A Micro-Spatial Analytics Approach to Predictive Crowd Risk Assessment in Indian Mass Gatherings
DOI:
https://doi.org/10.54489/ijcim.v6i1.587Keywords:
Crowd Safety, Interpersonal Distance, Business Analytics, Proxemics, Visual Analytics, Indian Mass Gatherings, Predictive Risk Assessment, Stampede PreventionAbstract
The study is conducted through a quantitative, ex post facto observational research design, where high-resolution historical visual records of real-world crowd disaster situations in India are analysed. The unstructured image data is systematically transformed into a structured analytical dataset by creating a data extraction pipeline with Python (OpenCV, Pandas, NumPy, Matplotlib). These micro-spatial variables are extracted and then analyzed using Pearson correlation, descriptive statistics and threshold identification. It shows that there is almost an inverse relationship between the IPD and the systemic risk score (r = -0.98), which means that the risk increases in an exponential manner with each violation of individual physical space. A critical "Red-Line" tipping point is set at IPD < 0.50 metres, which is the boundary of the Intimate Zone in Proxemic Theory. All crowd scenarios where IPD was below this value had risk scores of more than 80/100. Furthermore, it was identified that 100% of high-risk cases showed architectural bottlenecks, which is why they are considered as a structural amplifier of crowd danger. This study builds upon the concept of crowd science by demonstrating that business analytics tools developed and utilized in commercial and operational environments can be successfully applied as tools to assist in public safety management. The proposed Visual Analytics-to-Business Analytics (VA-BA) pipeline enables the "Red-Line Strategy", a data-driven approach for event managers to proactively intervene with crowds, which will help to shift the paradigm of reactive policing to predictive monitoring. The present study is one of the first studies to use the micro-spatial analysis with the theoretical background of Proxemic Theory and the Social Force Model and specifically for the religious crowd context in India. It is a methodology and an empirical dataset that is missing from the literature, as it is specifically oriented towards the Indian crowd that is inspired by the spirit and is a behaviourally and structurally different crowd from the Western crowd models that have been studied.
References
Dederichs, A., Köster, G., & Seyfried, A. (2020). Pedestrian crowd dynamics: Empirical results and modeling approaches. Transportation Research Procedia, 47, 3–10.
Gayathri, H., Aparna, P., & Verma, A. (2017). Crowd management strategies in India: A case study approach. International Journal of Disaster Risk Reduction, 21, 112–120.
Hall, E. T. (1966). The hidden dimension. Doubleday.
Iachini, T., Coello, Y., Frassinetti, F., & Ruggiero, G. (2020). Body space in social interactions: A comparison of reachability and comfort distance in immersive virtual reality. PLoS ONE, 15(11), e0238752.
Idrees, H., Tayyab, M., Athrey, K., Zhang, Y., Al-Maadeed, S., Rajpoot, N., & Shah, M. (2018). Composition loss for counting, density map estimation and localization in dense crowds. European Conference on Computer Vision (ECCV), 532–546. DOI: https://doi.org/10.1007/978-3-030-01216-8_33
Kang, J., & Kim, J. (2023). Effects of interpersonal distance on human perception and crowd behavior. Journal of Environmental Psychology, 85, 101920.
Qamar, A. M., Ali, M., & Khan, S. (2024). Analysis of crowd disasters and risk mitigation strategies in developing countries. Safety Science, 170, 106312.
Sindagi, V. A., & Patel, V. M. (2017). Generating high-quality crowd density maps using contextual pyramid CNNs. IEEE International Conference on Computer Vision (ICCV), 1879–1888. DOI: https://doi.org/10.1109/ICCV.2017.206
Still, G. K. (2014). Introduction to crowd science. CRC Press. DOI: https://doi.org/10.1201/b17097
Still, G. K., Bandini, S., & Manley, E. (2020). Modelling and managing crowd disasters: The DIM-ICE framework. Safety Science, 130, 104879.
Subramanian, V., & Verma, A. (2022). A data-driven approach to crowd risk assessment using analytics. Journal of Urban Management, 11(2), 145–158.
Testa, A., & Musse, S. R. (2025). Crowd simulation and risk prediction: Advances and challenges. Computers & Graphics, 110, 45–58.
Vanumu, L. D., Sano, K., & Reddy, T. S. (2020). Understanding stampede risk and prevention strategies in mass gatherings. Procedia Engineering, 180, 123–130.
Vandoni, L. (2019). Deep learning methods for crowd counting and density estimation. Pattern Recognition Letters, 128, 82–90.
Verma, A., Singh, R., & Mishra, S. (2022). Crowd disasters in India: Causes, impacts, and mitigation strategies. International Journal of Disaster Risk Science, 13(4), 567–580.
Wang, X. (2022). Personal space and crowding effects on human behavior. Frontiers in Psychology, 13, 845678.
Weng, W., Chen, T., & Yuan, H. (2022). Crowd dynamics and risk propagation in high-density environments. Physica A: Statistical Mechanics and Its Applications, 589, 126629. DOI: https://doi.org/10.1016/j.physa.2021.126629
Zhang, J., Klingsch, W., Schadschneider, A., & Seyfried, A. (2017). Transitions in pedestrian fundamental diagrams of straight corridors and T-junctions. Journal of Statistical Mechanics: Theory and Experiment, 2017(4), 043402..
