Abstract
Based on a simulated Non Volatile Memory fab, we employ data-mining to identify and quantify the apparent causes of work in process bubbles along the process. The chosen bubble formalization methods proved able to detect the phenomenon and enabled its occurrence frequency to be forecasted. In the chosen environment, bubbles seem to be highly correlated with the utilization patterns of the process segment considered.
Original language | English |
---|---|
Article number | 5357373 |
Pages (from-to) | 13-20 |
Number of pages | 8 |
Journal | IEEE Transactions on Semiconductor Manufacturing |
Volume | 23 |
Issue number | 1 |
DOIs | |
State | Published - Feb 2010 |
Externally published | Yes |
Keywords
- Congestion
- Production management
- Semiconductors
- Work-in-process