Intelligent Maintenance and Operations Systems

The Intelligent Maintenance and Operations Systems (IMOS) at EPFL focuses on developing intelligent algorithms to enhance the performance, reliability, and availability of complex industrial assets. Their research addresses challenges such as fault rarity, heterogeneous condition monitoring data, refurbishment issues, varying operating conditions, and unit specificities within fleets. IMOS collaborates with various partners on projects like Physics-informed Graph Neural Networks for Industrial IoT, thermal energy network validation, and Swiss railway traction grid state estimation. Their work includes domain adaptation for Remaining Useful Lifetime prediction, federated learning for fleet-wide fault diagnosis, and non-contact sensing for anomaly detection in wind turbine blades.