A Multi-Layer Navigation Approach for Interactive Pedestrian Flow Simulation in Digital Twins

Abstract

Pedestrian flow simulation is crucial for accurately depicting daily activities and dynamics of infrastructures, such as town halls, train stations, or airports. Current pedestrian flow models often lack the capability to interact with environmental changes in real-time or only focus on one-directional interactions via prescribed events. To address this limitation, we propose a hybrid approach that combines graph-based methods for large-scale navigation with the optimal steps model for small-scale navigation and locomotion of agents. This combination enables dynamic updates according to environmental changes provided by other simulations. We demonstrate the effectiveness of our proposed approach in an exemplary airport architecture where pedestrian simulation is coupled with an electrical simulation, resulting in a successful bidirectional coupling. Specifically, we consider a scenario where a saboteur agent meddles with an electrical circuit, causing a ripple effect that impacts ped estrian behavior.