Bridges are critical parts of the modern transportation network. In recent decades, the aging degradation on structural capacity makes the bridges very vulnerable, which keeps threatening the public safety. This project aims to overcome the above problem by proposing a fully automatic bridge structural health monitoring (SHM) system. The proposed bridge SHM system combines the structural mechanics, deep learning, computer vision, signal processing, and sensing technology. As a result, the system can automatically 1) detect and quantify the early structural damages, 2) model the damage on the digital twin (finite element model), 3) analyze the structural capacity for real-time structural state awareness, and 4) export the analysis result to the researchers and the bridge owners.