Collection of different types of milk with multi-tank tankers under uncertainty: a real case study

The milk collection problem can be basically defined as the daily collection of raw milk from different points (farms/milk collection centers) and delivering it to a dairy facility. To enhance final dairy quality, it is important to use the appropriate quality of raw milk for each dairy product. However, collecting different types of milk comes with additional logistics costs which significantly increase in an uncertain environment. In this study, we, therefore, propose a novel mathematical model for the collection of different types of milk from producers by multi-tank tankers with split deliveries, uncertain demand, service time and vehicle speed conditions. A real-life case study from a dairy company is solved under different risk assessment scenarios. Indeed, several brand-new benchmark instances for the core problem are presented and solved by utilizing an efficient heuristics approach called enhanced iterative local search. All the case study results show that considering the uncertainty is very critical for designing efficient collection networks. The findings of the study indicate that logistics decision makers should design their collection networks with low, but non-zero, risk levels.