AI Failure Prediction Model Updating With Test Feedback
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional predictive models for electronic devices are not updated after production, leading to a decrease in accuracy over time due to changing external factors and inconsistent testing results, resulting in inefficient prediction of product failures and normal operations.
Innovation Solution
An AI predictive model is generated and updated based on component input data, fail result data, and pass result data, allowing for real-time adjustments and improved prediction accuracy by transmitting data to a server for machine learning and model updates.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a conventional predictive model is used without updates, then the initial prediction accuracy is achieved, but the prediction accuracy decreases over time due to changing external factors
Solution Approach 1:
The patent implements a feedback mechanism where actual test results from complete products are fed back to update the predictive model. The system collects test data, compares predicted outcomes with actual results, and uses this feedback to continuously refine and update the model parameters, ensuring prediction accuracy is maintained over time despite changing external factors.
Solution Approach 2:
The patent transforms the static predictive model into a dynamic system that adapts to changing conditions. The model is designed to be continuously updated with new data, allowing it to evolve and adjust to changing external factors, manufacturing variations, and product designs while maintaining its predictive capability.
2Measurement precision
If the predictive model is updated continuously with new data, then prediction accuracy is maintained, but the complexity of the system increases
Solution Approach 1:
The patent implements a self-service mechanism where the system automatically collects test data, processes results, and updates the predictive model without requiring extensive manual intervention. The system autonomously manages the data collection, analysis, and model refinement processes, reducing operational complexity while maintaining continuous accuracy.
Solution Approach 2:
The patent creates a multi-functional system that handles multiple tasks within a unified framework: data collection from various sources, test result analysis, model parameter adjustment, and validation. This universal approach consolidates multiple functions into a single integrated system, managing complexity through functional consolidation.
3Ease of manufacture
If only component test data is used for prediction, then the testing process is simple, but the prediction result does not reflect actual complete product performance
Solution Approach 1:
The patent merges component-level test data with complete product test results into a unified predictive model. By combining data from both sources, the system captures both the detailed component behavior and the system-level interactions, providing a more accurate representation of actual product performance while maintaining the simplicity of component-based testing.
Data Source
AI summary
Various embodiments of the present disclosure disclose a method and apparatus, and comprise: a communication module comprising communication circuitry; a memory; and at least one processor comprising processing circuitry operatively connected to the communication module and/or the memory, wherein at least one processor is configured to: generate an AI predictive model based on component input data; acquire fail result data according to the AI predictive model;acquire pass result data according to the fail result data; and update the AI predictive model based on at least one of the component input data, the fail result data, and the pass result data.