Advanced Digital Twin and Machine Learning Technique for Sustainable Bridge Safety

The need for effective monitoring and early warning systems has become a priority in the management of infrastructure due to growing concerns about the structural integrity of bridges and their recurrence. Current methodologies lack the capability to provide real-time insights into critical parameters, thereby limiting the effective implementation of preemptive measures. This project addresses this gap by integrating machine learning and digital twin technology to enhance the safety and reliability of vital infrastructure. A Decision Tree Classifier is employed to analyze real-time sensor data, including strain, load, and displacement, enabling the accurate identification of structural anomalies. Simultaneously, a digital twin model is developed to visualize the bridge's condition and guide maintenance decisions. The proposed system achieves an accuracy of 92%, precision of 92.45% and recall of 94.70% demonstrating high reliability in detecting structural anamolies. The system is calibrated using historical data and validated with the current scenario. Combining machine learning with digital twin technology, this research highlights a novel approach in reducing the risk of catastrophic bridge failures, safeguarding critical infrastructure, and protecting human life.