Glossary · IIoT, data and AI
Predictive maintenance analytics (AI/ML)
Also known as: AI-based predictive maintenance, PdM analytics
German: Predictive-Maintenance-Analytik (KI/ML)
In maintenance, predictive maintenance analytics is the use of statistical and machine learning methods on condition, process and maintenance data to detect degradation, estimate remaining useful life and predict failures, so that maintenance can be planned before a failure occurs.
- IIoT
- AI
In one sentence
Predictive maintenance analytics applies statistics and machine learning to condition data to predict failures and plan maintenance.
Example
A model trained on vibration, temperature and past failure records estimates that a conveyor gearbox will need replacement within six weeks.
How it applies
- Engineering: Predictive models need failure history, which is often scarce. Many projects start with Anomaly detection and physical thresholds and add prediction as data accumulates.
- Maintenance: Predictions are useful when they give enough lead time and are linked to concrete actions in the maintenance system. False alarms quickly erode trust.
- Documentation: Document the model's inputs, the failure modes covered, the validation results and the recommended response. Maintenance manuals should state which inspections remain mandatory regardless of predictions.
Predictive maintenance analytics vs. predictive maintenance
Predictive maintenance is the maintenance strategy: acting on predicted condition. Predictive maintenance analytics is the analytical part that produces the predictions. For safety-related components, a prediction does not replace required inspections or proof tests unless this has been validated.