A Digital Twin (DT) serves as a virtual counterpart to real-world entities, including structures, systems, and components, created to interpret and provide insights about their physical counterpart, the Physical Twin (PT). These twins are dynamically updated via sensor networks and Internet of Things (IoT) technologies. In the DT, predictive capabilities are enabled through data-driven algorithms and modeling tools. Despite these advancements, accurately modeling system performance is complicated by numerous uncertainties present throughout their lifecycle. To address these challenges, the Risk-Informed Digital Twin (RDT) concept has emerged. RDT integrates probabilistic predictions informed by physics and data within the Sustainable Resilient Engineering (SRE) framework. SRE extends Performance-Based Engineering by incorporating Bayesian methods and leveraging Artificial Intelligence tools. This study explores the initial development of an RDT demonstrator, focusing on the butterfly-arch stress-ribbon pedestrian bridge in Fuzhou, Fujian, China. A tri-axial wireless sensor network with high synchronization was installed on the bridge deck to capture ambient vibration data. Utilizing Operational Modal Analysis, four distinct finite element models of the bridge were created. These digital models inform the prior configurations used in SRE, enhancing predictions of the bridge's performance, particularly for extreme scenarios and long-term evaluations.