Development of a Framework for Automated Through-Plate Girder Railroad Bridge Impact Detection and Condition Assessment
Advisor: Professor Billie F. Spencer Jr.
Abstract
Of the 100,000 railroad bridges in the United States, 50% are over 100 years old. Many of these bridges do not meet the minimum vertical clearance standards, making them susceptible to impact from over-height vehicles. These impacts can cause structural damage, service disruptions, and financial losses, necessitating a systematic approach for rapid detection and condition assessment. This dissertation develops a comprehensive framework for automated impact detection and condition assessment of railroad bridges using artificial intelligence, wireless smart sensors, and structural health monitoring techniques.
The first step in this framework is rapid impact detection, achieved through a machine learning-based event classification that distinguishes between impact events and train crossings with high accuracy. The detection step is enhanced by integrating artificial intelligence at the edge, deploying the classification model directly onto a wireless smart sensor platform. This edge computing paradigm eliminates the delays associated with centralized processing, enabling near real-time decision-making for infrastructure monitoring.
Beyond detection, the framework incorporates impact severity assessment, a critical step for prioritizing inspections. An artificial neural network model is developed to assess severity using key structural response metrics such as impact impulse, peak acceleration, and spectral energy. Validated with both simulated and field-collected data, this model provides an automated and scalable solution for prioritizing the allocation of limited inspection resources. The final step in the framework is post-impact condition assessment where a neural network-based approach estimates permanent displacements to determine the residual structural integrity of bridges.
By systematically integrating these components, this research establishes a robust and automated framework for railroad bridge impact detection and condition assessment. The proposed methodology enhances safety, minimizes operational downtime, and optimizes maintenance resource allocation for aging railroad bridges.