Next generation advanced process control: Leveraging big data and prediction

Advanced process control (APC), which includes run-to-run (R2R) process control and fault detection and classification (FDC) is pervasive in fabs today. While APC provides many benefits, it also presents many implementation and operational challenges including significant up-front configuration cost, lack of solution accuracy resulting in false/missed alarms, and lack of a consistent capability for prognostics leading to scrap, lost productivity and reduced quality. Next Generation APC (NG-APC) is an evolutionary solution that addresses these challenges and provides new capabilities for APC solutions. It leverages recent advancements in the traditional APC capabilities as well as benefits provided by big data and prediction capabilities. Key enhancements include single-threaded R2R control for improved high-mix process control capability, semi-automation of FDC model development with model and limits optimization through a move to "supervised" FDC, improved data management and analysis capabilities through leveraging of big data architectures, and incorporation of prediction capabilities in solutions for improved process quality and uptime with reduced cost. Additionally, NG-APC is structured as an extensible solution, thus it can continue to evolve as new APC and related technologies emerge.