Back to the Basics Part 5: Maintenance & Task Development

Chris Pepin (00:04): Ladies and gentlemen, welcome back to the manufacturing talent podcast. I’m your host, Chris Pepin, founder of Progressive Reliability. On the seven part series I’m going to be joined by Tim Goshert and Doug Plucknette. And we’re going to be going through back to the basics, the fundamentals and the importance of getting it right within manufacturing, maintenance, and reliability. Welcome to the conversation!

Chris Pepin (00:25): Well, hello again, ladies and gentlemen, and welcome to the manufacturing talent podcast with Progressive Reliability. My name’s Chris Pepin, founder of Progressive Reliability, and I’m joined again by Tim Goshert and Doug Plucknette. And I’m sure we’ve got the introduction covered elsewhere. I think what, five out of seven here. So, trying to make sure we don’t repeat ourselves too much to keep the information as tight as possible for you guys. And frankly, today we’re going to build upon the other four conversations and get into the maintenance task development conversation. What five types of maintenance tasks there are in the requirements for each type of task. And spearheading this conversation today, it’s going to be Doug. Doug, do you want to go ahead and get us started?

Doug Plucknette (01:16): I certainly can. So last we talked about developing failure modes and understanding failure modes. And from that point you get into developing maintenance tasks, right? So what you need to do is match the correct task to your failure mode. And there’s a couple of ways of how that works. You know, you have to think about, is our failure mode a wear based failure mode, or is it a random based failure mode? If it’s a wear based failure mode, preventative maintenance might apply. And there certainly are some factors that we have to save when we’re looking at, okay, whereby say items that are measurable, what type of PM we’d use there. It’s going to be a qualitative type of PM, where we can measure that where, and based on the, where and time that it took to get there, how much longer can it last, does it need to be replaced? That type of thing. And then, or is it a random based type failure?

Doug Plucknette (02:15): And often random based failures, we hope to either, that we can use some type of [inaudible: 00:02:18] condition tasks to detect that the failure is in the process of occurring, if it has a useful PF curve. If it does, then, then we talk about the different types of unconditioned tasks that we could do. So let me first get into talking about, there are five types of maintenance tasks that [inaudible: 00:02:37] traditionally look at when we’re building a complete maintenance strategy. The first is on condition maintenance. Some people you’re going to hear a lot of people call it predictive maintenance. I’m not fond of that name. Condition-based maintenance, I think is a better name for it. [inaudible: 00:02:56] with condition maintenance. That’s what Stan Nowlan and Howard Heap, the creators of RCM call it. You’ll also hear people call it condition monitoring.

Doug Plucknette (03:07): At any rate to condemn the intent of the unconditioned task is really detect failures or potential failures. So, the use of condition monitoring or inspection techniques so that the corrective work order can be planned, scheduled and then executed before the failure occurs. The second type of maintenance that we look at is preventative maintenance. And preventative maintenance is based on and only works and is only effective on wear based types of components of failure modes, right. And the interval of the PM tasks should be based on the known useful life of the component. You’ll often see people say, oh, it needs to be based on the MTBF. Well, meantime between failure, that means many have failed before that, and many are going to fail afterwards. So you really want to look at what’s the useful life of the component, as opposed to basing it on the MTBF which is often overused.

Doug Plucknette (04:03): The third type of tasks that we’d look at are failure finding tasks. Failure finding tasks or time-based inspections of a hidden function device that are really intended to ensure the device is still capable of performing its intended function, such as, an emergency stop switch, a high level switch, a low-level switch, a high pressure switch, a low pressure switch, a relief valve, a rupture desk, right? All those things during normal operating conditions, we can’t tell whether the item is in a failed state or not. It’s normal for it to be in the condition that it’s in, right? It’s normal for the emergency stop switch to be in the closed position, right? When you push it, it opens and it stops the machine. How do I know whether or not that’s going to work, right, at any given time, you have no idea.

Doug Plucknette (04:57): And it could be in a failed state where if it’s failed closed, somebody pushes it nothing’s going to happen. So what we need to do is identify those hidden function components and set up failure finding tasks for those. The fourth type of tasks that we consider is redesign. Redesign in the terms we’re talking about is any change in equipment process or procedures. So we look at our components and talk about how they could fail. Is there a way that we could do a simple redesign and eliminate this failure mode? That’s really the best thing we can do, right? Anytime we can eliminate a failure mode, we certainly should do that. Sometimes those changes in procedures are as simple as how we install something to make sure it’s installed properly so that we don’t see the failure mode down the road. The last type of task we’re going to discuss are consequence reduction tasks.

Doug Plucknette (05:52): So if we can’t detect the failure, reduce the failure through PM, if we can’t redesign that failure out, then we get to what we call a, in RCM, no scheduled maintenance. When we get there, we should talk about how can we reduce the consequences? So we’re going to accept that this failure mode can occur. What can we now do to reduce the consequences of that failure? Things like having spares in place to correct spares, having good job plans written so that we know how to replace the item in the most effective way, because some of these can be, you know, you get into some of these tight little packaging machines that can be [inaudible: 00:06:33] that’s packed away inside there. What are the three items that need to come out before he can get to this one? Those types of job plans saved a significant amount of time over years and years, right? So we want to make sure if we consider no schedule maintenance, what can we do to reduce the consequences? Now, when we look at these five types of tests, there’s certainly a balance that we should be seeing. And that’s where Tim comes in. He, at Cargill, his team looked at, saying, what’s the correct balance of these tasks? So, Tim, if you could speak to that, that’d be great.

Tim Goshert (07:08): Yeah. Well, going back, you know, 20 years or so when we first started understanding this at Cargill, what I found was we were doing basically two different things. We were doing what we call the dead time PMs, and they were probably not the right type of PM’s as you defined them, Doug. And we were just doing work-based when the equipment failed. So, this was a very enlightening process for us and understanding that equipment failed, we needed to have different tasks to do it. And what we did as we found is that, you know, we were probably doing 50 to 60% failure based maintenance. And then the balance was what people at that time called PMs, which were not quantitative in nature, but they were more qualitative. And they were just doing loose inspections based on a company, you know, like almost using the CMMS as a come up file and say, well, we need to look at this every six weeks or every six months.

Tim Goshert (08:31): So, it was a long process for us to attach, you know, understand failure modes and then attaching the right type of maintenance strategy to that. And what we found was the majority of our machines failed randomly. And therefore, we should be spending the majority of our time, you know, more than 50% of our time working on condition-based type maintenance activities, whether that be going and doing their surveillance and running the routes, and then the results from that work. And the smaller amount of work was preventive maintenance tasks based on equipment that failed due to abrasion or wear. And that should take up maybe 25% of our time. So, we were saying that 80% of our work, roughly 75 to 80% of our work should be condition-based and then preventative tasks, and then the balance or the other three categories of failure finding tasks, which we had what we call critical safety devices, and the critical safety device checks.

Tim Goshert (10:03): That’s where we checked high level alarms, low pressure alarms and those kinds of things. And our lowest category at the time when we started was the redesign, because we were just, we weren’t really focusing on finding ways to redesign our processes or machinery to eliminate the failure mode. Now, one good example that comes to mind was, we were struggling with pump, you know, dirty oil and pumps and various machines. And what we did was do a redesign of, instead of using lip seals, we went to inpro/seals, or more of a mechanical, tighter seal on, on the turning [inaudible: 00:10:52]. And we found that that helped us. So, this was revolutionary for us and getting us to a much higher level of production, as well as it helped us reduce our costs by doing the right type of maintenance for the right type of failure mode on our critical equipment.

Doug Plucknette (11:17): So the key with this, Tim, it differs from some of the other stuff we talked about earlier is you really need some people with experience to go to this level, right? When we start talking about differentiating tasks, that how the task should match a failure mode, right, that we get the correct task done. This takes a little bit of education.

Tim Goshert (11:42): Absolutely, just the area of condition-based maintenance takes quite a bit of education, because you need people that understand what the various condition-based monitoring tools can do when, what they can’t do. You need, just an example, you need people that understand vibration to the point that they understand it can find different failure modes and bearings, and even [inaudible: 00:12:18] and things like that. And you need to have people that are knowledgeable in those areas to be able to find the failure mode, give the maintenance person a task that was based on that failure mode of what to, you know, what to fix. So, educating the organization on the different condition-based tasks, educating them on what is really a qualitative, preventative maintenance plan and what is really a good failure finding task.

Tim Goshert (12:57): What are all these elements of these different categories of maintenance types, you know, what do they entail? And you needed an organization that understood that. In addition to that, what we did is we wanted to track at each facility how many work hours where we attaching to each one of these maintenance types on a weekly and monthly basis. Were we doing the right amount of condition-based maintenance, where we doing too much preventive maintenance, were we keeping up with our failure finding tasks that we knew were critically important to us? So we developed systems to measure the type of work that we were doing on a weekly, daily, monthly, and quarterly basis.

Chris Pepin (13:52): A question that may come up, and this is going back to our initial conversation, how is this going to be affected by the newer tool, the newer machinery, the newer processes, all the promises that are coming out right now with new technological development? Where does this process get mixed up and where does this get pulled aside or replicated or shortchange where it really needs a lot of focus and effort. And it’s really adaptable to have this before any new equipment or any kind of new technique,

Doug Plucknette (14:32): That’s a great question, because I could tell you right from the get-go, these new machines that are going to come with all those unconditioned bells and whistles already attached. So, they’re going to be there full time. All right? So if you’re a company that hasn’t done the foundation of this, and doesn’t understand this, right, there are going to be literally way more alarms than you’ve ever seen before, if you don’t handle this right from the start. Right? If we don’t look at it thoroughly during design, during the installation, if we don’t set up our equipment hierarchy properly and understand what came with our machine, what are those things monitoring? What do these alarms mean? Right? Because they’re going to come with various types of alarms. Right? Is it critical alarm, is it a warning alarm, is it a combination? Right? So, we’ve got two things, one’s looking at vibration, the other one’s looking at noise, right?

Doug Plucknette (15:29): You have two of them at the same time. What do they all mean? Right? And if your company doesn’t understand that upfront, the tentation is going to be to say, how do we turn these off? Right? I’ve seen this many, many times even going back to where these were rounds that were done by people once a month. Right? What did we do to ignore it? Right. Well, how much longer will it run? Can I just keep hitting reset? Right? So really, what did we buy, and what are these things mean, and what should we do about them has to come with that.

Tim Goshert (16:08): And I would add, Doug, that sustainability of those systems is going to be a challenge. What I’ve seen in the past year or two is that, and this has happened to be a wireless vibration sensors that were installed and retrofitted in existing equipment. And I didn’t really realize it first, but soon came to see it. Well, the sustainability of that system is very, very critical. You have to have your maintenance staff that’ll understand what these sensors are. They have to treat them as important. So, when they replace a component, they need to take off, maybe remove that sensor correctly, and then re-install it. And so, I find that that’s a challenge, especially when you’re putting it on. Well, the organization’s not used to it on new equipment or retrofitted equipment with these sensors. And so now, a wireless sensor on the unit [inaudible: 00:17:17] replaced, but instead of being, you know, the component was replaced, but the sensor wasn’t put back on, and now it’s hanging in the air, and it’s showing no vibration and you think you’re covered. But what really has happened is , you’re flying blind. You don’t know if that equipment is being monitored or not. So sustainability of these new systems, whether it’s vibration, ultrasound, oil, or whatever it is, it is critical importance, and the organization has to be attuned to it. It’s a huge challenge that I personally underestimated, and I thought, well, you put this on and it’s just going to happen. Well, no, there’s change of equipment and components and things like that. And everybody in the organization needs to know about it.

Doug Plucknette (18:13): I mentioned, I think early on one of these podcasts at a year and a half ago, I worked with a company on their new smart machine. And not only did I work with a company that was buying it, we’ve worked, did the RCM with a company that manufactured it. And they were really amazed when we got into it and started talking about failure modes. And we found, you know, a couple dozen things that they immediately said, holy smokes, that’s not right. We’ve got to change this. Right? And they made changes to it. Now, prior to doing that analysis, the company that was purchasing it actually made the manufacturer sign off that they would not give those changes to their other customers, right, that they would not give this package of a complete maintenance strategy to other customers that bought this machine. It kind of put them in a bit of a bind where they, you know, they had to take a couple of time outs during the week for more than an hour to say, all right, this is really something that needs to be there with the machine.

Doug Plucknette (19:09): And we missed it. You found it for us. And they ended up working out negotiating on price, as opposed to, could the new customers have that, you know, change or redesign. So these things are really important. It gets into, if you don’t understand this, you’re going to completely miss out on, right? It’s really that black and white when it comes down to it and the importance of on condition maintenance and how you should respond to it, right? Why it’s there, and what’s the proper way to respond? Because nine times out of 10, that response is going to be, how much longer will it go? And the truth, the proper response is, let’s start planning and scheduling and pick a date to get it done. If you’re not there before you purchase this stuff, don’t think that it’s just magically going to happen because arguing machine has the alarms.

Chris Pepin (20:07): Okay. Wow. Well, so, what else in terms of [inaudible: 00:20:12] coming to the end of this conversation, how can we, how do we tie together all the building blocks and really at this point in the journey, what are some of the things that may have been missed in prior points in other aspects of the organization? Like where did the mistakes start happening? Where does the process start coming apart? And what are the dangers to look out for kind of at this phase of getting your fundamentals in place?

Doug Plucknette (20:38): Right. It’s really, you know. I’m really a person that says you need to focus on education first, right? This training of this and understanding of it. If you’re not, you know, lots of you hear it all the time, you don’t know what you don’t know. Right? If you don’t understand this stuff, if you think you do, sit down with somebody that does, all right? And go through and have them explain it to you. And if you’re missing out on… You know, I hate to say, take a test, because there’s a million of them out there. Right? But the CMRP is a good example, right? If you’ve got a number of people that have taken that in the past, then you’ve got people that understand this, all right? Those are the right people to involve, to say, let’s start putting this together.

Doug Plucknette (21:29): If you don’t have that, then find somebody out there that does, you know, give your company a call. Because you obviously have people or offer people that can help with this type of information and how to get people on board with it. Because it really is not only a [inaudible: 00:21:44] education, it’s a sales pitch, right? We’re going to sell you that this is the best way to do maintenance. And although we haven’t been doing it this way, we understand that, that change is a culture shift. We need help with that. Right? That’s where management comes in. That’s where leaders come in.

Chris Pepin (22:03): Yeah. And with the CMRP, how does one figure out somebody that has the right communication tools to, I know there isn’t necessarily so much of a change management certification out there, but when you’re talking about changing the culture and really coming from the grassroots, what are some things to look out for? And what are some things that you guys have found [inaudible: 00:22:24] tremendously helpful or successful?

Tim Goshert (22:25): Well, I would echo with Doug, what Doug said, that education is the key. You have to go after educating your entire organization, not only the maintenance people, not only operations, but also engineering, procurement, even the commercial people, on the process. And then you have to expect and have expectations of use of what they learn. Expect that when we buy a new equipment, that we’re going to understand its failure modes and have maintenance strategies on the piece of equipment before it started up, not weeks afterwards, its expectations of implementing the new processes and procedures.

Tim Goshert (23:24): Now, everybody is not going to have the light bulb come on at the same time. So it’s continual reinforcements of the right process and processes, which we basically laid out in the last couple of podcasts. This is how work should be done in an industrial facility that has any critical pieces of equipment. This is how you evaluate it. You just have to use it and expect [inaudible: 00:23:58].

Chris Pepin (23:57): Well, gentlemen, I think, unless there’s anything else, that’s pretty well covered. One of the last big topics next, we’re going to be going into the Achilles’ heel. So, we want to close up the series with some of the more key points of really getting into the depth of your crew, your team, how to have level of precision, and really get everybody, you know, well off into the right trajectory. Tim, Doug, anything else?

Doug Plucknette (24:31): No, I think we’re good to go, you know, we’ll talk next week, you know, as you said, about the Achilles heel, and probably add some in there about, from what direction does this need to be led, right? Can it be bottom up? Can it be top down? What are the differences between the two, and which way is most successful? Blah blah blah… But this is one of those things that everybody needs to hear, I think, regardless of where you are with this. Those that have been through this have a lot to add, and those that haven’t, I encourage them to find those that have.

Chris Pepin (25:07): Wonderful. Thanks for your time today, gentlemen. And we’ll return to you guys shortly with our next one. Everybody have yourselves a great week, and thanks again for listening.

Chris Pepin (25:18): Well, thank you for joining Tim, Doug and myself again. We encourage you to download the white paper, which you can find at our website www.proreli.Com so as you can find myself, Tim, and Doug on LinkedIn, we look forward to joining you on the next one.