Information-Based Allocation Strategy for GRID-Based Transshipment Automated Container Terminal

In this paper, we introduce a storage allocation strategy for a transshipment container hub using a new automated container terminal called the hybrid GRID system. This study provides a novel approach of developing an efficient storage allocation strategy for new terminal concepts so that the complex formulations can be approximated by simple functions that can be quickly computed. Specifically, the storage allocation strategy is derived from an optimal allocation decision learned from a Mixed Integer Programming (MIP) model. Some input parameters of the MIP model are collected from a simulation model. An index measuring the storage location convenience is proposed and we regress this index with important variables to build an empirical model that provides recommendations on where to allocate containers to storage locations. The advantage of using the empirical approach is that it allows for fast computation which is expected in the dynamic and uncertain port environment. Numerical results show that four variables can significantly affect the performance of the allocation decision. This strategy is shown to be robust for different scenarios in terms of terminal configuration and container workload. In addition, the new strategy performs better than the commonly used strategies found in warehousing literatures, and the solution is close to the optimal allocation decision. The online appendix is available at https://doi.org/10.1287/trsc.2017.0736 .