Abstract for: Entering the Festival Grounds – A Computational Model of Crowd Dynamics during the Vierdaagsefeesten

Implementing crowd management to ensure safety and security for people attending public gatherings is highly challenging. However, decision-making is often based on intuition and this can fall short in high-risk and escalating situations. Therefore, there is a need for model-driven approaches that provide realistic and practical computational models to effectively support decision-making during events. Often, agent-based modeling (ABM) is used in this field to capture complex behavioral feedback mechanisms of people. This study combines ABM and system dynamics to model crowd behavior at the Vierdaagsefeesten, the case used in this research. Within the model, the focus is on inter- and intrapersonal factors that shape overall crowd dynamics at large-scale events. A simulation model is being built iteratively, using existing literature to identify relevant variables and mechanisms for crowd dynamics. The current proof-of-concept model includes visitor agents with characteristics like type of the visitor, group membership, and goal-seeking behavior, set within a real map of the Vierdaagsefeesten. Behavioral mechanisms are still being worked out, though initial results show realistic behavior. The current model will be developed further, using real visitor data from the Vierdaagsefeesten to validate behavior and measuring how accurately the model predicts crowd dynamics. Furthermore, next steps include enriching agents with detailed intra- and interpersonal mechanisms, integrating system dynamics in the model, and connecting visitor preferences to behavior. The last step will be to test various scenarios, and scaling up to simulate larger number of visitors.