That title should be a bit longer. What you need to find is the one that's best for each of your assets, the one that strikes the right balance between criticality and cost.
Maintenance, a working definition
Before anything else, we need a working definition of maintenance itself. What we need is a clear understanding of the destination before we can decide on the best route there. So, what is maintenance?
Reducing it to the bare essentials, maintenance is what we do to keep things up and running for the longest stretches possible while using the smallest amount of resources. Here resources are things like time, money, and manpower. Breaking resources into different types can be helpful, but remember that they're generally interconnected. Manpower costs money, for example. And everyone knows, time is money.
If maintenance were just about keeping things up and running for as long as possible, we could pour tons of resources into every situation and call it a day. Need to keep your family station wagon on the road? If you're not taking resources into account, hiring a NASCAR pit crew to hang out in your garage becomes a viable solution. Or, just buy a new station wagon every time the old one gets to a quarter tank of gas. Those are extreme examples, but they highlight how important it is to always consider resources when looking for the best maintenance strategy.
But before you can choose the one that's best for you, you need to understand your options.
It might be a bit surprising to see this one on the list of strategies; it's often presented as what happens when you don't have a plan at all. But it can be a viable strategy as long as you use it with the right assets. Generally, run-to-failure works well when assets are disposable or cheaper to replace than repair. The classic example of a disposable asset is the lightbulb. It makes sense to squeeze as much value out of each bulb as you can before throwing it away and quickly and easily replacing it. It's designed to be unrepairable. Some items are not designed to be disposable but are usually thrown out instead of repaired. By the time yours breaks, a newer model is already available for just a bit more than what it would cost to repair the old one. For example, what would happen if the computer screen you're using right now suddenly died? It's likely it would get replaced, not repaired.
The great thing about run-to-failure is that it doesn't require you to invest any time or energy leading up to an asset breaking. The tricky part is making sure you're only using it for the right types of assets.
This one is a personal favorite. Preventive maintenance schedules maintenance in advance to catch small issues before they become large problems. Schedules tend to first be based on manufacturers' recommendations, but over time they're fine-tuned using historical work order data.
A great way to start a preventive maintenance program is with a user-friendly, intuitive solution that makes setting and tracking PMs a snap. Cloud-based CMMS software ensures data security and accessibility.
Condition-based and predictive maintenance
Looking at these two together makes it easier to understand each one individually. In very broad strokes, they're two ways of using the same data collected from an asset.
The old-school version is the visual inspection. Departments schedule walkthroughs where technicians look for small issues before they become large problems. For the new-school version, imagine you have a vibration sensor on a fan inside an HVAC system. As soon as the vibrations get too strong, the maintenance department is notified. Or, you have a thermal imaging camera monitoring a motor. As soon as it gets too hot, alarm bells go off in the maintenance department office. The asset is locked out and checked.
In both cases, a maintenance work order is triggered when the asset falls out of a predetermined comfort zone. The fan is spinning too slowly or quickly. The motor is too hot or too cold. Condition-based maintenance is all about how the asset right now compares to a set standard.
Here you're using the data to look for trends, and then basing your PM schedule on those trends. Instead of worrying about what the fan is doing right now, you're interested in what it's been doing for the last year. You're not constantly comparing the current condition to a predetermined ideal. You're looking at the historical data and using it to make predictions. Predictive maintenance is all about how the past is going to tell you what to do in the future.
Choosing the right maintenance strategy
Now that you know the options, how can you choose the one that's best for each of your assets? You need to weigh two equally important factors: strategy cost and asset criticality.
From least to most expensive, the maintenance strategies are:
The last two can require significant upfront and ongoing investments. You need to buy sensors, have them properly installed, and then get the software to monitor them. You're also likely looking at additional training for your current staff. For predictive maintenance, you might have to bring in new people to push the data through sophisticated algorithms. And then there's the ongoing cost of replacing worn-out sensors, which makes sense because you're taking sensitive equipment and strapping it to things like high-pressure pipes under the hot desert sun.
Cost can't be your only consideration. If it were, run-to-failure would always be the best option.
So, what is criticality? It's the answer to the question, How bad would it be if this asset stopped working? What are the consequences to a breakdown? To start to think about the answer, we can look at categories of consequences and levels of severity. Remember, every industry is different, so you need to make your own industry-specific lists.
Let's look at some general ones. When assets break down, there can be negative effects on:
- Maintenance costs
Once you have the categories, you can think about levels of severity. To make things easy, we'll limit ourselves to three:
Now you need to match them up, drawing on historical data from your facility and industry standards. Take production, for example. You might decide that moderate is a production loss of less than an hour, severe is one to four hours, and catastrophic is five or more hours of lost production time. For each category of consequences, define every level of severity.
Once you've done that, you're ready to look at each of your assets and determine its criticality.
Let's walk through a quick example, an ice cream plant. One of your big assets is the refrigeration system in your warehouse. It's where you store all the finished products before shipping them to stores. What happens if that asset breaks down?
For production, you know that in the past, it's taken you a long time to fix the asset, between five and eight hours, which means you could lose everything in the warehouse. And you can't run the line if you don't have anywhere to store the new product. So, when the asset is down, production takes a catastrophic hit.
In terms of safety, it's only moderate. No one is going to get hurt if the warehouse hits room temperature. But, there are going to be giant puddles of ice cream, so there are some additional hazards.
Compliance is also where you have a lot of trouble. Because of all the rules and regulations concerning environment, failing to maintain the right temperature could cost you extra visits from an inspector.
The last one, maintenance costs, you determine is only moderate. The maintenance department can do the work itself and already keeps all the necessary parts and materials onsite. Even with some overtime, it's never cost you a lot in labor to get the asset back up and running.
Based on the above information, you'd likely want condition-based or predictive maintenance. Breakdowns of this specific asset are costly enough that avoiding them is worth the added expense.
But what about the forklifts that are used to unload delivery trucks? What if one of them breaks down? Production loss is below moderate, as is safety. There are no real compliance issues. But, forklifts tend to be expensive to repair down at the ice cream plant. You don't keep a lot of spare parts onsite, and the maintenance department often has to bring in a vendor. Criticality is relatively low, but still high enough to warrant an investment in preventive maintenance.
Before you can choose the right strategy, you need to establish criticality. And the way to do that is to start to think of categories of consequences and levels of severity. They are specific to your industry and facility, so you'll need to collect some data. Luckily, a good CMMS Software is a great way to easily and efficiently collect work order histories and crunch KPIs. If you don't already have a CMMS, now's the time to reach out and start talking with providers.