Industrial Machine Data Analysis Using Ranked Decision Trees
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Manufacturing sites face challenges in identifying signs of abnormalities in data collected before an event occurs, due to the need to compare vast amounts of normal and abnormal data, which is time-consuming and difficult.
Innovation Solution
An operation data analysis device that extracts feature amounts from industrial machine data and uses a decision tree to visualize differences, supporting users in analyzing data and identifying signs of state changes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If data about normal state and abnormal state are collected and compared to find signs before abnormality occurs, then the ability to detect abnormality signs is improved, but the time and effort required for analysis increases significantly due to the huge amount of data
Solution Approach 1:
The patent extracts only the necessary data portions related to abnormality signs from the huge amount of collected data. The extraction unit selectively extracts data portions from normal and abnormal states that are relevant to identifying abnormality signs, rather than analyzing all collected data, thereby reducing analysis time while maintaining detection accuracy.
Solution Approach 2:
The patent introduces a relevance determination unit that acts as an intermediary to determine the relevance between extracted data portions and abnormality signs. This intermediary component filters and prioritizes data based on its relevance to abnormality detection, enabling efficient analysis of large datasets without sacrificing detection precision.
2Measurement precision
If all collected data is analyzed to find signs before abnormality occurs, then the completeness of abnormality sign detection is improved, but the complexity of the analysis process increases
Solution Approach 1:
The patent segments the analysis process into distinct functional units: a data extraction unit that extracts relevant data portions, a relevance determination unit that assesses data relevance, and an abnormality sign identification unit that identifies signs based on relevant data. This segmentation reduces analysis process complexity by breaking down the comprehensive analysis task into manageable, specialized components.
3Reliability
If data from multiple time points before abnormality is compared with normal state data, then the accuracy of predicting abnormality is improved, but the amount of data to be processed increases
Solution Approach 1:
The extraction unit selectively extracts only the necessary data portions from multiple time points before abnormality occurs, rather than processing all available data. This extraction focuses on data that shows meaningful changes or patterns indicative of abnormality, reducing the volume of data to be processed while maintaining prediction accuracy.
Solution Approach 2:
The patent applies partial action by extracting and analyzing only the essential data portions needed for accurate abnormality prediction, rather than processing excessive amounts of data. This approach achieves reliable prediction results with a reduced dataset, improving efficiency without sacrificing accuracy.
Data Source
AI summary
An analysis device according to the present disclosure is provided with: a data acquisition unit that acquires data detected by an industrial machine; an operation state extraction unit that extracts, from the data, data detected during operation of the industrial machine; an annotation unit that creates a plurality of data set groups by giving, on the basis of a predetermined standard, annotations indicating motion states of the industrial machine to a plurality of data sets segmented by a predetermined standard from the extracted data during the operation; a feature amount extraction unit that extracts a feature amount of data included in each of the data sets; a learning unit that generates decision tree models respectively for the plurality of data set groups; and a display unit that ranks the generated decision tree models on the basis of correct answer rates regarding prediction of the annotations based on predetermined training data and that displays the decision tree models.