Physics IS Enhancing Machine Learning

Moving away from accurate-but-wrong predictions for bridges, wind turbines… and the climate.

  • Mon 27th May 2024
  • 19:30-21:00
  • Wolfson Lecture Theatre, Churchill College, Storey's Way, Cambridge, CB3 0DS and Zoom
  • Add to calendar

View location

You can book for in-person attendance at Bookwhen, where you will also find the link for Zoom viewing.

Machine Learning algorithms are revolutionising many scientific fields by enabling the development of models from observations – so called data-driven. However, in many engineering applications, we usually have access to a limited amount of “informative” data - hindering the applicability data-driven approaches – but a great deal of physics understanding and domain knowledge! This opens up opportunities to combine physics and domain knowledge with data-driven approaches for guiding high-consequence decision making in engineering applications.

This seminar will give a brief non-technical introduction to Machine Learning and an overview of recent research work carried out within the Data, Vibration and Uncertainty Group (https://sites.google.com/view/dvugroup) focusing on developing Physics Enhanced Machine Learning (PEML) strategies in applied mechanics. It will showcase recent PEML methods developed for tackling challenges in wind turbines, bridges and structural joints, and ongoing efforts for investigating climate repair strategies.

Dr Alice Cicirello, University Assistant Professor in Applied Mechanics

Dr Alice Cicirello is a University Assistant Professor in Applied Mechanics at the Cambridge University Engineering Department and a Fellow of Churchill College. She is the founder and head of the Data, Vibration and Uncertainty group (https://sites.google.com/view/dvugroup).

Alice obtained her PhD from the University of Cambridge in 2013. She was a Marie Curie Early Stage Researcher (2009-2012) and a Research Associate (2012-2014) at the same institution. Alice worked as a Senior Research Scientist at SLB (2014-2017) and returned to academia as a Lecturer at the University of Oxford (Engineering Science Department and Balliol College, 2017-2019), and then continued as an Associate Professor and Section Head of the section Mechanics and Physics of Structures at TU Delft (2020-2023).

Alice held visiting positions at several research institutions, including MIT and the Alan Turing Institute. Alice is currently an Alexander von Humboldt Experienced Research Fellow (2023- present), an Editorial Board member of the Data-Centric Engineering journal, and an Executive board member of the European Association of Structural Dynamics. Alice was the chair of the first (2022) and second (2023) workshops on Physics-enhancing Machine Learning in Applied Mechanics. Recently, Alice was a keynote speaker for the AIUK Workshop (March 2024): Latest Developments in Physics-Informed Machine Learning; a Panelist on "What roles can AI play in the future of design?" at the IABSE Future of Design event, London (Sept 2023), and will give a keynote address at European Nonlinear Dynamics Conference (ENOC 2024).

Alice enjoys exploring exciting new techniques based on physics-enhancing machine learning, uncertainty quantification, dynamic testing and advanced physics-based models… including those of spiders! She has experience working on research challenges related to energy, automotive, aerospace and civil engineering. She has also worked on cross-disciplinary research challenges related to Artificial Intelligence, Animal Vibration and more recently Climate Repair.

Attending lectures

The lecture will be preceded by a short presentation from a CSAR PhD Award Winner

Enzymatic blood group conversion of human kidneys for transplantation.

Miss Serena MacMillan, Department of Surgery, University of Cambridge

Thank you for your interest in CSAR and its programme. If you would like to help us maintain our activities at their current level, you can make a donation to CSAR here, via PayPal or a bank card. Your gift will by default go into our general income fund; if you would like it to be used for a specific purpose such as the PhD Students Awards scheme, please let us know at info@csar.org.uk. CSAR is a registered charity run by volunteers.