Cluster Tool State Estimation via Weighted Entity Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional techniques for assessing cluster tools in semiconductor manufacturing yield less accurate results due to reduced measurement accuracy and confidence, particularly when dealing with complex cluster tools and identical process entities, leading to unrealistic reliability and maintainability metrics.
Innovation Solution
A technique that defines combined states for cluster tools by establishing appropriate weighting or normalization factors for individual sub-states, allowing for a more accurate and reliable assessment of cluster tool characteristics, such as reliability, availability, and maintainability, by considering the influence of specific sub-states on the total state, and using a hierarchy to rank states for improved measurement efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional techniques are used to assess cluster tools by treating each entity independently, then the assessment process is simple, but the measurement accuracy and reliability of the metrics are reduced
Solution Approach 1:
The cluster tool is segmented into multiple entities, each assigned a specific state (productive, standby, engineering, scheduled downtime, unscheduled downtime, non-scheduled). This segmentation allows independent monitoring of each entity while maintaining overall system accuracy, resolving the contradiction by enabling precise measurement without requiring complex inter-entity analysis.
Solution Approach 2:
The states of multiple entities are merged into a combined cluster tool state through weighted aggregation. The state estimation unit combines individual entity states with their respective weights to determine the overall cluster tool state, achieving accurate system-level assessment while keeping individual entity assessment simple.
2Reliability
If individual entity states are monitored separately without weighting, then the monitoring process is straightforward, but the reliability and maintainability metrics become unrealistic
Solution Approach 1:
Weighting factors are assigned to different entities based on their importance and capacity. This parameter change transforms the simple state monitoring into a weighted state estimation system, where each entity's contribution to the overall cluster tool state is proportionally adjusted, resulting in more realistic reliability and maintainability metrics.
Solution Approach 2:
The state estimation unit continuously receives state information from each entity and automatically updates the combined cluster tool state. This automated feedback loop ensures that reliability metrics are continuously refined based on actual entity states, improving metric accuracy while maintaining automated operation.
3Measurement precision
If all entity states are given equal weight, then the calculation is simple, but the influence of critical sub-states on the total state is not properly reflected
Solution Approach 1:
Different entities are assigned different weighting factors based on their local importance and functional capacity within the cluster tool. Critical entities receive higher weights, ensuring their state has greater influence on the overall cluster tool state estimation, thereby improving measurement accuracy without requiring an excessive number of factors.
Data Source
AI summary
By using weighted entity states for representing a state of a cluster tool, a highly efficient technique for the measurement and monitoring of cluster tool characteristics, such as reliability, availability and maintainability, is provided. For example, individual entities of the cluster tool may be weighted according to their capacity and corresponding entity states may be ranked in accordance with a predefined hierarchy structure, thereby enabling an efficient combination of weighted entity states so as to represent the cluster tool state.