Stay Ahead with MOVUS Insights! Get the latest updates, tips, and industry news delivered straight to your inbox.

Luck is not a maintenance strategy

By Malcolm Schulstad and Aliesha Aden 

I was in a glass plant a few months ago when the Head of Maintenance told me a story about something that underwent what he called a “rapid disassembly process.” 

A fitter had heard something that wasn’t normal on a fan assembly. He approached the asset and placed his hand on the housing, using the old diagnose-by-feel technique. 

He only touched it briefly. 

The moment he stepped back, the fan assembly, which was running at the time, self-destructed. He had just enough time to get clear. 

The phrase was delivered with dry humour… But the reality behind it was not funny. If he had stood there one second longer, the outcome could have been catastrophic. 

In another case at a cement plant, a fan almost four metres in diameter failed violently. When it let go, it tore through its housing and travelled down a roadway, damaging equipment and putting lives at risk. 

In both cases, no one was injured. That was luck. 

The uncomfortable truth is this. Neither event needed to happen. 

 

From “Feel and Listen” to Data and Foresight 

For decades, maintenance relied on experience, sound, smell, and touch. Skilled tradespeople walking the plant floor, listening for subtle changes in pitch, feeling vibration through a housing. 

That expertise still matters. But standing beside a high-speed rotating asset that is already trending toward failure is not a safety strategy. It is exposure. 

Traditional condition monitoring improved detection, but often still required physical confirmation. People were sent to stand next to live equipment to check what the data suggested. 

Prescriptive maintenance changes that model. With sufficient sensor coverage and high-resolution data, modern industrial IoT systems can: 

  • Detect deviation from normal behaviour early 
  • Diagnose likely failure modes remotely 
  • Recommend specific maintenance actions 
  • Estimate urgency and remaining operating window 

The first response to abnormal behaviour no longer needs to involve someone walking up to a live machine. 

 

Keeping Humans Out of the Hazard Zone 

The goal is not to replace maintenance teams. It is to protect them. 

When imbalance, bearing degradation, looseness, misalignment, electrical anomalies, or process-driven overload are identified early, something powerful happens – it creates time. 

Time to plan.
Time to isolate safely.
Time to schedule maintenance properly. 

Instead of reacting to a machine that could be minutes from violent failure, work can be completed: 

  • During planned shutdowns 
  • With full lock-out and isolation procedures 
  • With appropriate tools and resourcing 
  • Without the pressure of escalating breakdown 

When maintenance becomes planned rather than reactive, repair quality improves, secondary damage reduces, spare parts usage becomes more predictable, and production disruption decreases 

Safety and efficiency stop competing. They align. 

 

Accuracy Is the Difference 

None of this works without accuracy. False alarms erode trust and missed failures undermine confidence. 

Effective prescriptive maintenance depends on robust sensing, correct placement, sufficient data resolution, and algorithms tuned to real industrial environments. 

When implemented correctly, the system does not just say something is wrong. It identifies what is wrong and what should be done about it. 

That clarity is what allows organisations to confidently keep people away from dangerous environments, rather than sending someone in just to check. 

 

Designing Out Exposure 

At its core, this is risk engineering. Every time someone stands beside a high-speed fan, a conveyor drive, a mill, or a crusher trending toward failure, there is exposure. 

If technology can reliably tell us: 

  • This bearing is degrading 
  • This rotor is becoming unstable 
  • This motor is drawing abnormal current 
  • This asset has days before critical failure 

Then we can design work differently. We move from reactive intervention to structured prevention, emergency response to controlled execution, and from luck to predictability. 

No one should have to rely on stepping back just in time as their primary safety control. 

If we can detect, diagnose, and prescribe action early enough, we minimise the chance of another so-called rapid disassembly process. And more importantly, we make sure the only thing coming apart is the machine, safely isolated and intentionally dismantled. 

 

If you are still sending people to stand beside live equipment to confirm what might already be visible in the data, there is a better way. 

Prescriptive monitoring is not about more dashboards. It is about reducing exposure, improving planning accuracy, and making sure interventions happen on your terms. 

If you’d like to understand how this could apply to your assets, we’re happy to walk through it with you.

Ready to start monitoring?

Talk to us and find out how you can get started for less than a cup of coffee a day

Sign up for the MOVUS newsletter and stay informed on how to maximise uptime and reduce unplanned failures.

Chat To Our Team Today

Contact MOVUS today to discuss a condition monitoring package tailored to support your critical assets and end-to-end operations. 

107 Milton Rd, Milton QLD 4064