Federated Learning-Driven Digital Twin: A Privacy-Preserving AI Approach for Crisis Logistics
Hafsa El Mouhsine, Rajaa Saidi, Walid Cherif · 2025
In emergency situations, rapid action is critical to save lives, yet humanitarian logistics often grapple with challenges like information dispersion, tight deadlines, and strict privacy regulations. This research introduces FL-DT-HSC, a novel approach integrating Federated Learning (FL) and Digital Twins (DTs). Federated Learning enables the management of sensitive data across multiple sites without centralization, while Digital Twins offer live simulations to guide decision-making. Tested through a fictional case based on the 2022 Pakistan floods, FL-DT-HSC shows promise for faster, more efficient, and privacy-conscious responses. Though still a concept, it leverages established ideas from healthcare and industrial applications, laying the groundwork for real-world experiments to transform crisis logistics.