Samsung Uses Data-Driven Approach to Manage Work-in-Process of Bottlenecks in Semiconductor Manufacturing Operations

Semiconductor manufacturing is a repetitive and time-consuming process, with cycle times ranging from one to three months. To improve market responsiveness and reduce operational costs, minimizing work-in-process (WIP) throughout the manufacturing process is crucial, especially for the bottleneck process because it determines the efficiency of the entire production system. To achieve this goal, we utilize optimization and statistical techniques based on real-world data to determine the optimal level and threshold bounds for managing the WIP of the bottleneck process. Our results show a 35% reduction in the average WIP level of the bottleneck process without compromising its productivity, demonstrating the effectiveness of our approach in practical applications.

History: This paper was refereed.