Decentralized Reliability and Resilience for Infrastructure Networks
Speaker: Leonardo Duenas Osorio
When: 11:00a-12:00p Monday, December 10, 2018
Where: Devon Energy Hall, 120
Leonardo Duenas-Osorio is currently an Associate Professor of Civil and Environmental Engineering at Rice University. He obtained his Master’s degree from the Massachusetts Institute of Technology in 2001, and Ph.D. from the Georgia Institute of Technology in 2005, both in Civil and Environmental Engineering. He develops analytical and computational methods, including quantum algorithms, for the performance assessment and upkeep decision for structural and Infrastructure systems. He chairs the Structural Engineering Institute’s Task Group on “Risk Assessment of Structural Infrastructure Facilities and Risk-Based Decision Making” of the American Society of Civil Engineers (ASCE), and the Dependency Relations Sub-Committee of ASCE’s Infrastructure Resilience Division. He is former Associate Editor for the ASCE Journal of Computing in Civil Engineering, current Associate Editor for Natural Hazards Review, and member of the editorial board of Structural Safety.
Dr. Duenas-Osorio’s work has garnered recognition, including the Best Ph.D. Thesis Award in Civil and Environmental Engineering from the Georgia Institute of Technology in 2006 and the National Science Foundation (NSF) CAREER award in 2008. More recently, he received in 2015 the Outstanding Earthquake Spectra Paper Award by the Earthquake Engineering Research Institute (EERI), the 2016 Wilson Tang Best Paper Award at the Asia-Pacific Symposium on Structural Reliability and its Applications (APSSRA), and the 2017 Early Achievement Research Award by the International Association for Structural Safety and Reliability (IASSAR), recognizing “contributions to and impact on the field of structural safety and reliability”.
Seminar Topic: Decentralized Reliability and Resilience for Infrastructure Networks
As automation and interdependence across critical infrastructure networks are becoming ubiquitous, we need trustable system state estimation methods to inform operational and contingency management decisions. However, current methods to quantify infrastructure performance under uncertainty are often rendered computationally infeasible, or are inadequate to capture essential features of engineered systems, such as hierarchical operation and coordinated decentralization. This talk introduces methods for the efficient reliability and resilience assessment of networked systems, built on algorithms consistent with their decomposable nature, and also offering exact bounds or a priori guarantees on the quality of their estimates. In particular, I present algebraic and game theoretic strategies that exploit the factorable structure of interdependent infrastructure networks, especially when they need to be optimally restored in minimum time at a minimum cost. Our methods explicitly recognize the decentralized processes for network restoration documented in post-disaster investigations. Probing resilience also reveals the key role of complementary system-level properties such as reliability. Hence, I introduce a logic-based constrained counting formulation that offers precision and scalability for network reliability assessment while harboring probably approximately correct (PAC) guarantees. Together, these new decentralized yet coordinated computational strategies contribute to the principled design and upkeep of modern engineered networked systems in a way compatible with the measurement sciences.