Posters
Come meet the poster presenters to ask them questions and discuss their work
We would like to invite you to come and see the posters at our upcoming conference. The posters will showcase a diverse range of research topics, and will give delegates an opportunity to engage with the authors and learn more about their work. Whether you are a seasoned researcher or simply curious about the latest developments in your field, we believe that the posters will offer something of interest to everyone. So please join us at the conference and take advantage of this opportunity to learn and engage with your peers in industry and the academic community.
PO204: Leveraging Reliability Features over Raw Sensor Data for Remaining Useful Life Prediction in Wind Turbines
Gustavo Grubler, Data Scientist, AQTech Power Prognostics
Abstract
The estimation of Remaining Useful Life (RUL) of wind turbine components through Machine Learning (ML) is a key challenge in predictive maintenance, with direct implications for reliability and cost reduction in the energy sector. This work explores the challenges of developing RUL prediction models, emphasizing the importance of having both failure records and operational data that correlate strongly with degradation processes. Initial experiments using only raw sensor data from wind turbines showed weak correlation with the RUL, resulting in poor predictive performance. To address this limitation, we engineered a set of features derived from failure events, such as failure counts, failure rates, rolling Mean Time Between Failures (MTBF), and time since last failure. These derived features provided more meaningful input for the ML models, leading to improved predictive accuracy compared to using sensor data alone. Despite these advances, the results indicate that further work is needed to refine feature engineering strategies and integrate domain knowledge to enhance model robustness. Overall, this study highlights the value of leveraging calculated reliability features over raw sensor data when direct correlations with RUL are weak, offering practical insights into how ML can be applied to predictive maintenance in wind energy systems.
