Reliability & Asset Health
Maximise asset availability, prevent run-to-failure events, and control degradation trajectories.
Industry Challenge
Gearbox failures are the most expensive unplanned failure mode in wind turbine O&M, with replacement costs of $250K–$500K per event plus 3–6 weeks of lost production for crane-required exchanges. Gear tooth pitting, micropitting, and bearing raceway spalling develop over months before manifesting as audible noise or vibration signatures detectable in standard CMS thresholds. Oil particle count analysis can detect metallic wear debris weeks before vibration signatures emerge, but most operators sample oil manually on 6–12 month cycles — a frequency that routinely misses the acceleration phase of gearbox degradation.
Virentis Initiative
Virentis fuses SCADA temperature and speed signals with CMS high-frequency vibration data and oil analysis records to build a composite gearbox health index per turbine. Anomaly detection models trained on historical fault precursors identify spectral changes in gear mesh frequencies and bearing defect frequencies that precede macroscopic damage. Oil sample results are integrated as discrete health observations that update the Bayesian health estimate. Maintenance scheduling agents propose optimal intervention timing by minimising the expected cost of premature action versus run-to-failure risk.
Interested in deploying agentic analytics across your asset portfolio?