Risk-Controlled Product Mix Planning in Semiconductor Manufacturing Using Simulation Optimization
Kuo-Hao Chang
Abstract
Kuo-Hao Chang
Abstract
Production ramp-up refers to the period between the completion of product development and full capacity utilization. In semiconductor industry, engineering lots are given higher priority in the manufacturing process in order to improve the manufacturing process and/or to facilitate the new product development. However, the implementation of engineering lots can disrupt the smoothness of the manufacturing process and thus increase the cycle time of normal lots. Because semiconductor manufacturing is a very complicated manufacturing process, the determination of the appropriate product mix is a difficult task for many semiconductor companies, especially when risks are taken into consideration. We propose a model to characterize this hard tradeoff and apply an efficient simulation optimization method to obtain the optimal product mix where risks are taken into consideration. An extensive numerical study shows that the proposed method has satisfactory performance and is capable of identifying the nearly optimal product mix in reasonable computing time. An empirical study conducted in collaboration with a semiconductor company further validates the viability of this research in practical settings.
OpenAlex reports 4 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
Production ramp-up refers to the period between the completion of product development and full capacity utilization. In semiconductor industry, engineering lots are given higher priority in the manufacturing process in order to improve the manufacturing process and/or to facilitate the new product development. However, the implementation of engineering lots can disrupt the smoothness of the manufacturing process and thus increase the cycle time of normal lots. Because semiconductor manufacturing is a very complicated manufacturing process, the determination of the appropriate product mix is a difficult task for many semiconductor companies, especially when risks are taken into consideration. We propose a model to characterize this hard tradeoff and apply an efficient simulation optimization method to obtain the optimal product mix where risks are taken into consideration. An extensive numerical study shows that the proposed method has satisfactory performance and is capable of identifying the nearly optimal product mix in reasonable computing time. An empirical study conducted in collaboration with a semiconductor company further validates the viability of this research in practical settings.
Key concepts: Semiconductor device fabrication, Product mix, Manufacturing engineering, Product (mathematics), Production planning, Semiconductor industry, Computer science, Engineering