Using Public Transport in a 2-Echelon Last-Mile Delivery Network

In this paper, we investigate the integration of public transport services into last-mile delivery networks. The free capacity of an already established public transport system that operates according to a given timetable on predetermined lines is used to carry goods from outside of the city into the city center. Dedicated bus or tram stations of the public transport network serve as satellites. From these satellites, city freighters pick up the goods and deliver them to the final customers. The main contribution is an extensive computational study, in which we consider various scenarios of the shared part of the public transport network and investigate different hierarchical objectives minimizing two of the three key indicators (number of city freighters, routing costs, and number of trips). We quantify the tradeoff between the key indicators for different objectives and the impact of limiting the capacity of public transport vehicles. The key instrument to conduct all these experiments are effective exact and heuristic branch-price-and-cut algorithms that we develop and evaluate in detail. In a further study, we analyse the algorithmic performance and evaluate the cost increase when customer time windows are reduced, e.g., as a reaction to increased customer dissatisfaction.