Anomaly Detection

Anomaly detection uses statistics or machine learning to flag when equipment or process behaviour deviates from its normal pattern, catching problems that fixed alarm limits miss. In industry it gives early warning of developing faults and efficiency drift.

Rather than waiting for a value to cross a fixed threshold, anomaly detection learns what 'normal' looks like across many variables and operating states, then flags meaningful deviations. This catches subtle, multivariable problems — a slow drift in a turbine's behaviour, an emerging exchanger fouling trend — well before a single-sensor alarm would trip.

Related terms

Related guides

Software