Share this post
Hippo provides a comprehensive and easy yo use CMMS for streamlining your asset management
Although the theoretical concepts behind prescriptive maintenance are easy to understand, setting up a program comes with more than a few practical challenges.
Which means that, for some, implementing this maintenance strategy is the logical next step in their digital evolution, while for others, it's the sort of thing they should know about but not worry about having to adopt just yet, if ever.
For a lot of maintenance professionals, prescriptive maintenance is much more than they currently need, and the strategy would have a negative return on investment: You'd be paying more for the maintenance than you'd be getting back in added uptime and cut costs.
That said, it represents impressive developments in bleeding-edge maintenance technology, and as prices come down over time, more and more departments could find room for it in their toolboxes.
So, after all the hedging and qualifying, let's get down to solid definitions and examples.
Prescriptive maintenance takes predictive maintenance one step further. Instead of telling you when assets or equipment are likely to fail, prescriptive maintenance can show you how specific changes affect the outcome. And because it knows how each possible change can affect the outcome, the software can make solid recommendations about what you should do to get the most out of your assets.
Just like a doctor prescribes medicine, the software behind prescriptive maintenance can show you your best next maintenance move.
Remember, predictive maintenance relies on a series of asset-attached sensors collecting large, continuous streams of data from assets, which then feeds into sophisticated software, which then tells you when failures are most likely to happen. How does the software know? It works with sophisticated algorithms and a lot of both historical and current data. Prescriptive is doing a lot of the same things, but then also more. The software is constantly pulling in data and then using it, with artificial intelligence and machine learning, to make itself smarter.
So, predictive shows you the future. But prescriptive can show you multiple futures, each tied to a distinct set of operational conditions. Instead of determining that your asset is likely to fail in X number of days, with prescriptive, the software determines different failure days using different parameters.
What do all these hypotheticals look like in practice?
Imagine you have a pump that cycles X number of times an hour. You have it covered in sensors, catching data related to temperature and vibration.
With prescriptive maintenance, you can see when it's most likely to fail. But with prescriptive, you can see multiple possible failure points based on specific changes to the number of cycles per hour or the number of hours it's online a day or some other variable. So, at your current rate, the failure is three weeks away. But if you cut the cycles by 25%, you get an extra week. If you run an hour less each day, you get a few extra days. You can also use the software to look at the effects of different maintenance tasks.
Another example: based on sensor data, the software figures out that a motor in your manufacturing plant is set to fail within the next week. It then checks to see when you have that motor scheduled for preventive maintenance inspections and task because it wants to know if the motor can make it to the PMs without disrupting production. Once it decides that it can't, it then looks at other options. Does it make more sense to stop the line now and replace the motor? Or, would you lose less production time overall by slowing down the line to 75% of regular output, decreasing the strain on the motor so it can limp its way to the PMs?
Once you've worked out all the possible futures using the software, you can decide which steps to take, based on the software's recommendations, to move yourself down the right path.
Prescriptive maintenance is a bleeding-edge maintenance technology that takes you even further than predictive maintenance. Just like predictive, it used asset-connected sensor data and sophisticated software to determine when your next failure is most likely to happen. But it also takes things a step further, giving you a view into multiple possible futures, each connected to a different set of operational conditions. So, while predictive maintenance can help you see when as asset is set to fail, prescriptive maintenance can help you see how changing how you run and maintain an asset affects predicted failure dates. And using those insights, you can determine how best to both run and maintain your assets. Although this maintenance strategy promises to revolutionize modern asset management, it is still more than most maintenance departments actually need. Because of the large upfront and ongoing investments in sensors, software, and training, for a lot of maintenance departments, it would create a negative return on investment.