Using transfer learning technique to define seismic vulnerability of existing buildings through mechanical models

The paper presents a methodology to derive seismic vulnerability of existing reinforced concrete buildings by using mechanical models derived from data obtained by transfer learning technique. In detail, the methodology aims to exploit the combination of some structural building features obtained from the analysis of photographs within the VULMA framework and additional general information obtained from other sources, such as year of construction, localization and covered area. These data are joined and matched allowing to derive the input data for simulating an ideal mechanical model, which is elaborated and analyzed under seismic actions. The method is firstly presented in its general features and then applied to a couple of existing buildings, with the objective to derive seismic vulnerability. The results of the analyses show have highlighted that the methodology can be applied to define risk mitigation strategies in a focused area, relying on two automatization processes.