UTSA professors receive funding for work using AI to improve disaster recovery

MARCH 4, 2024 — A 7.8 magnitude earthquake devastated parts of Syria and Turkey on Feb. 6, 2023 killing more than 3,700 people. Roads and hospitals were destroyed. Internet outages made communication difficult hampering damage assessments and rescue efforts. It was difficult to prioritize next steps or get resources to the areas that were hit the hardest.

UTSA researchers have found a way to address these issues and have received funding to further their research. Assistant professors Ao Du and Jiannan Cai in the UTSA Department of Civil & Environmental Engineering, and Construction Management received a $400,000 grant from the National Science Foundation (NSF) and National Institute of Standards and Technology (NIST) for their research using artificial intelligence (AI) to rebuild transportation infrastructure after an earthquake.

The grant is part of the $7.1 million Disaster Resilience Research Grant program, which is funding research to improve the ability of communities, infrastructure, and buildings to withstand severe natural disasters. UTSA is one of just 15 universities to receive a grant.

“Recent advances in AI allow for adaptive decision-making to help restoration teams prioritize which roads, bridges or rail lines need fixing first.”

Ao Du

Du, who teaches civil engineering for the Margie and Bill Klesse College of Engineering and Integrated Design, said that their research can help speed up decision making related to repairing transportation infrastructure— making it easier to bring in relief supplies, search and rescue teams and reconstruction crews.

"Making restoration decisions for regional transportation networks in a post-earthquake environment is highly complicated due to many uncertainties and resource constraints throughout the entire restoration time horizon," Du said. "But recent advances in AI allow for adaptive decision-making to help restoration teams prioritize which roads, bridges or rail lines need fixing first."

Jiannan Cai

"This research will introduce risk-aware and explainable AI techniques such that stakeholders will be offered more transparent and risk-informed explanations of the AI-derived policies for trustworthy post-earthquake restoration planning,” said Cai, who teaches civil & environmental engineering and construction management.

Du said the final restoration decisions are always left to the stakeholders. The AI is there to offer, “risk-based and quantitative measurements of the long-term benefit of the options and then suggest the optimal actions,” he said. “Such information can be very helpful, especially in a post-disaster situation where the stakeholders need to make rapid and strategic decisions under huge stresses."

Du and Cai have chosen to focus on transportation networks given the highly uncertain nature of earthquakes and their hazardous impact on an area’s physical infrastructure. They will integrate AI models with civil engineering domain knowledge to make sure it reflects the characteristics of the transportation infrastructure network for reliable decision making.

“I believe such applications and projects can attract students from different backgrounds,” Cai said. “For instance, civil engineering students may find it important to have AI and programming skills to solve complex problems that would be challenging using traditional methods.”

“This has the potential to be adapted and generalized to deal with the restoration of other types of infrastructure systems affected under other types of natural disasters,” Du said.

— Sean M. Wood