Finding the sweet spot in your PM program

As I age, I find that my body's talking to me more. A twinge in the knee, stiffness, and a pop in a joint that I had not noticed before. I can ignore the signals and hope they fade, or I can pay attention. My doctor tells me to take preventive action rather than wait for worsening conditions or injury. It's the same thing that I tell plant managers. Catch it early, and the fix is likely small. Wait, and it compounds.

In maintenance, ignore the bearing noise, and you just don't lose the bearing. The shaft or the housing may be damaged. The unplanned downtime impacts production. Now it’s the downtime cost, plus the bearing and the secondary damage costs. Reactive maintenance is expensive, as is ignoring your health. Small problems manifest into larger ones.

But this is not new. Maintenance and reliability articles stress the need to become more proactive. That said, I want to bust a myth, one that may make you uncomfortable. The myth is that more maintenance is better maintenance.

With preventive health approaches, there are subscription services that run 160-plus lab tests over the course of a year to screen for 1,000-plus medical conditions. It sounds proactive. Who wouldn't want to know everything? Run enough tests on a healthy person, and you'll find something. A marker slightly out of range. A value the algorithm flags. Likely, it's mostly noise, meaning nothing. But it's on the report, and you can't unsee it. You chase it, more tests, more appointments, more worry. You've traded one risk for another, the risk of doing too much.

Odds are your PM and PdM programs do the same thing.

Walk into most plants, and you'll find the PM programs have grown like weeds. Tasks are added after every failure, every audit, and every “just to be safe” comment. Nothing gets removed. The maintenance group prides itself on 80 to 90% of its work being PM or PdM activities. The result is a schedule stuffed with tasks that don't map to how the equipment fails or the likely failure modes. We disturb otherwise stable (healthy) systems to look for problems or to overhaul them. Every intrusive PM task introduces the opportunity for infant mortality. Maybe it's a reassembly error, a contaminated bearing, or an overtightened bolt. Realize that upwards of 40% of self-induced failures are due to human error. 

It's common for maintenance managers to tell me they need more technicians. Yet, every hour that a technician spends on non-value-added tasks in a PM is an hour not spent on work that matters. Until you optimize your PM program, it's difficult to say that you have a technician shortage.

The sixth edition of the Society of Maintenance and Reliability Professionals (SMRP) Best Practices Guide recommends that the percentage of PM and PdM activity be about 30% of your work. Not 60 or 90%. It's like running 160 lab tests on a machine that doesn't need them.

Where is the sweet spot? It's not a number that you adopt. The 30% is evidence telling you that “more” is a myth. 

The good doctor doesn't run every test on every patient. They weighed the likelihood against the consequences. Their goal is to intervene when it changes the outcome. Right-size the diligence to manage the risk.

Your plant assets deserve the same approach. Ask some questions for every task on the PM and PdM sheets, such as: 

1.    Does this task address a failure mode that's reasonably likely? In the case of safety or environmental consequences, we consider less likely failure modes.
2.    Is it technically feasible to do the task using the approach or tool that you have identified? Can you identify the potential failure condition?
3.    Is the consequence of that particular failure worth the effort to prevent it?
4.    If the answer to questions one through three is yes, and the task already exists in the PM plan, have we ever found a potential failure condition, or have we not found it soon enough?

If a task can't survive questions one to three, then it's noise. Remove it. Question four determines the task Interval. Are we doing it too frequently or not frequently enough? There is an old rule of six inspections for every corrective action. In the RCM world, we do inspections on one-half of the estimated PF interval.

The goal was never more maintenance or less maintenance. It's the right maintenance for the way the asset actually fails. Your body taught you that lesson a long time ago. Listen to the signals that matter, and don't let a thousand that don't run your life.

The same plant-floor wisdom applies to the machine. The trick, in your health and in your reliability program, is knowing the difference.