PSO-Based Optimization of Shipping Box Configurations: An Empirical Study with South Korean Enterprise Data

Background: The rapid growth of e-commerce has intensified the need for packaging strategies that reduce logistics costs and environmental impact. Traditional box recommendation methods select the best-fitting box from a fixed set of options, which limits their ability to minimize unused space and total costs. Methods: This study formulates the Shipping Box Configuration Problem (SBCP), which aims to determine an optimal set of box types and dimensions for multi-product orders. To solve this problem, we propose a Particle Swarm Optimization (PSO)-based heuristic that dynamically designs box configuration rather than selecting from predefined sizes. Results: The proposed method is evaluated using real order data from two South Korean e-commerce companies with different product characteristics and existing box configurations. Computational results show that the PSO-based approach reduces total packaging and shipping costs and improves space utilization compared to current box configurations. The analysis also indicates that increasing the number of box types and reducing safety ratios generally lead to cost savings, although these effects must be balanced against operational complexity. Conclusions: The results suggest that adaptive box configuration design can improve both economic efficiency and environmental performance, providing practical guidance for e-commerce logistics managers seeking to optimize packaging strategies under operational constraints.