AI-Enabled Decision Support for Volunteer Allocation in Disaster Management: A Three-Stage Optimization Framework

Abstract

Recent advances in artificial intelligence (AI) are transforming disaster management through intelligent decision support, predictive resource allocation, and adaptive operational planning. Volunteer management — a critical yet underexplored challenge in humanitarian logistics — represents a domain where AI-driven optimization can deliver substantial operational gains. This chapter presents a three-stage AI-enabled decision support framework for volunteer allocation in large-scale disaster response. At its core is a mixed-integer linear programming (MILP) model that constitutes the optimization layer of a broader AI-assisted humanitarian logistics ecosystem. The model assigns volunteers to disaster management centers while maximizing demand satisfaction, respecting capacity and travel-time constraints, incorporating individual preferences, enforcing group cohesion to improve long-term retention, and preventing scheduling conflicts. Computational experiments using a synthetic scenario with 100 volunteers, six volunteer types, four disaster management centers, and a seven-day response horizon confirm the model’s feasibility: the MILP solver reaches a solution in under 8 seconds and generates schedules that substantially reduce unmet demand. Sensitivity analysis on the preference–demand trade-off reveals important operational dynamics. The chapter concludes with a roadmap for extending the framework with machine learning demand forecasting, reinforcement learning-based dynamic reassignment, digital twin simulation, and predictive retention modeling — positioning the system as a foundation for next-generation AI tools in disaster humanitarian logistics.