Semiconductor Equipment Troubleshooting With Sensor Health Graphs
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing semiconductor manufacturing equipment lacks a comprehensive system for monitoring the overall health of sub-systems, leading to undetected deteriorating conditions and increased downtime due to reliance on individual sensor monitoring.
Innovation Solution
A system that generates manufacturing graphs from sensor data, applies machine-learning models to diagnose sub-system failures, and provides corrective actions, enabling adaptive troubleshooting and predictive maintenance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If individual sensors are monitored to detect problems during deposition, then specific sensor issues can be identified, but the overall health of sub-systems cannot be assessed and deteriorating conditions go undetected
Solution Approach 1:
The patent combines multiple sensor readings from the same sub-system into a single aggregate metric (e.g., average temperature, overall pressure trend). This merging approach preserves the detection capability of individual sensors while simultaneously providing an overview of the entire sub-system's health status, thus resolving the contradiction between specific sensor monitoring and overall system assessment.
Solution Approach 2:
The system creates multi-functional monitoring that serves dual purposes: individual sensor-level diagnostics and sub-system-level health assessment. Each sensor continues to provide its specific measurement function while collectively contributing to aggregate metrics that indicate overall sub-system health, enabling both specific problem detection and general system monitoring simultaneously.
2Reliability
If comprehensive monitoring of all sensors is implemented to detect deteriorating conditions, then early detection of faults is achieved, but system complexity and data processing requirements increase
Solution Approach 1:
The monitoring system is segmented into hierarchical levels: individual sensor monitoring, sub-system aggregate monitoring, and overall system health monitoring. This segmentation allows comprehensive fault detection at each level without requiring the complexity of analyzing all sensor data simultaneously, as each level processes only the data relevant to its scope.
Solution Approach 2:
Aggregate metrics serve as intermediaries between individual sensor readings and overall system health assessment. Instead of directly analyzing complex multi-sensor data, the system uses simplified aggregate values (average temperature, overall pressure trend) as intermediate representations that capture sub-system health while reducing data complexity.
3Ease of operation
If manual monitoring and troubleshooting procedures are used, then detailed analysis of specific issues is possible, but downtime and repair time increase significantly
Solution Approach 1:
The system provides self-service troubleshooting by automatically generating diagnostic information and guiding operators through remediation steps. When a fault is detected, the system autonomously identifies the affected sub-system, retrieves relevant sensor data, and presents targeted troubleshooting guidance, eliminating the need for manual data collection and analysis while maintaining detailed diagnostic capability.
Solution Approach 2:
The system performs preliminary diagnostic actions automatically before operator intervention is needed. By pre-processing sensor data, identifying trends, and preparing diagnostic information in advance, the system reduces the time required for operator analysis and decision-making, enabling faster response to faults while maintaining thorough diagnostic capability.
Data Source
AI summary
A system includes a processing device, operatively coupled to the memory device, to perform operations comprising obtaining a plurality of sensor values associated with a deposition process performed, according to a recipe, in a process chamber to deposit film on a surface of a substrate; generating a manufacturing data graph based on the plurality of sensor values; receiving, via a user interface, a selection of a data point on the manufacturing graph; receiving failure data associated with the data point; and storing, in a data structure, the failure data to be accessible via the user interface presenting the manufacturing data graph.