A hybrid bi-objective optimization approach for joint determination of safety stock and safety time buffers in multi-item single-stage industrial supply chains. (June 2022)

• Record Type: Journal Article
• Authors: Silva, Pedro M.; Gonçalves, João N.C.; Martins, Tiago M.; Marques, Luís C.; Oliveira, Miguel; Reis, Marcelo I.; Araújo, Luís; Correia, Daniela; Telhada, José; Costa, Lino; Fernandes, João M.
• Is Part Of: Computers & industrial engineering. Volume 168(2022)
• Publication Date: 2022-06
• Subjects: Safety stock -- Safety time -- Multi-objective optimization -- Decision support; Engineering -- Data processing -- Periodicals; Industrial engineering -- Periodicals; 620.00285
• DOI: 10.1016/j.cie.2022.108095
• Languages: English
• ISSNs: 0360-8352

Abstract

Highlights: Safety stock and safety time inventory buffers are jointly optimized. Inventory holding costs are minimized while maximizing service level. A decision support system is implemented in a major automotive electronics company. The practical applicability of our approach is demonstrated on real-world data. Abstract: In material requirements planning (MRP) systems, safety stock and safety time are two well-known inventory buffering strategies to protect against supply and demand uncertainties. While the role of safety stocks in coping with uncertainty is well studied, safety time has received only scarce attention in the supply chain management literature. Particularly, most previous operations research models have typically considered the use of such inventory buffers in a separate fashion, but not together. Here, we propose a decision support system (DSS) to address the problem of integrating optimal safety stock and safety time decisions at the component level, in multi-supplier multi-item single-stage industrial supply chains under dynamic demands and stochastic lead times. The DSS is based on a hybrid bi-objective optimization approach that simultaneously optimizes upstream inventory holding costs and β -service levels, suggesting multiple non-dominated Pareto-optimal solutions to decision-makers. We further explore a weighted closed-form analytical expression to select a single Pareto-optimal point from a set of non-dominated solutions, thereby enhancing the practical application of the proposed DSS. We describe the implementation of our approach in a major automotive electronics company operating under a myriad of components with dynamic demand, uncertain supply and requirements plans with different degrees of sparsity. We show the potential of our approach to improve β -service levels while minimizing inventory-related costs. The results suggest that, in certain cases, it appears to be more cost-effective to combine safety stock with safety time compared to considering each inventory buffer independently.