Artificial Intelligence in Crowd Disaster Management: A Comprehensive Review of Technologies, Applications, and Future Directions
Abstract
Large crowds—such as those at religious events, concerts, or sporting venues—are prone to risks of crowd disasters including stampedes, panic-induced chaos, and congestion-related incidents, which constitute serious threats to public safety. Real-time monitoring, predictive analysis, and timely decision-making are key requirements for effective crowd disaster management to minimize risk and improve safety measures. In this field, Artificial Intelligence (AI) has emerged as a powerful solution, utilizing advanced technologies including machine learning, computer vision, and simulation models to assess and control crowd behaviour efficiently. This paper reviews the application of AI in crowd disaster management, including risk assessment, anomaly detection, evacuation planning, and emergency response. Live video feeds are analysed by AI-powered surveillance systems, which also predict potential hazards based on movement patterns. Through IoT devices, data can be instantly processed, enabling dynamic evacuation routing through knowledge-based evacuation systems. Robotic and drone technology combined with AI ensures that emergency responders can respond to situations as they unfold. Beyond disaster prevention, AI in crowd management improves emergency management processes and resource utilization. The role of AI is discussed in this paper, leading to the finding that it can improve safety, make evacuations more efficient, and reduce loss of life. By utilizing AI-driven intelligent systems, authorities can significantly enhance crowd control measures, creating safer and more organized environments for large gatherings.
Keywords
Download Options
Introduction
Globally, crowd disasters—namely stampedes, crushes, and attendant panic—have over the years caused heavy loss of human life and material damage. Such risks can only be mitigated if public safety can be assured in densely populated areas such as stadiums, religious gatherings, concerts, and urban events [1]. With predictive analytics, real-time monitoring, and automated decision-making, Artificial Intelligence (AI) has emerged as a powerful tool for disaster management to prevent and control crowd-related incidents. Situational awareness through AI-driven technologies, specifically computer vision, machine learning, and simulation models, enables crowd behaviour analysis, detection of anomalies, and prediction of hazards. From live footage, surveillance systems using AI recognize overcrowding, while predictive algorithms assess risks and recommend preventive measures. Drones and robots powered by AI can assist during emergencies to help evacuate people quickly. This paper examines the role of AI in crowd disaster management—including risk assessment, early warning systems, and crisis response applications. Based on these insights, authorities can improve crowd control strategies, evacuation procedures, and minimize casualties. The integration of AI in disaster management offers a revolutionary approach to public safety, ensuring safe and regulated large assemblies through data-driven intelligent solutions.
Conclusion
Advanced technologies comprising sensing, Internet of Things (IoT), social media (SM), big data analytics, and artificial intelligence (AI) have the potential to significantly reduce casualties and infrastructure loss during natural or human-made disasters. This survey examines the development of SM and AI-based solutions for prediction, detection, response, and emergency management, providing a comprehensive overview of existing disaster management technologies and assessing their effectiveness in disaster scenarios. Each approach is systematically categorized and evaluated using various performance metrics. While these technologies offer significant benefits—including faster response times, improved situational awareness, and better resource allocation—substantial challenges remain in their implementation, including data reliability, scalability, real-time decision-making, and ethical considerations. Overcoming these limitations requires continued research and development.
12.1 Micro Blogging System: Beyond improving accuracy and precision in detecting relevant messages, AI-based emergency management models face additional challenges. Most existing SM crowd management research relies primarily on Twitter as a data source. Information synchronization and response patterns may differ across platforms, and some platforms like Facebook restrict data extraction. Incorporating diverse SM platforms, considering varying data quality across platforms, remains an open challenge. Data type and quality may vary significantly based on typical user demographics for each platform.
12.2 User Participation: Effective crowdsourcing, big data, and SM applications require high levels of user engagement. While incentives can encourage participation, research suggests that users may not require additional motivation to engage in socially beneficial collective actions [47]. Questions regarding optimal incentive quantity and quality have been partially addressed in literature but require further investigation.
12.3 Individual Privacy: Public safety during emergencies must be balanced against individual privacy rights to protect the dignity and wellbeing of affected populations. Disaster management sometimes requires collection of personal and sensitive information to support rescue and recovery operations. Such data raises privacy and security concerns regarding potential misuse or unauthorized access. Ensuring that information collected from crisis zones remains protected from malicious actors is essential. Balancing privacy protection with effective emergency response presents significant challenges requiring careful examination and optimization of the privacy-efficiency trade-off. Each scenario must be evaluated individually to design systems achieving optimal balance between efficiency and privacy protection. Continued research is needed to develop secure, adaptive frameworks supporting both public safety and data protection in crisis situations.
12.4 Cost Reduction: Global researchers are working to reduce equipment and software deployment costs for disaster management technologies while maintaining or improving system performance. Given that emergency systems may remain idle for extended periods but must perform reliably during rare critical events, questions about cost-effectiveness arise. Developing methods to utilize emergency equipment during non-emergency periods without compromising emergency response capabilities represents an important research direction.
References
[1] Still, G. K. (2014). Introduction to crowd science. CRC Press.
[2] Boin, F., & Schulman, P. (2008). Institutional amnesia and disaster preparedness. Journal of Contingencies and Crisis Management, 16(1), 20–29.
[3] Indian Institute of Science. (2016). Computational modelling of crowd dynamics at religious gatherings (Technical Report). IISc.
[4] Johansson, A., Helbing, D., & Shukla, P. K. (2007). Specification of the social force pedestrian model by evolutionary adjustment to video tracking data. Advances in Complex Systems, 10(2), 271–288.
[5] Tirachini, A., Hensher, D. A., & Rose, J. M. (2013). Crowding in public transport systems: Effects on users, operation and implications for the estimation of demand. Transportation Research Part A: Policy and Practice, 53, 36–52.
[6] Singh, J. (2022). Necro politics of devotion: The political economy of pilgrimage disasters. Routledge.
[7] Gupta, A., & Rathore, N. (2023). Real-time crowd risk assessment using AI: A case study on the Kumbh Mela. IEEE Transactions on Neural Networks and Learning Systems, 34(6), 4321–4332.
[8] Kumar, S., Jain, P., & Agarwal, R. (2021). Sentiment analysis of disaster reports: Uncovering patterns in post-stampede narratives. IEEE Access, 9, 145673–145688.
[9] Hasanuzzaman, M., Hossain, S., & Shil, S. K. (2023). Enhancing disaster management through AI-driven predictive analytics: Improving preparedness and response. International Journal of Advanced Engineering Technologies and Innovations, 1(1), 533–562.
[10] Yin, J., Bi, Y., Zheng, X.-M., & Tsaur, R.-C. (2019). Safety forecasting and early warning of highly aggregated tourist crowds in China. IEEE Access, 7, 119026–119040. https://doi.org/10.1109/ACCESS.2019.2936245
[11] Hong, Z., Xiao, D., & Wenxuan, D. (2016, December). A crowd pre-warning system based on mobile locators and behaviour prediction. In Proceedings of the 6th International Conference on Digital Home (ICDH) (pp. 122–127). IEEE. https://doi.org/10.1109/ICDH.2016.035
[12] Yang, P., Xi, L., Ji, H., & Zhang, H. (2017, October). A novel mobility prediction scheme for outdoor crowded scenario using fuzzy C-means. In Proceedings of the IEEE 28th Annual International Symposium on Personal, Indoor, and Mobile Radio Communications (PIMRC) (pp. 1–5). IEEE. https://doi.org/10.1109/PIMRC.2017.8292536
[13] Wang, Y., Hou, X., & Chau, L.-P. (2021, July). Dense point prediction: A simple baseline for crowd counting and localization. In Proceedings of the IEEE International Conference on Multimedia & Expo Workshops (ICMEW) (pp. 1–6). IEEE. https://doi.org/10.1109/ICMEW53276.2021.9455954
[14] Sundaram, V., Tripathy, A. K., Deshmukh, R., & Pawar, A. (2020, April). A crowd density estimation approach using GPS mobility for its dynamics and predictions. In Proceedings of the 3rd International Conference on Communication Systems, Computing and IT Applications (CSCITA) (pp. 67–72). IEEE. https://doi.org/10.1109/CSCITA47329.2020.9137804
[15] Maclean, C. (2008). Pilgrimage and power: The Kumbh Mela in historical perspective. Oxford University Press.
[16] Jenelius, E. (2020). Data-driven metro train crowding prediction based on real-time load data. IEEE Transactions on Intelligent Transportation Systems, 21(6), 2254–2265. https://doi.org/10.1109/TITS.2019.2914729
[17] Drabicki, A., Kucharski, R., Cats, O., & Fonzone, A. (2017, June). Simulating the effects of real-time crowding information in public transport networks. In *Proceedings of the 5th IEEE International Conference on Models and Technologies for Intelligent Transportation Systems (MT-ITS)* (pp. 675–680). IEEE. https://doi.org/10.1109/MTITS.2017.8005598
[18] Dong, R., Li, L., Zhang, Q., & Cai, G. (2018). Information diffusion on social media during natural disasters. IEEE Transactions on Computational Social Systems, 5(1), 265–276. https://doi.org/10.1109/TCSS.2017.2786545
[19] Priyanga, K. M., & Alamelu, M. (2020, June). User request emotion prediction approach in a crowdsourcing platform. In Proceedings of the 4th International Conference on Trends in Electronics and Informatics (ICOEI) (pp. 23–29). IEEE. https://doi.org/10.1109/ICOEI48184.2020.9143021
[20] Goh, G., Koh, J. Y., & Zhang, Y. (2018, November). Twitter-informed crowd flow prediction. In Proceedings of the IEEE International Conference on Data Mining Workshops (ICDMW) (pp. 624–631). IEEE. https://doi.org/10.1109/ICDMW.2018.00097
[21] Aswathy, A., Prabha, R., Gopal, L. S., Pullarkatt, D., & Ramesh, M. V. (2022, February). An efficient Twitter data collection and analytics framework for effective disaster management. In Proceedings of the IEEE Delhi Section Conference (DELCON) (pp. 1–6). IEEE. https://doi.org/10.1109/DELCON54057.2022.9753627
[22] Yuan, K., Wu, J., & Zhao, Z. (2016, June). Burst prediction from Weibo: A crowd-sensing and tweet-centric method. In Proceedings of the 13th International Conference on Service Systems and Service Management (ICSSSM) (pp. 1–6). IEEE. https://doi.org/10.1109/ICSSSM.2016.7538627
[23] Krishnan, J., Purohit, H., & Rangwala, H. (2020, December). Unsupervised and interpretable domain adaptation to rapidly filter tweets for emergency services. In Proceedings of the IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining (ASONAM) (pp. 409–416). IEEE. https://doi.org/10.1109/ASONAM49781.2020.9381480
[24] Chen, X., Zhou, X., Chan, J., Chen, L., Sellis, T., & Zhang, Y. (2022). Event popularity prediction using influential hashtags from social media. IEEE Transactions on Knowledge and Data Engineering, 34(10), 4797–4811. https://doi.org/10.1109/TKDE.2020.3048428
[25] Tian, Y., Fan, R., Ding, X., Zhang, X., & Gan, T. (2020). Predicting rumor retweeting behaviour of social media users in public emergencies. IEEE Access, 8, 87121–87132. https://doi.org/10.1109/ACCESS.2020.2989180
[26] Jiang, R., Cai, Z., Wang, Z., Yang, C., Fan, Z., Chen, Q., Tsubouchi, K., Song, X., & Shibasaki, R. (2023). DeepCrowd: A deep model for large scale citywide crowd density and flow prediction. IEEE Transactions on Knowledge and Data Engineering, 35(1), 276–290. https://doi.org/10.1109/TKDE.2021.3077056
[27] Hao, Q., Zhang, L., Zhou, R., Zhang, D., Xu, T., & Chen, E. (2021, June). Urban crowd density prediction based on multi-relational graph. In Proceedings of the 22nd IEEE International Conference on Mobile Data Management (MDM) (pp. 39–48). IEEE. https://doi.org/10.1109/MDM52706.2021.00018
[28] Liu, C. H., Piao, C., Ma, X., Yuan, Y., Tang, J., Wang, G., & Leung, K. K. (2021, April). Modeling citywide crowd flows using attentive convolutional LSTM. In Proceedings of the IEEE 37th International Conference on Data Engineering (ICDE) (pp. 217–228). IEEE. https://doi.org/10.1109/ICDE51399.2021.00026
[29] Gupta, T., Nunavath, V., & Roy, S. (2019, October). CrowdVAS-Net: A deep-CNN based framework to detect abnormal crowdmotion behaviour in videos for predicting crowd disaster. In Proceedings of the IEEE International Conference on Systems, Man and Cybernetics (SMC) (pp. 2877–2882). IEEE. https://doi.org/10.1109/SMC.2019.8914152
[30] Punn, N. S., & Agarwal, S. (2019, August). Crowd analysis for congestion control early warning system on foot over bridge. In Proceedings of the 12th International Conference on Contemporary Computing (IC3) (pp. 1–6). IEEE. https://doi.org/10.1109/IC3.2019.8844927
[31] Punn, N. S., & Agarwal, S. (2019, August). Crowd analysis for congestion control early warning system on foot over bridge. In Proceedings of the 12th International Conference on Contemporary Computing (IC3) (pp. 1–6). IEEE. https://doi.org/10.1109/IC3.2019.8844927
[32] Ma, Y., & Bai, G. (2010, November). Short term prediction of crowd density using ν-SVR. In Proceedings of the IEEE Youth Conference on Information, Computing and Telecommunications (pp. 234–237). IEEE. https://doi.org/10.1109/YCICT.2010.5713088
[33] Ma, J., Chan, J., Rajasegarar, S., Ristanoski, G., & Leckie, C. (2020, November). Multi-attention 3D residual neural network for origin-destination crowd flow prediction. In Proceedings of the IEEE International Conference on Data Mining (ICDM) (pp. 1160–1165). IEEE. https://doi.org/10.1109/ICDM50108.2020.00142
[34] Sharafat, A. R., & Bayati, M. (2021). Patient Flow Net: A deep learning approach to patient flow prediction in emergency departments. IEEE Access, 9, 45552–45561. https://doi.org/10.1109/ACCESS.2021.3066164
[35] Yao, L., Leung, K., Hong, J., Tsai, C., & Fu, L. (2020, October). A system for predicting hospital admission at emergency department based on electronic health record using convolution neural network. In Proceedings of the IEEE International Conference on Systems, Man, and Cybernetics (SMC) (pp. 546–551). IEEE. https://doi.org/10.1109/SMC42975.2020.9282952
[36] Han, W., Gan, Y., Chen, S., & Wang, X. (2020, October). Study on earthquake prediction model based on traffic disaster data. In Proceedings of the IEEE 11th International Conference on Software Engineering and Service Science (ICSESS) (pp. 331–334). IEEE. https://doi.org/10.1109/ICSESS49938.2020.9237667
[37] Dogan, A., & Demir, E. (2020, October). Earthquake prediction in Turkey with structural recurrent neural networks. In Proceedings of the 28th Signal Processing and Communications Applications Conference (SIU) (pp. 1–4). IEEE. https://doi.org/10.1109/SIU49456.2020.9302400
[38] Bande, S., & Shete, V. V. (2017, August). Smart flood disaster prediction system using IoT & neural networks. In Proceedings of the International Conference on Smart Technologies for Smart Nation (SmartTechCon) (pp. 189–194). IEEE. https://doi.org/10.1109/SmartTechCon.2017.8358367
[39] Hou, H., Yu, S., Wang, H., Xu, Y., Xiao, X., Huang, Y., & Wu, X. (2020). A hybrid prediction model for damage warning of power transmission line under typhoon disaster. IEEE Access, 8, 85038–85050. https://doi.org/10.1109/ACCESS.2020.2992528
[40] Tang, S., Xu, X., Yang, Y., Yu, Z., Liu, Z., & Guan, C. (2020, September). Research on the damage class prediction method in typhoon disaster for power system. In Proceedings of the IEEE 3rd International Conference on Information Systems and Computer Aided Education (ICISCAE) (pp. 72–76). IEEE. https://doi.org/10.1109/ICISCAE51034.2020.9236927
[41] Tiwari, D., Kumar, R., Yadav, M., Gupta, G. K., Singh, S. K., Dhapekar, N. K., & Sharma, R. (2025). Holistic analysis of Ganga basin water quality: A statistical approach with WQI, HMCI, HMQI and HRI indices. RSC Advances, 15(5), 3290–3316.
[42] Joseph, J. K., Babu, N., Dev, K. A., & Pradeepkumar, A. P. (2016). Identification of potential health risks in mass gatherings: A study from Sabarimala pilgrimage, Kerala, India. International Journal of Disaster Risk Reduction, 17, 95–99.
[43] Still, K., Papalexi, M., Fan, Y., & Bamford, D. (2020). Place crowd safety, crowd science? Case studies and application. Journal of Place Management and Development, 13(4), 385–407.
[44] Owaidah, A., Olaru, D., Bennamoun, M., Sohel, F., & Khan, N. (2019). Review of modelling and simulating crowds at mass gathering events: Hajj as a case study. Journal of Artificial Societies and Social Simulation, 22(2), Article 9.
[45] Taibah, H., & Arlikatti, S. (2015). An examination of evolving crowd management strategies at pilgrimage sites: A case study of 'Hajj' in Saudi Arabia. International Journal of Mass Emergencies & Disasters, 33(2), 188–212.
[46] Pechlaner, H., Innerhofer, E., & Erschbamer, G. (Eds.). (2022). Overtourism: Tourism management and solutions. Routledge.
[47] Abualsaud, K., Elfouly, T. M., Khattab, T., Yaacoub, E., Ismail, L. S., Ahmed, M. H., & Guizani, M. (2019). A survey on mobile crowd-sensing and its applications in the IoT era. IEEE Access, 7, 3855–3881.