Macroscopic directional analysis of pedestrian flow

Incidents at mass gatherings all over the world continue to raise concern about pedestrian safety. This rising concern calls for researchers to understand the underlying dynamics behind pedestrian and mass movements. The literature offers several approaches that reproduce reality to varying degrees, simulate it, and model it under different circumstances. Across these approaches, limited effort has been invested in looking at the projection of flow measures along different directions of travel (e.g., longitudinal versus lateral movement in omnidirectional environments) at the macroscopic level. To address this gap, this paper looks at directional flows and speeds and how they relate to each other at the macroscopic level. The results from analysis of experimental three-dimensional trajectory data show that a macroscopic analysis of directional travel, particularly of lateral travel, helps explain when and to what extent sidestepping occurs in the corridor. An increase in areawide density results in a decrease in flow in the direction of travel but increases the lateral direction flow. Furthermore, the relationship between speed and density in the lateral direction delineates two regimes in the corridor, whereby pedestrians do not increase their speed in the lateral direction until areawide density levels reach magnitudes that cause an areawide capacity drop in the system. The work in this paper therefore suggests that there is useful insight to be gained by complementing nondirectional flow analysis with directional flow analysis.