The Walking Behaviour of Pedestrian Social Groups and Its Impact on Crowd Dynamics

By Mehdi Moussaïd, N. Perozo, Simon Garnier, Dirk Helbing, Guy Theraulaz. PLoS ONE, 2010

1,124 citations recorded by the source corpus.

Human crowd motion is mainly driven by self-organized processes based on local interactions among pedestrians. While most studies of crowd behaviour consider only interactions among isolated individuals, it turns out that up to 70% of people in a crowd are actually moving in groups, such as friends, couples, or families walking together. These groups constitute medium-scale aggregated structures and their impact on crowd dynamics is still largely unknown. In this work, we analyze the motion of approximately 1500 pedestrian groups under natural condition, and show that social interactions among group members generate typical group walking patterns that influence crowd dynamics. At low density, group members tend to walk side by side, forming a line perpendicular to the walking direction. As the density increases, however, the linear walking formation is bent forward, turning it into a V-like pattern. These spatial patterns can be well described by a model based on social communication between group members. We show that the V-like walking pattern facilitates social interactions within the group, but reduces the flow because of its “non-aerodynamic” shape. Therefore, when crowd density increases, the group organization results from a trade-off between walking faster and facilitating social exchange. These insights demonstrate that crowd dynamics is not only determined by physical constraints induced by other pedestrians and the environment, but also significantly by communicative, social interactions among individuals.

Subfields: Control and Systems Engineering, Ocean Engineering, Safety, Risk, Reliability and Quality

View Paper: https://www.semanticscholar.org/paper/8cc8bda43875b6a74993a4264d4a7233ff9554f0

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Control and Systems Engineering

In Ocean Engineering

In Safety, Risk, Reliability and Quality

Citations in the corpus

Human mobility: Models and applications (2017, background): “This controlled environment can provide insight into questions such as how pedestrians walk in social groups [263], the quantification of pressure and density in different types of bottlenecks and junctions [264], the perception of personal space [265], the clogging effects and instabilities in narr”
SR-LSTM: State Refinement for LSTM Towards Pedestrian Trajectory Prediction (2019, methodology): “Succeeding methods take more potential factors into account, such as pedestrian attributes [48, 52], walking group [30, 45], stationary group [48, 49].”
Crowd disasters as systemic failures: analysis of the Love Parade disaster (2012, background): “Walking in groups reduces the capacity further [91].”
Single-Pedestrian Detection Aided by Multi-pedestrian Detection (2013, background): “On the other hand, sociologists have found that pedestrians walking together show a few particular spatial patterns [22].”
Social-STGCNN: A Social Spatio-Temporal Graph Convolutional Neural Network for Human Trajectory Prediction (2020, background): “According to [19], 70% of pedestrians tend to walk in groups.”
Human motion trajectory prediction: a survey (2019, methodology): “…methods model group relations by introducing additional attractive forces between group members (Yamaguchi et al. 2011; Pellegrini et al. 2010; Singh et al. 2009; Qiu and Hu 2010; Karamouzas and Overmars 2012; Seitz et al. 2012; Moussaı̈d et al. 2010; Choi and Savarese 2010; Robicquet et al. 2016).”
The impact of social networks on animal collective motion (2011, background): “…reticulata, prefer to shoal with familiar conspecifics (Griffiths & Magurran 1999), the majority of pedestrians walk in small social groups (Moussaïd et al. 2010), and group fission in macaques (Macaca tonkeana and Macaca mulatta) is strongly influenced by individual affiliations (Sueur et…”
Learning socially normative robot navigation behaviors with Bayesian inverse reinforcement learning (2016, methodology): “For simulation, we use an open-source1 pedestrian simulator [22] in which agents are guided by potentials governed by the social force model [23].”
Empirical study on social groups in pedestrian evacuation dynamics (2017, background): “Larger groups tend to split up to smaller groups of two or three members [3, 5, 13, 14, 15].”
A Survey of Recent Advances in CNN-based Single Image Crowd Counting and Density Estimation (2017, background): “topic with researchers from di↵erent communities (such as sociology [68, 10], psychology [5], physics [13, 38], biology [72, 110], computer vision and public safety) have addressed”