How modern condition monitoring helps prevent motor failures before they even begin
Each year, roughly 7% of the world’s nearly 300 million industrial electric motors fail. With the rise of machine learning and the internet of things, it’s become possible to catch much of that failure in advance. But that’s only half the story. With the ability to monitor motor health in real time comes the ability to catch problems before they even inflict damage.
By monitoring the physical signals your motors emit, you can closely track their health. Temperature, vibration, oil, current and voltage are among the signals that can reliably detect developing motor problems.
But until the digital age, using these signals required both manual data collection and skilled human analysis. That put a practical limit on how often engineers could query a motor’s health, and thus on how quickly developing damage could be caught.
AI + IIoT = continuous real-time insight
Enter the 21st century: the age of artificial intelligence (AI) and the industrial internet of things (IIoT). We now have the ability to apply automated, expert real-time analysis to vast amounts of continuous data collected by wireless sensors. As a result, modern condition monitoring systems can often detect problems that will lead to wear and tear before any actual damage has occurred—like an industrial version of Minority Report.
In Philip K. Dick’s 1956 short story “The Minority Report” (and the 2002 film based on it, shown here), the police identify would-be killers before they commit their crimes.
Here are three ways modern condition monitoring systems can help you keep your motors in excellent health:
1. By comparing a motor’s incoming data with a vast library of “failure fingerprints.”
That long history of human analysis hasn’t been tossed by the wayside; it’s been used to create a digital library of known patterns that indicate not only developing damage (such as bearing brinelling) but also situations known to cause damage (such as a sharp increase in load). Today’s condition monitoring systems can continuously compare a motor’s incoming data to these known fingerprints of failure in real time, so you can act early.
2. By comparing a motor’s incoming data with its own past record of health.
Even a 100-year collection of fingerprints can’t catch every culprit. So modern condition monitoring systems use machine learning algorithms—a.k.a. neural networks—to discern when a machine is starting to act strangely, even if the new pattern still falls within normal tolerances.
3. By providing an up-to-date picture of machine health with no gaps in the data.
Modern condition monitoring systems can sample their underlying signal, then analyze it and flag changes, thousands of times an hour, all day every day, year in year out. The advantage over periodic inspections is enormous. It’s like the difference between wearing a continuous glucose monitor and manually pricking your finger: just as three-a-day finger pricks can miss recurring spikes in your blood sugar level, weekly or monthly motor health measurements can miss spikes in torque, voltage unbalance, and other recurring but transient events that silently stress the motor until irreversible damage occurs.
Thanks to these three factors, modern condition monitoring systems can help you significantly reduce downtime, use less energy, save money and extend motor lifetime.
So which condition monitoring signal should you use to monitor your critical electric motors?
The perfect-world answer is “all of them.” Each technique has its strengths and weaknesses, related not only to the type of failure you want to prevent but also to the motor’s environment and the process it operates in. In practice, of course, you don’t want the cost and complexity of using ten tools where two would suffice. So we’ve written a comparison guide to help you start choosing your best technologies. It’s a detailed guide to five of the most reliable 21st century condition monitoring techniques. We hope it helps you stop your electric motor killers before they even get started!
The condition monitoring comparison e-book for induction motors
Learn how current, voltage, oil, sound, vibration and heat can give you early insight into developing induction motor failures—and which techniques work best in which situations to keep critical induction motors 100% up and running.
This e-book covers:
- the anatomy of an induction motor
- condition monitoring in induction motors
- examples of when different techniques will detect different failures
- energy and performance insights
Fill in the form to download the e-book.
Resources83 See all resources
We wanted to help our clients keep their pumps operating at their best efficiency points in the first place. So we’ve developed a tool to do just that. Meet SAM4’s new real-time pump performance dashboard!
In this report, we explain exactly how SAM4 detects common faults in industrial pumps using actual results from anonymized SAM4 data, to help engineers evaluate SAM4’s value for their own operations.
- Technical documents
- White papers
Learn how SAM4 reduces unplanned downtime in critical water and wastewater processes.