Can artificial intelligence transform the practice of structural engineering?
Imagine a future where structural engineers not only rely on their own knowledge and experience, computational tools, or evaluation equipment, but also utilize powerful AI assistants to significantly boost their productivity. This vision is becoming a reality through a groundbreaking collaboration between the University of Houston and Purdue University, supported by the National Science Foundation (NSF).
The newly planned Center for Visual Structural Expertise for Resilience (C-ViSER) aims to develop and equip engineers with cutting-edge AI and machine learning tools, transforming traditional workflows. This center is an Industry-University Cooperative Research Center (IUCRC) — a partnership structure designed to generate breakthrough research by enabling close and sustained engagement between industry innovators, world-class academic teams and government agencies.
The new center spans two sites: Purdue University and the University of Houston. A University of Houston workshop scheduled for October 21-22 of this year will disseminate the goals of the center, including fostering the new partnerships between industry and academia mentioned above, and will engage prospective center members in discussions around C-ViSER's focus, value propositions and potential research projects to develop a research roadmap that aligns with the needs and interests of industry.
“My team at UH and I have been developing deep learning, computer vision, and artificial intelligence solutions for varied structural engineering problems over the past decade," said Director of the Structures and Artificial Intelligence Lab Vedhus Hoskere, assistant professor of civil and environmental engineering (CEE) and electrical and computer engineering (ECE).
Hoskere currently leads projects with the Texas Department of Transportation on developing digital twins of bridges and the U.S. Navy for improving structural inspections, and leads research on Large Language Models as part of the university's recent Department of Defense contract to revolutionize U.S. Army decision-making.
Now, he is looking to bring these technologies to the structural engineering industry through C-ViSER.
Structural engineers draw on centuries of experience captured in detailed documents like codes and specifications to create designs, develop drawings, capture photos, and produce reports that guide construction and maintenance decisions.
“By using AI to tap into insights from past projects, our goal with the Center is to help engineers realize efficiencies that ultimately result in more sustainable, cost-effective, and resilient construction and maintenance of civil infrastructure,” said Hoskere.
Deepank Kumar Singh, a civil engineering Ph.D. student under Hoskere's advisement and a member of the C-ViSER team, also works on a project sponsored by the NASA Commercial SmallSat Data Acquisition (CSDA) Program to assess buildings in the aftermath of disasters.
“In a recently published paper, we introduced a state-of-the-art semi-supervised Transformer model trained on vast amounts of unlabeled data to categorize buildings into one of five damage states automatically," Singh said.
This past summer, Singh was able to apply his unique skills during an internship at a structural engineering firm, experiencing first-hand the impact that AI can have on real-world challenges in structural engineering. “During my internship, I developed 2 tools that utilize similar AI models: a project search tool and a report generation tool specifically for building fire inspection.”
Muhammad Taseer Ali is another PhD student advised by Hoskere and working towards realizing the Center, enrolled at UH after 10 years of experience as a structural engineer designer in industry.
“I have directly encountered the challenges that hinder efficiency in structural engineering and design processes," Ali said. "This experience drives my strong interest in developing novel solutions that streamline these processes."
“Another major goal the center is to mentor our students to be leaders at the forefront of technology and help them directly work on pressing industry-relevant problems. The center will provide students opportunities for networking and building relationships with companies interested in their work," said Hoskere.
Other UH collaborators include Roberto Ballarini, Thomas and Laura Hsu Professor and Chair of the Department of Civil and Environmental Engineering, and Craig Glennie, Hugh Roy and Lillie Cranz Cullen Distinguished Chair and director of the National Center for Airborne Laser Mapping. Glennie also leads the project team funded by UH's $63.5M U.S. Department of Defense Contract to help modernize U.S. Army decision-making.