 RK: We’re always ready. [chuckle]
0:03:31 Ryan: Alright. Alright, so the very first question is, “what is today’s real impact, advantages and disadvantages of machine learning in predictive maintenance?”
0:03:43 GW: Pass.
0:03:45 Ryan: Who wants to take… “pass” [laughter]
0:03:49 RK: I can take this one. So in terms of machine learning, there’s a lot there. So I think that the benefits obviously the benefits are, you can sort of hand off a lot of the analysis to the models, so you don’t have to spend your person’s time. Like I developed a machine-learning algorithm at my previous job to look at oil samples and so we had people on my team that their full-time job was to look at samples and alert customers on whether there was issues or not. We were able to show just on a pilot from someone who learned machine learning on the internet was about 50%-time savings. So that’s the good side. Now, if we’re gonna move over to the disadvantages side, I think there’s a lot there. So a lot of companies right now, they’re able to sell these types of products because they sound cool and whatever, but a lot of companies also buy those products and don’t see success. So a few of those things lack of process not understanding that when you’re installing sensors, when you first turn those like that switch on, there’s a lot of maintenance that you’re gonna have to do on just the sensors no matter the equipment that’s paired with the sensor. And in that process, so when this thing… When you turn it on, there’s gonna be… If you’re in a plant that has kind of lacked maintenance for a while, you’re gonna have a lot of alarms to deal with and you gotta get over that overwhelming hump at the beginning.
0:05:30 Ryan: So I wanna take it one step back as well, because we’ve talked… I hear this term being thrown out a bunch in the industry of machine learning, predictive maintenance, AI, how much has it infiltrated actual industry? Are you guys seeing 1% of facilities that you guys consult with and work with implementing some sort of predictive maintenance machine learning algorithms? Or is it like 10% or 50%? 100%?
0:06:01 RK: Joe’s shaking his head. [chuckle]
0:06:07 JA: Not to say that it’s not out there, because I know it is, but I haven’t seen it anywhere at all.
0:06:16 RK: I’ve seen it on the big companies. So we’re using it here at my current role. There was… My previous company in mining, they’re using it. Some of these companies have also hired data scientists to do their own in-house platforms. We’ll get in… We could talk about that, but I haven’t seen anything really successful, though.
0:06:42 Ryan: Right, so this question was…
0:06:44 JA: Nothing really specific as well. For example, your oil and gas, they always spend a lot of money when prices are good, they’ll invest a lot of money in technologies where food manufacturing, they won’t spend any money. Hardly any, doing any… Especially on the AI side, we can barely get them to acquire predictive maintenance tools, let alone machine learning. So I think a lot of it is industry specific, as well. Plus, if you’re a top 1% company, they’re probably doing it anyway because they can afford to.
0:07:22 Ryan: Right. So, I think to sum it up, what I’m hearing is the question actually centered around what is today’s real impact and it sounds like the impact that it’s having today is actually very small and it’s only the top 1%, and it’s also industry specific, of companies that can actually afford to use this, have the personnel and staff to be able to analyze, let’s call it machine learning, predictive maintenance, that centers around sensors and all of these fancy terms. It’s actually very interesting because I think, again, a lot of people get thrown into this industry hearing all of these fancy terms but then when we take a step back, and we realize how much… Are people actually doing this? It’s only really… But I think you mentioned Joe, the top 1%.
0:08:12 JA: Well, it’s when I started getting into this more and becoming more familiar with everything, it was discouraging for me. You hear all these people talking about all these cool things and they’re doing all this stuff when I’m not doing any of it and I’m going, “Man I’m so overwhelmed trying to just understand how to develop a maintenance strategy,” that all these people are doing these cool things. And the more you get involved in the industry, the more you realize that sexy things are cool to talk about. And it’s a perception game. In the end, all that matters is what are the results. No matter what you’re doing, what initiative you’re doing. Am I gaining extra throughput, am I driving waste out of my system, are my skillsets advancing? Am I doing things that we see improvement on. And really that’s where the rubber meets the road, I think.
0:09:10 Ryan: Awesome, So we’ve got a few guests in here as well. I just wanted to let you guys know, this is a live session, so if you guys have any questions, throw them in the chat box at the bottom right-hand corner. I’m looking at them, I’ll review them, and we’ll help answer those questions, too. So the next question that we have up on this list: Any advice on dealing with union crafts to create and generate buy-in from the crafts on maintenance reliability, and DE activities? I’m gonna hold my hand up and ask you guys, what does DE activities stand for?
0:09:47 JA: Defect elimination.
0:09:49 Ryan: Defect… Okay, got it, got it, got it. There we go. Alright.
0:09:53 JA: Who knows? I didn’t write the question, so I don’t know. I don’t think… I could be way off. Well, coming from a union background honestly, it never made a difference to me whether it was union or non-union, it was about leadership and how you deal with people. So I would do it the same way I would do it with a non-union shop, if I were to lead a defect elimination initiative, and that’s one bringing awareness: Training, fully providing support and me removing obstacles, whatever the obstacles are that they face, it’s my job to remove those as the leader to make sure that they’re successful. And then creating a minor budget when it comes to defect elimination, giving them a little bit of my budget and some leeway to go out and to solution some of these problems. I don’t think it really… At least, in my opinion, it never made a difference whether it was union or non union.
0:10:58 GW: I would agree. I think that ultimately, you’ve gotta earn the respect of the people at the shop floor and the way you do that is by listening and removing problems. And so, if you intently listen, listen to understand, not listen to solution, but understand exactly what issues they’re facing on a day-to-day basis, and you can make some of those things go away to make their day-to-day work easier, you’ll earn some respect. But like Joe said, I don’t think that’s different, whether it’s union or non union.
0:11:29 JA: Yeah. The only thing you have is some things in the contract that you have to watch out for, like some strong unions were… If I pick up a wrench as a manager, I’m stealing work from the… Just some of those little things, those nuances that you gotta watch out for. But honestly, there’s not a whole lot different. We’re still dealing with people. And people need direction, they wanna be led and they wanna know that you care. And it’s as simple as that.
0:12:00 GW: Yeah, absolutely.
0:12:02 RK: You make it sound so easy, Joe.
0:12:04 JA: It’s not that it’s easy, it’s just it’s the same. If I meet you in person, I wanna meet the same person that you see as we’re interacting here. I’m not gonna change who I am as a leader because we have union here or non union here. It doesn’t make a difference. I think the major reason the unions exist today is because of the lack of leadership. If we took care of our people, there would be no need for a union. And so I think that leadership key is the main significance of the whole issue.
0:12:40 Ryan: So Rob, any experience? I have no personal experience dealing with unions, but I think where the questions…
0:12:49 JA: You live in California?
0:12:50 Ryan: Sorry?
0:12:51 JA: You live in California?
0:12:52 Ryan: Yeah. Yeah.
0:12:53 JA: It’s a union state, isn’t it?
0:12:58 Ryan: I think where the question’s coming from, it sounds like someone might be struggling with trying to generate buy-in, and it sounds like they might be part of a union company, any advice for them as they might be struggling to create this buy-in?
0:13:20 JA: Number one, know the contract. Right? You’ve got to know the contract ’cause you have to understand the operating context, you have to operate within it. So my advice would be to understand that contract understand how you have to operate to engage with that workforce and then second, do everything you can to earn their respect.
0:13:40 Ryan: Yeah. Yeah, absolutely. Alright, definitely I learned a bunch about about unions just now. [chuckle] So the third question that we have here on this list is: Can you talk about the importance of BOMs? B-O-M bill of materials? Rob, you wanna take that one? [chuckle]
0:14:04 RK: I mean, yeah for me, I don’t really have too much experience with it. However, other than the obvious things like bill of materials helps you procure parts understand, what you’re actually running, like you can hook up your spare parts management system to it, right? And so it’s like actually one of the issues we’re having with right now, with one of the extremely old equipment we’re trying to deal with, is that we don’t have drawings, we don’t have this bill of materials, we don’t really know what other than what’s there we don’t really know how to replace it with something that’s new because it’s obsolete. And so, if you have that information in your system, it makes your job a lot easier. Maybe not today, but in 50 years.
0:14:58 Ryan: Yeah, gives way more visibility into if something were to go wrong, what do we need? What are the parts? Give us that, you know, the entire bill of materials. George, Joe, any insight there?
0:15:14 GW: So the bill of materials serves a couple of purposes, one; it allows you to procure easier, so you can make selections directly inside most TMMS systems and say, “Create my purchase requisition.” The largest advantage though, is in the efficiency of the planner itself. So the planner today may be having to go walk things down and go find the spare parts drawings or maybe even go investigate or write a work order for somebody else to open something up before they can figure out what parts they have to order to truly fix something. Part of a planner’s job at that point is to then put those parts on that bill of material list, So their job becomes efficient the next time they have to go do a similar job for that same asset and so the bill of material list really serves a much higher purpose of making the planner more efficient in the future.
0:16:10 Ryan: Yeah, absolutely, absolutely. It’s making all parts and components of the entire business more efficient, more effective, gives you better visibility into your equipment, parts, spare parts and also helps with planning and scheduling. So absolutely. So we got a question actually, from Prem let’s see if we could help answer this one: How do you see career progression, all the wolves being in process of being replaced, by fresh blood? Oh no. [chuckle] so I think this question is career progression. We’re starting to see a shift in the industry, we’re seeing, I guess the old wolves, being… Maybe not replaced, but we’re starting to see more younger, more of the younger generation come into this industry.
0:17:03 JA: Yeah. Yeah, boomers are retiring right? And so they have to be replaced with somebody. So what’s the question then, how do you fill the gap or?
0:17:20 Ryan: I think the question from Prem is really, how do you see career progression for let’s say the boomers who maybe aren’t about to retire just yet, but still want to stay in this role and not be replaced?
0:17:38 JA: I’d shift them to a more of an advisory coach and mentor for everybody that’s younger. Right? I mean they have all the tribal knowledge. We don’t do a very good job capturing the knowledge that these people are leaving behind when they leave the organization, so I would use them to help the younger generation, develop standard work procedures develop troubleshooting guides, try to get everything out of their heads that you possibly can to help the transition, so that the learning curve is smaller and not so steep.
0:18:14 Ryan: Yeah, I think you brought up a really good point there because I think as the younger generation comes into these businesses, I think that there’s a tendency to feel fear of like, “Oh shit, my job is gonna get replaced”, but I think what you’re saying is you and us, being in this business, have so much industry knowledge that you could really go into a very impactful role by advising and helping train the younger engineers.
0:18:44 GW: All that boils down to adding value right, Ryan? So, whether you’re gonna continue as the reliability engineer, or whatever that position is you’re in today, you’ve gotta be able to add value, so continue to learn and figure out how you’re gonna add value to the organization, which may be in coaching and mentoring the next… I’ve always looked at my job as… When I was a practitioner of, my job was to run myself out of work. So somebody else could do what I’m doing, so I can go some place else. And so, I don’t think that changes with your age, I think you’ve just gotta find where that niche is and make sure that you’re continually learning and looking for a way to add value to your organization. I don’t think that changes based on your age, your strategy for doing that may change, but you still have to add value.
0:19:35 Ryan: Absolutely. Alright, we got… Thank you for that question, Prem, that was a really good question. We have another question from Sudrati. How can we set minimum and maximum values for bill of materials and reorder points if we don’t have a consumption history? Alright, that sounds like a chicken or the egg problem right there. [chuckle]
0:19:58 JA: Well, there’s a formula too, but if you don’t have a history, you’re just gonna have to guess.
0:20:06 GW: You have to get the data points you can get. Somebody knows the history of the machine that’s on the site. And they can give you a relative idea as to how often components are failing. You may not have good data inside the COMS to be precise with that, but somebody knows it, plus you’re purchasing it, so maybe you don’t have the data in your COMS system but maybe it sits in your procurement system, if they’re not one and the same. And so, it’s a little bit more leg work, it’s a little bit more analysis that has to happen, but I think you can gather some data points that will put you in a relatively good position to create that initial reorder point [0:20:46] ____.
0:20:51 JA: It’s not like it’s gonna lead you to stocking 12 and 20 of something either. I mean, you might be off by one, you might stock three and five when you really need two and four or whatever for your min-max and it’s not gonna hurt you, you can improve upon that through time. So it’s not like you’re gonna be ordering and setting min-maxes for a ton of stuff and being way off base by doing that.
0:21:17 GW: All that being said, if you’ve 20,000 components in your store, this is not a small exercise. But you can get relative information and make some assumptions.
0:21:28 RK: I think, just for me, not specifically on this question, but it’s a broader thing too. So it’s not just min-maxes, this is everything. I hear all the time, “Hey, we don’t have failure histories, what do we do?” “Hey, we don’t have this history, what do we do?” It’s really the same answer. You gotta do some work. You gotta go find the old guy who’s maintained it for 20 years, buy him a box of donuts and talk to him. That’s really how it is. [chuckle]
0:21:56 Ryan: I completely agree with that. You’re not having the consumption history, we find that so many times that yes, we don’t have this beautiful, clean history that we would hope to have in order to make these big decisions, but I think what Joe and George, you guys mentioned was, you just gotta start somewhere. The impact is gonna be way worse if you don’t start than if you’re one or two off.
0:22:26 GW: Yeah, absolutely.
0:22:28 Ryan: Alright, great question. So we got one more question coming up here. So, what are best practices to prepare a plant changeover in order to align the generation of PM activities for rotating assets with redundancy in COMS? Alright, let’s see if we could boil that question down. [laughter] Who got that one?
0:22:52 JA: A plant changeover. Well, you gotta understand some context. So plant… Is this an automotive facility where they shut down and do a complete changeover? Is it…
0:23:03 GW: It depends on the… I think the question is about redundancy… So let’s say you have a pump skid, and you got a primary and a secondary pump. I think the question is about how do you manage the PM strategy and when do they switch? In other words, from lead-lag. If I understand the question but I could be way off.
0:23:25 Ryan: Let’s go with that. [laughter]
0:23:28 GW: But there’s a lot of debate about should you do a half and half scenario, should you run one-third, two-thirds? And I’m sure Rob probably has way more statistical analysis in this arena than I do, but because most values fall in a random pattern, I’m not a believer that it can’t be a one-to-one ratio, you’re not gonna see the exact same wear out pattern in pump 1 versus pump 2. And so, for people to take so much initiative in trying to figure out the right ratio between, “Should it be one-third, two-thirds between lead-lag?” I don’t think it makes much of a difference. I think if what’s generally happening in most plants is probably they run the lead until it fails and then they turn the lag on and it fails right away because it’s been sitting there so long. So if you’ve already got some kinda strategy, you’re way ahead of most people.
0:24:23 GW: And I would say that creating a PM that reminds the plant to change over lead-lag is a good thing to do. The other thing to consider is, well, if you would normally PM that pump every, let’s say, 1000 hours of run time or 5000 hours of run time, well, since it’s gonna be a lag half of that time do you then make… It’s still 5000 of run time but most plants are based on a time frequency, so instead of a quarterly, maybe it’s every six months, maybe it’s every four months, instead of every three months. So you’ve gotta be able to alter that as well. So I think, [A] just like with the min-maxes, have some strategy. But I wouldn’t put a whole lot of weight in the argument of, “They shouldn’t be equally run,” I think that’s a non-sense argument.
0:25:16 JA: Well, let’s ask the reliability [0:25:16] ____.
0:25:17 GW: Let’s ask the [0:25:17] ____. [laughter] Are you gonna tell me how wrong I am? [chuckle] Go ahead, Rob.
0:25:26 RK: Actually, I don’t really have a good answer for this one, to be honest.
0:25:30 GW: Alright next.
0:25:31 RK: It’s really… I kind of fall in the same camp. I think a lot of people, they kind of underestimate the damage that just equipment just sitting there does to the equipment. And I mean, I liken it to this one time, I left my car at the airport for I think it was four or six weeks, I forget how long it was, but when I came back, the battery was shot. I had to get a new battery, change the oil, the whole bit. And it’s like people don’t think about that in their plant. And then, also, given that, let’s say you have two pumps next to each other, like the one that’s not running, it’s vibrating because…
0:26:10 JA: Right, [0:26:10] ____.
0:26:12 RK: Right. So you’re causing failures there so you have to flip it on.
0:26:16 JA: Yeah.
0:26:19 Ryan: Yeah, that’s a tough question and I think it depends on industry, it depends on criticality, it depends… So many different factors. That’s a tough one. [chuckle] Alright, so next question we’ve got, What are best practices to prepare weekly look ahead plans? How do you update them on a daily basis with a new event? I think this is best practices for scheduling a week ahead or look ahead plans, and then also, best practices on updating them every single day, because we know we create plans, things always change. You guys have any tips, advice? [chuckle]
0:27:08 GW: Yeah, number one, don’t alter that plan based on today’s events. [chuckle] Your supervisors manage daily work, not the plan and not the schedule. After that schedule is set, it should be set in stone and you either achieve it or you don’t achieve it based on what those daily events and how they impact things, and you let supervision level folks alter their day-to-day operations. In terms of developing that weekly schedule, I’m a big believer in getting the voice of your customer and understanding what work they want out of the backlog. And this can turn into a really long conversation and a really long debate, but I’ll break it down as quickly as I can. Number one, it’s not your work, you don’t own the asset. If you did, you’d be able to turn it off whenever you want.
0:27:56 GW: Since it isn’t your work, you don’t own the backlog. Find out who owns the backlog and make them tell you what’s important. Once you know what’s important, get it on on a schedule, then find similar work in similar areas from similar crafts people, and schedule as much work as you can in the areas they wanted you to work already. That way you get two work orders done instead of one. Then you’ll become much more efficient. You base that schedule on what your customer wants. You fill it up based on logistics. UPS doesn’t deliver packages based on the time stamp, they deliver packages based on where you live, so you wanna load that truck up with as many work orders as you can and let them go get that work done. That’s kind of my strategy for a weekly schedule. And of course, the PMs go on first and you wanna have a sequence in terms of how you put that work on there, but find out what’s important to your stakeholder. And then find similar work based on logistics and based on crafts, skill sets, and things of that nature, so…
0:28:58 JA: The only change to the schedule with an event is that something lower in priority drops off. Because you’ve now lost time to be able to execute that work. Your lower priority work, the paint the fence and the mop the floor work, gets pushed off of the schedule. If you’re having a problem prioritizing work, then I would work on your prioritization system. If everything is a priority, then nothing’s a priority. [chuckle] Your weekly schedule has to be set up to where you can drop off some of that work, because events will happen, but that doesn’t mean that you adjust the whole schedule for the week. You just drop off whatever it was that wasn’t as important, because that works not gonna get done.
0:29:45 Ryan: I’m curious. A lot of our customers, they work in that backlog, they do a weekly backlog review, they do a retrospective at the end of every single week. We’ve also talked to a few of our customers that said that what they do is they allocate, let’s call it, 10% of the time, as just un-allocated. No, I’m… No, don’t do it. [chuckle]
0:30:07 GW: No, no.
0:30:08 JA: That’s not a best practice. It’s a good starting point because you’re so reactive, it’s hard to gain control, but as you get better at it, you shouldn’t allot for any of that.
0:30:18 GW: You don’t plan to have emergencies.
0:30:20 JA: Yeah.
0:30:20 Ryan: Yeah. Fair point, you don’t plan for emergencies, but they happen, right?
0:30:28 JA: They can.
0:30:28 GW: They do, so here’s why, Ryan. You’re gonna schedule… Let’s say you’re scheduling 100% of your time, and you only are getting 85% done. That 15% you can run a report on what didn’t I do that I said I was gonna do? And what did I do that meant I couldn’t do this? And now I analyze this and determine were they really emergencies or not? If I’m only scheduling 85 and I’m hitting 85, am I ever gonna look at that?
0:31:00 Ryan: Probably not.
0:31:00 GW: So, I’m never getting better.
0:31:01 Ryan: Absolutely. I think what I’m hearing from you, is that push towards 100% is the push towards excellent. You might not always hit that 100% compliance, but at least that’ll give you the reporting structure to analyze what happened so that you can improve.
0:31:21 JA: It’s a progression. So, one plant I was at, we were reactive about 80% of the time. And when I left two years later, we had two breakdowns a month on 15 lines. So there’s no reason not to schedule 100% of your work, right. But as you get better, you’re gonna start to see the gaps. And then you’re gonna ask questions around some of the work that your maintenance guys are getting called out to, like, “I couldn’t get the machine started, I have to push the button.” That’s not a maintenance issue, that’s an operations issue. When the machine breaks and it doesn’t work anymore, that’s when my guys should be out there executing work. So what you find is a lot of maintenance mechanics are doing the work of operations, trying to fill that gap, and your time is wasted and you’re not getting any of the work that’s important to the asset done. And so, being able to analyze all this stuff, you’re not gonna be able to do it if you don’t schedule 100% of the time.
0:32:22 GW: And that brings up another really good point that everything you do must be a work order. If it’s not a work order, you can’t do it. So it didn’t happen.
0:32:32 Ryan: What do you think Rob? Similar sentiment there.
0:32:37 RK: Well, I’ve been digging into work orders, like thousands of work orders lately, and what George just said, I’m getting fired up now.
0:32:49 Ryan: Yeah, if it’s not documented, you didn’t do it.
0:32:53 JA: Right.
0:32:55 Ryan: Very fair point, doing the work is only as important as being able to analyze why it was done so that we could prevent it from happening again, especially living in this industry. So the next question, actually somewhat relevant to what we’ve been talking about, How do you start and apply a reliability maintenance program for a reactive plant with no relevant asset data history? It seems like a big one. Maybe where do we start? Yeah. [chuckle]
0:33:26 GW: I mean there’s a lot to start but it’s like that… We answered part of that question with the bombs, in terms of finding the data but that’s really… I wouldn’t start at the data, what I would really start with is: What are my maintenance practices? What does my plant look like? Is it clean? Is it tidy? Are the tools where they’re supposed to be? When a PM starts, do the guys have all the tools they need? Do they have the spare parts they need? Do they know what work they’re trying to do? Like basic, basic, basic stuff before I worry about doing a Pareto on my failure history or RAM modeling, type of stuff.
0:34:12 Ryan: Do you see… Are there certain checkpoints within to getting a plant to, let’s call it a reliability… Full-on reliability program? Do you see it as first we need to, let’s call it, get our asset history or hierarchy in as step number one, the next thing that we need to do is move everything in a digital system, the next thing that we need to do is run PMs and before we actually get to doing a RCM pro… Like a full-on RCM program? Do you see separate…
0:34:44 JA: No, step one is you gotta stop the bleeding.
0:34:47 Ryan: Stop the bleeding, alright.
0:34:48 JA: You’ve gotta stop the bleeding first. You are a reactive organization. So the way I’ve done it and I don’t recommend this for everyone but the way I’ve done it is I don’t even look at my PM program, I walk straight up to the floor, I ask every operator what is the issue with this machine and then I bring in my maintenance guys and we solve the issue so it doesn’t come back, and we just systematically move through every piece of equipment on that line until that line is now reliable. And I’m telling you it takes a week or two weeks maybe, depending on the complexity of your line. So there are some complex lines where it’s hard to solve those issues, but most of those problems, within a few weeks, you can have your line up and running to where it’s manageable. And now, you can start shifting to maintenance strategy and focusing on maintenance practices but if you’re that reactive and you don’t stop the bleeding, it doesn’t matter. Anything else you do isn’t gonna change the fact that you’re bleeding. Alright, you’ve got a cut clear across the main artery and you’re putting band-aids on scratches on your elbows, it doesn’t make any sense.
0:35:52 RK: No, no, no, no, Joe, you gotta start by AI and…
0:35:58 RK: Hold on, but this is basics.
0:36:00 GW: That’s not it. [0:36:00] ____. So you guys are both right. Part of stop the bleeding is, can I be a good reactive organization and I think some of that is what Rob is talking about. And I think once we get to a point where they can create or alter the way they do business in a more reliable fashion, I think what they’re kind of looking for is: What do I own? How critical is it? How did it fail? What am I gonna do about it? But you can’t do those things until you get the bandwidth to do that work. And since we can’t just hire 12 more people to continue to be reactive while we go build a strategy, you’re both right, you’ve gotta become organized and you’ve gotta stop the bleeding and then you barter that resource time, the resources that are no longer reacting to the same thing day-to-day or day-after-day can now go help focus on and collect your equipment hierarchy, can now begin to help you with criticality analysis. So it becomes a bartering mechanism, you’ve gotta eliminate some inefficiencies first and then barter that time to become effective.
0:37:15 JA: And then, depending on: Are you set up for planning and scheduling? Do you have a planner and scheduler? Most… A lot of organizations don’t. And so, once you get that established and you become more effective at that, that frees you up even more time to go out and continue to improve the process.
0:37:34 GW: Yeah. If you’re a completely reactive organization a lot of times you have to do what I term growing flat which means you can add head count but you’re gonna alter your organization so as you gain efficiencies you hire a planner. That planner creates efficiencies and now you have free resources and you convert one to a reliability engineer or vocation mechanic or a reliability technician that’s collecting data and so as you gain those efficiencies you transition that to become more effective.
0:38:03 Ryan: Yeah, George that’s a really good point there because I feel like people’s first inclination is typically like we don’t have enough resources, let’s go hire two more people that can help us run a…
0:38:14 JA: Yeah, nobody ever has enough resources, ask every maintenance manager out there. No, they don’t ever have enough resources.
0:38:23 GW: Meanwhile the rent climbs at 20%
0:38:25 JA: Instead of 60 we are…
0:38:27 GW: Yeah, yeah.
0:38:28 JA: They could have four times the amount of work getting done with the same head count and so we all need more resources.
0:38:37 Ryan: Yeah, so I’m gonna try to summarize what I heard here from you all; The first one before you start a RCM program stop the bleeding. The next thing that I heard is, alright let’s get that asset register and let’s start building that history, start focusing on maintenance planning and scheduling, build more efficiencies into the organization which would then allow you and your organization, to then start thinking about maintenance reliability programs. The one thing that I’ll add into this section here in this question here is, we often get asked this question too, when does it make sense to apply an entire strategy for my organization? What we always say is don’t try to boil the ocean, pick one asset, pick the most critical asset, and run a full RCM analysis on that one before you try to do it for the entire plan, because to us, you’re gonna learn so much. Would you agree with that approach?
0:39:46 JA: Doing an RCM?
0:39:48 Ryan: On a single asset before…
0:39:52 JA: It depends on if you have a criticality analysis done, doing an RCM on the whole plan doesn’t make sense.
0:40:02 GW: Yeah.
0:40:02 GW: But I agree with the approach of not boiling the ocean. And I talk about this a lot in the training classes for planning and scheduling. A lot of times we’ll have planners in there that say, well I planned for 40 people, I planned for 40 people, and they just can’t be effective. That you cannot effectively plan for 40 people. So what I typically tell them to do is instead of writing a job plan for 40 people for the work orders, but not getting any parts for them, which means you’re not gaining any efficiency, get that down to five or 10 people, and pick just one area and focus on getting them all the parts they need before they go out into the field because that’s where the efficiency gain is. And so I think that approach works whether it’s we’re gonna go through a FMEA or RCM process, or whether that’s even just creating an equipment hierarchy inside the CMS for the first time go pick a small area. ’cause you’re gonna have lessons learned. So I think that advice that you’re giving Ryan applies to a lot of things that deal with asset management and the liability.
0:41:10 Ryan: Yeah, absolutely. All right. So we’ve got another question here. How often should Instrument equipment be calibrated? So this is probably a very broad question, maybe we could boil it down into a few high level categories.
0:41:32 JA: There’s a lot around that. That’s a loaded question. Is it FDA regulated? Is it not FDA regulated?
0:41:38 GW: Is that Jason’s question?
0:41:41 Ryan: Yeah.
0:41:41 GW: It is. This is a tough… So that the simplest answer is that the instrument requires calibration on a specific frequency. Some of them don’t require calibration at all, some of them require calibration every five years, some of them require a calibration every six months. Your industry, however, may decide to take a risk-averse approach, which means, let’s say you do a batch operation and you’re running product that takes a month to create, and it’s worth millions and millions of dollars, you may not be willing to wait five years to find out your instrument was out of tolerance. And then I have to investigate five years worth of product. And so, there’s a balance between the risk part of it, inside your business, and what the instrument looks at. And I think one of the things…
0:42:34 RK: What are you doing?
0:42:37 GW: Yeah, so one of the things we have to do is really balance that risk approach with what the instrument is supposed to do, and so, some of the ways you can do that is by monitoring what’s called drift. So over time, you monitor drift of the instrument to determine whether or not in between frequencies of calibration it becomes more and more out of tolerance, and if so you can begin to predict at what point you’ll get an out of tolerance event, and so monitoring drift and monitoring the risk of your business helps you start to narrow down specific instruments. But I think the basis of the question is really around that the whole calibration rationalization piece, because a lot of folks can’t just go with the embedded frequency from the vendor because of the risk to the business.
0:43:28 GW: And so, first and foremost, make sure that your validated equipment really only says which instruments are actually critical, ’cause they’ll call everything critical and then that’s the biggest problem you’re having. So this is a big giant loaded question, but the only critical instruments are critical and then start monitoring drift to see if you can stretch the frequencies.
0:43:54 JA: Yeah.
0:43:56 Ryan: Yeah, that’s a very interesting question because what I heard is it’s not necessarily about the schedule that the vendor gives you, it’s also what’s inherent to your business, it’s also very custom-tailored to what’s important to your business, is the mis-calibration by.001 gonna be detrimental to your business. Or is it gonna be okay? And the vendor doesn’t know that, but you as the asset owner, might.
0:44:30 GW: The interesting thing about that, Ryan is that you’ll get an out of tolerance event and then your organization will do an investigation and 99.99999% of the time, the output of that is… Well, there was no risk to the product we could ship it out, but we spend all this money doing the investigation anyway. If at the end of that investigation, they were required to now not called that a critical instrument. Since it didn’t matter that much. Things’ll be a lot of different.
0:44:58 Ryan: Yeah. And it goes back to what’s a priority? If everything’s a priority, nothing’s a priority.
0:45:04 GW: Exactly right.
0:45:07 Ryan: Alright, so we got one more question here. Where do you start from in an organization where there’s no reliability practice? Where do we start?
0:45:20 RK: Why don’t we just… Didn’t we already answer this one?
0:45:24 Ryan: Yeah, yeah. [chuckle] That’s a fair question. But it’s interesting because a lot of these questions stem from, “we know this is what we should be doing, we know that these are best practices, our facility is not doing it, where should we start?”
0:45:42 JA: But the hard part for them is that they don’t know where to begin. They know what the goal is. But getting started is hard. And so, like I said, I would go out to the equipment with my maintenance guys and talk to the operators, “What’s the top five things on this machine that’s eating your lunch?” Yeah, and then I would solution, those out. And make ’em go away, and continue to move through my lines to continue to improve it, before I come back and try to focus on anything else. And so that’s the easiest way to stabilize your organization and allow you to focus on developing the maintenance practices necessary to achieve that goal. But you gotta have some stability.
0:46:30 Ryan: Yeah.
0:46:32 GW: I think you focus on is education.
0:46:34 JA: Yeah.
0:46:34 GW: I think if folks are in an organization today, that has no approach to reliability, for me, step one would be; go understand what the heck that means, ’cause if you’re not doing an FMEA or an RCM, for the sake of a piece of equipment without understanding its value proposition back to your business is a waste of time. And so educate yourself on what reliability really is all about and what asset management is about and educate yourself on your business so that you can understand what the value proposition of your equipment is.
0:47:11 Ryan: Yeah. It’s very fascinating, very interesting. It sounds like there’s so many people who want to get kick-started in this industry. Learn more. But we go back and I think, Joe, you mentioned this, that there’s never time in the day to focus on education. There’s never any extra time in the day for maintenance manager to go out into the plant floor and just talk about what’s wrong, how do you balance that?
0:47:42 JA: Quit going to meetings.
0:47:44 GW: Yeah. [chuckle]
0:47:46 JA: Number one should be your job. 90% of the meetings on your calendar are an email. And so, what I… Honest, like I said, I probably wouldn’t do it if I were you, but the way I went about it is I quit going. I would talk to the plant manager and say, “Hey is there really value in me going here? ’cause I think my value’s out on the floor getting you the throughput that we need to hit the numbers so we hit our sales projection.
0:48:10 GW: Yeah.
0:48:11 JA: So do I really need to be in here or should I be out there? And every time you need to be out there. Alright, thank you. Right. And there’s no reason to go to all these meetings. And so what happens is, as a maintenance manager, your entire 8 to 5 is packed full of meetings. That means, you either come in at five in the morning to spend three hours doing some work that needs to get done, and stay until 9-10 o’clock at night to try to spend more time to get work done. And so now your work-life balance is a mess, because you’re putting in 20 hours a day, to keep a plant running, and it doesn’t make sense. After six months, you’re burned out, and you lose interest and you’re just like, “This isn’t gonna work.”
0:48:54 GW: Yeah.
0:48:54 JA: So why not be effective in the 8 to 5 by not attending the meetings and going out and getting the work done? But again, that’s me. Not everybody else. I’m a little more hard headed and straightforward than most, but it works. I have a tracker that provide me notes. I can back up that with my tracker. And so, having 7, 8, 9, 10 plant turn arounds. I can tell that that’s what I’ve done. And every one of them, I walked in, was just the same as everybody else. Reactive as can be, even the clients we’re working with now, it’s the same thing. I’m not doing anything different for my clients that I did when I was a maintenance manager.
0:49:41 RK: Except you get paid more now.
0:49:43 JA: I’m getting paid more.
0:49:47 Ryan: Oh man. Well I’m curious how many hours a day were you spending in meetings?
0:49:54 JA: When I first started… Any job, it’s almost every hour of the day. They put you in meetings. I walk in, I don’t even know where the bathroom is and I’ve got 15 meetings on my calendar for the first day. Right? And I haven’t even met… It takes two weeks before I get to meet all my maintenance guys. And it’s been that way just about everywhere I’ve been. You walk in and here, you gotta go, you gotta go to this meeting at 8, you go to that meeting at 08:30, there’s this one at 9. And I’m sitting back going, “Hang on here. Wait I use the bathroom first, and then we’ll figure out what my calendar needs to look like.” And really talking to especially your maintenance guys, sitting down and having a conversation with them and say, “How’s life treating you here? And would you like to see it better? And if I can help you become better would you support me in pushing this through?
0:50:52 GW: Yeah.
0:50:52 JA: Right? And most of them, there’s always the 10%ers that they don’t wanna do anything but for the most part, people are gonna support you, and then it’s one of the obstacles that you face from keeping you from doing your job. And most of the time it’s tools and time. They don’t give us time on the equipment and they don’t give us tools. And that’s [0:51:15] ____ anything. The reason you don’t have time on your equipments, because you have no credibility. Because every time we do a PM and they start up the line, the machine fill. Right? And so you lose credibility as a maintenance guy. Okay, so we’ve gotta work on our credibility. When we say we’re gonna do something, we’re gonna do it 110%. That’s we know possible. And then it’s my job to blockade all the negative criticism and just keeping them going. It’s…
0:51:44 Ryan: Alright.
0:51:45 JA: That was too long man.
0:51:47 Ryan: No, that was good. I think the recap, there is, it’s, maybe I’ll say look at your calendar, figure out which meetings that you might not need to attend, versus maybe straight up not going and not telling anyone. [chuckle]
0:52:04 JA: I would always get my boss’s approval. I didn’t just not show up, but I always had a way of selling it to where the morning meeting where they’re talking about blaming maintenance for the night before the production, I’m always in that one ’cause I bring data with me, right? You guys say, “Okay, I’ll take the 20-minute breakdown we had, but based on our designer to this equipment, we’re still using six to eight hours of downtime, so where’s the rest of that time going?”
0:52:34 GW: Yeah.
0:52:35 JA: And so they quit blaming me for that. So I’m always in that meeting. That’s an important meeting, but for the most part it’s, “Hey, we got this supplier coming in today, we wanna look at this stuff. Oh, we have this weekly call to go over our 6000 key performance indicators.” [chuckle] And I’m like, “I don’t need to be in here. I don’t need to be in here.” You have 6000 KPIs you’re not measuring any of them.
0:53:04 Ryan: Yeah.
0:53:04 JA: So, you should have maybe eight tops…
0:53:06 JA: Which means you’re not measuring anything, ’cause what are you gonna do with all that data? So, it’s those types of things where it was just getting so out of hand, especially as you gain more knowledge and people lean on you as a subject matter expert, they wanna pull you more in and so it’s hard, but it’s like “I’ll share with you my opinion, but we can do that offline. I don’t need to come to this meeting to do that. I’m more than happy to stop by for a few minutes, give you some feedback, and let you know what I think, but I really need to be focused out here.” And so it all depends on how you handle yourself, I guess.
0:53:50 GW: Yes.
0:53:50 RK: The astute worker.
0:53:52 JA: Yeah.
0:53:52 Ryan: Alright. Well, we’ve actually got one more question from Jason Verly, let’s see if we could go ahead.
0:53:57 JA: Oh, here we go.
0:54:00 RK: Oh no! [chuckle]
0:54:01 JA: He already stumped us. We haven’t even heard the question yet.
0:54:04 Ryan: Alright, there’s been a lot of talk about digital twins. I understand the concept of digital twins goes beyond just having a bill of material when an asset is deployed. For new projects and assets, a lot of this information is owned by the CapEx engineering team. What are some sources you’ve used to educate engineering on digital twins to show the value of passing information on to maintenance and/or reliability teams?
0:54:32 GW: Wow, so I think that’s a good question, Jason. For me, the digital twin gives us an opportunity to close a gap that has always existed in the engineering turnover package. Its existence wouldn’t be a necessity, it’d be a luxury, not a necessity, if we were actually getting those engineering turnover packages over the last 50 years, but we don’t, so in the maintenance side, the asset drops on the floor, we get told, “Hey, have a nice day, go operate, maintain, that for the next 40 years”, and you’ve got no information. In addition to that, the next time they need to do an engineering project, no one knows where the drawings are. So now we’re hiring an engineering firm to come out and assess all of this stuff and make new drawings. So there’s the whole benefit that’s associated with the fact that you didn’t get the information when it was on paper. The other part of the digital twin is you can begin to do modeling, so you can then say, “Well, what happens if my capacity or my demand for chilled water goes up?”, and you can begin to model all of that inside the space of I guess the building information modeling and all that stuff, where the digital twin resides.
0:55:55 GW: So I think there’s a piece that closes the gap that existed on the paper trail side, and then there’s the piece that helps engineering and [0:56:04] ____ The trouble I see is, I don’t necessarily know that the creation of or utilization of a digital twin is gonna actually cause them to put the data in the system that will get you the turnover package. I can create a digital twin that just has the process parameter information and I still didn’t go get all the other data that should have been part of the turnover package to begin with.
0:56:29 JA: I think that’s gonna be the gap, it’s the same, it’s gonna create the same problem that we have today and not getting the turnover packages and not getting over that.
0:56:38 GW: Just like any other system, if you have a process that’s currently chaos putting in an electronic version of that simply automates the chaos. [chuckle] And so I think you still have to have sound requirements on the engineering side, on the CapEx side, of what must be included in the digital twin.
0:57:00 JA: And I think that’s where maintenance and reliability have to provide those requirements to the engineering department. I think that’s where you start to get that buy-in is helping them to develop specifications, ’cause a lot of times they don’t even have them, and so if you can come in and provide them with that data, it allows you to develop the credibility as being a go-to person for every engineering project that they come up with, and I think that’s kind of how you would help close the gap there.
0:57:35 GW: What’s your experience here, Rob?
0:57:37 RK: Yeah, it’s one of those things, right? Like when we talk about data, usually it’s about that strategic asset management data policy and so this is part of it. So what data needs to be transferred over from your hand over, what does that include? Like I was in a meeting actually last week, and we were talking about that and defining what’s the data that is at this step, what’s the data required, what decision does that data help inform and what’s the value to the organization? Those are kind of those pieces, right? Like George, you mentioned with the calibration side, it’s about risk. Well, everything, like risk is another way of saying dollars and cost or benefit and cost. And so that’s what everything we do really boils down to. Like how do you make your organization more profitable, more profitable, safer, better for the environment, better for the community and help them achieve their mission, vision values, right?
0:58:44 Ryan: Absolutely. Well, it’s such an interesting and great question that we had from Jason and for a matter of fact, all the questions that we had were super insightful, I think we talked a lot about how to get started, some unique insights into technology as well. I had a great time, thank you everyone for joining, thank you to all of our experts on our panel for sharing so many amazing insights. To wrap us all up, I know we’ve got maybe one more minute, can we share… Do you guys wanna just go around in a circle and basically, we could share with all the folks listening on how they could get in touch with you guys in the future?
0:59:30 JA: Yeah, sure, for Joe and I, we can be reached or anyone at ReliabilityX at [email protected]
0:59:38 Ryan: Alright, let’s for sure do that. And Rob?
0:59:42 RK: Yeah, for me, you can email me at [email protected] or subscribe to the podcast as well, Rob’s Reliability Project on your favorite podcast platform. [chuckle]
0:59:53 Ryan: Alright, well, thanks everyone for tuning in for today’s live Q&A. I’m Ryan, I’m the CEO and founder of UpKeep, I’m pretty active on LinkedIn, so you could find me there. Thanks everyone. This has been a ton of fun.
1:00:05 RK: Thank you. See you around.