In this project we want to develop a digital twin of a real bridge (Nibelungenbrücke). The bridge is equipped with a variety of sensors that continuously measure temperatures, strains, accelerations or inclinations. In parallel, we have developed a virtual representation of the bridge using a finite element model. In a first test, this is a thermo-mechanical model implemented in FEniCSx. At regular intervals, the data is queried from the data server to update the simulation model using Bayesian inference approaches including uncertainties. Based on the updated simulation model, additional quantities of interest can be computed, ranging from temperatures in any part of the structure (not only at measured locations), but also deformations due to thermal expansion, stresses. In addition, we aim to integrate the weather forecast to allow virtual sensors to "measure" or predict the temperature distribution inside the bridge in the future. The final goal is to automatically detect the onset of damage or structural degradation and automatically take actions ranging from warning messages to the administration to immediate measures such as a complete closure of the bridge to prevent fatalities. The goal of the project using the de.NBI cloud service is to test how we can provide a user interface to our simulation models to other partners in the SPP, with the option to directly run their queries based on a shared computing infrastructure. The results can be made available via a REST API that can be accessed by any partner without the need to maintain individual compute infrastructures.