Data-driven dynamic stacking strategy for export containers in container terminals

Abstract

This study investigates a method for improving real-time decisions regarding the storage location of export containers while the containers are arriving. To manage the decision-making process, we propose a two module-based data-driven dynamic stacking strategy that facilitates stowage planning. Module 1 generates the Gaussian mixture model (GMM) specific to each container group for container weight classification. Module 2 implements the data-driven dynamic stacking strategy as an online algorithm to determine the storage location of an arriving container in real time. Numerical experiments were conducted using real-life data to validate the effectiveness of the proposed method compared to other alternative stacking strategies. These experiments revealed that the performance of the proposed method is robust, and therefore it can improve yard operations and container terminal competitiveness.