DiOMera PREDINSIGHT is ANDRITZ Hydro’s advanced predictive and prescriptive maintenance analytics solution for hydropower plants. Using machine learning methods applied to historical and continuously acquired monitoring data, PREDINSIGHT derives Key Diagnostic Indicators (KDIs) that provide a consolidated and continuously updated representation of the condition of each monitored asset. By identifying subtle deviations from expected equipment behavior, KDIs enable developing anomalies to be detected at an early stage, before conventional alarm thresholds are reached.

KDIs combine information from relevant monitoring signals using machine learning models trained on historical operational data. They are continuously evaluated against reference conditions and predefined diagnostic thresholds, allowing changes in equipment behavior to be identified while degradation is still at an early stage. Beyond condition assessment, PREDINSIGHT analyzes the evolution of KDIs over time to derive Key Trend Indicators (KTIs). These indicators extrapolate observed trends and estimate when a predefined maintenance threshold is expected to be reached, transforming condition information into a time-referenced maintenance forecast.

This predictive insight is formalized through automated maintenance notifications that consolidate the current condition, the observed trend, and the predicted time to reach the relevant maintenance threshold. Maintenance teams can use this information to anticipate interventions and integrate them into maintenance planning before equipment condition becomes critical. Through the complete chain from signal monitoring and KDI derivation to KTI forecasting and maintenance notification, PREDINSIGHT supports the transition from reactive and calendar-based maintenance toward condition-based and predictive maintenance, with prescriptive guidance supporting maintenance decision-making.