This study presents an integrated approach toward developing a digital twin of the KW-51 bridge, combining real-time structural response prediction and machine learning-based damage identification. This framework effectively addresses the challenges associated with traditional finite element (FE) modeling and damage assessment in civil engineering, providing a comprehensive structural monitoring and analysis solution. The first part of the framework employs Deep Operator Networks (DeepONet) for real-time prediction of structural static responses. Structural balance laws guide this physics-informed network and eliminate the need for repetitive FE modeling, allowing for near-instantaneous predictions across the entire structural domain. The method demonstrates significant computational efficiency, achieving high accuracy with less than 5% error in structural response predictions. By leveraging hybrid loss functions that incorporate principles of energy conservation and static equilibrium, this approach enhances the reliability of predictions and minimizes training time, confirming its potential as an efficient tool for real-time structural analysis. The second part of the framework focuses on a machine learning-based damage identification method (CMLDI), which integrates modal analysis and dynamic analysis strategies. This method is applied specifically to the KW-51 bridge, utilizing signal processing techniques alongside modal inputs and acceleration data to inform damage detection, localization, and magnitude assessment. The CMLDI employs advanced machine learning techniques, including stacked gated recurrent unit (GRU) networks for detecting damage existence, k-nearest neighbor (kNN) classifiers for estimating damage magnitude, and convolutional neural networks (CNNs) for identifying damage locations. The results indicate high accuracy, efficiency, and robustness in damage identification, demonstrating the method’s effectiveness in utilizing minimal monitoring data. By combining these two methods, this research establishes a robust pathway for real-time structural analysis and damage identification, laying the foundation for a fully integrated digital twin of the KW-51 bridge.