Modeling and Mitigating Uncertainty in Dynamic Graph Routing for Disaster Evacuation Systems

ResQNet is a disaster-response routing and resource-allocation system designed to improve emergency decision-making during natural disasters. Traditional navigation systems prioritize shortest distance or fastest travel time, but often fail during emergencies when roads become hazardous, infrastructure is damaged, and rescue resources are limited. This project models transportation networks as weighted graphs and combines risk-aware pathfinding with capacity-constrained resource allocation. ResQNet integrates Dijkstra's shortest-path algorithm and the Edmonds-Karp maximum-flow algorithm to identify safer rescue routes while maximizing emergency response coverage. The system evaluates factors such as disaster severity, weather conditions, population exposure, and road accessibility to generate risk-weighted routing decisions. Experimental testing on simulated disaster scenarios demonstrates significant improvements in rescue efficiency, including reduced response times, increased coverage, and elimination of unmet rescue demand. ResQNet was developed independently by a Grade 11 student as a research and software engineering project exploring the application of graph theory and optimization algorithms to disaster management.