报告题目：Towards Intelligent Operation and Maintenance of Complex Systems
报 告 人：Dr. Olga Fink - ETH Zürich（苏黎世联邦理工学院智能维护系统SNSF教授）
邀 请 人：李巍华教授（Prof. Weihua Li）
联 系 人：陈祝云博士（Dr. Zhuyun Chen, firstname.lastname@example.org）
Olga Fink has been assistant professor of intelligent maintenance systems at ETH Zürich since October 2018, being awarded the prestigious professorship grant of the Swiss National Science Foundation (SNSF). She is also a research affiliate at Massachusetts Institute of Technology and Expert of the Innosuisse in the field of ICT. Her research focuses on Intelligent Maintenance Systems, Data‐Driven Condition‐Based and Predictive Maintenance, Hybrid Approaches Fusing Physical Performance Models and Deep Learning Algorithms, Deep Learning and Decision Support Algorithms for Fault Detection and Diagnostics of Complex Infrastructure and Industrial Assets. Before joining ETH faculty, she was heading the research group “Smart Maintenance” at the Zurich University of Applied Sciences (ZHAW) between 2014 and 2018. Olga received her Ph.D. degree from ETH Zurich with the thesis on “Failure and Degradation Prediction by Artificial Neural Networks: Applications to Railway Systems”, and Diploma degree in industrial engineering from Hamburg University of Technology. She has gained valuable industrial experience as reliability engineer with Stadler Bussnang AG and as reliability and maintenance expert with P?yry Switzerland Ltd. In 2018, she was selected as one of the “Top 100 Women in Business, Switzerland”, in 2019, she was selected as young scientist of the World Economic Forum and in 2020 and 2021 as young researcher of the World Laureate Forum.
The amount of measured and collected condition monitoring data for complex infrastructure and industrial assets has been recently increasing significantly due to falling costs, improved technology, and increased reliability of sensors and data transmission. However, faults in safety critical systems are rare. The diversity of the fault types and operating conditions makes it often impossible to extract and learn the fault patterns of all the possible fault types affecting a system. Consequently, faulty conditions cannot be used to learn patterns from. Even collecting a representative dataset with all possible operating conditions can be a challenging task since the systems experience a high variability of operating conditions. Therefore, training samples captured over limited time periods may not be representative for the entire operating profile. The collection of a representative dataset may delay the implementation of data-driven fault detection and isolation systems. Moreover, some of the current limitations include the limited scalability, generalization ability and interpretability of the developed models. The talk will give an overview of the currently ongoing research at the chair of Intelligent Maintenance Systems at ETH Zürich, including 1) research in the field of domain adaptation and unsupervised transfer learning for fault detection and diagnostics at fleet level, 2) research on algorithms combining deep learning and physics-based approaches; 3) research in prescriptive operations and 4) research on multi-agent systems for decision support and maintenance scheduling.