Production Process Error Discovery Using Z-Score Cluster Changes

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing methods for discovering errors in metal production processes, such as those described in 'Exploring due date reliability in production systems using data mining methods adapted from gene expression analysis,' are limited in identifying underlying mechanisms and deriving patterns for process improvement, leading to insufficient reduction in production time and costs.

Innovation Solution

The method involves generating modified clusters by removing or adding process parameter values, applying performance enrichment analysis to determine changes in Z-score values, which indicate the influence of parameters on performance indicators and identify potential errors, thereby enabling mechanistic cause analysis and optimization strategies.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If enrichment analysis is applied to identify clusters with high influence on performance indicators, then error discovery capability is improved, but understanding of underlying mechanisms and derivation of patterns is insufficient

Engineering Contradiction:
Improveerror discovery capabilityVSAvoidunderstanding of underlying mechanisms
Core Design Contradiction:
Measurement precisionVSLoss of information

Solution Approach 1:

The method segments the error discovery process into multiple analytical stages: initial cluster identification through enrichment analysis, followed by systematic modification of clusters by adding/removing parameters, and subsequent re-evaluation. This segmentation allows each stage to address specific aspects of error discovery while preserving mechanistic understanding through the iterative modification process.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The method performs preliminary actions by systematically modifying clusters before final error identification. By pre-modifying clusters through addition and removal of parameters, and evaluating Z-score changes in advance, the method prepares comprehensive data structures that enable both precise error discovery and retention of mechanistic information for pattern derivation.

Inventive Principle:
Principle #10Preliminary action

2Reliability

If traditional enrichment analysis is used to identify influential clusters, then clusters with high influence on performance indicators can be identified, but the method does not enable derivation of patterns for process improvement

Engineering Contradiction:
Improveidentification of influential clustersVSAvoidprocess improvement capability
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The method implements feedback by continuously evaluating Z-score changes when clusters are modified. The change in Z-score values provides feedback information about the influence of individual parameters on performance indicators, enabling not only identification of influential clusters but also derivation of actionable patterns for process improvement through systematic analysis of parameter contributions.

Inventive Principle:
Principle #23Feedback

Solution Approach 2:

The method systematically changes parameters by adding and removing process parameters from clusters, then measures the impact on Z-score values. This parameter change approach enables identification of which specific parameters drive cluster influence on performance indicators, thereby enabling pattern derivation for process improvement while maintaining reliable cluster identification.

Inventive Principle:
Principle #35Parameter changes

3Quantity of substance

If cluster analysis is performed without systematic modification, then initial Z-score values can be determined, but the influence of individual parameters on performance indicators cannot be precisely measured

Engineering Contradiction:
Improvenumber of process parameters in clusterVSAvoidinfluence measurement of parameters
Core Design Contradiction:
Quantity of substanceVSMeasurement precision

Solution Approach 1:

The method extracts information about individual parameter influence by systematically removing parameters from clusters and measuring the resulting Z-score changes. By taking out individual parameters one at a time and evaluating their specific contribution to cluster performance, the method precisely measures the influence of each parameter while maintaining the overall cluster structure for context.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The method applies partial action by modifying clusters incrementally through addition or removal of single parameters rather than analyzing all parameters simultaneously. This partial modification approach enables precise measurement of individual parameter influence on performance indicators while maintaining computational feasibility and interpretability of results.

Inventive Principle:
Principle #16Partial or excessive action

Data Source

PatentUS12182750B2Method and computer programme for discovering possible errors in a production process
Publication Date: 2024.12.31 SMS GROUP GMBH

AI summary

A method and a computer program for discovering possible errors in a production process for manufacturing metal products. The method involves removing at least one process parameter value from the cluster or adding at least one process parameter to the cluster. Second Z-score values, which are compared with the first Z-score values, are then determined for the thus-altered cluster. The changes in the Z-score values provide suggestions for troubleshooting and process optimization which are direct and can be implemented immediately.