Integrating Supply, Production, and Demand Uncertainties in Manufacturing Inventory Systems

Effective inventory management in manufacturing systems is vital for enhancing production efficiency and reducing costs, as multisourcing uncertainties pose a significant challenge. This study addresses the often-overlooked issue of production uncertainty by extending the classic newsvendor model to integrate uncertainties in production, supply, and demand within a multiperiod framework. A novel multiperiod newsvendor model is developed to determine the optimal order quantity, minimizing total costs, including ordering, production, holding, and shortage costs. Given the lack of distributional knowledge for uncertain parameters, we adopt a robust optimization approach, constructing two distinct uncertainty sets: 1) box and ellipsoidal and 2) budget-based, to model the uncertainties in supply, production, and demand. The model is reformulated into a tractable second-order cone programming problem. Computational experiments demonstrate the effectiveness and robustness of the model, showing strong resilience to parameter variations and price fluctuations. Managerial insights drawn from numerical experiments highlight the strategic advantage of leveraging early-stage supply to build inventory buffers in multiperiod, multisource uncertainty scenarios. The findings emphasize prioritized raw material acquisition in initial periods to counter cumulative risks, coupled with responsive order adjustments guided by real-time demand–production fluctuations and critical evaluations of supplier reliability. These findings underscore the practical applicability of the model in addressing real-world challenges within complex and uncertain manufacturing environments.