Date of Award
Thesis - Open Access
Master of Science in Civil Engineering
Rapid development in trucking technology and increasing demands in freight transportation has led to longer and heavier vehicles traveling on Florida’s highway system. Vehicles with gross vehicle weight (GVW) over 80,000 pounds, or permit vehicles, have significant effects on infrastructure, thus requiring an approved permit prior to departure. The combination of these increasing loads and harsh environmental conditions that Florida is subject to requires an enhanced infrastructure management program. Additionally, there is a need to eliminate inconsistencies in permit applications and derive a uniform maintenance practice for Florida’s infrastructure. In this research, the focus was to develop an analytical procedure for the characterization and prediction of superload (GVW ≥ 150,000 pounds) and overweight (80,000 pounds ≤ GVW < 150,000 pounds) vehicles using gradient boosting machine (GBM) learning algorithms. The characterization of permit vehicles was performed for Florida Weigh-in-Motion (WIM) sites and the prediction of GVW, maximum axle weight, and individual axle weights were accurately predicted using limited configuration parameters. A database that combined traffic input from WIM stations, environmental information, human factors, and bridge general condition ratings (GCR) from the National Bridge Inventory (NBI) database was applied to analyze the effects of various parameters on bridge deterioration trends. Prestressed concrete bridges were selected within 25 miles of each WIM site. Subsequently, a big data set was formulated considering all bridge deterioration modeling factors and a simulation was conducted for a standard bridge built in 1990. Results were then compared to the predictions with no maintenance and the current practice for a bridge in that environmental and loading category. It was concluded that bridges in coastal regions, especially those subject to high live loading, must be given special consideration in the management and maintenance procedures.
Scholarly Commons Citation
Jesso, Julian, "Artificial Intelligence Driven Infrastructure Management and Maintenance Plan" (2021). PhD Dissertations and Master's Theses. 582.