A semiconductor prices fluctuate due to changing supply and demand for semiconductors. Research for better productivity has become important as well as research for semiconductor technologies as production processes have become complicated. In this regard, this paper suggests a bottleneck detection framework to enhance productivity in semiconductors.
The semiconductor wafer FAB has various features such as the use of re-entrant production methods or batch equipment. Improving bottlenecks is crucial for better productivity in the semiconductor wafer FAB. This paper defines bottleneck problems and classifies actual bottlenecks for improvement. Furthermore, it proposes a bottleneck detection framework based on a simulation to detect actual bottlenecks.
This paper investigates actual bottlenecks in a production process perspective and aims to detect bottlenecks to improve productivity. Actual bottlenecks do not refer to those producing a small number of products or having a large number of WIP, but rather refer to those failing to meet targets in spite of their potentials. They also can have better productivity than other bottlenecks when problems are resolved. Equipment capacity is fixed. Facilities are able to meet target move rates but processes sometimes fail to keep up with facilities leading to missing target moves. Therefore, this study aims to investigate bottleneck steps rather than to bottleneck facilities.