AI– DRIVEN AUTONOMOUS BRIDGE HEALTH MONITORING USING UAVS AND DEEP REINFORCEMENT LEARNING
DOI:
https://doi.org/10.62643/Abstract
The safety and serviceability of bridge infrastructure are essential for maintaining reliable transportation networks. Conventional bridge inspection practices rely heavily on manual visual assessment, which is laborintensive, time-consuming, and often exposes inspectors to hazardous environments. This paper presents an intelligent bridge health monitoring framework that combines Unmanned Aerial Vehicles (UAVs), deep reinforcement learning, and deep learning–based damage assessment to automate structural inspections. The proposed framework enables a UAV to navigate bridge surfaces autonomously using a Proximal Policy Optimization (PPO) agent that continuously adapts its flight path in the presence of dynamic obstacles such as moving vehicles. High-resolution images acquired during inspection are processed through a transfer learningbased convolutional neural network to identify structural defects, including surface cracks, with high precision. Image enhancement and edge-based preprocessing techniques improve feature extraction under varying environmental conditions. A simulated inspection environment is employed to train and validate the navigation strategy before deployment. Experimental evaluation demonstrates that the integrated framework achieves faster inspection coverage, reliable crack localization, and improved operational efficiency compared with conventional inspection approaches. By combining autonomous navigation with intelligent visual analytics, the proposed system supports continuous structural assessment, minimizes human intervention, and provides a scalable solution for predictive maintenance of transportation infrastructure. The framework has the potential to enhance inspection safety, reduce maintenance costs, and improve the long-term reliability of critical bridge assets. Keywords: Bridge Health Monitoring, Structural Health Monitoring (SHM), Unmanned Aerial Vehicle (UAV), Deep Reinforcement Learning, Proximal Policy Optimization (PPO), Transfer Learning, Convolutional Neural Network (CNN), Crack Detection, Computer Vision, Predictive Maintenance, Infrastructure Inspection, Autonomous Navigation.
Downloads
Published
Issue
Section
License

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.













