“Civil engineering is not limited to constructing buildings and bridges. It’s also highly computational and works closely with other disciplines. I’m doing more computational physics now than ever before.”
In the last year alone, Leonardo Dueñas-Osorio, assistant professor of civil and environmental engineering (CEE) at Rice University, has co-authored scholarly articles with statisticians, physicists and computer scientists devoted to topics as various as power grid failure, urban water security and the likelihood of homes in Houston being damaged by high winds or flooding during a hurricane.
Dueñas-Osorio was born in Bogota, Colombia, where he earned a B.S. in 1996 from the Universidad de La Salle, and M.S. in CEE in 1998, from Universidad de Los Andes. His father was an operations engineer who worked for Colombia’s national railroad, and Dueñas-Osorio sometimes accompanied him on inspection, maintenance and repair trips into the country’s mountainous interior.
“As a boy I was most interested in physics. Now I find myself flirting with quantum computing and quantum algorithms. It is one way to help us look at the reliability of civil engineered systems,” he said.
In 2001, Dueñas-Osorio earned his professional master’s degree in CEE with an emphasis on high-performance structures from MIT, and four years later his Ph.D. in CEE, with a minor in industrial engineering, from Georgia Tech. After five months there as a postdoc, Dueñas-Osorio joined the Rice faculty in 2006.
Much of his research focuses on developing methods to measure and predict the effects infrastructure systems have on each other during normal operation and such extreme events as earthquakes and hurricanes. What happens when stress is applied to power, telecommunication, water and gas networks? Dueñas-Osorio’s preliminary findings suggest that cascading failures across civil infrastructure systems are preventable.
“But it will require planning, including infrastructure interface topology and operational control, plus strategically locating back-up and redundant system components. What we need is better network-level analysis and optimization tools” he said.
In recognition of his research, Dueñas-Osorio was the recipient of the 2017 IASSAR Early Achievement Research Award, given every four years by the International Association for Structural Safety and Reliability.
His commitment to civil engineering and society has led him to work with colleagues from various departments at Rice. With Devika Subramanian, professor of computer science, and Robert Stein, the Lena Gohlman Fox Professor of Political Science, he has looked at the factors that contribute to the likelihood of Houston residences being damaged by high winds or flooding. They discovered that some of the customary criteria for predicting damage to houses, such as roof shape and type of framing, were not sufficiently predictive and that new factors were: building and land value, quality of construction, years lapsed since remodeling, among others.
“We built machine learning models directly from the data we took from NOAA about wind speeds and direction. We used LIIDAR data and data from the Harris County Appraisal District, and fused it all with engineering models of structural damage for higher-resolution assessment,” Dueñas-Osorio said.
“We learned from Hurricane Ike that people are not too sensitive to various risk types, such as rainfall, storm surge, wind and outages. Instead, people are more concerned with the seriousness of the risks they perceive as relevant to them, highlighting an appetite for high-resolution damage assessment models.”
The research suggested that there was a “mismatch,” in Dueñas-Osorio’s words, in every major flooding event in recent years. Flooding was not confined to the flood plains, but involved a significant percentage outside of those designated areas. This was especially true during Hurricane Harvey.
More recently, he has collaborated with Moshe Y. Vardi, Karen Ostrum George Distinguished Service Professor in Computational Engineering at Rice, and Javier Rojo, professor of statistics at Oregon State University (formerly of Rice). Their paper, “Quantum Boolean States for Engineering Network Reliability,” will be published soon in Structural Safety. It sets out the first quantum algorithm for engineered network reliability applications.
This summer, Dueñas-Osorio begins a year-long sabbatical and plans to further investigate quantum computing and its applications in gauging structural reliability, resilience and risk assessment during natural disasters and adversarial disruptions. The work is funded, in part, by the U.S. Department of Defense.
“The field is always changing, always growing more complex and sophisticated. We need all the tools available,” he said.