In order to maintain the efficiency and safety of transportation networks, Ensuring the structural integrity of bridges is of utmost importance. This study presents a data-driven framework for detecting and localizing structural damage based on anomaly detection in the bridge’s structural responses. The methodology is applied to the Old ADA Bridge, represented by a Finite Element model refined through Bayesian updating via the Transitional Markov Chain Monte Carlo (TMCMC) technique. The interaction between passing vehicles and the bridge is characterized using a Vehicle-Bridge Interaction (VBI) model, facilitating a more accurate assessment of structural behavior. To detect and localize damage, the Matrix Profile technique is employed to identify anomalies in sensor-acquired signals. The proposed approach demonstrates a high degree of efficacy in detecting structural changes using a minimal dataset, thereby eliminating the need for prior supervised learning. This feature enhances its practical applicability, making it a viable tool for real-world bridge health monitoring.