SYNTHETIC DATA GENERATION FOR THE CREATION OF BRIDGE DIGITAL TWINS WHAT-IF SCENARIOS

Abstract

The Digital Twin (DT) concept, as understood nowadays, appeared in the early 2000s as an attempt to create virtual replicas of physical assets, such as bridges, that can be used to examine, monitor and manage their performance. Up to this day, it has been successfully applied in the fields of aeronautics, manufacturing, medicine, and more recently, in the architecture, engineering, and construction industry. The DT of a bridge requires the creation of a virtual replica of the real-life asset, along with the connection and feedback of information channel between the two of them. This connection is currently achieved through the generation of real-time data by the placement of sensors in the real bridge and the application of structural health monitoring techniques to analyze such data. This connection could result in a complex, time-consuming, and expensive process which would hinder the creation of DT prototypes for development purposes in the bridge engineering field. This paper aims at exploring the currently available synthetic data generation methodologies and tools, which could be used as a faster and a more economically feasible alternative to real monitoring, for the creation and development of DT prototypes of bridges, for both industry and research-oriented purposes. A synthetic data generation framework is proposed that can produce FAIR benchmark databases that are based on Findability, Accessibility, Interoperability, and Reuse, which could be used in the prototyping of bridge DTs. Finally, tentative future improvements in this topic are discussed.