Term of Award

Summer 2025

Degree Name

Master of Science, Civil Engineering

Document Type and Release Option

Thesis (open access)

Copyright Statement / License for Reuse

Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.

Department

Department of Civil Engineering and Construction

Committee Chair

Marcel Maghiar

Committee Member 1

Gustavo Maldonado

Committee Member 2

Soonkie Nam

Abstract

Mechanically Stabilized Earth (MSE) retaining walls are durable, settlement tolerant, easy to install, and present cost-effectiveness, in highway bridges. Nevertheless, the cost of repairing MSE walls is high, and it is crucial to have an accurate, quantitative, and scalable inspection method for early detection of potential problems. This study aims to analyze and evaluate undesired displacements that may affect these walls over time. For this, two remote-sensing technologies are considered: Light Detection and Ranging (LiDAR) and 3D Close Range Photogrammetry (CRP). Their performances are compared against an accurate Robotic Total Station (RTS) device, serving as the benchmarking instrument. This work investigated two distinct case studies, at different sites, B1-Old River and B2-Sandersville. At the first site, both LiDAR and RTS were used and analyzed at two different times. At B2–Sandersville, RTS, LiDAR, and CRP were each implemented at the same time. At both sites, several Ground Control Points (GCPs) were established, also spatial data was collected and processed for noise removal, alignment, and georeferencing. LiDAR data was processed via Leica’s Cyclone Core, while photogrammetric data was handled via Agisoft’s Metashape. Finally, at both sites, C2C and M3C2 distance analysis were performed using CloudCompare. First, At B1-Old River, 12 checkpoints were considered, and the resulting comparisons and statistics assisted in estimating the noise level associated with three different measuring/modeling techniques. Those levels are: 9.7 mm for RTS, 11.9 mm for target-based (TB) LiDAR, and 14.3 mm for visual-alignment LiDAR. However, the CloudCompare analyses of a comparable pair of full point clouds, considering millions of points, indicated a 95% probability of attaining absolute positional discrepancy magnitudes, between two TB LiDAR models, equal or less than 8.0 mm, via the C2C scheme, and equal or less than 4.4 mm when using the M3C2 scheme. That is, accurate TB LiDAR and CloudCompare is an auspicious combination to capture small displacements. Additionally, the analyses completed at B2-Sandersville indicate that two other LiDAR techniques, with modified procedures, provide levels of accuracy close to TB LiDAR, and one of them only requires targets for georeferencing. These findings provide a framework for scalable monitoring of MSE walls.

Research Data and Supplementary Material

No

Download

Files over 10MB may be slow to open. For best results, right-click and select "Save as..."

Share

COinS