Drive-by damage detection in bridges
Development of methodologies based on Artificial Inteligence for damage dection in railway bridges
Abstract
One of the major challenges in managing the bridges that are part of the national railway infrastructure is ensuring their safety and integrity throughout their service life. In this context, the construction of a Digital Twin (DT) is crucial to assist in the decision-making process for defining optimized maintenance policies, rather than the traditionally employed time-based approaches. Indeed, having an up-to-date digital replica of the physical asset can bring various benefits throughout its lifecycle, including monitoring its structural integrity and performance. A DT requires the acquisition of data from structural monitoring to feed the digital model (information flow from the real system to the virtual model). However, it is unfeasible to directly monitor all the bridges in a railway line. A scientifically and practically interesting alternative is indirect monitoring, which is based on installing sensors on a service train to capture responses from the dynamic interaction of the vehicle-track-bridge system. The main advantage of this approach is the scalability of the solution, enabling the monitoring of all the bridges in a railway line by instrumenting only one vehicle. In this context, the present project aims to propose a methodology for constructing DTs of railway bridges using indirect monitoring data, based on artificial intelligence (AI) techniques. The developments proposed in this project will use AI techniques to: (i) develop a method for detecting, locating, and quantifying damages using indirect monitoring data, which will enable the information flow from the real system to the virtual model, and (ii) propose a methodology for constructing a digital twin that incorporates this method and uses the information from the virtual model to make decisions about structural maintenance. The resources from this project will enable the acquisition of real data from a railway bridge, which is indispensable for the practical validation of the proposed developments. It is expected that these developments will assist in the decision-making process for managing special structures, bridging the gap between theory and practice, and contributing to the advancement of the efficiency and safety of the national infrastructure.
Team
Prof. Rafael Holdorf Lopez
Prof. Diogo Ribeiro (ISEP)
Prof. Leandro Fadel Miguel
Thiago Moreno Fernandes
Gabriel Padilha Alves
Reports
17/10/2024: Relatório_FTC__Ferrovia_Tereza_Cristina_171024