AI-driven decision support for disaster management and humanitarian logistics: Models, applications, and research challenges

As disasters grow more frequent, complex, and destructive, the demand for intelligent and responsive logistics decision-making has intensified. This systematic review examines the integration of Artificial Intelligence (AI) into Decision Support Systems (DSS) with a focus on disaster response and humanitarian logistics. Drawing from peer-reviewed literature published between 2008 and 2024, the review identifies major AI techniques, including machine learning, deep learning, natural language processing, reinforcement learning, and AI–IoT integration, and evaluates their application in key humanitarian logistics functions such as demand forecasting, facility location and prepositioning, transportation routing, last-mile delivery, and resource allocation under uncertainty. The findings reveal a concentration of research on natural disasters and early-phase emergency response, while recovery, mitigation, and long-term supply chain resilience remain underexplored. Persistent challenges include fragmented and incomplete datasets, limited explainability of AI-driven models, siloed system designs, and ethical concerns around fairness and equity in resource distribution. The review concludes by outlining a research agenda that calls for interoperable, interpretable, and context-sensitive AI-DSS frameworks capable of enhancing supply chain coordination, equity in aid distribution, and decision quality across the full humanitarian logistics lifecycle. This review is the first to systematically categorize existing studies by both AI technique and disaster phase and introduces an integrative conceptual framework that links AI methods to humanitarian logistics decision functions.