MISE: Predictive Maintenance
Using predictive maintenance, we develop solutions to identify system maintenance requirements at an early stage and to determine and predict optimum times for such work to be carried out. The goal is to detect problems in systems before malfunctions or failures occur.
Unlike time-based maintenance strategies, predictive maintenance solutions provide maintenance recommendations based on the system’s actual condition. This requires constant monitoring of the system and its critical sub-systems using suitable sensor technology. Predictive maintenance can achieve higher availability and cost savings as work is only carried out when justified by the system’s actual condition.
Our current research project with WinJi AG is working on machine-learning algorithms for automatic early fault identification in wind turbines. Here we combine sensor data from high-frequency vibration meters and SCADA system control technology. The machine-learning algorithms created are used on the WinJi TruePower platform, which provides asset managers of wind and solar power plants with a cost-effective monitoring solution that enables the early detection of imminent faults or malfunctions as well as condition-based system maintenance.
Network: Smart Maintenance Network
The Swiss Smart Maintenance Network is made up of asset managers, data scientists and maintenance specialists who develop and deploy intelligent maintenance systems and software for the condition-based maintenance of systems and infrastructure or focus on new developments and trends in predictive maintenance. The networking events provide a forum for discussion between experts in data-based maintenance and help to forge links within this group in Switzerland.
Meyer, A., 2021, Multi-target normal behaviour models for wind farm condition monitoring, Applied Energy, Elsevier, doi: 10.1016/j.apenergy.2021.117342.
Meyer, A. and B. Brodbeck, 2020, Data-driven Performance Fault Detection in Commercial Wind Turbines, Proceedings of the 5th European Conference of the Prognostics and Health Management Society 2020, Turin, Italy, ISBN 978-1-93-626332-5
B. Brodbeck and A. Meyer, Patent application PCT/EP2020/058768 “Device, Method and Computer Program for Evaluating the Operation of a Power Plant”, 2020