Date of Award

2026

Degree Type

Thesis

Degree Name

Master of Science in Systems Engineering

Department

Mechanical, Industrial and Systems Engineering

First Advisor

Valerie Maier-Speredelozzi

Abstract

Service-parts logistics requires high product availability and short delivery times, while companies are increasingly expected to reduce the environmental impact of transportation activities. This creates a strategic problem in the design of warehouse networks. Specifically, there is a need to position inventory in close proximity to customer demand in order to support service requirements. However, decisions regarding warehouse location and customer assignment also influence transport distances, transport modes, and related carbon emissions. The problem that this thesis addresses is how an additional warehouse location and optimized customer-warehouse assignments can affect service coverage and transport-related CO2e emissions in a U.S. service-parts distribution network.

To resolve this issue, this thesis proposes and implements a lexicographic multi-objective warehouse location-allocation model. The model prioritizes demand-weighted service coverage as the primary objective, with a secondary objective of minimizing modeled transport-related CO2e emissions. This approach is indicative of a service-first decision context, wherein emission reductions are pursued without compromising the optimal service performance that is achievable. The model is applied to historical shipment data from an industrial partner, using ZIP-code-level customer demand, existing warehouse locations, and a predefined set of candidate locations derived from demand patterns and logistics-infrastructure considerations.

Several service-distance scenarios are evaluated, including strict ground-compatible one-day thresholds, a relaxed two-day ground-compatible threshold, and an empirical all-mode one-day benchmark derived from historical one-day shipment distances. The findings indicate that augmenting the existing infrastructure with an additional warehouse, in conjunction with the optimization of customer-warehouse assignments, can enhance service accessibility and reduce modeled transport-related emissions. The implementation of strict ground-compatible thresholds has been identified as a key factor in the limitation of attainable service coverage. This limitation is attributed to the geographic dispersion of demand and warehouse locations. In contrast, the empirical all-mode one-day scenario achieves full demand-weighted service coverage within a historically observed one-day distance benchmark. The emission analysis indicates that modeled CO2e reductions are primarily driven by lower total tonne-kilometers rather than substantial changes in average emission intensity.

Further analyses provide the operational and managerial context. The sensitivity analysis is a methodical assessment that evaluates the stability of the lexicographic solutions. The inventory analysis is a data-driven approach that demonstrates that optimized customer assignments require aligned product availability at the assigned warehouses. A time-zone diagnostic indicates that time-zone effects are not the dominant driver of network performance. Furthermore, a Midwest-only future-state scenario illustrates potential trade-offs in alternative strategic network configurations.

The thesis proposes a service-oriented decision-support approach for evaluating warehouse location and allocation decisions under environmental considerations. The study proposes a structured basis for analyzing trade-offs between service accessibility and modeled transport-related emissions in service-parts logistics by combining demand-weighted service coverage with tonne-kilometer-based emission estimation in a lexicographic optimization framework.

Download

Available for download on Sunday, September 10, 2028

Share

COinS