Diagnostic System for Industrial Process Sub-Domains
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current diagnostic systems for industrial processes, such as lithographic processes, are limited in their ability to consider a wide range of context variables, leading to sub-optimal performance and the need for subjective human intervention to combine results from disconnected diagnostic sub-systems, which hampers the discovery of new causes and relationships between variables.
Innovation Solution
A diagnostic system that implements a network of sub-domains with probabilistic connections, allowing for the influence of diagnostic information from one sub-domain on another, using Bayesian inference to propagate belief and integrate knowledge across sub-domains, thereby enabling a more comprehensive and automated root cause analysis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If multiple disconnected diagnostic sub-systems are used to analyze industrial processes, then each sub-domain can be analyzed independently, but the ability to discover new causes and relationships between variables is limited and subjective human intervention is required to combine results
Solution Approach 1:
The patent combines multiple disconnected diagnostic sub-systems into a single integrated diagnostic system that processes data from multiple sub-domains simultaneously. The system merges independent analysis capabilities while automatically integrating results through a unified processing framework, eliminating the need for subjective human intervention to combine results from separate sub-systems.
Solution Approach 2:
The diagnostic system is designed with universal processing capabilities that can handle multiple types of data from different sub-domains through a common analysis framework. The system performs multiple functions including data collection, processing, and integration across various industrial process domains within a single unified platform.
2Measurement precision
If a comprehensive diagnostic system integrates multiple sub-domains with probabilistic connections, then the accuracy and completeness of root cause analysis is improved, but the computational complexity and data processing requirements increase
Solution Approach 1:
The diagnostic system segments the industrial process into multiple sub-domains that can be analyzed independently and then integrated. Each sub-domain is processed separately through dedicated processing modules, and results are combined using probabilistic connections. This segmentation allows comprehensive analysis while managing computational complexity through modular processing.
Solution Approach 2:
The system introduces probabilistic connections as intermediaries between different sub-domains to facilitate integrated analysis. These probabilistic models act as mediators that combine information from multiple sub-domains in a computationally efficient manner, enabling accurate root cause analysis without requiring exhaustive computation across all possible variable interactions.
3Adaptability or versatility
If subjective human intervention is used to combine results from diagnostic sub-systems, then flexibility in analysis is maintained, but the efficiency and objectivity of the diagnostic process deteriorates
Solution Approach 1:
The diagnostic system performs automated self-service by independently collecting, processing, and integrating data from multiple sub-domains without requiring human intervention. The system automatically establishes probabilistic connections between sub-domains and generates diagnostic conclusions through self-contained processing algorithms, maintaining both objectivity and efficiency.
Solution Approach 2:
The system implements automated feedback loops that continuously process diagnostic data and adjust analysis based on probabilistic relationships between variables. This feedback mechanism enables the system to adapt to new information and refine diagnostic conclusions automatically, maintaining analytical flexibility while eliminating the time-consuming nature of manual result integration.
Data Source
AI summary
A diagnostic system implements a network including two or more sub-domains. Each sub-domain has diagnostic information extracted by analysis of object data, the object data representing one or more parameters measured in relation to a set of product units that have been subjected nominally to the same industrial process as one another. The network further has at least one probabilistic connection from a first variable in a first diagnostic sub-domain to a second variable in a second diagnostic sub-domain. Part of the second diagnostic information is thereby influenced probabilistically by knowledge within the first diagnostic information. Diagnostic information may include, for example, a spatial fingerprint observed in the object data, or inferred. The network may include connections within sub-domains. The network may form a directed acyclic graph, and used for Bayesian inference operations.