Agentic AI-Enabled Cognitive Digital Twin for Urban Flood Prediction and Disaster Management: A Multi-Agent Framework for the Indian Urban Context

We described a cognitive digital twin for urban flood prediction in which seven agents sense, predict, simulate, explain, alert, coordinate and converse. Its distinguishing feature is that prediction confidence controls what the system does next, with bounded retries and escalation to a human. The framework is built around ISRO data products with open fallbacks, and the prototype presents the twin in a browser client that adapts to the device. A demonstration on zones in Bengaluru and Mumbai showed the views, the explanation text and the alert log agreeing with one another. The next step is a quantitative evaluation: running the pipeline on live sources for labelled flood events, reporting precision, recall, F1 and IoU, and counting what the retry rule changes against a version without it. Further work includes a field trial with disaster-management staff, calibration of the confidence score and alert depth, adversarial testing of the chat refusal behaviour, additional cities, and coupling with coastal surge models, as the authors of [5] suggest.

1. Hlal, M., Baraka Munyaka, J.-C., Chenal, J., Azmi, R., Diop, E.B., Bounabi, M., Ebnou Abdem, S.A., Almouctar, M.A.S., Adraoui, M.: Digital twin technology for urban flood risk management: A systematic review of remote sensing applications and early warning systems. Remote Sens. 17(17), 3104 (2025) 2. Roy, P.: It's been 40 years! Update India's flood map now. Down To Earth. https://www.downtoearth.org.in/news/climate-change/it-s-been-40-years-update-india-s-flood-map-now-78166 (accessed 11 Oct 2026) 3. The Quint: As Bengaluru loses lakes, green cover, impact of urban floods takes a toll. https://www.thequint.com/climate-change/as-bengaluru-loses-lakes-green-cover-impact-of-urban-floods-takes-a-toll (accessed 11 Oct 2026) 4. Grieves, M.: Origins of the digital twin concept. White paper (2016) 5. Mandal, D., Zou, L., Wadhwa, A., Wilkho, R.S., Cai, Z., Zhou, B., Ye, X., Newman, G., Gharaibeh, N., Güneralp, B.: FlowsDT: A geospatial digital twin for navigating urban flood dynamics. Comput. Environ. Urban Syst. 126, 102414 (2026) 6. Fakhri, F., Gkanatsios, I.: Quantitative evaluation of flood extent detection using attention U-Net case studies from Eastern South Wales Australia in March 2021 and July 2022. Sci. Rep. 15, 12377 (2025) 7. Ge, C., Qin, S.: Urban flooding digital twin system framework. Syst. Sci. Control Eng. 13(1), 2460432 (2025) 8. Li, B., et al.: Disaster management in the era of agentic AI systems: A vision for collective human-machine intelligence for augmented resilience. arXiv:2510.16034 (2025) 9. Fan, C., Zhang, C., Yahja, A., Mostafavi, A.: Disaster City Digital Twin: A vision for integrating artificial and human intelligence for disaster management. Int. J. Inf. Manag. 56, 102049 (2019) 10. Ronneberger, O., Fischer, P., Brox, T.: U-Net: Convolutional networks for biomedical image segmentation. In: Proc. MICCAI, pp. 234–241 (2015)