Wafer-Centric Database and Neural Network Scheduler for Semiconductor Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Semiconductor manufacturing environments face challenges in real-time control due to proprietary software from various vendors and the need for efficient data processing across heterogeneous systems, leading to inefficiencies in scheduling and monitoring processes.
Innovation Solution
A wafer-centric database combined with a real-time scheduler using a neural network and a graphical user interface for simulated operation, enabling improved usability and computational efficiency in semiconductor manufacturing processes, particularly through a state machine and neural network-based scheduling system.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If proprietary software from various vendors is used in heterogeneous semiconductor manufacturing systems, then system compatibility and integration are improved, but data processing efficiency and real-time control capability deteriorate
Solution Approach 1:
The patent introduces a central controller as an intermediary component that mediates between heterogeneous process modules with proprietary software and the overall manufacturing system. The central controller implements a standardized interface layer that translates between different vendor-specific protocols and a unified control architecture, enabling data aggregation and processing without requiring modification of the underlying proprietary software. This intermediary approach maintains system compatibility while centralizing data processing functions to improve overall efficiency.
2Loss of information
If chronological logs of sensor data are compiled for all process modules, then complete process information is captured, but the time and computational resources required to find and process relevant data increase
Solution Approach 1:
The patent segments the monolithic chronological log data into module-specific data streams that are processed and filtered at the source. Each process module's sensor data is organized into structured records with predefined schemas that identify critical parameters. The central controller then selectively aggregates only the relevant segmented data needed for specific control decisions, avoiding the need to search through all chronological logs. This segmentation approach maintains complete process information while dramatically reducing data retrieval and processing time.
3Ease of operation
If simple chronological logs are used to record process data, then data storage and retrieval are simplified, but real-time scheduling decisions and process control capability are limited
Solution Approach 1:
The patent implements a dynamic data structure that evolves from simple chronological logging to structured real-time databases. The system maintains chronological records for historical analysis while simultaneously organizing data into structured formats with hierarchical relationships that support real-time querying. The data structure dynamically adapts based on the control needs, providing both the simplicity of chronological storage and the sophistication of structured data access for automated scheduling decisions. This dynamic approach enables both easy data storage and advanced real-time control capabilities.
Data Source
AI summary
Software for controlling processes in a heterogeneous semiconductor manufacturing environment may include a wafer-centric database, a real-time scheduler using a neural network, and a graphical user interface displaying simulated operation of the system. These features may be employed alone or in combination to offer improved usability and computational efficiency for real time control and monitoring of a semiconductor manufacturing process. More generally, these techniques may be usefully employed in a variety of real time control systems, particularly systems requiring complex scheduling decisions or heterogeneous systems constructed of hardware from numerous independent vendors.