Robust facility location selection frameworks for service systems
No Thumbnail Available
Authors
Meeting name
Sponsors
Date
Journal Title
Format
Thesis
Subject
Abstract
[EMBARGOED UNTIL 08/01/2026] This dissertation develops a unified robust optimization framework for strategic facility-location decisions in service logistics under uncertainty. Motivated by the growing need for infrastructure resilience against demand volatility, budget constraints, and environmental targets, this research integrates multi-criteria decision-making (MCDM) methodologies with stochastic, fuzzy, and regret-based optimization techniques across increasing levels of network complexity. Study 1 focuses on robust single-facility location under uncertainty, proposing a two-stage potential-assessment and siting approach for regional airports. The methodology combines a Robust Slack-Based Measure Data-Envelopment Analysis (SBM-RDEA) to assess existing facilities, regression modeling to identify influential factors, and GIS-based weighted overlays for suitability mapping. Applied to Sistan-and-Baluchestan Province, Iran, the model significantly narrows the candidate location area, pinpointing Zahedan County as optimal, outperforming conventional entropy--AHP methods. Building upon this approach, Study 2 expands the scope to multiple-facility network location under uncertainty, specifically addressing a closed-loop dry-port network. It employs a hybrid Fuzzy SWARA-COPRAS method to derive expert-validated weights for economic, environmental, infrastructural, and socio-political criteria. These weights guide a multi-stage fuzzy-stochastic chance-constrained program utilizing a novel robustness metric that balances expected benefit, worst-case profit, and bounded relative regret. An Iranian national case illustrates that establishing four strategically located inland terminals interconnected by rail substantially reduces costs, lowers CO₂ emissions, and maintains profitability under significant demand fluctuations. Study 3 further advances the model by integrating routing considerations into multiple facility location decisions, focusing on depot and battery-swapping networks for delay-sensitive drone deliveries. The research formulates a lexicographic, regret-aware location--routing MILP that simultaneously maximizes on-time service reliability, extends spatial coverage, and minimizes capital investments amidst drone endurance uncertainties. Computational experiments using adaptive large-neighborhood search achieve near-optimal solutions for metropolitan-scale problems. A case in Dallas--Fort Worth demonstrates the practical benefits of optimized depot layouts in meeting critical service deadlines for medical deliveries at reduced infrastructure investment. Collectively, the dissertation reframes robustness as a comprehensive design principle, bridging strategic location planning with tactical network optimization and operational routing decisions. By coupling data-driven criteria weighting with robust and tractable optimization models, it provides actionable frameworks for planners in air transport, inland freight logistics, and drone delivery services. These tools enable informed decision-making, effectively balancing cost, resilience, and sustainability across diverse future scenarios.
Table of Contents
PubMed ID
Degree
Ph. D