Adaptive Intelligent Process Control (AIPC)
- Category: Research project
- Duration: 2012-2013
- Funded by: ELMOS Central IT Services GmbH
- Project holder: Institute of Knowledge Based Systems & Knowledge Management
The aim of the cooperation is to counter the existing problem in semiconductor manufacturing for intelligent control of complex process steps such as SOG-Coating, Plasma Etching and CMP. Also is to apply methods of AI for learning, reasoning and problem solving with consideration of human intervention. By this, a classical gap between process and development is bridged, which in conventional approaches constrains or blocks the emerging of new ideas and sustaining production excellence. The kind of intelligent process controlling provides an opportunity to include process engineer’s expertise in the controller algorithm in a hybrid way (i.e. combination of heuristics and mathematical methods).
Adaptive Intelligent Process Control (AIPC) is software for modeling and controlling SOG-Etching. The software uses the principles of traditional and intelligent controllers to handle the complex process of Etching. The software is designed to be extendable and adaptable for controlling other processes in future. Therefore AIPC is designed as a component based software. The Equipment Manager (EM) and the Controller Manager (COM) are organizing units of the AIPC Framework. The EM is gathering the Equipments and their attributes (chambers, variables, etc.) from the AIPC Package which is the Interface to the ELMOS Database. In this way the characteristics of the Front-end Equipments are also extensible for a future interface to add new characteristics of available equipment, or to integrate new equipment to the AIPC Framework.
The COM is a separate manager to enable and initiate the analyze process which is available and chosen from the Front-end Personal (End user). The information about the possible analyzing method gets the COM from the Virtual Controller Storage (VCS) which is a structure of administrative tables. This structure is also extensible and common for each of the equipments. This kind of architecture preserves the point of Adaptiveness of the AIPC Framework.
The currently developing software consists of two controllers for modeling and controlling of the SOG-Etching:
- Proportional–Integral–Derivative (PID) Controller
- Intelligent Controller (Artificial Neural Network based controller)
Project partner: ELMOS Central IT Services GmbH