From Manual to Machine: Leveraging Deep Learning for Safer Bridge Inspection

Disciplines

Civil Engineering

Abstract (300 words maximum)

The United States faces a critical challenge with its aging infrastructure, as nearly one in three bridges is considered structurally deficient. Millions of daily trips depend on these deteriorating structures, highlighting the urgent need for efficient and reliable inspection methods. Traditional manual bridge inspections are time-consuming, costly, and prone to human error, motivating the development of automated approaches. This study investigates the application of artificial intelligence (AI) and deep learning for detecting structural deterioration in steel bridges. A convolutional neural network (CNN) model was developed to identify cracks in steel components using visual data. A labeled dataset of crack images was collected from publicly available sources, and data augmentation and transfer learning techniques were employed to improve model generalization and robustness. Model performance was evaluated through accuracy and precision metrics, with iterative refinement to optimize results. The CNN demonstrated strong performance and effectively detected cracks in test images, indicating its potential to support faster, more accurate, and cost-efficient bridge inspections.

Use of AI Disclaimer

no

Academic department under which the project should be listed

SPCEET – Civil and Environmental Engineering

Primary Investigator (PI) Name

Da Hu

This document is currently not available here.

Share

COinS
 

From Manual to Machine: Leveraging Deep Learning for Safer Bridge Inspection

The United States faces a critical challenge with its aging infrastructure, as nearly one in three bridges is considered structurally deficient. Millions of daily trips depend on these deteriorating structures, highlighting the urgent need for efficient and reliable inspection methods. Traditional manual bridge inspections are time-consuming, costly, and prone to human error, motivating the development of automated approaches. This study investigates the application of artificial intelligence (AI) and deep learning for detecting structural deterioration in steel bridges. A convolutional neural network (CNN) model was developed to identify cracks in steel components using visual data. A labeled dataset of crack images was collected from publicly available sources, and data augmentation and transfer learning techniques were employed to improve model generalization and robustness. Model performance was evaluated through accuracy and precision metrics, with iterative refinement to optimize results. The CNN demonstrated strong performance and effectively detected cracks in test images, indicating its potential to support faster, more accurate, and cost-efficient bridge inspections.