Podcast: How Bush Brothers & Company deployed a three-phase PM optimization strategy

TW: All right, let's move to Phase Two, this is where Jeff comes in. You said the team was making some progress, but there were some places where you were getting kind of bogged down. Maybe you can talk about that and how you brought Jeff and his team into the picture.

TP: Yeah, so we were coming off of Phase One, started seeing some really good results. Within maybe 18 months of starting Phase One, we went to about 65% on time completion rate. So again, you start doing another dive, everything's updated in the system; second dive, we noticed that we had a hard time scheduling some of these PMs as only that person knew it. Or we were getting inconsistent results from them. 

So we knew there's another opportunity to take it to the next level. We said, oh, we're just going to get more detail there, so we kind of knew what we were going to do. The struggle is I brought the same team together. We started on a phase and we were chasing rabbit trails, going to places we shouldn't have went. Wordsmithing was horrible, that group there, they were always caught on words. And lastly, we defaulted writing such amount of detail in there, right, thinking that we just needed a person off the street to be able to do this. 

So at least we were mature enough to sit there and say that, hey, “we need to time out here, we need to regroup.” And at the time, Jeff and the People And Processes team was working with the main plant here in Dandridge, TN, and they were having everybody go through their PMO exercise. They were having all their techs go through it for an education piece. So we said, ah, we’ll go there, and then through that workshop, they answered a lot of questions for us. And we figured out a very disciplined approach for Phase Two because it needed it.

TW: Jeff, what were some of the things that you observed when Tony brought you on board, and you saw how Phase One had gone and you saw where Phase Two was pointing?

JS: Yeah, so as we talked about, it was a much larger implementation. It started with the original plant here in Dandridge. We began with an assessment and we identified all the opportunities. We went through the planning and scheduling process. We put together the storeroom steering team or storeroom team to go after that world, and at the same time, the reliability engineering aspects for the PM optimization, if you will. 

One of the sites, the local site, they began their PM optimization activities early on. But then one of the things we were trying to get them to do is begin with the education piece first so they didn't have the misstep, if you will, of Phase One. Not that Phase One didn't have value for Tony's world, but if he could have got it all in one path, one motion, that would have been a whole lot better for him from a timing perspective. 

What happened was, we were doing an RCM2 intro class for that particular site here. And they said, we're going to go ahead and start PMO. And we said, okay, are you sure you want to do that? Because you're going to learn a lot of things that you probably going to make you rethink the way you approach PMO. And so they said, well, we're going to wait till September. (This is a December-January time frame.) So they went down that path and realized, just like Tony did, that they, you know, had done a lot of great things. They found PMs, for example, that didn't exist, assets that didn't exist in the CMMS, things like that. 

So they were able to go do that kind of work, but they had to go back and revisit the failure modes, for example, and things like the operating context and other pieces so that they could truly understand, okay, what is the probable failure condition that those technicians are looking for? And what you often find, too, in organizations, it's very common across the world, really, 46% of the PMs that we're sending technicians out to do now actually had no value because they failed to address the likely failure modes of the asset. 

Many of you, I'm sure, on the podcast, would raise your hand and say, hey, I could use more technicians, Jeff. But think about this. If 40 to 60% of your PMs had no value, we're sending people out there, we require them to do the PMs, we want 95% PM compliance, for example. Those PMs are triggering out every week, every month, every six months, whatever, and we are consuming resources, but it doesn't prevent us from failing. What I'm always amazed by is I go to sites and you'll see the supervisors sit around the table, hey, they're celebrating that 95% PM compliance, but yet I can pass the plant manager in the hall and he says, yeah, they do that, Jeff, but the equipment's still failing. 

So it highlights the point of understanding the exact reasons why the equipment's failed. What are the likely failure modes that we can expect? And then addressing that from the PM program and then what indicates the potential failure condition. For example, if you use a simple example like a brake pad, at what point do we want to identify the P, the potential for failure, as opposed to the functional failure portion of that, and have enough time to go in and proactively plan and schedule the repair, get the materials, and execute it when it's convenient for operations. 

So those are some of the opportunities that we saw. The other thing that you get into as well is getting everybody on board with the process. And I'll tell you that people buy into what they helped create. And that's one of the things that Tony leveraged well.

TW: Yeah, the team approach you described involved everyone that was key in the team, but also key influencers, it sounds like, too, where the word would sort of roll out among everyone what was going on.

TP: Yeah, we brought in our maintenance techs to help with that across all shifts and in Phase Two we actually brought maintenance technicians from our other plant in because it was very similar. We had the same processes but very similar equipment and say, hey, what have you had? We both learned quite a bit through that. So now when they read a PM and engineer didn't write that, I help write that. And then at the end, we're using data to help us drive “is it right or not” and we'll adjust accordingly as long we go.

TW: In the spirit of recognizing the impact these would have on people on the wider teams, one of the things you mentioned in the presentation was that part of Phase Two was to identify whether PM tasks were intrusive or non-intrusive. And you mentioned to you that was a real game changer because, #1 it helped people plan the work better, but #2, what I'm picking up with the subtext is people felt more respected. Is it going to intrude on someone's time or responsibility? How can we finesse that?

TP: Yeah, the driver of the big driver of intrusive and non-intrusive, where it was, again, another aha moment for us, we started digging into, why are these PMs late? So I might have, say, six tasks on there. There's one task that they can't do because I need the line to stop, so that PM would just sit there and sit there and sit there. So through Phase Two, more awareness, it was just a, when can you get that done? Not just the frequency, but can I do it while the line's running? 

So we made an intent to say all, it's more PMs in the system, technically, but we started separating tasks that says we can do these tasks. Now my planner can schedule it, and they know the production schedule, but it's just, like you said, we're up to about 85% on time completion right now for that, because we're just, and maintenance techs know, when I get this PM, I can actually complete it because it's non-intrusive.

JS: So to add a little bit more around that, your listeners will probably recognize that as well with regard to running versus shutdown. You know, we think about intrusive versus non-intrusive, but really the goal is to schedule the technicians 100% of the available hours. So that means what are we doing when the line's running? As part of that, you know, what are these PMs that we can do, these checks that we can do while the equipment's running that actually indicate the potential for failure?

TW: Well, I've got two more areas to cover, both of them fairly brief, I think. First of it is the wrap-up metrics. What are some of the numbers that you and your team are pointing to right now where you know, okay, we're succeeding in this project?

TP: The main driver was on-time completion rate. And so that was helping us, and it's PMOs active. It's not a project. It doesn't have an end, right? And so as I keep pushing for that, my goal is 95% on time completion rate, and right now we're about 85%. So we're well on the way. 

The other one is the work order compliance. Work orders complete, in our world, is completely separate from PMs. But our work orders are now at about 73% on time completion rate. It's allowing us more time to plan the work, because we're catching this stuff higher on the PF curve now, because the way they're written, through that process. 

And lastly, to really look at, we're using a six-to-one ratio. So we're fortunate with our CMMS with the analytics, we were able to adopt the analytics to run a program in the background. It's just a report. When we fail a PM, it's flagged in the system. And then we’ll know “PM 123” has not had any follow-up work or has not been failed in that six-to-one ratio. We can get a report, and now we don't have to go chase everyone – the report's always there if you want to chase for the next opportunity. We don't have to do much digging, so we're trying to get some proactive analytics now to help us with that, but we're really using a six-to-one ratio, and all the techs are on board now, too, with that. They want to make sure their work's value-added.

TW: You've also gone to a state where you have a backlog of projects now, too, right? Where before it was a mostly break-fix.

TP: Absolutely. The way we used to do things, if they find something on a PM, they feel obligated, oh, it's down, I'm going to go ahead and fix it. Well, two things with that, they would never fail the PM, right? They would just create the work order and do it, and they'd just check things off on a PM or make a scribble note on the bottom. But they're also, by them fixing it, they couldn't get all their other work done because we scheduled all this work to get done. 

So now the rule is that, and we kind of use a rule that Jeff and team shares, if you can't fix it in 15 minutes with tools on hand, then you got to do a follow-up work order, right? And so what that does, it actually starts creating a backlog. Either they have to fail the work order, or fail that PM, or say hey that PM's working. And then the other thing is now I've got work to actually plan and get done on schedule and I'm actually keeping my techs busier, feel more organized.

JS: One of the things if we could talk about a little bit for clarity, we talked about the RCM, we talked about two really two different approaches. We talked about an intro for RCM and then we talked about PMO. And what Tony's world was talking about really is PMO. 

I mentioned doing an intro class. The intro class would be for management. When we talk about the technicians, the goal is, we have a PMO class specifically, and then immediately after class, we roll into coaching with those technicians. So the goal is not to do RCM. The goal is actually how do we improve our new strategies. And with regard to the PMO, that's better at a technician level. So we take the RCM concepts, we build them into the PMO world from the standpoint of the course exercises, the education, if you will. And then what we do is we teach them things like, you know, what are the potential failure modes, how to operate in context, all these things, relative to that. 

Then take that and say, okay, let's take your PMs the next two days, two days on the education roughly, and two days on coaching to start with. And what happens is you actually use in this case, Tony's world, his PMs, and the technicians to go through those within the instructor and the coach and step through those, okay, is this really written to a level of specifications that's necessary, is it intrusive / non-intrusive, and so on. And you start to educate the technicians. 

One of the things that we do too is we come back and we do additional coaching. So we can actually go back and evaluate the competency relative to PMO and structures for each technician. And then the ones that need additional coaching, we can provide that, so that way we're spooling everybody up, trying to get them to the same level, where that way it becomes a very effective organization moving forward. And as I said earlier, one of the benefits is the techs own it – they buy into it because they built it.

TW: That paves the way for Phase Three then, which is where we'll get you out of here on this one. Once you've done all the uplifting of the technicians and getting them on board with management, you're pointing towards Phase Three of this project too, right?

TP: Yeah, Phase Three, and we started tapping into it a little bit in Phase Two. I got a couple people who are really forward thinkers, and the question's been coming up is, can we eliminate the need for somebody to physically do that PM? So you start looking at go / no-go gauges or visual indicators, or of course there's just predictive technologies. So my goal is to continue to reduce the amount of PMs, because if your PMs are not organized correctly or efficiently, it is a resource hog. For example, I was able to pull my planner, scheduler, and my maintenance coordinator out of my maintenance pool without replacing those heads. So it is just that amount of work in there. 

Phase Three, I really want to go from calendar-based to condition-based. And that doesn't necessarily mean PdM. It could be run hours, meter hours, units through there. And our CMMS has all that capability along with our PLC controls team. I really want to get out and let the equipment start telling me high enough in the PF curve that, hey, we got to start initiating some of these PMs out there. 

Now, you're not going to get rid of all of them, you know, because we're regulated by FDA, so we got to make sure we do all those stuff accurately. But that's our next phase here to really get away from monthly, because right now even the techs are saying, I don't need to do this every month. They're kind of seeing that calendar starting to really not make sense for us, so we really want to move the condition based.

TW: Thank you guys for being here today. I know you just came out of the presentation. This was like a second recap, but I really appreciate you sharing your knowledge with our readers. And I'll put links to your LinkedIn profiles in the show notes, if that's all right with you guys. So if anybody wants to contact Tony or Jeff for more information, you can go ahead and do it. Okay. So guys, thank you so much.

TP: You bet.

JS: Thank you as well. We appreciate it, Tom.