Process Chamber Hazard Prediction With AR Temperature Overlay
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for assessing hazards in manufacturing equipment, such as semiconductor manufacturing equipment, are cumbersome, expensive, and difficult to implement, particularly for surfaces that are not easily measurable, posing risks to users during service operations.
Innovation Solution
Utilizing machine learning models trained with sensor data and physics-based models to predict hazardous conditions, and presenting these predictions through augmented reality overlays to users, enabling safer service operations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional methods are used to assess hazards on manufacturing equipment surfaces, then measurement accuracy may be adequate for accessible surfaces, but the method becomes cumbersome, expensive, and difficult to implement for inaccessible surfaces
Solution Approach 1:
The patent introduces machine learning models as intermediary systems that indirectly assess hazards on inaccessible surfaces by analyzing data from accessible sensors and components. Instead of directly measuring inaccessible surfaces, the system uses trained models to predict surface temperatures and hazard conditions based on correlated measurable parameters, thereby resolving the contradiction between measurement accuracy and assessment complexity.
Solution Approach 2:
The system creates virtual representations (digital twins) of equipment surfaces and hazard conditions through machine learning models. These digital copies allow hazard assessment of inaccessible surfaces without physical measurement, replicating the hazard information in a accessible virtual form that can be viewed through augmented reality interfaces.
2Loss of information
If physical measurement methods are used for hazardous surfaces, then direct hazard data can be obtained, but user safety is compromised during measurement operations
Solution Approach 1:
The system performs preliminary hazard assessment using machine learning models before users approach or contact equipment surfaces. By predicting hazard conditions in advance and displaying them through augmented reality, the system provides complete hazard information without requiring users to physically measure or expose themselves to hazardous surfaces during the assessment process.
Solution Approach 2:
The machine learning model acts as an intermediary that obtains hazard data remotely through sensor inputs without requiring direct user contact with hazardous surfaces. The model processes sensor data and delivers hazard information to users through safe interfaces, eliminating the need for users to expose themselves to harmful conditions while maintaining complete hazard data collection.
3Object-affected harmful factors
If comprehensive hazard assessment is performed on all equipment surfaces, then user safety is improved, but the cost and time required for implementation increases
Solution Approach 1:
The machine learning models continuously self-assess hazard conditions by automatically processing sensor data from the equipment. This automated self-monitoring provides comprehensive safety coverage without requiring manual intervention or time-consuming physical inspections, as the system performs hazard assessment continuously and autonomously.
Solution Approach 2:
The system maintains continuous hazard assessment through ongoing processing of sensor data by machine learning models, rather than performing periodic manual inspections. This continuous automated monitoring ensures constant user safety protection while eliminating the time loss associated with repeated manual assessment operations.
Data Source
AI summary
A method includes obtaining first data indicative of a temperature of a first component of a process chamber. The method further includes processing the first data using a trained machine learning model. The trained machine learning model generates an output. The output includes second data, indicative of a temperature of a surface of the process chamber. The method further includes displaying an augmented reality overlay including a visual indication of the temperature of the surface to a user.