MORAF: A Multi-Objective Framework for Resource Allocation with Fairness for Efficient Disaster Management

Abstract

After large-scale disasters, emergency responders must allocate limited resources across affected areas within critical time windows. This creates a fundamental challenge: concentrating resources maximizes total lives saved but leaves some areas underserved, while distributing resources equally ensures coverage but reduces overall effectiveness. This work presents MORAF (Multi-Objective Resource Allocation with Fairness), an optimization framework that balances both objectives simultaneously. MORAF incorporates diminishing returns (the first ambulance saves more lives than the tenth), context-specific effectiveness (concrete breakers work for earthquakes but not floods), and resource synergy (teams, vehicles, and equipment work better together). Since this optimization problem is non-convex, the framework employs Sequential Least Squares Programming with multiple starting points. MORAF is validated on a hurricane scenario using the Hurricane Michael road network across 100 zones with 12 resource types. Results demonstrate 19.4% higher rescue utility and 69% improved fairness (Fairness score: 0.910 versus 0.537) compared to baseline approaches.