Human‐Centered Infrastructure Restoration: An Integrated Framework for Demand Estimation and Resource Allocation

Given the critical importance of lifeline infrastructures in maintaining society functioning, the main objective of infrastructure restorations following disasters is to satisfy community demand in a rapid and effective manner. In existing literature, community demand on infrastructure services is often assumed to remain constant before and after disasters, which might lead to a mismatch between restored infrastructure serviceability and actual community demand. To address this gap, this study proposes an integrated demand-oriented infrastructure restoration framework. The integrated framework is designed to (1) estimate community demand using a Bayesian-based method, allowing for the integration of multiple information sources and the rapid updating of demands as new data becomes available; (2) develop a demand-oriented optimization model that prioritizes resource allocation to the infrastructure components serving communities with higher levels of demand; and (3) create a reliable solution method using an iterative process to accommodate the dynamics of disaster situations, complemented by a hybrid simulation-optimization approach to manage demand uncertainty. For illustrative purposes, the restoration of interdependent power and water infrastructure networks in Shelby County, TN, is studied. The results demonstrate that the proposed framework significantly improves the satisfaction of community demand, and meanwhile decreases the penalty costs associated with unmet demands. Beyond post-disaster restoration, the framework is capable of assisting decision-makers in simulating various disaster scenarios, enabling more community-centered resilience planning.