All ETDs from UAB

Advisory Committee Chair

Nasim Uddin

Advisory Committee Members

Ashraf Z Al-Hamdan

Mohammad R Haider

Muhammad M Sherif

Christopher J Waldron

Document Type


Date of Award


Degree Name by School

Doctor of Philosophy (PhD) School of Engineering


Bridges, as long-life large structures, are considered a challenge in both their construction and maintenance. Since they are subjected to frequent cyclic loads normally and accidents occasionally, they are prone to possible damages during their lifetime. There-fore, Structural Health Monitoring (SHM) is a must to avoid collapse or undesired closing. Various methods are used for that purpose that differ in their cost, efficiency, and ac-curacy. To obtain a better understating of the acceleration records of the bridge, the response of the vibrating beam subjected to stationary or moving mass is first studied by solving the equations of motion of both the beam and the mass. The methodology goes through three stages for both cases, a constant value load, a harmonic load, and a sprung mass. Existing damage detection techniques are reliant on monitoring the anomalies in the structure behavior. This approach depends on the application of Laplacian, second de-rivative, to the structure measured accelerations to localize the damages signature in the measurements. As the possibilities of damage are large, for location and severity, a statistical analysis is usually used to achieve the optimum scenario of inspection. Bayes’ Theorem is used to optimize accelerometers’ configuration for damage detection. As the implementation of Bayes’ Theorem requires the evaluation of numerous cases, this usually makes the problem impossible to be solved within a reasonable time. Therefore, Genetic Algorithm is usually used to obtain the optimum solution or a solution close to it. The last part is using drones with cameras to monitor the traffic over bridges. Traffic monitoring is the centerpiece of congestion mitigation and traffic management. Instead of using Convolutional Neural Network (CNN) to detect vehicles, a faster technique called YOLOv3 is used. This technique requires huge computational power. Therefore, GPUs are used to process the videos recorded by drones’ cameras. By calibrating the camera using real measurements compared to their apparent values in images, the detected vehicles can be tracked. The targeted features (herein, features correlated to traffic congestion) are re-produced utilizing a traffic simulation model. The proposed methodology was tested by collecting and investigating video images from drones.

Available for download on Friday, May 09, 2025

Included in

Engineering Commons