“Without data, you’re just another person with an opinion,” said W. Edwards Deming, the widely regarded pioneer of modern quality management. The quote captures an important principle of modern predictive maintenance, but collecting data is only part of the challenge. Here, Katarina Mitchell, business development manager at CP Automation, explains why the focus of our industry should move from predicting failure to understanding what the data is telling us about electrical system health.
Most predictive maintenance strategies are built around the assumption that the asset itself is the source of the problem, whether it’s a motor, drive or another piece of industrial machinery.
However, many failures are symptoms of a wider system issue, with electrical conditions potentially being a contributing factor. For instance, a bearing fault is typically treated as a mechanical issue, yet electrical phenomena can act as underlying contributors long before any vibration signature appears.
Predictive maintenance has traditionally focused on the health of individual assets, but equipment does not operate in isolation.
Looking beyond the individual asset
Consider a motor showing signs of deterioration. The natural response is to investigate the motor itself. Is the bearing wearing? Is it overheating? Is vibration increasing? Is lubrication becoming an issue?
These are all important questions, but the conditions the motor is operating under tend to be forgotten or overlooked.
Electrical conditions such as voltage imbalance, harmonic distortion and transient activity can place additional stress on connected equipment and may be a contributing factor in equipment deterioration. They may not immediately cause a failure, and they may not always be visible through conventional mechanical monitoring.
Instead, their effects can accumulate over time, gradually affecting equipment performance and reliability, thereby creating a potential blind spot in predictive maintenance strategies.
If an asset is repaired or replaced without addressing the conditions that may have been a contributing factor in its deterioration, the same problem may eventually return. The component has been treated, but the wider conditions affecting it have not necessarily changed.
From asset health to electrical health
This is why the conversation around predictive maintenance needs to extend beyond the individual asset.
Industrial organisations already understand the value of monitoring mechanical health. Vibration, temperature and other indicators can provide an early warning that something is changing, and electrical infrastructure deserves the same level of attention.
Power quality and electrical health are often treated as separate technical concerns, rather than as part of the wider reliability picture. Yet the electrical environment in which equipment operates is constantly changing.
Loads change, new equipment is installed, drives are upgraded, production patterns shift, and so on. A system that was operating within acceptable conditions several years ago may therefore be experiencing very different electrical conditions today.
The challenge is not simply collecting more data, because most industrial facilities already have plenty of it. It’s now about understanding what that data is telling us about the health of the wider electrical system and, importantly, whether those conditions could be a contributing factor in equipment problems.
Measure, Analyse, Report and Solve
This does not mean replacing conventional predictive maintenance techniques. Mechanical monitoring remains essential, and technologies such as vibration analysis and temperature monitoring provide valuable insight into asset condition.
The opportunity is to combine that information with a better understanding of electrical conditions. This creates a more complete picture of asset health, where mechanical performance is considered alongside the electrical environment in which equipment operates.
At CP Automation, we introduced our Mission to M.A.R.S. to help customers tie everything together: Measure, Analyse, Report and Solve. Measurement establishes baseline behaviour, Analysis identifies deviations and emerging stress, Reporting turns data into operational insight and Solving addresses root causes. Crucially, M.A.R.S. is a continuous cycle, with measurement beginning again to verify results and monitor changes in equipment, loads and facilities.
While predictive maintenance has rightly transformed how industries approach reliability, the next stage is understanding equipment deterioration more deeply, including the conditions that may be contributing to it.
Remember, “it’s not enough to do your best, you must know what to do, and then do your best.” If a programme can tell you what is failing, but not why, then the factors contributing to the problem may remain unidentified.
Sometimes, the factors contributing to that deterioration may not be mechanical at all. For more information on industrial power quality, visit the CP Automation website.
