AI-Powered Digital Twins of Concrete Bridges for Enhanced Asset Management

Abstract

Ensuring the long-term sustainability of bridge assets is crucial, but manual inspections are often time-consuming, laborious, and can be hazardous. These issues can be addressed by leveraging Artificial Intelligence (AI) with Digital Twin technology for enhanced monitoring and maintenance. This approach automates defect identification, reduces field time, and simplifies generating repair documentation, shifting from traditional to proactive, data-driven asset management. In this study, the condition assessment of concrete bridge decks involves three steps: data acquisition, interpretive analysis, and visualization. Autonomous Unmanned Aerial Vehicles (UAVs) are utilized to capture high-quality images of the bridge, which are then stitched into 3D models using photogrammetry. Deep Learning algorithms detect defects such as cracking and spalling in these images, significantly reducing manual inspection time. The Digital Twin is visualized in an Asset Management Portal (AMP), providing stakeholders with a clear, interactive representation of the bridge's condition. This enables timely interventions and historical tracking of defects. By leveraging UAVs, photogrammetry, and ML/AI, the transformative potential of Digital Twins in Structural Health Monitoring can be fully realized.