Standard vs Statistical Alarms in Condition Monitoring
Standard vs Statistical Alarms in Condition Monitoring
Alarm thresholds are a fundamental part of condition monitoring. They help maintenance and reliability teams identify when measured behaviour has moved beyond an expected range and when further investigation may be needed. But not every machine behaves in exactly the same way. Two assets performing similar duties can have different normal vibration levels because of their design, installation, loading or operating conditions. This is where statistical alarms can add useful asset-specific context.
What are standard alarm thresholds?
Standard alarm thresholds are predefined limits used to indicate when a measured value deserves attention. In vibration monitoring, these limits may be based on recognised standards, engineering guidance, manufacturer information or established site criteria.
Their main strength is consistency. They provide a common reference across a population of assets and are particularly useful where historical data for an individual machine is limited. That makes them a valuable first line of defence, but a fixed threshold cannot always reflect the normal behaviour of every asset.
Why asset history matters
Condition monitoring is most useful when it identifies meaningful change. An asset can remain below a broad standard threshold while still developing a fault if its normal vibration level is much lower. Equally, another machine may operate consistently at a higher baseline without that level representing deterioration. Repeated measurements build a picture of how an individual asset normally behaves. That history gives engineers another reference point for recognising when behaviour changes.
How statistical alarms work
Statistical alarms use historical asset data to establish alarm levels that reflect the measured behaviour of the individual machine. Within Machine Sentry®, the Statistical Alarm Tool can use an asset’s established history to support alarm levels that are more specific to that machine than a generic threshold alone. The purpose is not to replace recognised engineering limits. Statistical alarms complement them by adding asset-specific context.
The distinction is simple:
A standard threshold asks: Is this measurement beyond an established general limit?
A statistical alarm asks: Is this measurement unusual for this particular asset compared with its established behaviour?
Those questions are related, but they are not the same.
For example, a machine with a historically stable low vibration level may begin trending upwards while remaining below a generic alarm threshold. Statistical analysis can help highlight that change earlier.
Conversely, a machine with a consistently higher but stable baseline may repeatedly approach a generic threshold without showing evidence of deterioration.
Why both approaches are useful
A strong alarm strategy does not need to choose between standard thresholds and statistical alarms.
Standard thresholds provide recognised baseline criteria and consistent screening. Statistical alarms add sensitivity to changes from an individual asset’s normal behaviour. Used together, they can help teams distinguish genuine changes in condition from normal variation and focus engineering attention where it is most useful.
The quality of the data still matters
Statistical analysis is only as useful as the data behind it. Measurements collected under very different operating conditions can create normal variation that needs to be understood before an alarm is interpreted as a fault.
Reliable statistical alarms therefore depend on repeatable measurement points, consistent sensor installation, representative historical data and awareness of operating state. Most importantly, an alarm is an indication that something deserves attention. It is not a diagnosis.
From alarm to engineering decision
Whether an alarm is triggered by a standard threshold or statistical behaviour, the next step is interpretation. Engineers may review trends, vibration spectra, temperature, operating conditions and maintenance history to understand what has changed and why.
That is where condition monitoring becomes more than simply generating alerts. The objective is to turn measurements into informed maintenance decisions.
Standard thresholds provide the baseline. Statistical alarms add asset-specific insight. Together, they allow the alarm strategy to become more informed as useful asset history develops.
