AI-Driven Rapid Condition Assessment for Masonry Arch Bridges

ongoing

Masonry Arch Bridge (MAB) still plays a critical role within railway and roadway networks in Europe and worldwide. MABs constitute the majority of the European bridge stock. These structures were mainly built following empirical rules and were not designed to resist current traffic loading. Furthermore, nowadays, MABs are confronting the challenges of climate change and severe environmental conditions that degrade their performance. An accurate assessment of these complex structural systems represents a crucial step to preventing future failures and preserving them for the next generations. AIdaMArch introduces a tailored structural health monitoring approach that leverages a two-step data-driven methodology based on Long Short Term Memory (LSTM) models. The LSTM model enhances structural health monitoring by rapidly analyzing large amounts of sequential sensor data for real-time, continuous and adaptive monitoring of structures. MAB sensor measurements are not able to be analyzed by data-driven methods due to a lack of proper training data. To simulate the structural behaviour, a novel numerical model will be developed, and the data set will be augmented by a cyclic generative adversarial network for a higher level of applicability. AIdaMArch possesses the versatility of using a two-step data-driven method to not only detect anomalies (damage detection) but also extend its application to identify the severity of damage in MABs by the small number of measurements. The main goal of this proposal is a continuous, rapid, and accurate methodology for the health monitoring of MABs. To ensure the successful implementation of the project, Dr Emadi will conduct research at the Imperial College of London, with the guidance of Prof. Macorini, a leading scholar in the field, as a main supervisor. This project will prepare Dr Emadi to become a tenured scholar and obtain advanced funding, strengthening his academic profile and allowing him to establish his own research group.