Strengthening post-earthquake logistical resilience: a fuzzy-optimization framework for emergency route planning in Türkiye

Introduction Post-earthquake humanitarian routing is challenging because transportation conditions may be incomplete, uncertain, and direction-dependent. This study develops a reproducible fuzzy-optimization framework for emergency route planning in earthquake-affected regions of Türkiye. Methods Hexagonal fuzzy numbers (HFNs) are audited and processed using deterministic data-cleaning rules and a bounded ranking operator to obtain generalized cost units (GCUs). The resulting directed cost matrix is formulated as an asymmetric traveling salesman problem (ATSP). Held–Karp dynamic programming provides an exact benchmark, against which five metaheuristic algorithms—ant colony optimization (ACO), genetic algorithm (GA), particle swarm optimization (PSO), cuckoo search algorithm (CSA), and bee colony optimization (BCO)—are evaluated. Independent computational runs, statistical tests, and heterogeneous disruption scenarios are used to assess solution quality and robustness. Results The framework provides a reproducible procedure for transforming fuzzy transportation data into a directed routing model. The exact solution serves as a reference for evaluating the metaheuristic algorithms, while run-level statistical comparisons and sensitivity experiments examine algorithm performance and the effects of transportation-cost increases, road closures, and node-level disruptions. Discussion The study demonstrates the value of explicit fuzzy-data auditing, bounded cost transformation, exact benchmarking, and reproducible computational experiments in emergency route planning. The proposed framework offers methodological support for post-earthquake logistics and can be extended to more comprehensive humanitarian routing models incorporating vehicle capacities, multiple vehicles, time-dependent conditions, and prioritized demands.