Modeling Passenger Boarding Times Using Sumonity’s Sub-Microscopic Pedestrian Simulation

Abstract

Short dwell times at bus stops are crucial for efficient public transport operations, yet existing traffic simulation tools commonly simplify passenger boarding. In this paper, we extend the SUMO-based co-simulation framework, Sumonity, to incorporate a sub-microscopic pedestrian model for city bus boarding. Our approach simulates real-time passenger flow, pathfinding, and door congestion in a Unity-based environment. We conduct a full-factorial simulation experiment with four bus door configurations, ranging from fully open double doors to partially closed options. We also consider different passenger loads between 1 and 50, yielding 200 unique scenarios. Detailed spatiotemporal data on passenger movements and boarding times are generated for each scenario. Analysis of crowding behaviors and door usage reveals significant sensitivity of boarding times to both passenger volume and door availability. These findings demonstrate the importance of accurately modeling pedestrian interactions for reliable dwell-time forecasts and underscore the potential of sub-microscopic pedestrian simulations.