Emergence of collective oscillations in dense crowds

Building upon a combination of quantitative observations and theoretical insights, we unveil and elucidate the emergence of collective oscillations within densely packed pedestrian crowds.

Our comprehension of the dynamics within dense crowds has long been hindered by the scarcity of quantitative data. This limitation is rooted in two key factors. First, conducting controlled, reproducible experiments involving crowds with densities exceeding five individuals per square meter would necessitate the recruitment of thousands of volunteers and the implementation of safety protocols that have yet to be established. Second, the occurrence of dense crowd events in modern cities is infrequent and often unpredictable, as they are generally avoided by city officials and security forces.

To overcome this fundamental challenge, we conducted our study during the massive gathering at the Chupinazo, the opening event of a lively annual festival in San Fermin, Spain. Several hours prior to the festival's start, thousands of individuals congregate on the city's main square, creating ultra-dense crowds within a safe environment.

Utilizing quantitative video analysis, we reveal that beyond a critical density threshold, this active crowd exhibits slow, periodic oscillations that manifest across system-spanning scales. We will demonstrate the emergent nature of these collective dynamics and offer a theoretical explanation by employing a mean-field description of crowds, treating them as an active continuum.