BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:ENAC Seminar Series by Prof. O. Fink
DTSTART:20210330T090000
DTEND:20210330T094500
DTSTAMP:20260506T015743Z
UID:acb6debdafb435d8a0b7376537a4dd00cca61fbf2e5fd00d436f4ae6
CATEGORIES:Conferences - Seminars
DESCRIPTION:Prof. Olga Fink\n09:00 – 09:45 – Prof. Olga Fink\nAssistan
 t Professor at ETHZ\, Zürich\n\nDomain adaptation for intelligent mainten
 ance systems\n\nThe amount of measured and collected condition monitoring 
 data for complex infrastructure and industrial assets has been recently in
 creasing significantly due to falling costs\, improved technology\, and in
 creased reliability of sensors and data transmission. However\, faults in 
 safety critical systems are rare. The diversity of the fault types and ope
 rating 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 collectin
 g a representative dataset with all possible operating conditions can be a
  challenging task since the systems experience a high variability of opera
 ting conditions. Therefore\, training samples captured over limited time p
 eriods may not be representative for the entire operating profile. The col
 lection of a representative dataset may delay the implementation of data-
 ​driven fault detection and isolation systems.\nThe talk will provide in
 sights into potential solutions that enable to transfer models and operati
 onal experience between different units and between different operating co
 nditions also in unsupervised setups where data on faulty conditions is no
 t available. Moreover\, a synthetic-to-real framework for domain adaptatio
 n will be presented\, where only knowledge of the healthy class is needed.
  The healthy class is augmented with synthetic faults to generate the rare
  fault examples based on expert understanding of how faults affect the mea
 sured signals. A domain adaptation (DA) is proposed to adapt the model fro
 m synthetic faults (source) to the unlabeled real data (target) overcoming
  the different imbalance levels in source and target datasets.\n\n\nShort 
 bio:\nOlga Fink has been assistant professor of intelligent maintenance sy
 stems at ETH Zürich since October 2018\, being awarded the prestigious pr
 ofessorship grant of the Swiss National Science Foundation (SNSF).\nOlga i
 s also a research affiliate at Massachusetts Institute of Technology and E
 xpert of the Innosuisse in the field of ICT. Olga’s research focuses on 
 Intelligent Maintenance Systems\, Data‐Driven Condition‐Based and Pred
 ictive Maintenance\, Hybrid Approaches Fusing Physical Performance Models 
 and Deep Learning Algorithms\, Deep Learning and Decision Support Algorith
 ms for Fault Detection and Diagnostics of Complex Infrastructure and Indus
 trial Assets. Before joining ETH faculty\, she was heading the research gr
 oup “Smart Maintenance” at the Zurich University of Applied Sciences (
 ZHAW) between 2014 and 2018. Olga received her Ph.D. degree from ETH Zuric
 h with the thesis on “Failure and Degradation Prediction by Artificial N
 eural Networks: Applications to Railway Systems”\, and Diploma degree in
  industrial engineering from Hamburg University of Technology. She has gai
 ned valuable industrial experience as reliability engineer with Stadler Bu
 ssnang AG and as reliability and maintenance expert with Pöyry Switzerlan
 d Ltd. In 2018\, Olga was selected as one of the “Top 100 Women in Busin
 ess\, Switzerland” and in 2019\, she was selected as young scientist of 
 the World Economic Forum.\n 
LOCATION:Zoom https://epfl.zoom.us/j/83675937537
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR