Podcast: Reliability reality check—why more PMs and high wrench time don’t guarantee reliability

Now, you may think that's extreme, but I'm telling you, I find something like that in every plant I go to. They'll make a PM and just inflate it to eight hours when it's really four hours, and then they're supposed to get emergency work done in between doing the planned work, and that fills up the eight-hour window.

So managing KPIs is really a common practice. I hate to say that it's a common practice, but, you know.

If we go on to increasing PMs—let's say they're precision, let's say they're at the right time—I was actually trying to push down my PMs. I was trying to do more condition monitoring because this rebuilding of a gearbox—why are you even doing that? Why can't you find that from vibration, or infrared, or oil sampling?

I wasn't really trying to manage PMs. I was trying to manage planned work, but not PMs.

So I'd also say you've got to have problem solving. If you get really good at doing PMs, you're going to be horribly disappointed with the impact you're going to have on the bottom line, because you're not doing problem solving.

PS: Okay, KPIs—I know we have an episode way back talking about how people fake their KPIs and some of the ways you can lie through the numbers. So just want to throw that one out there for everyone. But onto scenario number two.

So my client has done all the necessary PMs. We're not doing too much planned maintenance, we're not doing too little. So Joe—myth or reality: if the PMs are completed, the equipment will always be reliable?

JK: This has to be a myth. You never arrive at the perfect maintenance plan. You never arrive there. You're constantly, like I said, trying to replace PMs with predictive maintenance. You're trying to solve problems and not do more PMs.

You’ve got to have precision and problem solving inside of a program, and you must get production involved. So just doing more PMs and completing more PMs and being at the perfect spot—that's a false reality. You're never there. You never arrive at the top of the mountain and can sit in a lawn chair because everything's perfect—that’s not reality.

PS: That makes sense. We're never going to find perfection. Alright, scenario number three: my plant has finally graduated to the digital future. We've let go of our paperwork orders, and we’ve got a brand-new CMMS, or computerized maintenance management system.

Joe—myth or reality: a CMMS will automatically improve maintenance performance?

JK: Yeah, it's kind of a trick question here. I would say it's going to improve maintenance performance, so I would say true—but to what degree?

A digitized CMMS system is going to eliminate some waste in your plant. Just how you get work orders in, how you track those, how you close them out—you're going to be naturally better than with a manual system. So yes, this is true, but I'll put a small “t” on that.

The problem is, people think or it is sold to you as the end-all, be-all. You're going to be horribly disappointed with the overall impact. It does allow better problem solving as well. I think it's a start, and it's not necessarily—I hate to say this—this is not where I would start if I had a 100% reactive plant.

I would not start with a CMMS system. I would try to get some early wins around doing better lubrication, better motor management. What are my problems? Trying to get control of reactive work. And then this is how to get to another level—maybe a year or even two down the road.

So many people want to start with something they can buy, okay, instead of working on culture. Culture is always hard to work on, but culture is where all the results are. This is nice to have. It's true with a small “t.”

PS: And this is still just a tool. It may give you some initial efficiency, but long term it’s going to come down to how you use it and what is the culture behind it, as you said.

Okay, so last up—scenario number five. My plant is very focused on KPIs. Our executive team is almost solely focused on numbers, so we track everything. Joe—myth or reality: high wrench time equals high productivity?

JK: High wrench time equals high productivity—I would say no. I, again, am very cautious of KPIs. KPIs, I would say, would be a great thing for the viewers to audit. Do the KPIs you're tracking match your reality that you see on the shop floor?

I've had more bad experiences—maybe I'm too negative on KPIs. Maybe they're a little better today, but not in my experience, not in the 41 plants that I've been into.

When you mentioned wrench time specifically, I can't tell you—I wish I tracked this—but it's almost 100% of the 41 plants I've been into, they'll tell me that the wrench time is 40%, 45%, or 50%. And then I organize an audit of that, and I've never, never, never come out of a plant on day one with a higher wrench time than 15%. It has never happened.

So they think they're at 45–50%, and I happen to audit them over one or two days, and it's less than 15%. Now, they usually give excuses for that—this was an unusual day, you came on a bad day—but I always come on bad days. So a lot of times people don't know how to measure wrench time. They assume that this person was getting ready to work, or if they would have had that tool, that would have saved them an hour. And I say, this is what we know. What did you do today?

Wrench time is a scary one for me because almost universally, people have been wrong in their calculation and estimate of wrench time—100% of the time. So I'm very skeptical of that one.

PS: Yeah, or they're inflating those PMs to give them better numbers. Okay, and I'm sorry I skipped one there—we have five today, and I said that was number five, but we do have one more.

So our last scenario today: my plant is doing effective planned maintenance. We're tracking it appropriately in a CMMS. But no matter what, we still have a lot of emergency work. Joe—myth or reality: emergency work is unavoidable in manufacturing?

JK: Yeah, technically, this one is absolutely true. You're going to have it, and it's smart, actually, at times.

The analogy people like to use is the lights in your building—you really wait till they go out. If you have 10 lights in a room and one goes out, then that's run-to-failure. So that's when you replace it. That's an easy example that everybody understands.

But where it gets a little more complicated—this happened to me at our power plant—we had a pump. We had five pumps pulling out of the Ohio River here for cooling water, condensing water, and the pump went out after 30 years’ worth of use. Okay, is that acceptable? Yeah. Do you need to root-cause something that's happened once every 30 years? Probably not, especially if you have an inline spare.

We didn't need all five of these pumps to work, so that's an acceptable answer. We've got a great design here It's highly robust, and it fails every 30 years. That's a good strategy. That's the best practice, I think.

But don't take that as an excuse to have 90% emergency work. You must relentlessly pursue it. But if you've got a pump that fails every 30 years, you probably have lower-hanging fruit—something that happens once a month, once a week, twice a year. Go after those things.

It's relentlessly pursued, but acceptable in a lot of cases.

PS: Yeah, a 30-year life cycle is pretty good. If you've got your equipment lasting that long, you'll take that emergency for sure.

JK: Absolutely. But our reliability program—the rules we had—triggered that to do root cause analysis on it. Now, I had to stop it because it made no sense. That's where a reliability program can go wrong, when you try to create firm rules and you just didn't think of this scenario.

PS: Yeah, make sure you're using common sense too, as we all say.
Well, Joe, thanks for playing our new game. I think it was a success.

JK: It was fun doing the rapid-fire questions.

PS: Right, I was going to say—great job on your succinct answers. I know we like to go very deep sometimes, but we tackled a lot of really great topics today, so I think that was good.

About the Podcast
Great Question: A Manufacturing Podcast offers news and information for the people who make, store and move things and those who manage and maintain the facilities where that work gets done. Manufacturers from chemical producers to automakers to machine shops can listen for critical insights into the technologies, economic conditions and best practices that can influence how to best run facilities to reach operational excellence.

Listen to another episode and subscribe on your favorite podcast app