Ecological Data Reveal Imbalances in Collision Avoidance Due to Groups' Social Interaction

Abstract

The relative dynamics in collision avoidance between individual pedestrians and dyads has been recently studied, and it was shown that individuals may intrude dyads that are not socially interacting. Building on this, our current study examines how much each party contributes to collision avoidance by measuring deviations from their intended paths. Our findings suggest that individuals prioritise trajectory efficiency in undisturbed situations, but prioritise safety when encountering dyads, deviating more from their intended path. Non-interacting dyads present a similar behavior, although their trajectories appear to be even more efficient than those of individuals in undisturbed situations, and their deviations during encounters less pronounced. On the other hand, socially interacting dyads are not very efficient in undisturbed situations, and their behavior is mostly unaffected by encounters. These results strongly suggest that group dynamics affects in two ways the behavior of pedestrians, namely it has a dynamical and a social effect. The dynamical effect stabilises their trajectory, while the social one decreases the ability to focus on the external environment, leading to reduced efficiency and safety. Another finding concerns the tendency of individuals to avoid in a more prominent way the interacting dyads as compared to non-interacting ones. This suggests that individuals may assess others' contribution to collision avoidance. An impact parameter analysis reveals that collision risk influences path deviations in pedestrian encounters. For individuals, larger behavioral differences between low and high interaction levels of the dyad occur both when the collision risk is high and during less critical encounters. For dyads, the deviation differences between low and high interaction levels are most pronounced when the individual is on course to pass close to the dyad.