C3.ai Predictive Maintenance™ provides maintenance planners and equipment operators with comprehensive insight into asset risk, enabling them to maintain higher levels of asset availability across their entire portfolio.
C3.ai Predictive Maintenance aggregates petabyte-scale data from individual sensors, smart devices, enterprise systems (e.g., asset management, work management, outage management) and operational systems (e.g., SCADA, OMS, GIS) to generate accurate predictions of asset failure.
C3.ai Predictive Maintenance uses advanced machine learning algorithms to compute asset risk scores. The algorithms are trained using historical failure data and can be configured to estimate probability of failure over different operating horizons (e.g., 14 days, 30 days, or 6 months).
In addition to supervised machine learning techniques that require historical failure data to train algorithms, C3.ai Predictive Maintenance also includes unsupervised learning techniques to identify and predict anomalous operating states without the use of historical failures. C3.ai Predictive Maintenance provides closed-loop work order integration that enables continuous improvement of the underlying machine learning models.