Maintenance And the Butterfly
Klaus M. Blache | June 15, 2018
Q: Why is standardized work in maintenance so important?
A: Maintenance professionals pride themselves on being able to do a job without direction. They are the experts in machinery and equipment repair, so what’s the concern? As I say often, “It’s all related.”
Some of you may remember the 1990 movie “Havana,” in which Robert Redford states, “a butterfly can flutter its wings over a flower in China and cause a hurricane in the Caribbean.” That statement suggests that, if a weather system is in perfect balance, a slight flutter can decide direction.
Although memorable and poetic, scientists know that the butterfly example, involving such a small air movement, is just bad physics. On a more comparative scale, this starts to make more sense. For example, if a large boulder is perfectly balanced on top of a mountain, throwing a rock at it can decide the direction of fall.
Mathematician Edward Lorenz developed a model in the 1960s that, when mapped, looked like a butterfly. Lorenz found that the resulting shapes were extremely sensitive to initial conditions (minor change in starting point). Minor shifts in any direction mapped a completely different butterfly. This mystery of the butterfly effect evolved to Chaos Theory, which studies the behavior of very sensitive dynamic systems. Because of the sensitivity to these very small differences, it’s extremely difficult to predict the outcome of these complex systems.
Chaos Theory, like reliability and maintainability, is about getting ready for the unexpected. If you don’t follow standardized work processes, it’s much more difficult to attain stability. What presumed little things go on in your maintenance department that can have a “butterfly effect” on your organization? Consider trades/technicians who:
• take verbal job requests that should be formal work orders
• sometimes get to the job site and discover that operations personnel won’t allow the machine to stop
• receive unclear maintenance tasking direction, so they improvise
• see most jobs prioritized as High Priority or Safety as a way to get them to the top of the list
• skip less-critical PM checks because many can’t be finished in time and/or are on the monthly schedule (with these tasks often reported as completed, thus making the KPIs look good).
• struggle to find the correct part because all parts are not coded with a standardized system, a problem that also affects reordering/procurement
• learn that only some of the asset history is captured because work orders are not closed with enough detail, or at all
• hoard parts because they do not trust stock-room data
• do not perform root-cause, trending, and reliability-growth tracking analyses because they lack quality asset data
• see continuous backlog growth, assuming you are counting all work orders not performed
• are unable to predict costs, other than that they continue to go up
• doubt data validity/accuracy, though KPIs look good (for the most part looking at daily plant-floor practices reinforces their distrust).
Here’s a good example of such problems: PM compliance is reported at 95%, but those that are doing the PMs and repair know that proper PMs are only being done on 60% of the assets. Are you reporting compliance on 100% of your assets? The same is true for predictive technologies, meaning only routes identified are being tracked (so the numbers look good). Nobody talks about what percentage of assets have proper technologies and routes.
Done once or a few times, any of these behaviors may not have a significant impact. However, cumulatively (done individually or by different persons), they can cause a significant negative outcome in reliability and maintenance performance. It’s all related.
Maintenance, just like production, needs standardized work processes to enable best practices. For example, three trades/technicians (at various levels of experience) should be able to perform the same work order and achieve the same outcome. If not, more clarity is needed in your PM tasking.
“Doing the same thing over and over again and expecting different results is the definition of insanity,” (author unknown, although Einstein often gets credit for this quote). With maintenance needs being so unpredictable, and 75% of North America not doing enough predictive/condition-based monitoring (you can’t fix what you don’t find), you are just fixing the same stuff over and over again, at different times.
If this is your operation and you think that it’s going to make things better, you’re feeding insanity.
Many of the sociotechnical systems (people and processes) in which we live and work are chaotic and complex. By better understanding how it’s all interconnected, we can avoid major issues. By doing a better job of performing standardized work, accountability, precision maintenance, and so on, some of the unpredictability (and unwanted consequences) can be removed. But you first need to understand how it’s all related. EP
Based in Knoxville, Klaus M. Blache is director of the Reliability & Maintainability Center at the Univ. of Tennessee, and a research professor in the College of Engineering. Contact him at firstname.lastname@example.org.