Managing the flow and density of people at international conferences, exhibitions, and large-scale events is crucial to maintaining a lively atmosphere while minimizing the negative effects of excessive congestion. Quantitative evaluations of crowd perception are necessary for this purpose. However, how attendees perceive congestion at large-scale events has remained unclear. This study is the first in the world to quantitatively compare perceived congestion with measured crowd levels at events attended by over 10,000 people.
Over three years, the research team from the University of Tokyo, Hokkaido University, and Osaka University conducted surveys and Bluetooth-based mobility analyses at large-scale events held at Tokyo Big Sight, one of Japan's largest venues for exhibitions and conferences. The surveys collected responses in three ways: participants reported the times they felt congestion through open-ended answers, marked the locations on maps divided into specific areas, and selected photos representing perceived crowd density. Meanwhile, trajectory analyses, conducted using beacons distributed to participants, tracked when, where, and to what extent the participants experienced congestion.
The comparison between survey responses and beacon-based mobility data revealed that participants accurately identified the most crowded times and locations but showed a weak correlation between selected crowd-density photos and actual congestion. This indicates that perceptions of crowd intensity are shaped by factors beyond density itself. Additionally, reported congestion times often aligned with departure times rather than peak crowding, highlighting the influence of recent memories on perception.
This study provides key insights into how people perceive and remember crowding, offering practical applications for managing crowd flow and density in event planning. By addressing congestion closer to participants’ departure times, organizers can improve event experiences while maintaining safety and efficiency. These findings provide a foundation for practical strategies to optimize crowd dynamics while preserving the advantages of large gatherings.
(left) Illustration of congestion at large-scale event. This image was also used as part of the survey to evaluate perceived crowding. (right) An overview of this study's approach to analyzing actual and perceived congestion.
Papers
Journal: Scientific Reports
Title: Quantifying crowding perception at large events using beacons and surveys
Authors: Sakurako Tanida*, Hyerin Kim, Claudio Feliciani, Xiaolu Jia, Akira Takahashi, Tetsuya Aikoh, Katsuhiro Nishinari