Choosing the right strategy for each of your assets may seem like a daunting task, but all you really need to make the right choice are a couple of of core concepts. And then all the time and effort you invest will pay off in reduced costs and less stress.
There's always more than one right maintenance strategy
And that's because you have more than one type of asset, and it's the type of asset that determines the maintenance strategy. You need to pick the strategy that best matches each of your asset types, so you almost always end up with more than one strategy. It's the same as matching the medicine to the illness. Chicken soup is great when you have a cold, but it's useless against a broken arm. My personal favorite is thinking of maintenance as a type of insurance. Both are about risk management. The higher the risk, the more insurance you need.
Before deciding which strategies you need, you first need to know your options. Let's do a quick rundown of the four main maintenance strategies.
Run to failure
This one tends to have a bad reputation because on the surface it looks like it's exactly the same as having no strategy at all. When you look closely, though, you see that it involves plenty of careful decisions and lots of planning, both for before and after breakdowns.
Generally, it's great for assets that are cheap and easy to replace or repair. Also, when these assets breakdown, there's no safety risk and they don't really hamper production. The classic example is a light bulb. As long as you have extras in inventory, you can quickly and cheaply replace the burned out ones. Run-to-failure is also good for assets that have a random failure pattern. This means that the chances of them failing are equal throughout the life-cycle, making it very difficult to predict and schedule maintenance, and looks like this:
For these types of assets, there's always an equal chance of them failing. They're the opposite of the tires on your car, where the longer you use them, the more likely they are to fail. With tires, you can set up a schedule to rotate, repair, and replace them. Tires match the fatigue failure pattern, which looks like this:
You can also use run-to-failure for assets that are expected never to fail. Back to the example of a car, you would use run-to-failure for the car's frame, which is not really designed to be maintained. Hopefully, you never have an accident, but if you do, that's when you fix it. Otherwise, there's nothing you can really do to maintain the frame.
It's important to once again clarify that run-to-failure is not the same as not having a strategy at all. To use this strategy properly, you need to carefully decide which assets to use it on. Then you need to make sure you always have the right parts and materials in inventory. On top of that, you need solid plans for what to do when there's a failure. Basically, you need prewritten work orders that get assigned as soon as failures occurs. Therefore, unlike a complete lack of strategy, run-to-failure, when done correctly, greatly benefits from CMMS software. You need it to track and maintain your parts and materials inventory and hold data-packed work orders that can be assigned within a few menu scrolls and clicks. If your CMMS has a lot of historic work order data, you can even use it to decide which, if any, of your assets you want to switch over to run-to-failure.
Easy to explain to technicians; something breaks and they replace or repair it
Relatively lower upfront and ongoing costs
- Hard to schedule around. It's challenging to have the right people on hand because you can't predict when breakdowns will occur.
- Added cost of always carrying the required parts and materials; because you don't know when you're going to need the inventory, you always need to keep it on hand.
Quick warning about the above pros and cons: they apply only when the strategy is being used correctly and for the right types of assets. If you use it where you shouldn't, all the cons are going to be much worse and the pros disappear completely.
Just like the name implies, this strategy is all about preventing issues from developing into larger problems. Instead of waiting for an asset to fail, you set up a regular maintenance schedule which allows you to get all the necessary work done ahead of time, when it's more convenient, cheaper, and easier for you. Instead of replacing a busted fan belt midway through the third shift, which means pulling in technicians for overtime hours and having to explain to the front office why a bunch of operators were idle for for an hour during a busy production run, the belt gets changed before it breaks, in the afternoon between shifts.
- Reduced downtime thanks to catching small issues before they become budget-busting problems
- Easy to schedule around. Because preventive maintenance is planned in advance, you have the right people and parts when you need them.
- Requires upfront and ongoing investments of time and money (but they pay off, in the end, and you should be actually enjoying a nice ROI)
- Possible to stray into over-maintenance, because you're working on assets based on time or usage, not on whether they really need it or not.
To find out how a good CMMS makes preventive maintenance possible, read 5 Ways CMMS Gets You Laser-Focused on What Matters.
Condition-based and predictive
Because they're closely related, it's easiest to look at these two together. For condition-based, you start by setting up a way to consistently collect data from an asset. The classic walk-through is a good example. At set intervals, technicians walk through the facility looking for early signs of problems, anything that's out of place. The key is they need to know what everything looks like when it's "in place." The hydraulic press might be making noise at the end of every cycle, but it's only going to be investigated if the technician has never heard that particular before. In the new-school version of condition-based maintenance, different types of sensors are attached to assets, and you've moved from data collection that's consistent to data collection that's constant. Data about vibrations, temperatures, fluid levels, and acoustics stream into the CMMS, and as soon as data from an asset slips outside of a predetermined safe zone, a work order is generated.
For example, your wielding station ventilation relies on a large central fan. As soon as the sensor detects too much vibration, the CMMS generates a work order to have it checked and rebalanced. Because you're only fixing things when they need it, you're not wasting time and energy over-maintaining them. Assets are fixed when they need it. Or, to put it more accurately, they're fixed right before they need it. The fan hasn't failed yet. It's still running. But the change in vibration means a current small issue is heading toward being a larger problem.
For predictive maintenance, you start off with the same constant collection of data. Sensors pour data into the CMMS, but this time it's not looking for variations outside a predetermined safe zone. Instead, the CMMS pushes the data through complex algorithms to then make accurate predictions on when the assets are going to fail. Your CMMS becomes a crystal ball and you're able to predict each asset's future. And because you know when each one is going to fail, you can fix them right before they do.
- Sensor-based condition-based maintenance can make technicians' lives at lot safer. It's not easy climbing to the top of a stack to check on pipes, so a well-placed sensor is going to make life a lot easier. Companies are also looking at drones as a way to inspect hand-to-reach places.
- Having a really solid idea of when your assets are going to fail means being able to run right up until just before they do. You're only doing the maintenance that's needed.
- Condition-based can be as cheap and simple as using a dipstick to check the oil on a motor, but it can also be as complicated and expensive as buying and installing delicate sensors, with initial costs for training and installation followed by ongoing costs for recalibration and replacement.
- Because of the associated costs, only makes sense for mission-critical assets.
Now that we know our options, let's look at how to choose the right combination.
How to choose the right maintenance strategy
A big part of choosing the best strategy for each asset will be determining its criticality. The more critical the asset, the more it makes sense to invest time and money in protecting it. Remember the analogy from earlier: maintenance strategies are like insurance. They are ways of mitigating risk. And just like with regular insurance, the item's value dictates the best type of coverage.
But what, exactly, is criticality?
Criticality summed up in one small chart
Let's stop using the word "value" and instead use "criticality," which is more accurate. On a basic level, criticality answers the question "How bad is bad?" And we can answer it, on a basic level, by looking only at the associated costs of a failure, including:
- Lost production
- Spoiled or damaged feed stock
- Wages for idle operators
- Overtime for inside technicians and payment for third-party vendors
- Rush shipping for materials and parts
The more it costs you when the asset fails, the higher its criticality.
By looking at cost and criticality, choose your strategy
And then based on the criticality, you choose your strategy. But, is criticality really that simple?
Criticality summed up in one bigger chart
On a basic level, criticality as the amount of money we lose when an asset fails. The larger the cost, the higher the criticality. But we can expand that to include all kinds of other variables, giving us a much deeper understanding of an asset's criticality. Look at the chart below.
We can see it's more than just money. A part might not cost a lot to repair or replace, but its failure may lead to serious safety risks for your team. Think about the brakes on a forklift. They're not especially expensive and any junior technician can likely change them. But what happens when they fail? Certain failures may lead to environmental damage, a safety risk to your whole community. Once you start including more variables, you might find assets' criticality moving up or down. But the general idea remains the same. The larger the risk, the more insurance you need.
To make the process of establishing criticality easier, get yourself a good CMMS. It makes collecting and leveraging data much easier. When you want to see the maintenance big picture, there's no more shuffling stacks of papers or randomly clicking through conflicting versions of spreadsheets. A good CMMS keeps the data safe for you, and then helps you use it with automatically generate reports packed with KPIs. Instead of guessing how much an asset costs you when it breaks down, you can find the exact numbers. Historic work orders are full of useful data, including how much time technicians spent working on assets and the costs of the associated materials and parts. Then once you've made your data-backed decision about which combination of maintenance strategies to use, the CMMS makes applying them a breeze. Track inventory for run-to-failure. Set and never forget PMs for a robust preventive maintenance program.
If you don't have a CMMS, start reaching out to providers. When watching demos, make sure to pay attention to ease-of-use. All the bells and whistles don't matter when they're hidden behind a frustrating UX design. If you can't find a feature, what's the point of paying for it? When you can't find the keys, even the best car becomes the world's most embarrassing lawn ornament. Also, make sure you look at what they deliver for on-boarding and ongoing support. A good provider has a long list of ways they're invested in your success. Make sure to ask them about it.