Mechanical and Civil Engineering Seminar
Machine Learning and Big Data Analytics Enabled Integrity Assessment of Urban Buildings
Abstract: Buildings, as a key component of urban areas, are the basic carrier for human activities and economy prosperity. As time goes, urban buildings are deteriorating and becoming vulnerable under the threat of natural extreme events such as earthquake and hurricane. How to effectively assess, maintain and manage our urban buildings, more broadly, our living communities and cities, under normal operating conditions, intermediate stress conditions and natural hazards become a critical question. Recent advances in emerging sensing, data science and computation technologies enable an information-driven perspective to address such a question. In this talk, I will discuss how to use state-of-the-art big data analytics, machine learning and computational modeling to realize integrity assessment and resilience quantification of buildings. More specifically, I will introduce data interferometry for building monitoring and characterization, data-driven quantification of environmental effect on building's property, Bayesian inference-based model uncertainty quantification, deep learning enabled fragility analysis, and 3D fragility map for resilience assessment of urban buildings cluster. A proof-of-concept study, which has been performed with datasets from simulation, lab tests and field monitoring, will be discussed. The objective of this approach is to help improve maintenance planning for our urban buildings, reduce risk and management cost, provide quantitative information for disaster response planning and decision making, as well as enhance the resiliency of our living communities and cities.
Biography: Hao Sun is an Assistant Professor in the Department of Civil and Environmental Engineering and Director of the Lab for Infrastructure Sensing and Data Science at the University of Pittsburgh (Pitt). He obtained his PhD and MPhil in Engineering Mechanics and MS in Civil Engineering from Columbia University, and received his BS in Civil Engineering from Hohai University in China. Prior to joining Pitt, he was a Postdoctoral Associate at MIT. His research focuses on developing innovative methodologies to address safety, resilience and sustainability issues of the built environment. His specific interests include advanced sensing, big data analytics, machine learning, system identification, uncertainty quantification and inverse computational mechanics, for structure and infrastructure health management and resilience assessment. He has co-authored over 20 peer-reviewed journal publications, and his work has been recognized and reported by various media coverages more than forty times over the past two years, including MIT News, Fox News, Forbes Magazine, ASCE Civil Engineering Magazine, Science Daily, Economic Times, Pittsburgh Business Times, etc. He is the receipt of multiple scholarships and awards, including the 2018 Forbes "30 Under 30": Science, the 2017 Hewlett International Grant Award, two poster awards from EMI 2014, Boeing Fellowship, NSF Workshop Travel Award, China National