FROM CROWD DYNAMICS TO CROWD SAFETY: A VIDEO-BASED ANALYSIS

JOHANSSON, ANDERS; HELBING, DIRK; AL-ABIDEEN, HABIB Z.; AL-BOSTA, SALIM · 2008 · Crossref

DOI: 10.1142/s0219525908001854

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This paper addresses the critical lack of empirical data regarding pedestrian dynamics under extreme crowding conditions, a gap that hinders the safe design and capacity assessment of facilities for mass events. While previous research established that pedestrian flows follow a "fundamental diagram" where speed decreases to zero at maximum density, these models were largely based on moderate densities (up to 4–6 persons per square meter). The authors investigate whether these fluid-dynamic assumptions hold at higher densities, specifically examining the transitions to dangerous states such as stop-and-go waves and "crowd turbulence," which were previously unexpected due to the rapid acceleration capabilities of humans. To analyze these extreme conditions, the researchers conducted a video-based study of pilgrim flows entering the Jamarat Bridge in Mina, Saudi Arabia, during the 2006 Hajj pilgrimage. They utilized 12 fixed cameras mounted on high poles to record over 2 terabytes of data, focusing on the entrance area where densities were highest. Because commercial tracking software failed to handle the scale and density of the crowd, the team developed a custom automated tracking algorithm. This method employed digital filters, adaptive histogram equalization, and Artificial Neural Networks to identify pedestrian heads and calculate trajectories. The system was validated against manual counts, showing high reliability. The authors also developed specific methods to correct for data distortions caused by umbrellas, which obscured pedestrians, and defined local density, speed, and flow metrics using weighted averaging techniques to account for spatial variations. The study revealed that average individual speed does not drop to zero even at local densities of 10 persons per square meter, contradicting traditional fluid-dynamic models. Instead, the data showed that as congestion increases, flow drops significantly, triggering stop-and-go waves. These waves further increase local density until critical conditions are reached, leading to "crowd turbulence"—a state of highly irregular motion caused by coordination failures among pedestrians competing for space. The authors identified that neither density nor speed alone are sufficient indicators of criticality; rather, the dynamic behavior of the crowd, specifically the onset of turbulence, signals imminent danger. The significance of these findings lies in their implications for crowd safety and facility design. The results demonstrate that maximum density and flow rates vary significantly based on cultural contexts and body size distributions, meaning data from one region cannot be directly applied to another. Consequently, relying on incorrect fundamental diagrams can lead to inadequate facility dimensioning and increased disaster risk. The authors conclude that pedestrian facilities must be operated well below their maximum capacity to avoid the onset of crowd turbulence. They emphasize the need for improved safety measures and accurate, context-specific empirical data to prevent crowd disasters in mass gatherings.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-prismaquant on 2026-06-26; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• traffic density