Optimal Transport in Urban Logistics: A City Municipality Case Study

Thesis Abstract

Urban freight logistics in metropolitan areas faces rising congestion, emissions, and delivery delays, threatening service levels and urban livability. This study investigates how optimal transport (OT) frameworks can enhance urban logistics efficiency within a mid-sized city municipality by integrating municipal policy constraints with private sector delivery operations. The aim is to develop and validate an OT-based decision-support model that reduces total transportation cost and emissions while maintaining or improving service quality. Specific objectives are (1) to characterize the current urban freight network, demand patterns, and regulatory constraints; (2) to formulate a mixed-integer OT model that combines regional distribution with last-mile optimization under constraints such as time windows, loading/unloading regulations, and vehicle quotas; (3) to calibrate and validate the model using real-world data from the city’s transportation authority and local courier firms; (4) to compare the OT-based approach against baseline heuristics and route-planning practices in terms of total cost, emissions, and delivery reliability; and (5) to develop policy recommendations for scalable implementation and to identify barriers to adoption. The study adopts a quantitative-empirical research design anchored in transport economics and operations research theory, complemented by a policy integration perspective drawing on sustainable mobility and the Theory of Planned Behavior to explain practitioner adoption of OT-based solutions. The population comprises urban freight operators, municipal planners, and logistics coordinators operating within the city. A purposive sampling strategy selects 15 courier firms, 8 retailers, and 6 municipal departments, with data from 3 representative weeks of operations yielding a dataset of approximately 2,400 daily route records and 900 distinct delivery nodes. Data collection instruments include (i) operational records from courier firms (vehicle type, load, departure and arrival times, geocoded routes); (ii) municipal regulation databases (loading zones, curfews, emissions zones); (iii) traffic and road network data from the city’s GIS; and (iv) semi-structured interviews with 12 frontline logistics managers to capture constraint details not evident in quantitative data. The analytical framework employs a canonical OT model augmented with capacity, time-window, and policy constraints, solved via a bespoke column-generation algorithm implemented in Python with Gurobi for linear programming subproblems. Model validation uses hold-out cross-validation on 20% of the data and out-of-sample testing for seven peak periods. Comparative analyses include deterministic and stochastic Scenario Analysis, paired t-tests for performance differentials, and sensitivity analyses on key parameters (fleet composition, curbside restrictions, and carbon pricing). Descriptive statistics reveal that the existing system incurs higher total cost and average delivery delay than OT-based plans under identical demand, with estimated annual CO2-equivalent emissions reductions ranging from 12% to 24% under policy-enabled routes. Expected findings include (i) quantifiable reductions in total transport cost and distance traveled without compromising service levels; (ii) substantial improvements in delivery reliability and predictability due to integrated routing with time-window constraints; (iii) measurable emissions reductions attributable to consolidation and modal shifts. The study contributes to knowledge by empirically linking OT theory to real-world urban logistics within a municipal governance context, demonstrating how grid-based network optimization can coexist with policy constraints, and by operationalizing a replicable OT framework for city-scale freight planning. Theoretical contributors include an applied synthesis of OT with urban policy design and the application of the Theory of Planned Behavior to elucidate adoption drivers among operators. Practically, the research provides a decision-support tool ready for pilot testing in municipal planning departments and a set of evidence-based guidelines for expanding curbspace management, incentive schemes for consolidation centers, and data-sharing protocols. The main conclusion is that OT-based routing and consolidation can deliver significant efficiency and environmental benefits when municipal constraints are explicitly embedded in the optimization model, and the recommendations advocate staged pilot implementations, continuous data collection for model recalibration, and policy alignment to sustain improvements beyond initial deployments. Limitations include data gaps from smaller operators and potential generalization constraints to cities with markedly different regulatory environments; future work suggests integrating dynamic traffic forecasting and real-time data streams to enable adaptive, near-real-time OT optimization.