Substrate Processing ML Scheduling for Yield and Downtime
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Manufacturing facilities face inefficiencies in scheduling and routing substrate processing due to manual planning, which is time-consuming and often results in suboptimal schedules and decreased yield, exacerbated by equipment interruptions.
Innovation Solution
A machine learning platform that utilizes historical and simulated data to generate efficient schedules and routes by training a machine learning model to predict corrective actions, improving dispatching, scheduling, and maintenance planning.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If manual planning is used for scheduling and routing substrate processing, then flexibility in handling equipment interruptions is maintained, but time consumption increases and yield decreases
Solution Approach 1:
The patent replaces manual mechanical planning processes with an automated machine learning-based scheduling system. The ML model analyzes historical data, equipment status, and process parameters to generate optimized schedules automatically, eliminating the time-consuming manual planning process while improving yield through data-driven decision-making.
Solution Approach 2:
The scheduling system performs self-service by automatically generating and adjusting schedules based on real-time equipment status and historical performance data. The machine learning model continuously learns from new data and autonomously optimizes routing and scheduling decisions without requiring manual intervention, thereby reducing time consumption and improving productivity.
2Productivity
If manual scheduling is used, then system complexity remains low, but scheduling efficiency and accuracy deteriorate
Solution Approach 1:
The patent introduces a machine learning model as an intermediary between raw process data and scheduling decisions. This intermediary layer processes historical data, equipment status, and process parameters to generate optimized schedules, thereby improving scheduling efficiency and accuracy while managing system complexity through modular architecture.
Solution Approach 2:
The system performs preliminary actions by pre-processing historical data and training the machine learning model in advance. This preliminary training phase enables the system to make rapid, accurate scheduling decisions in real-time operations, improving scheduling efficiency without adding complexity to the operational phase.
3Reliability
If equipment maintenance is performed periodically, then equipment reliability is maintained, but production interruptions increase
Solution Approach 1:
The machine learning model performs preliminary analysis of equipment data to predict potential failures before they occur. By identifying equipment that requires maintenance in advance, the system can schedule maintenance during optimal times that minimize production interruptions while maintaining equipment reliability, thus resolving the contradiction between preventive maintenance and production continuity.
Data Source
AI summary
A method includes identifying at least one of historical data associated with historical substrate lots processed by substrate processing tools in a substrate processing facility or simulated data for simulated substrate lots processed by simulated substrate processing tools. The method further includes generating features from the at least one of the historical data for the historical substrate lots or the simulated data for the simulated substrate lots. The method further includes training a machine learning model with data input comprising the features to generate a trained machine learning model. The trained machine learning model is capable of generating one or more outputs indicative of one or more corrective actions to be performed in the substrate processing facility.