Constructing daily delivery and pickup tours for postmen is a critical yet complex task for Swiss Post. Although a dedicated planning system is used to generate these tours, last-minute disruptions may occur and render them inaccessible. This project, carried out in collaboration between Swiss Post and the University of Fribourg, aims to develop backup algorithms capable of rapidly generating high-quality tours to ensure operational continuity in such situations, while also ensuring that their implementation is compatible with deployment in a mobile application.
Beyond its practical importance, the project raises research challenges related to designing algorithms for NP-hard combinatorial optimization problems. It also opens avenues for handling scenarios in which some information may be partially or completely unavailable, thus motivating the exploration of hybrid methods that combine machine learning with optimization—an area that is gaining increasing prominence in Operations Research.
The project team from the University of Fribourg includes Dr. Nour Elhouda Tellache (project leader) and Prof. Dr. Bernard Ries, and an intern will be hired to support the project.
