Optimizing Post-Conflict Recovery: An AI-Driven Multi-Criteria Decision Framework for Prioritizing Structural Interventions

Post-conflict reconstruction demands transparent and defensible decisions about where to intervene first when numerous structures are damaged and resources are scarce. This paper proposes an integrated decision framework that links an artificial-intelligence diagnostic layer with a multi-criteria decision-making (MCDM) layer to support resilient and people-centered prioritization. The AI diagnostic layer processes aerial and ground imagery to quantify each asset's physical condition as a normalized Structural Integrity Score (SIS), while the MCDM layer integrates this score with socio-humanitarian and feasibility criteria such as service accessibility, population served, repair cost, logistics, and environmental impact. A Taiz-based illustrative scenario demonstrates the complete workflow - from diagnostic scoring to aggregated prioritization, including sensitivity testing (±10% criteria variation) and geospatial visualization for stakeholder interpretation. Results show that the framework achieves a 0.95 Spearman rank correlation across weighting scenarios and prioritizes life-safety assets above less critical ones, optimizing resource allocation for faster and more equitable recovery. The study's contribution is a reproducible, data-driven mechanism that bridges AI diagnostics and decision science, transforming damage assessment into strategic, evidence-based decision-making for humanitarian and engineering logistics in fragile environments.