The premium service of express integrators is the overnight delivery of packages within regions as large as Europe. Designing an efficient network of flights to support such service is a task known as the Express Shipment Service Network Design problem (ESSND). Normally, when the ESSND problem is modeled, the time frame considers one day of operations, using the expected commodity demands for that period as input. Then, the resulting operations plan will be repeated every day for months. However, the literature does not consider demand variabilities, while, in practice, the commodity demands change every day. With certain frequency, some peak commodity demands prevent the express integrators to deliver all the packages on time, which is costly and affects negatively the customers’ perception. In this work, we propose a robust model for solving the tactical ESSND problem with demand uncertainty. Based on the Light Robustness approach, we develop a model that maximizes the demand uncertainty that can be absorbed with a specific budget, while ensuring a minimum service level for each commodity. By solving a set of experiments on realistic instances, we evaluate how much our model can reduce the unmet demand caused by the demand variability, compared to a deterministic approach.
Quesada Perez, J. M., Lange, J.-C., & Tancrez, J.-S. (2018). A robust optimization approach to solve the multi-hub express shipment service network design problem. EURO 2018, Valencia, Spain. https://hdl.handle.net/2078.5/173143