April 25, 2025 1:00 pm
Bjorn Birgisson, Ph.D., Principle Investigator, Chair, School of Environmental, Civil, Agricultural, and Mechanical Engineering; Sajab Saha, Ph.D., Assistant Research Scientist; Mi Geum Chorzepa, Ph.D., P.E., Professor of Civil Engineering; Adeyemi David Sowemimo, Ph.D. Candidate
Presentation Documents:
Presentation Slides: UGA slides for IBT ABC-UTC webinar on 2025 Apr 25
Description: Traditional bridge asset management practices often fall short in adequately capturing the interactions among bridge elements to predict service life, as well as in assessing the sensitivity of the freight network to disruptive events such as natural disasters and automated trucking. To address these limitations, this project has introduced a proactive bridge asset management system that employs modern technologies, including artificial intelligence (AI), geographic information system (GIS), and graph-based network modeling, leveraging massive data repositories. To quantify the structural and topological criticality of bridges at a network level, a graph-based network model of roadways and bridges within Georgia’s National Highway Freight Network (NHFN) has been developed based on geospatial coordinate datasets and roadway intersections. The structural quality of individual bridges is quantified by predicting bridge condition ratings using advanced recurrent neural network (RNN) models that employ Long Short-Term Memory (LSTM) and Gated Recurrent Unit (GRU) architectures, featuring a Time Distributed output layer. The model enables both current and future forecasting of bridge deck condition ratings based on recorded element data, age, operating rating, surface type and traffic datasets. The outcome of this analysis reveals the complex relationships that influence individual bridge performance as well as the topological vulnerabilities within the freight network, prioritizing maintenance, repair, and replacement activities. IBT/ABC-UTC is undertaking the development of the Next Generation Bridge Asset Management System. This presentation will briefly elaborate on the topic.
Presenters:
Bjorn Birgisson, Ph.D.,
Principle Investigator
Chair
School of Environmental, Civil, Agricultural, and Mechanical Engineering
University of Georgia
Email: birgisson@uga.edu
Dr. Birgisson, is the former J.L. “Corky Frank/Marathon Ashland Petroleum LLC Chair in the College of Engineering at Texas A&M University. He is a Foreign Fellow of the Royal Swedish Academy of Engineering Sciences and a Fellow of American Society of Civil Engineers (ASCE) and the Institution of Civil Engineers (ICE). He has published over 240 technical papers focused on transportation infrastructure, construction materials, additive materials for infrastructure, civil systems-of-systems and resilience.
Sajab Saha, Ph.D.,
Assistant Research Scientist
School of Environmental, Civil, Agricultural, and Mechanical Engineering
University of Georgia
Dr. Saha was a postdoctoral research associate in the Center for Infrastructure Renewal (CIR) and the Texas A&M Transportation Institute (TTI) at Texas A&M University (TAMU). He received his Ph.D. from the Department of Civil and Environmental Engineering at TAMU. His research spans a broad range of topics, including multiscale characterization of pavement materials, pavement design and analysis, artificial intelligence, nondestructive pavement evaluation, and infrastructure resilience.
Mi Geum Chorzepa, Ph.D., P.E.,
Professor of Civil Engineering
School of Environmental, Civil, Agricultural, and Mechanical Engineering
University of Georgia
Dr. Chorzepa is a Professor of Civil Engineering at UGA with extensive experience leading various bridge projects and teaching a Bridge Engineering course.
Adeyemi David Sowemimo,
Ph.D. Candidate
School of Environmental, Civil, Agricultural, and Mechanical Engineering
University of Georgia
Mr. Sowemimo is a PhD candidate at the University of Georgia specializing in bridge data analytics and the development of RNN models, with significant contributions to multiple research projects.
Presentation Graphics:
Figure 1
Figure 2
Figure 3