Smart Health Monitoring of Concrete Bridges Using Digital Twin and Ai Applications

Safety and sustainability of reinforced concrete bridges may be increased by observing their condition during operation and thus accurately predicting their behaviour under various load conditions. This can be achieved through a monitoring system and automatic error detection based on the measured data. By detecting potential issues early on, significant damages can be prevented before they occur. Despite extensive data collection from many monitored bridges, this data often remains unprocessed and uninformative in its raw form. We aim to transform this data into a format that can help to estimate a bridge’s health condition. This approach is presented through a case study of an existing reinforced concrete box girder bridge in Hungary. Digital twin (DT) technology was used to simulate the bridge’s behaviour and to verify structural conditions under any given traffic load arrangement. Static calculations and verification of load-bearing and serviceability conditions were performed on a validated 3D finite element (FE) model. Different traffic load scenarios were randomly generated using Monte Carlo simulation, and the bridge’s condition was evaluated for each case. The actual condition was quantified by parameters such as the bridge’s utilization for different USL and SLS limit values, especially for deflection and crack width. In the FE model, the physical characteristics that are recorded on the real bridge by the actual measuring instruments were also recorded at the locations corresponding to the monitoring points on the actual structure. The relationship between the virtual bridge’s condition and the virtual monitoring data was determined using artificial intelligence (AI) applications, particularly artificial neural networks (ANN) . Based on this relationship, the monitoring data measured on the real bridge can be processed, and predictions about the bridge’s actual condition can be made to support maintenance and improve the safety and sustainability of the structure. This approach demonstrates the potential of DT and AI in structural health monitoring techniques.