Silva and Marcos' Digital Data Collection Approach for Unconventional Superstructural Components

Retrofitting of superstructural components (SCs) is very important in order to strengthen and prolong their useability however, the appearance of SCs becomes unconventional. Due to the retrofitting results, the strength of unconventional superstructural components (USCs) could not be predicted by the codes and standards. This study gives insight into how to predict the strength of USCs using an artificial neural network (ANN). The dataset of ANN training was digitally collected from the model simulation of the finite element method (FEM). Moreover, sensitivity analysis (SA) was performed to evaluate the variables. This study presents a generalized methodology and results. Utilizing a synthesis of the FEM for a set of data, the ANN for forecasting, and the SA for assessment would benefit to the Industry 4.0 transition and digital transformation.