Substrate Processing Apparatus Fault Detection via Computational Inference
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In semiconductor manufacturing, accurately monitoring the temperature of gas pipes is challenging due to the displacement of temperature detecting units during maintenance, leading to inaccurate temperature measurements and potential faults in substrate processing.
Innovation Solution
A substrate processing apparatus with a monitored item table that collects and compares device data, including temperature and electrical power, to diagnose faults by setting reference values and executing a device status monitoring program, allowing for fault detection without direct temperature measurement.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If temperature detecting units are installed on gas pipes to monitor temperature, then temperature monitoring capability is improved, but the detecting units may be displaced during maintenance leading to measurement inaccuracies
Solution Approach 1:
The patent introduces an intermediary computational model that relates pipe temperature to heater power consumption and ambient temperature. Instead of directly relying on the potentially displaced temperature detecting unit, the system uses the temperature detection as one input among others (power consumption, ambient temperature) to computationally infer the actual pipe temperature, thereby mediating the unreliable direct measurement
Solution Approach 2:
The system implements feedback by continuously monitoring the relationship between heater power consumption, ambient temperature, and detected pipe temperature. When discrepancies arise (indicating possible detector displacement), the system adjusts by relying more on the computational model derived from power consumption and ambient temperature data, creating a self-correcting feedback loop
2Reliability
If multiple parameters are monitored to improve fault detection accuracy, then diagnostic capability is improved, but data management complexity increases
Solution Approach 1:
The patent creates a universal computational model that serves multiple functions: it can predict normal temperature behavior, detect anomalies, identify fault types, and guide maintenance decisions. This single multi-functional model handles various diagnostic tasks without requiring separate complex systems for each function
Solution Approach 2:
The system transforms multiple monitoring parameters (power consumption, ambient temperature, detected temperature) into a unified diagnostic framework by changing the parameter representation from raw values to relationships and deviations from expected behavior patterns, simplifying the management complexity while improving detection accuracy
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enables accurate detection of temperature-related faults, preventing degradation in film uniformity and reducing downtime by identifying issues through statistical data analysis and diagnostic rules, thus improving substrate processing quality and efficiency.
Implementation Method 1
a pipe heater (310) configured to heat the gas pipe (10)
Implementation Method 2
a temperature detecting unit provided at the pipe heater and configured to detect a temperature of the gas pipe
Data Source
AI summary
There is provided a technique for detecting a fault of a device from an error in device data. According to the technique described herein, there is provided a substrate processing apparatus including: a pipe heater configured to heat a gas pipe; a temperature detecting unit provided at the pipe heater and configured to detect a temperature of the gas pipe; a control unit configured to control the pipe heater based on device data representing the temperature of the gas pipe measured by the temperature detecting unit by executing a process control program to adjust an electrical power applied to the pipe heater; a memory unit configured to store a monitored item table; and a device status monitoring unit configured to execute a device status monitoring program.