Article Transportation Research Record 1–12 Ó National Academy of Sciences: Transportation Research Board 2018 Article reuse guidelines: sagepub.com/journals-permissions DOI: 10.1177/0361198118796714 journals.sagepub.com/home/trr Traffic Flow of Merging Pedestrian Crowds: How Architectural Design Affects Collective Movement Efficiency Zahra Shahhoseini 1 and Majid Sarvi 1 Abstract The need for developing reliable and rigorous models that can replicate and make predictions of pedestrian crowd evacua- tions has necessitated an understanding of the impact of architecture on individuals’ interactions with their surroundings and the behavioral rules that govern their movements. Due to the challenges of providing such behavioral data from natural eva- cuations and previous crowd incidents, simulation-based and laboratory-based evacuation experiments have recently been employed as innovative data-provision approaches to study crowd behavior notably under emergency conditions. This study explores pioneer experiments of emergency escape with a view to investigating the relationship between spatial constraints and collective behavior of human crowds. Here, we make use of two types of empirical and analytical data obtained from a large number of well-controlled laboratory and evacuation simulation experiments. This study presents findings correspond- ing to how and to what extent the presence of conflicting layouts in egress areas, particularly merging corridors, affect the collective motion of pedestrians. The focus of attention will be on measures of performance at macroscopic level derived from both observations. Our results suggested that the movement patterns observed in both types of experiments are sensi- tive to the angle between the two merging streams and the symmetry/asymmetry of the merging layouts, with symmetric lay- outs almost invariably outperforming the asymmetric counterparts. Also, within each symmetry/asymmetry structural type, the angle at which the flows combined with each other affected the efficiency of discharge. Our findings provide further evi- dence as to the significant role of the architectural structure of the movement area in facilitating the traffic flow of heavy crowds of pedestrians. A great deal of attempt has been made recently to under- stand how humans behave when fleeing a confined, crowded space. The answer to this question has impor- tant implications for planning safe and efficient evacua- tions as part of the risk assessment process essential for any mass event (1). The risk assessment conception can apply to a wide range of applications such as organizing sports events, music concerts, to name a few. In the sim- plest form, one might raise these basic but vital questions as to the time that it is going to take for the entire crowd to be safely evacuated in case of an emergency and if there is anything we can do to expedite this process. This question has guided a rapidly growing area of research focusing on developing prediction modeling frameworks (2–27) based on laboratory experiment (human and non- human subjects) and simulation experiments that have proved to be practical for assessing the emergency per- formance of built environments. A major challenge that analysts face towards provid- ing a reliable answer to such questions is that the impact of the design of escape areas as important determinants of crowd dynamics in evacuation scenarios are poorly understood. As is suggested by Pinter-Wollman, Fiore and Theraulaz (28), the architecture of built environ- ments, especially the layout of egress areas, strongly influences the collective behavior of crowds especially at mass gatherings. Since human interactions with each other and their interaction with the physical features of escape areas can considerably impede the collective movements of the panicked crowd and result in danger- ous bottlenecks and loss of time, it might culminate in injuries or loss of life. In social science, Kabo et al. (29) have found that the design of buildings, particularly 1 Centre for Disaster Management and Public Safety, Department of Infrastructure Engineering, The University of Melbourne, Australia Corresponding Author: Address correspondence to Zahra Shahhoseini: zahra.shahhoseini@unimelb.edu.au