Modelling Empty Container Repositioning Logistics

Getting empty containers back to the places that need them is a logistics challenge, and this volume approaches it through the discipline of inventory control. The author examines the problem within regional transport systems, where demand differs from one location to another, shifts over time, and is hard to predict. Uncertainty in transport and the practice of leasing containers are also treated as part of the picture.

The book first situates this issue among related problems, surveying the redistribution of empty freight vehicles, empty passenger vehicles, empty bicycles, empty container chassis, and empty containers. It explains what the container case shares with these others and where it departs from them.

Stochastic dynamic programming is the central analytical method. It provides an algorithmic route to identifying optimal policies and models decisions made in sequence, anticipating uncertainties that unfold across time and location. For relatively simple systems, the author derives closed-form solutions and structural properties of the best policies. These findings then guide threshold-based policies for more intricate networks. Such policies resemble the (s, S) and (s, Q) rules familiar from inventory control theory. Their stated strengths are that they are decentralized, simple to understand and operate, responsive to random events, and require minimal computation and communication during operation.

For complex stochastic optimization problems, the book introduces a range of advanced techniques: approximate dynamic programming, simulation-based metaheuristics, stochastic approximation, perturbation analysis, and ordinal optimization. It is aimed at researchers and professionals in logistics, transport, supply chain, and operations research.