Date of Award

Fall 2026

Abstract

The importance of bridge infrastructure is significant within the world of transportation. They provide quick and safe passage over obstacles that would otherwise impede the movement of goods and people. As bridges begin to age, it is important to perform inspections to assess the continued capabilities of the bridge’s performance. In some cases, lack of historical data and/or difficulty in accessibility can make it difficult to determine the reliability of the bridge. An efficient way to get information about bridge components and dimensions is to use Unmanned Aerial Vehicles (UAV) fit with cameras and Light Detection Ranging (LiDAR) sensors. This study provides the development of a methodology for the assessment of the safety level of wide-flange steel girder bridges using data collected from UAV LiDAR The variables used in the prediction models were selected to be bridge span length, number of spans, deck width, deck thickness, total number of girders, and depth of girders. A Taguchi Design of Experiments was performed to create a synthetic database for bridge demand prediction models. The bridges within the database were modelled in CSiBridge software to determine the maximum girder demand for both shear and moment in positive and negative cases for both dead and live loads. The use of linear regression in the creation of the predictive models was found to be ineffective, so nonlinear predictive models were investigated. Specifically neural networks provided predictive models that were satisfactory. The reliability of the methodology for safety level assessment was evaluated through Monte Carlo Simulations. The predictive models were then used to create a MATLAB code that can be used to determine if the controlling girder of a bridge is safe enough to cross given the inputs of the predictive model.

Document Type

Master's Thesis

First Advisor

Philippe Kalmogo

Second Advisor

Erin Bell

Third Advisor

Eshan Dave

Department or Program

Civil Engineering

Degree Name

Master of Science

Download

Share

COinS