Digital Twin: a Hybrid Approach for Structural Health Monitoring

The scope of this work is to illustrate the advantages that can be obtained in the context of structural health monitoring (SHM) when data-driven and model-based approaches are combined through the construction of a numerical twin of the structure. While the strategy isn't entirely novel per se, the use of wholly integrated technologies and software developed by the same company for all parts of the workflow, preventing data loss and ensuring interoperability, is where the originality lies. ASDEA S.r.l. provides products designed to perform each part in the SHM cycle, spanning from data acquisition to alert emission and damage management. The paper describes how the pieces are put together inside the coherent environment provided by the STKO software, initially designed as a powerful interface to the OpenSees solver for finite element methods (FEM). Data is acquired through a network of MonStr sensors (produced by ASDEA Hardware), managed using artificial intelligence (AI). The data is then exposed to near-real-time analysis to obtain an accurate picture of the structural conditions, and the numerical model is updated continuously to reflect present conditions. When anomalies are detected by the AI-based classifier, they are compared to the output provided by the FEM analysis to ensure reliability.