Lay Eng TeohJin Wang ChangKhoo Hooi Ling @ Lai Hooi Ling2026-10-132026-10-132026-0110.1016/j.ssci.2025.107027https://dspace-cris.utar.edu.my/handle/123456789/11720Floods have emerged as a worldwide challenging disaster primarily due to global warming. Thus, ensuring an adequate heterogeneous evacuation fleet to promptly evacuate affected victims is of utmost importance. Correspondingly, properly coordinated evacuation fleet planning should explicitly unify both aspects of demand (number of victims) and supply (provision of evacuation vehicles). However, the demand and supply aspects are rarely explored simultaneously in past studies. By embedding a novel 3-step victim demand modeling framework, this study develops a scenario-based heterogeneous fleet planning model to maximize the total number of victims evacuated from disaster areas and minimize the composition of heterogeneous evacuation vehicles. The proposed approach can capture evacuation demand and supply in a single unified framework to solve flood evacuation problems under different flooding scenarios, including low/moderate and high water levels. By analyzing an illustrative case study for the context of Malaysia, several insightful results are revealed: (1) the probable demand level of victims for evacuation would vary greatly under different flood severity due to numerous influential factors; (2) varying compositions of evacuation vehicles (buses, vans, lorries, and boats) would be optimally required to evacuate different clusters of affected victims; (3) incorporating fleet adjustment strategy would impact the total evacuation time to a certain extent. (4) the proposed approach is computationally efficient to yield the optimal fleet planning decisions. Concisely, it is anticipated that the proposed bi-objective fleet planning approach, which possesses beneficial managerial implications, could assist relevant stakeholders, especially emergency planners and rescue teams, in implementing a smart evacuation strategy. © 2025 Elsevier Ltd.en-USFleet planningFlood evacuationHeterogeneous vehiclesSafety sciencesScenario analysisVictim demand modelingDisastersEconomicsFleet operationsFlood controlFloodsVehiclesDemand and supplyDemand modellingHeterogeneous fleetSafety scienceScenario-basedScenarios analysisArticleclustering algorithmcomputer modelcontrolled studydecision makingdisaster planningdisaster responseenvironmental parametersfemaleflood evacuationfloodinghealth serviceheterogeneous fleet planninghigh water levelhumanlow water levelMalaysiamalemoderate water levelmotor vehiclestakeholder participationunsupervised machine learningvictimGlobal warmingtext::journal::journal article