A Monitoring Strategy for Recurrent Structures

Most of the infrastructural heritance will be monitored, in this very close future. Hence, one challenging aspects of Structural Engineering is the development of a successful monitoring strategy. However, a general solution that can provide concrete tools is not yet defined, since Structural Health Monitoring (SHM) is now studying “healthy” or “pseudo-healthy” existing structures. A second critical aspect that is now limiting the spreading of SHM is that most bridges and viaducts existing nowadays are old structures that were not inspected for most of their service lives and that now may show some criticalities whose cause is unknown. The present work marks the importance of monitoring recurrent structures that were designed according to similar criteria and that are exposed to similar weather conditions. This type of constructions can be luckily found in the structural patrimony since constructions that belong to the same site and period were usually built by the same company. The method proposed can build a robust mathematical model of a group of structures and alert in case of not-correspondent behaviour. In the pre-processing stage, data is cleaned and filtered, while in the post-processing stage the analysis is carried on with Gaussian Processes Regression. Lastly, if the data is in accordance with the original mathematical model, a process of data cleansing is carried on strengthening the model using the latter data of monitoring. The process can be applied on local or global scale, and in the present paper it is finally provided an application to a case study.