Production Line Causal Modeling for Abnormality Root Cause Estimation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Identifying the causal factors of abnormal events in production lines is complex due to the increasing number of mechanisms and changing operating conditions, making it difficult for both skilled and unskilled workers to accurately determine the cause-and-effect relationships, which existing technologies have not adequately addressed.

Innovation Solution

An information processing device and method that estimates abnormality causal factors by analyzing cause-and-effect models, using a display to visualize the results, allowing users to select models based on their input, and utilizing an event-related matter registration unit and event estimator to calculate contribution ratios and similarity degrees for simplified estimation.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If skilled workers use professional expertise to analyze cause-and-effect relationships, then accuracy of abnormality detection is improved, but complexity of operation increases and unskilled workers cannot perform the task

Engineering Contradiction:
Improveaccuracy of abnormality detectionVSAvoidease of maintenance operation
Core Design Contradiction:
Measurement precisionVSEase of operation

Solution Approach 1:

The patent introduces an information processing device as an intermediary between the complex production line systems and the maintenance workers. This device automatically analyzes cause-and-effect relationships among mechanisms using collected operation data, generating visualized cause-and-effect models that guide workers. The intermediary handles the complex analysis work, allowing unskilled workers to perform maintenance operations with the same effectiveness as skilled workers would achieve through expertise.

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent replaces the mechanical system of human expertise and experience-based analysis with an automated information processing system. Instead of relying on workers' professional knowledge to interpret complex mechanism relationships, the system automatically collects data from multiple mechanisms, analyzes correlations, and presents simplified cause-and-effect models. This substitution transforms the dependency from human skill to automated computational analysis.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

2Productivity

If the number of mechanisms in production line increases, then productivity is improved, but difficulty of detecting and measuring cause-and-effect relationships increases

Engineering Contradiction:
Improveproductivity of production lineVSAvoiddifficulty of detecting cause-and-effect relationships
Core Design Contradiction:
ProductivityVSDifficulty of detecting and measuring

Solution Approach 1:

The patent segments the complex analysis task into distinct functional modules: data collection from individual mechanisms, correlation analysis between mechanisms, cause-and-effect model generation, and visualization. By dividing the overall problem of analyzing multiple mechanisms into these manageable segments, the system can effectively handle increasing numbers of mechanisms without overwhelming complexity in any single analysis step.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent changes the parameters of data representation and analysis by collecting standardized operation data from multiple mechanisms and transforming it into cause-and-effect models with quantified relationship strengths. The system uses contribution rates and probability values as parameters to represent causal relationships, making complex multi-mechanism interactions measurable and analyzable even as the number of mechanisms increases.

Inventive Principle:
Principle #35Parameter changes

3Measurement precision

If more abnormality data is collected to identify causal factors, then accuracy of abnormality identification is improved, but device complexity increases

Engineering Contradiction:
Improveaccuracy of abnormality identificationVSAvoidcomplexity of information processing device
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent extracts only the essential data elements needed for cause-and-effect analysis from the broader set of available abnormality data. Instead of processing all possible operational parameters, the system focuses on collecting and analyzing specific operation data that directly relates to mechanism interactions and abnormal events. This selective extraction maintains accuracy while reducing the complexity of data processing requirements.

Inventive Principle:
Principle #2Taking out (Extraction)

Data Source

PatentEP4202584B1Information processing device, information processing method and information processing program
Publication Date: 2024.11.13 OMRON CORP
  • EP4202584B1 patent drawingFigure 1
  • EP4202584B1 patent drawingFigure 2
  • EP4202584B1 patent drawingFigure 3

AI summary

An information processing device (1) includes a cause-and-effect model generator (114 to 116), an abnormality detector (117), a contribution ratio calculator (118), and an estimator (119). The cause-and-effect model generator (114 to 116) generates a plurality of cause-and-effect models between variables based on a relationship between a plurality of mechanisms in a production process of a production line, a dependency between a plurality of events associated with the plurality of mechanisms, or a control-based relationship between the plurality of mechanisms. The abnormality detector (117) obtains a result of abnormality detection in the production line. The contribution ratio calculator (118) calculates a contribution ratio of a variable contributing to abnormality in the result of abnormality detection. The estimator (119) estimates an abnormality causal factor for at least one of the plurality of cause-and-effect models based on the calculated contribution ratio of the variable contributing to abnormality.