People often ask us if we do predictive maintenance. It’s understandable—who wouldn’t want to predict the future and know exactly when something is going to break?
However, predictive maintenance tools are often too good to be true, producing little of the value they promise and diminishing trust in AI technology. That’s why we avoided the buzzword and decided to call our solution asset optimization.
Like predictive maintenance software, Fero relies on machine learning, analyzing the plethora of data generated in a plant to come up with the most likely outcomes. But rather than predicting abnormal events, our goal is to tell you how to optimize the long-term health of your equipment so your plant can run more efficiently. Within 3 months of installing Fero software, the average plant increases asset efficiency by 11%.
What plants and cars have in common
If you own a car, you know that oil and wiper fluid must be changed from time to time. While these procedures can be a hassle, they’re required for the car to keep functioning.
Your manual might instruct you to change the oil every six months and the wiper fluid annually. Based on these facts, you might try to optimize for efficiency by thinking—Okay, since my oil change is due soon, I might as well change the wiper fluids, even though that change doesn’t have to happen for a few more months. That way I only have to drive to the auto shop once.
But those time periods in the manual are just estimates. They don’t take into account your actual car usage and conditions. Maybe you barely drive, in which case you only need to change the oil once a year. If someone told you with confidence that you could skip your next oil change, wouldn’t it be a huge time saver?
Similarly, in a factory, every piece of equipment must eventually be taken down for maintenance. Fero analyzes the long-term health of your machines and tells you months in advance when each one will need maintenance, allowing you to plan more efficiently.
After noting a heat exchanger degrade over time, for instance, the software might suggest that the machine will need to be serviced in six weeks. If you have a scheduled maintenance shutdown in two weeks, you know that the machine can be safely ignored until the next shutdown, letting you focus on other high-priority tasks.
You can also use Fero to learn how to extend the lifetime of equipment, delaying the need for costly repairs. Many factors contribute to degradation. Think of Fero as prescriptive maintenance—it will tell you what makes your specific equipment age faster and recommend actions to take to prolong its usefulness.
What about predictive maintenance?
The promise of “predictive maintenance” is that it alerts you when a unit will soon break down. However, the reality is that much technology marketed as predictive maintenance doesn’t give you much advance notice, and it’s often wrong.
Imagine an alert that pops up five minutes before your car needs an oil change. This is obviously not helpful. Predictive maintenance solutions, similarly, employ a technique called anomaly detection. When a machine exceeds a specified range—for example, spiking to a temperature of 250 degrees, in contrast to the usual 200—the software sends you an alert.
As anyone who’s ever seen a complex production environment knows, factories are rarely that simple. Machine temperature fluctuates based on hundreds of factors, from production volume to the season. As a result, machine learning software trained to learn "normal" behavior will detect numerous anomalies that mean nothing.
The biggest cost of a false positive (in addition to wasting your time) is that it erodes trust in the technology. In extreme cases, this can result in a cry wolf scenario like the BP leak, which happened because there were so many false positives, the engineers shut down the alarms.
Benefits of efficiency
By deploying an asset optimization solution like Fero and following its recommendations, you can expect to make your plant as much as 11% more efficient. You’re probably thinking about how this will save you money (no more reactive maintenance!) It’s also great for overall sustainability.
Think of changing the air filter in your AC unit. When you don’t do this, your unit runs less efficiently, and therefore you end up wasting more electricity. Asset optimization solutions forecast when the factory equivalent of your air filter will be problematic. As a result, you can produce more with the same amount of resources, driving sustainability improvements as well as increased profitability.