Uncorking efficiency: A network Data Envelopment Analysis agent-based modelling approach for optimising collaborative wine supply chains
提出一种结合网络数据包络分析与智能体建模的新框架，通过模拟阿连特茹葡萄酒产区供应链，分析协作如何在不同风险情景下提升整体效率，并揭示资源权衡与绩效变化。
The wine industry faces growing challenges associated with risk management, sustainability, and supply chain (SC) efficiency. This study proposes a novel methodological framework that integrates Network Data Envelopment Analysis (NDEA) and Agent-Based Modelling (ABM) to measure SC efficiency under varying risk and sustainability conditions. The framework extends conventional NDEA by explicitly modelling inter-decision-making-unit (DMU) product exchanges, thereby capturing the collaborative dynamics frequently observed in real-world SCs. Using the Alentejo wine region as a reference context, synthetic data are generated through ABM to simulate a range of plausible SC configurations. Results illustrate how collaboration between SCs can enhance overall SC efficiency, particularly in high-risk scenarios, while also revealing trade-offs and resource reallocations between global and local performance across SC stages. Although illustrative, the simulation-based results provide insights into the implications of inter-DMU collaboration, offering a foundation for future empirical validation and policy design. The study contributes methodologically by extending NDEA to incorporate inter-organisational exchanges and demonstrates how simulation can support complex efficiency measurements under uncertainty. Exploratory correlation analyses further reveal that collaborative flows are significantly associated with scenario-dependent efficiency gains, especially in upstream stages, reinforcing the explanatory power of the proposed model. • Proposes a network DEA model with inter-DMU collaboration in supply chains. • Uses Agent-Based Modelling to generate synthetic data under uncertainty. • Shows how collaboration affects efficiency in simulated risk scenarios. • Reveals correlations between collaborative flows and efficiency gains. • Provides a methodological basis for future empirical supply chain studies.