Emergency departments provide immediate medical care to patient with acute conditions. Uncertainties in patient flows are a major challenge in emergency departments. Patients arrive unplanned and their needs are not clear upon their arrival. The prediction of the patient’s length of stay and system load is barely possible. The concept of patient (product) and resource visualization in the digital shadow can reduce uncertainty by providing real-time prediction. However, in literature there are barely concept implementations available. This work contributes by proposing a concept for real-time prediction, that is based on process mining and data-driven simulation. The paper describes the implementation of the concept and provides an illustrative case study. Patient and resource virtualization enhance system control by providing predictive information, that were not available before. Further, this paper evaluates the feasibility of transferring to an operating emergency department and the required performance.