Honorary Lecture - Prof. Ian Smith
Date: 4 November 2022
Time:
Title:
“Structures, sensors, populations and sustainability”
Abstract
Creation and management of civil infrastructure, as well as buildings, are activities that are one of the worst offenders in terms of climate impact, energy consumption and non-renewable material use. Since these activities are core business for one of the largest of all industries, it is no longer possible to carry out “business as usual” for construction. Fortunately, there are many ways to do better. This talk examines the use of sensing (including full-field techniques) to improve knowledge of real structural behavior. The first part describes the challenges associated with asset management in the context of uncertainty, particularly systematic uncertainty which is unavoidable for large full-scale structures. A data-interpretation methodology is described that allows uncertainties to be represented clearly and understandably through populations of candidate behavior models. The population methodology also allows use of a sensor-system design strategy through use of a joint-entropy criterion. Results have revealed much reserve capacity above code requirements (over-design). The next part of the talk is an overview of research into adaptive structures. Three laboratory structures are described. Performance attributes such as learning, self-diagnosis, adaptation and deployment are discussed. The last structure was used to validate a new design concept based on whole-life energy use. Finally, a study on occupant detection and location is used to demonstrate the ability of the data-interpretation population methodology to be used more generally in situations of uncertainty. The appropriate use of measurements and populations have much potential to improve sustainability through whole-life energy design, replacement avoidance, well designed retrofitting and possibly, less conservative design codes.
About the speaker
Ian Smith came to EPFL in 1982 (for two years) after receiving his PhD from Cambridge University. He continued his research into fatigue of steel structures at ICOM and in 1991, he moved to the Artificial Intelligence Laboratory in the Computer Science Department. Five years later, he returned to civil engineering at IMAC as an Assistant Professor. He became an Associate Professor in 2000 and a Full Professor in 2005. His research interests are on intersections of computer science with the built environment. In 2003, he co-authored the text book Fundamentals of Computer-Aided Engineering (Wiley) and the 2nd edition appeared in June 2013. In 2004, he was elected to the Swiss Academy of Engineering Sciences and in 2005, he received the Computing in Civil Engineering Award from the ASCE. He has been an Adjunct Professor at Carnegie Mellon University, USA since 2011. From 2010 to 2020, he directed a second research group in Asia as Principal Investigator at the ETH Future Cities Laboratory, CREATE, Singapore. In 2022, he was elected to the National Academy of Construction, USA. From March 2022, he is the Founding Director of the Georg Nemetschek Institute AI for the Built World at TU Munich.
Time:
- 17:15-17:25: Welcome address by Prof. Katrin Beyer, Associate Dean ENAC
- 17:25-18:20: Lecture
- 18:20-18:30: Honorary Diploma and closure
- 18:30: Aperitif
Title:
“Structures, sensors, populations and sustainability”
Abstract
Creation and management of civil infrastructure, as well as buildings, are activities that are one of the worst offenders in terms of climate impact, energy consumption and non-renewable material use. Since these activities are core business for one of the largest of all industries, it is no longer possible to carry out “business as usual” for construction. Fortunately, there are many ways to do better. This talk examines the use of sensing (including full-field techniques) to improve knowledge of real structural behavior. The first part describes the challenges associated with asset management in the context of uncertainty, particularly systematic uncertainty which is unavoidable for large full-scale structures. A data-interpretation methodology is described that allows uncertainties to be represented clearly and understandably through populations of candidate behavior models. The population methodology also allows use of a sensor-system design strategy through use of a joint-entropy criterion. Results have revealed much reserve capacity above code requirements (over-design). The next part of the talk is an overview of research into adaptive structures. Three laboratory structures are described. Performance attributes such as learning, self-diagnosis, adaptation and deployment are discussed. The last structure was used to validate a new design concept based on whole-life energy use. Finally, a study on occupant detection and location is used to demonstrate the ability of the data-interpretation population methodology to be used more generally in situations of uncertainty. The appropriate use of measurements and populations have much potential to improve sustainability through whole-life energy design, replacement avoidance, well designed retrofitting and possibly, less conservative design codes.
About the speaker
Ian Smith came to EPFL in 1982 (for two years) after receiving his PhD from Cambridge University. He continued his research into fatigue of steel structures at ICOM and in 1991, he moved to the Artificial Intelligence Laboratory in the Computer Science Department. Five years later, he returned to civil engineering at IMAC as an Assistant Professor. He became an Associate Professor in 2000 and a Full Professor in 2005. His research interests are on intersections of computer science with the built environment. In 2003, he co-authored the text book Fundamentals of Computer-Aided Engineering (Wiley) and the 2nd edition appeared in June 2013. In 2004, he was elected to the Swiss Academy of Engineering Sciences and in 2005, he received the Computing in Civil Engineering Award from the ASCE. He has been an Adjunct Professor at Carnegie Mellon University, USA since 2011. From 2010 to 2020, he directed a second research group in Asia as Principal Investigator at the ETH Future Cities Laboratory, CREATE, Singapore. In 2022, he was elected to the National Academy of Construction, USA. From March 2022, he is the Founding Director of the Georg Nemetschek Institute AI for the Built World at TU Munich.
Links
Practical information
- Informed public
- Registration required
Organizer
- SGC - Céline Dupuy