Semiconductor: Fab-wide Control and Disruption Management in High Volume Semiconductor Manufacturing

University Of Texas At Austin, Austin TX

Investigators

Abstract

The semiconductor industry has advanced from 200mm to 300mm wafers and will transition to 450mm by 2012 according to the roadmap of the Semiconductor Industry Association. Along with this transition is the doubling or tripling of capital expenditure. The aim of this project is to develop new control and optimization methods to improve the yield and reduce operational cost in semiconductor manufacturing. This grant provides funding to develop a new hierarchical control and optimization framework for high volume manufacturing and a novel disruption management methodology for real-time operations recovery in an uncertain manufacturing environment. The new hierarchical control framework makes use of the multilevel measurement available in a modern fab to achieve on-target manufacturing of electrical properties and geometric specifications. The real-time disruption management provides real time recovery to ensure on-time delivery when unexpected tool shutdowns or uncontrollable disruptions occur. The proposed tasks are tightly coupled in that the disruption management system is integrated with fab-wide control to provide control-friendly production schedules and sequences. The proposed methods deal with complex manufacturing issues with the power of sophisticated nonlinear optimization theory to help industry make intelligent real time decisions and control. This project, if successful, will provide a fab-wide control framework and real time disruption management system to improve competitiveness of the U.S. semiconductor industry as it is transitioning to 300mm technology and 450mm technology in the foreseeable future. On-target product quality will be optimally controlled with minimal waste of raw material and resources. Manufacturing equipment, which is very expensive now, will have improved utilization and reduced idle time. The fab operations will be less dependent on the manual operation of the fab personnel and the overall competitiveness of the U.S. chip manufacturing will be improved. The intellectual merit of the project is highlighted by the creation of a new framework and method for semiconductor manufacturing control and operations. The proposed fab-wide control framework and disruption management will contribute significantly to the knowledge of operations and control in semiconductor manufacturing.