(2025) Automated Assessment of Defect in Bridge Structures, SPR-RE22(014)-8H-00, 2025. Transportation, Department of
|
PDF
SPR_RE22_014_8H_00_automated_assessment_of_defect_in_bridge_structures_final_report.pdf File Size:11MB |
|
|
PDF
SPR_RE22_014_8H_00_automated_assessment_of_defect_in_bridge_structures_tbrief.pdf File Size:3MB |
Abstract
Effective and timely bridge inspections are crucial for extending bridge lifespans and preventing catastrophic failures. Traditional inspection methods often involve manual visual assessments and can be time-consuming, labor-intensive, and prone to human error. Recent technological advancements in unmanned aerial vehicles (UAVs), artificial intelligence (AI), and machine learning (ML) offer promising solutions to these challenges. When high-quality images captured by UAVs are analyzed using AI and ML algorithms, structural defects can be detected and quantified with greater precision and efficiency than manual inspections. The primary objective of this research was to enhance the accuracy and efficiency of structural inspections by integrating UAV technology for image capture and AI-based detection models for analysis. High-resolution images of bridge components were collected using UAVs operating at various distances and angles and were then processed through a custom-developed convolutional neural network (CNN) to detect critical defects such as cracking and spalling. The model’s performance was assessed through multiple case studies, and its ability to detect and quantify defects under different conditions was validated against field data. This approach yielded significant improvements over traditional bridge inspection methods in terms of the precision with which structural vulnerabilities were identified and accurately quantified defect dimensions. Furthermore, the research incorporated the development of three-dimensional (3D) models of bridge structures using commercially available software to enable detailed structural assessments. High-resolution UAV imagery was successfully integrated into 3D modeling software to generate detailed models of bridge structures enabling comprehensive structural assessments and allowing for the quantification of detected defects. The results demonstrate the potential of UAV-based inspections combined with AI-powered detection models to revolutionize bridge inspection practices by offering a more reliable, efficient, and cost-effective approach to infrastructure maintenance and supporting more informed decision-making for infrastructure safety and longevity.
| Item Type: | Departmental Report |
|---|---|
| Keywords: | artificial intelligence—bridge inspection—convolutional neural network—unmanned aerial vehicles—machine learning |
| Subjects: | Transportation Transportation > Bridges and tunnels Transportation > Research Transportation > Data and Information Technology |
| ID Code: | 53343 |
| Deposited By: | Iowa DOT Research |
| Deposited On: | 10 Jun 2025 20:58 |
| Last Modified: | 10 Jun 2025 20:58 |
| URI: | https://publications.iowa.gov/id/eprint/53343 |