Predictive Maintenance

Predictive maintenance (PM) is a cost-saving and performance-based strategy that predicts when an equipment failure is likely to occur and performs planned maintenance to prevent the equipment failure.  

Essentially, predictive maintenance involves direct monitoring of the operating condition, efficiency, heat distribution and other indicators of assets to determine the actual loss of efficiency that would be detrimental to the operations of critical systems in the facility.

The goal of predictive maintenance is to decrease the frequency of maintenance, maximize the useful life of assets, and reduce unplanned reactive maintenance that is often performed at a high cost.

Consequently, you save equipment maintenance time, cut down the cost of parts and supplies, and reduce production hours lost to maintenance.  However, these cost-saving benefits come progressively after your initial spend on, sometimes, expensive condition monitoring techniques–including needing a specialist for data analysis.

Predictive Maintenance Techniques

Predictive maintenance requires a technique that accurately predicts failure, offers a significant warning period before planned maintenance. Determining the accurate method to monitor the condition of your equipment is critical and should be done by your maintenance team in consultation with condition monitoring experts and equipment manufacturers.

Common PM techniques include: wear particle analysis, infrared thermography, oil analysis, vibration analysis, equipment observation, and motor circuit and signature analysis.

When to do predictive maintenance?

Predictive maintenance produces the best results for equipment that have failure modes that can be condition monitored cost-effectively, and for equipment critical to operations.   

However, predictive maintenance is unsuitable for equipment that is not critical to operations and with no failure mode, or those with failure modes that are expensive predict.

Checklist for effective predictive maintenance

• Is the asset critical to operations?
• Does the asset have predictable failure modes?
• Is the condition monitoring tools for the asset’s failure modes cost-effective?  
• Do you have highly skilled and experienced personnel to analyze condition monitoring data accurately?
• Overall, does the asset pass the reliability centered maintenance assessment?

If you checked yes to all, then predictive maintenance is a good fit for that equipment.