“But we have always done it like that!” That is what I recently heard from a maintenance strategist, as we were discussing concepts for optimising maintenance activities. I also often hear the following statement or something similar: “I know the new maintenance strategies. But I already know how to keep my equipment in good condition. It is common sense for people and engineers after all.” Fair enough, but can one really leave it standing like that in all cases? Here is an overview of the maintenance concepts that are currently circulating through the industry:

Boredom as a Goal

If nothing happens, we are doing a good job. The best infrastructural operator is invisible. This way, one could reduce the target objective of our industry to a common denominator. It is, however, not possible to directly measure the influence of a good maintenance strategy. The reliability of the equipment depends on many factors. Mr. Coincidence is also on board and is trying his hardest to make life hard for the statisticians.

If one were to for example, measure the SADI – meaning the average outage duration – in the electric power grid, the reaction speed of the interference suppression plays a big role. And this has nothing to do with maintenance after all.

In the end, only two factors matter. Firstly, what do I do to keep it from impacting? In other words: how much budget do I stump out for the maintenance fun.  And secondly, what happens when it does bang? How big are the consequences and what do I want to allow myself in terms of malfunctioning?

Types of Maintenance – Courage Counts

There are various approaches, which differ in how much courage one is ready to muster for them. “Courage in the infrastructure? Mister Forester, you have courage on a roller coaster, but not in our electric supply mains!” I can already hear the commentaries.

Be it the laissez-faire approach, with which one only PUTS the equipment in good condition, but does not KEEP it in it. Or if one predicts the future condition with algorithms, so that outages should be impossible. Basically, there is no limit to the creativity. But that is exactly what makes it both difficult and fun, right? Because the perfect maintenance strategy that is identical for every company or piece of equipment does not exist. Cookbooks are there for cooking shows, but now for the maintenance strategy. One has to combine the ingredients anew all the time.

Oriented on Outages – the MacGyver of Asset Management

So, there is the laissez-faire approach, which really only reacts to malfunctions, but does not do anything else. The superhero jumps off with the parachute and fixes everything. That this seems rather uncoordinated and is not predictable is obvious. The condition of the mains is not known to the infrastructural operator, so the error tracking can turn into a search for a needle in a haystack. In the case of several simultaneous malfunctions, this can easily become a bigger problem, which means it is obvious that one does not use this method for any old piece of equipment. Especially compartmentalised mass merchandise, for which an outage or a longer unavailability does not carry any weight, would seem the equipment to use this strategy on. Just think of your blender. There, you only think about a repair if it conks out. And MacGyver also does not put on the parachute for the blender.

Time-based – the Derrick of Asset Management

Boring but enormously effective – that is time-based maintenance. In many places, an outage-oriented maintenance “Light” is often used in the form of regular maintenance work in practice. In the end, this only means that maintenance measures and, if necessary, repairs (almost) always take place on fixed dates. This can be described best with the temperature sensitivity of a typical John Doe in our region. The whole year around he does not care about the heater, but when it malfunctions on Christmas, when the in-laws are visiting, he gets really frustrated. To keep exactly that from happening, preventive maintenance contracts are concluded with the heating engineer of your trust, who then comes over once a year to maintain the system. If it then still malfunctions on Christmas Eve, at least you will know who is at fault.

Condition Oriented – Big Bang Theory in the Asset Management

Ultimately, one does not know how the equipment is doing with the last two maintenance strategies. The expert of the world calls that condition. Therefore, from time to time, it does make sense to know when something needs to be done. An outage, or rather an unavailability, cannot be put up with for every piece of equipment, after all.  However, the survey and regular monitoring of the condition is elaborate, because inspections have to be done on site or in modern times even via remote diagnostics through sensors. Just think about smart fridges that can tell you if the milk has gone bad or if someone left the door open again. Today, equipment in the infrastructure can do that as well: the transformer lets you know if it is too hot or if it feels overloaded. This needs a bit of thinking and a lot of experience. Nevertheless, Sheldon Cooper achieves good quality values at controllable costs with this strategy.

Reliability Oriented – Now It is Getting Spacey

So, not every maintenance strategy can be used par for par for every piece of equipment. Some are simply more important than others. An unavailability has far stronger consequences and can, in an extreme case, lead to a complete outage of the overall system or to high follow-up costs. Just think about your kitchen with its many electrical gadgets. The outage of the fridge with perishable food in it is worse than the outage of the oven. Its repair, on the other hand, is preferable during a shortage of resources (or the other way around depending on taste). You can see my point: Under certain prerequisites, certain pieces of equipment can be prioritised according to their importance. And that is a good thing. In some infrastructures, the assessment of importance is even more extensive than in one’s own kitchen.

Forecasting – Madame Seraphinas Crystal Ball

In times, in which the computer is smarter than Sheldon Cooper, maintenance can, of course, also be viewed mathematically. For this, one does not necessarily have to know the condition of a piece of equipment at all times. It is hard to believe, but such statistic processes work even better than the gut feeling of some experienced masters. “Mister Foerster, that’s enough! Nothing against our assemblers!” No, nothing against the masters and assemblers; I truly believe they are the backbone of our infrastructure. But maths is not bad either. Predictive Maintenance is a promising trend in the maintenance. And it does not only have to find its place in the infrastructure: from the intelligent pacemaker to the middle-class industrial enterprise up to the municipal works, Predictive Maintenance can find its entry. The idea is to collect a lot of data and calculate many correlations. This way, it is possible to precisely predict the equipment condition. However, one does not know, which correlation was the deciding one in the end – but it is not important, either. Just put it all into the pot, add some seasoning and somehow it is going to taste well. We are only at the beginning, but I am sure that we are going to see a lot more here.

And the End of the Story?

There are many different maintenance strategies and thus many different possibilities. Some of these strategies are used in a private environment, without anyone even really noticing it. This can obviously also be carried over to the infrastructure. Only because it has always been done one way, does not mean that one has to be satisfied with the status quo. Soon I will speak more about how I can find the right maintenance strategy. Together with my company meliorate, I lend my support for these questions to owners and operators of infrastructure since 2011.

FURTHER BLOG ENTRIES BY OLIVER D. FÖRSTER

Back to the Future: Predictive Maintenance

Indiana Jones and the Optimal Maintenance Strategy