Inverse network DEA for Industry 4.0: identifying pitfalls and proposing solutions

Purpose This study aims to develop the application of data envelopment analysis (DEA) models in network structural systems to assess the performance of production systems. It introduces the inverse DEA (IDEA) model to assist decision-makers in optimizing resource allocation and efficiency through sensitivity analysis while addressing challenges like infeasibility and inaccurate estimations. Design/methodology/approach A theoretical framework is developed to integrate DEA with network structures and the IDEA model, including mathematical formulations for efficiency measurement and parameter estimation. Case studies demonstrate the practical application of these models in real-world production systems, comparing their performance with benchmarks and proposing protocols to resolve infeasibility and estimation challenges. Findings The study considers the significance of network structural systems in the production process and introduces the IDEA model to improve efficiency scores and resource optimization. By addressing infeasibility and estimation inaccuracies, the research suggests protocols that improve decision-making and theoretical contributions to DEA and inverse analysis. Originality/value This work provides innovative solutions for applying IDEA to network systems, bridging theoretical advancements with practical applications in Industry 4.0, thus contributing to more effective production management.