Pedestrian evacuation simulation in the presence of an obstacle using self-propelled spherocylinders

Social aspects of collision avoidance: a detailed analysis of two-person groups and individual pedestrians

Pedestrian groups are commonly found in crowds but research on their social aspects is comparatively lacking. To fill that void in literature, we study the dynamics of collision avoidance between pedestrian groups (in particular dyads) and individual pedestrians in an ecological environment, focusing in particular on (i) how such avoidance depends on the group's social relation (e.g. colleagues, couples, friends or families) and (ii) its intensity of social interaction (indicated by conversation, gaze exchange, gestures etc). By analyzing relative collision avoidance in the "center of mass" frame, we were able to quantify how much groups and individuals avoid each other with respect to the aforementioned properties of the group. A mathematical representation using a potential energy function is proposed to model avoidance and it is shown to provide a fair approximation to the empirical observations. We also studied the probability that the individuals disrupt the group by "passing through it" (termed as intrusion). We analyzed the dependence of the parameters of the avoidance model and of the probability of intrusion on groups' social relation and intensity of interaction. We confirmed that the stronger social bonding or interaction intensity is, the more prominent collision avoidance turns out. We also confirmed that the probability of intrusion is a decreasing function of interaction intensity and strength of social bonding. Our results suggest that such variability should be accounted for in models and crowd management in general. Namely, public spaces with strongly bonded groups (e.g. a family-oriented amusement park) may require a different approach compared to public spaces with loosely bonded groups (e.g. a business-oriented trade fair).

A novel social distance model reveals the sidewall effect at bottlenecks

Intermittent and periodic outbreaks of infectious diseases have had profound and lasting effects on societies throughout human history. During the global spread of SARS-CoV-2 and the resulting coronavirus disease (COVID-19), social distance has been imposed worldwide to limit the spread of the virus. An additional deliberate intention of keeping a minimum safety distance from neighbors can fundamentally alter the "social force" between individuals. Here, we introduce a new "social distance" term inspired by gas molecular dynamics and integrate it into an existing agent-based social force model to describe the dynamics of crowds under social-distanced conditions. The advantage of this "social distance" term over the simple increasing of the repulsive range of other alternatives is that the fundamental crowd properties are precisely described by our model parameters. We compare the new model with the Helbing and Molnar's classical model and experimental data, and show that this new model is superior in reproducing experimental data. We demonstrate the usability of this model with a bottleneck motion base case. The new model shows that the bottleneck effect can be significantly alleviated through small wall modifications. Lastly, we explain the mechanism of this improvement and conclude that this improvement is due to spatial asymmetry.