Purpose This paper investigates the complexities of a multimodal dry port transportation network, operating under the challenging conditions posed by the availability of transportation modes. Design/methodology/approach Our study focuses on a network structure that includes multiple foreign seaports, local seaports, dry ports and manufacturers. We identify a significant research gap in the existing literature on dry port–based multimodal transportation networks. We first develop a mixed-integer non-linear programming (MINLP) model that adeptly integrates freight departure and demand scheduling, inventory management and backlog, while considering the availability of transportation modes in every node. To account for disruptions from uncertain events like extreme weather, we extend this deterministic foundation to a two-stage stochastic programming model. This stochastic model explicitly incorporates uncertainty in transportation mode availability, allowing proactive and adaptive strategies. Findings The computational study, conducted across two distinct setups, demonstrates our models effectively avoiding substantial backlog penalties as well as delivering robust, cost-balanced solutions to handle uncertainty and complexity. In conclusion, the paper presents a comprehensive analysis including sensitivity analysis, dynamic planning evaluation and cross-regional validation, providing insightful guidance for strategic decision-making and demonstrating the model’s adaptability across different operational contexts. Originality/value The insights gained from this analysis not only underline the practicality and strategic significance of our framework but also suggest key strategies for businesses.
Zhang et al. (Tue,) studied this question.