Sensor Fault Pattern Detection for Zero-Defect Manufacturing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods struggle to precisely detect and attribute errors in manufacturing processes, leading to difficulties in improving production quality and achieving error-free production.
Innovation Solution
A method and system for generating a parameter table based on historical sensor data to identify error patterns, using historical curves and assigning fault patterns, and performing binary logistic regression to determine error causes, enabling real-time detection and correction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional error detection methods are used in manufacturing processes, then the detection process is simple, but the measurement precision and ability to attribute errors to specific causes is insufficient
Solution Approach 1:
The patent segments the manufacturing process into multiple subprocesses and divides error detection into hierarchical levels. Each subprocess generates its own curves and error patterns, which are then integrated into a comprehensive parameter table. This segmentation allows precise error attribution to specific subprocesses while managing system complexity through modular organization of detection mechanisms.
Solution Approach 2:
The patent transitions from traditional single-dimensional error detection to multi-dimensional analysis by generating curves with at least two dimensions for each subprocess. The parameter table stores multiple parameters (at least three) for each error pattern, enabling errors to be characterized and attributed across multiple dimensions simultaneously, thereby improving measurement precision.
2Reliability
If comprehensive sensor data collection is implemented across all subprocesses, then error pattern detection capability is improved, but the loss of time and computational resources increases
Solution Approach 1:
The patent performs preliminary action by pre-generating and storing characteristic error patterns in the parameter table before actual error detection occurs. Historical sensor data is processed in advance to create reference curves and error patterns, which are then used for rapid comparison during production. This eliminates the need for complex real-time analysis, significantly reducing data processing time while maintaining high reliability.
Solution Approach 2:
The system implements feedback by continuously comparing actual sensor data against the stored parameter table and using the results to refine error detection. The parameter table is updated with new error patterns discovered during production, creating a self-improving system that becomes more reliable over time without proportionally increasing processing time.
3Loss of information
If multiple parameters are monitored for each subprocess, then the ability to identify error causes is improved, but the difficulty of detecting and measuring errors increases
Solution Approach 1:
The patent merges data from multiple subprocesses and multiple parameters into a unified parameter table structure. Each error pattern in the table integrates information from various sources (different subprocesses, different parameters) into a single coherent error description with attributed causes. This merging approach maintains complete error information while simplifying detection by providing a unified view rather than requiring separate analysis of each parameter.
Solution Approach 2:
The parameter table serves multiple functions simultaneously: it stores characteristic error patterns, provides reference curves for comparison, attributes error causes, and enables cross-subprocess error analysis. This multi-functionality reduces the need for separate detection mechanisms for each function, thereby reducing the overall difficulty of error detection and measurement while maintaining information completeness.
Data Source
Figure 1
Figure 1
Figure 2~3
AI summary
Provision is made of a method for monitoring at least one process and for determining fault patterns of faults occurring in the at least one process, wherein a parameter table with characteristic fault patterns is generated for a number of partial processes of the at least one process, wherein the parameter table is generated on the basis of historical sensor data, wherein the historical sensor data describe a number of historical curves which have at least two dimensions and are respectively assigned to a partial process, and wherein the historical curves for each partial process comprise historical OK curves (okay) and historical NOK curves (not okay), wherein the historical NOK curves represent faulty partial processes.