Analiza ruchu ludzi w tłumie w celu efektywnego zarządzania tłumem
[Analysis of human crowd motion for efficient crowd management]

Pratik Mullick (Katedra Badań Operacyjnych i Inteligencji Biznesowej, PWr, Wrocław)


The collective motion of social agents arising from their interactions has been the 
subject of intense scientific research. Understanding the collective dynamics of 
human crowds is crucial for improving pedestrian traffic flow, ensuring crowd safety, 
effective urban planning, and preventing crowd disasters. When two streams of 
pedestrians cross at an angle, striped patterns spontaneously emerge due to local 
pedestrian interactions. Such spontaneous pattern formation is an example of 
self-organized collective behavior, a topic of intense interdisciplinary interest. In 
this talk, I will present numerical strategies developed to study the geometric 
properties of striped patterns that arise as a consequence of two crossing flows.

From the perspective of crowd management, it is crucial to understand the relationship 
between crowd density and velocity, known as the fundamental diagram (FD). This 
relationship helps to study the capacity of spaces where traffic moves, such as roads 
for vehicles and sidewalks for pedestrians. Constructing a realistic FD requires the 
utilization of an effective method for density estimation. While existing literature 
offers several methods, determining the 'best' method remains unresolved and may depend 
on the crowd situation. Most research has focused on situations where the moving crowd 
is constrained within a physical boundary, such as a corridor or sidewalk. However, 
there is no well-defined method of density estimation for groups in an unbounded space. 
I will present our recently developed voronoi-cell-based density estimation method that 
can estimate crowd density in a wide variety of situations, irrespective of the presence 
of spatial constraints.