Ph.D. Thesis Defense
Lloyd Teta
(Faculty advisor: Professor Dimitri Mavris)
"A Framework for Concurrent Design and Fleet Operations Optimization of Unmanned Aerial Vehicle-Based Last-Mile Delivery Networks"
Tuesday, November 10
12:00 p.m.
Weber, CoVE
Abstract:
Last-mile delivery is one of the most costly and operationally complex stages of parcel distribution. Unmanned Aerial Vehicles (UAVs) offer a promising complement to ground delivery by bypassing road congestion and providing rapid service for lightweight packages. However, their limited payload capacity, battery energy, and endurance constrain their operational capabilities. A delivery network also requires aircraft to perform a wide range of missions with different distances, payloads, and energy requirements. Conventional aircraft design typically sizes a vehicle for a representative mission, while logistics planners optimize delivery networks around an existing or assumed aircraft. Decoupling aircraft design from delivery network planning can lead to undersized or oversized aircraft, compromising network feasibility, increasing fleet requirements, and raising overall delivery costs.
This dissertation develops Concurrent Optimization of Network, Design, and Operations (CONDOR), a framework that integrates Conceptual Vehicle Design (CVD) and Fleet-Level Operations (FLO) planning for UAV-based last-mile delivery. The framework is developed and evaluated through three research areas. The first formulates a Location-Fleet-Routing Problem (LFRP) that jointly optimizes depot placement, fleet sizing, and delivery routes and evaluates its advantages over sequential planning. The second extends fleet-level planning to uncertain demand and examines the trade-offs between cost, infrastructure investment, and service reliability. The third develops a multidisciplinary design optimization (MDO) architecture that couples aircraft design with delivery network planning and evaluates how aircraft characteristics and operational requirements influence system performance.
The results demonstrate that coordinating depot, fleet, and routing decisions can reduce operating costs and improve resource utilization. Planning under demand uncertainty reveals the additional infrastructure required to achieve reliable service and the limits of feasible coverage. The overarching demonstration shows that concurrent aircraft and network optimization can achieve lower delivery costs than conventional design-first and network-first approaches while maintaining the same service coverage.
The findings establish a methodology for evaluating aircraft designs through their effects on network feasibility, fleet requirements, and total system cost, providing a foundation for more efficient UAV-based delivery systems.
Committee:
Dr. Dimitri Mavris (advisor), School of Aerospace Engineering
Dr. Daniel P. Schrage, School of Aerospace Engineering
Dr. Kai A. James , School of Aerospace Engineering
Dr. Raphaël H. Gautier , School of Aerospace Engineering
Dr. Marco Ceze , Prime Air, Amazon