Abstract :The imbalances in world trade also affect container traffic and lead to large differences in import and export rates of many locations. As a consequence of this, the surplus containers are repositioned to locations where they are required. This causes extra costs for the container shipping companies. Therefore, one of the main objectives of all shipping companies is to reduce empty container repositioning (ECR) costs. Since empty ECR decisions involve too many parameters, constraints and variables, the plans based on real- life experiences cannot be effective and are very costly. For this purpose, this study introduces two mathematical programming models in order to make ECR plans faster, more efficient and at the lowest cost. The first mathematical programming model developed in this study is a mixed-integer linear programming (MILP) model and the second mathematical programming model is a scenario-based stochastic programming (SP) model which minimize the total ECR costs. Unlike the deterministic model, the SP model takes into account the uncertainty in container demand. Both models have been tested with real data taken from a container shipping company. The numerical results showed that, in a reasonable computational time, both models provide better results than real-life applications of the shipping company. Keywords : Empty Container Repositioning (ECR), Mathematical Programming, Container Logistics
