BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Physics-Enhanced Machine Learning for Monitoring &amp\; Twinning |
  An Exercise in Balance - Professor Eleni Chatzi\, ETH Zürich
DTSTART:20250314T140000Z
DTEND:20250314T150000Z
UID:TALK221203@talks.cam.ac.uk
CONTACT:46601
DESCRIPTION:Modern engineering systems—ranging from bridges to wind ener
 gy structures—operate under complex loading and evolving environmental c
 onditions. Ensuring their resilience requires understanding their real-tim
 e performance\, a goal addressed by Structural Health Monitoring (SHM). SH
 M follows a hierarchy from damage detection to prognosis\, but higher-leve
 l tasks demand more than purely data-driven methods. Achieving reliable in
 sights necessitates balancing physics-based models with operational data a
 nd expert knowledge while maintaining intuitive system representations.\nT
 his talk explores how critical infrastructure can be modeled as cyber-phys
 ical systems\, integrating sensing\, modeling\, control\, and networking t
 o create closed-loop digital twins. We emphasize the role of intuitive rep
 resentations\, such as those driven by physics principles and graph-based 
 representations\, in capturing system topology\, dependencies\, and evolvi
 ng states. By balancing data-driven augmentation\, physics-based modeling\
 , expert insights\, and system-wide considerations\, we develop augmented 
 twins that accurately represent structures\, predict responses beyond meas
 ured points\, anticipate future performance\, and support proactive decisi
 on-making across various engineering assets.\n
LOCATION:Department of Engineering - LR4
END:VEVENT
END:VCALENDAR