The Digi4RailBridges project aims to develop an innovative digital platform for assessing the condition of railway bridges, combining data acquired from structural health monitoring systems with data collected through computer vision systems on UAVs. The data analysis is fully automated and based on artificial intelligence techniques, promoting a paradigm shift towards predictive maintenance. The proposed methodology will be validated on a national railway bridge, ensuring the practical applicability of the results.

The project consortium brings together expertise in the fields of structural monitoring, artificial intelligence, computer vision and numerical modeling, ensuring a multidisciplinary approach focused on the digitalization of the railway sector.

Main objectives of Digi4RailBridges:

• Deploy a distributed and non-intrusive wireless system to collect and transfer bridge data to the cloud.
• Establish procedures for remote inspection and reality capture using UAVs equipped with cameras and LiDAR sensors.
• Develop machine learning strategies based on time-series from structural monitoring for early and automatic damage detection.
• Apply computer vision techniques to images from remote inspections for the automatic classification of multiple anomalies.
• Enhance experimental data through advanced numerical modeling using the finite element method.
• Design a digital twin architecture for railway bridges capable of integrating the various developed technologies and methodologies.
• Create methods to fuse information from time-series and images, supported by a structural condition rating system to assess anomaly severity.
• Develop a tool for predicting the Remaining Useful Life (RUL) of bridges.
• Implement and validate the proposed digital framework on a real railway bridge.