Dynamic optimization for same-day delivery operations

Abstract

Same-day delivery (SDD) is a service where consumers place orders online on the same day that these are processed and delivered to the customer. Providing a delivery service requires order management at the stocking location (depot), including request acceptance and processing; and order distribution from the depot to final customer locations, including order dispatch and delivery via vehicle routes. In SDD these processes are highly interrelated, are simultaneously executed, and have a high degree of information dynamism. In this thesis, we formulate and solve dynamic optimization problems to improve the operation of SDD systems. We study how our approaches perform over computationally simulated instances and provide managerial insights for SDD practitioners, including structural solution properties that any common SDD service should have, the trade-offs between common SDD objectives, and the logistics cost of operational constraints in SDD.