Explainable AI Quality Prediction for Real-Time Defect Prevention
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing quality inspection methods fail to ensure real-time monitoring of quality during production and analyze defects effectively, leading to increased costs and reduced quality due to delayed detection of product defects.
Innovation Solution
A quality cause analysis system that combines explainable AI with the Taguchi method, using an input device to collect machine and product data, a processing device to establish a quality prediction model, and an output device to provide real-time predictions and optimal machine data for immediate adjustments, thereby reducing waste and optimizing production processes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional sampling inspection is used after batch manufacturing, then labor cost and time for quality control are reduced, but real-time quality monitoring is lost and defective products increase
Solution Approach 1:
The system performs preliminary quality prediction during the manufacturing process by collecting machine status data in real-time and using the quality prediction model to forecast product quality before completion. This allows early detection of potential defects and proactive adjustments to machine parameters, preventing defective products rather than detecting them after batch manufacturing.
Solution Approach 2:
The system establishes a closed-loop feedback mechanism where real-time machine status data is continuously fed into the quality prediction model, and the predicted quality results are fed back to guide adjustments in machine parameters. This feedback loop enables dynamic optimization of manufacturing processes to maintain high quality while preserving productivity.
2Reliability
If real-time quality monitoring is implemented during production, then quality control is improved, but device complexity and cost increase
Solution Approach 1:
The system uses a multi-functional quality prediction model that can handle multiple quality attributes and machine parameters simultaneously. The same neural network framework predicts various product quality characteristics based on different machine status inputs, reducing the need for separate monitoring systems for each quality parameter and thereby limiting the increase in device complexity.
Solution Approach 2:
The system introduces an intermediary processing layer that collects and integrates data from multiple machine sources, then feeds consolidated information to the quality prediction model. This intermediary layer simplifies the complexity by standardizing data formats and filtering redundant information before quality analysis, making the real-time monitoring system more manageable.
3Manufacturing precision
If explainable AI and Taguchi method are combined for optimal machine evaluation, then manufacturing precision is improved, but device complexity increases
Solution Approach 1:
The system segments the quality prediction process into distinct functional modules: data collection from machine status, feature extraction from raw data, quality prediction using neural networks, and explanation generation using SHAP values. This segmentation allows each module to be optimized independently and facilitates easier maintenance and updates, limiting the impact of algorithm complexity on overall system management.
Solution Approach 2:
The system dynamically adjusts model parameters and hyperparameters based on the specific manufacturing context and data characteristics. The Taguchi method is used to optimize neural network parameters such as learning rate, batch size, and network architecture, achieving high manufacturing precision while keeping the parameter tuning process systematic and manageable rather than exponentially complex.
Data Source
AI summary
A quality cause analysis system is disclosed, including an input device, a storage device, a processing device and an output device. The storage device stores machine status data, product measurement data, and a plurality of algorithms. The processing device accesses the storage device to establish a quality cause analysis model. The quality cause analysis model includes a quality prediction module, a network explanation module and an optimal machine prediction module. The quality prediction module predicts whether the product measurement data meets the quality inspection regulations. The network explanation module uses an explainable AI algorithm to measure the Shapley value of the state variables. The optimal machine prediction module calculates the capability of accuracy value and selects a plurality of machine status data closest to the standard center value as the optimal machine data.