Manufacturing Data Classification Using Proxels Across Facilities
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional data processing devices for manufacturing processes are time-consuming and costly, lacking versatility as they require rebuilding models for each manufacturing facility, even for similar processes, due to the need for improved simulation accuracy.
Innovation Solution
A data processing system that collects and analyzes data from multiple business offices, calculates 'Proxels' (minimum data units) by grouping data with the same effect, allowing for highly versatile data processing across different manufacturing facilities and processes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If models are built for each manufacturing facility individually to improve simulation accuracy, then manufacturing precision is improved, but device complexity and loss of time increase
Solution Approach 1:
The patent segments the manufacturing process into discrete steps and identifies key process parameters (KPPs) at each step. By dividing the complex manufacturing process into manageable segments and focusing only on critical parameters, the system reduces model complexity while maintaining simulation accuracy for each facility.
Solution Approach 2:
The patent creates a universal data structure and analysis framework that can be applied across different manufacturing facilities and process types. The standardized approach to collecting, organizing, and analyzing process data enables the same methodology to serve multiple facilities, reducing the need to rebuild models from scratch for each new facility.
2Manufacturing precision
If models are rebuilt for each manufacturing facility to improve simulation accuracy, then manufacturing precision is improved, but loss of time increases
Solution Approach 1:
The patent establishes a preliminary standardized framework for data collection, organization, and analysis that can be applied to any manufacturing facility. By preparing the analytical structure and methodology in advance, the system eliminates the time-consuming process of building models from scratch for each new facility while maintaining the ability to achieve accurate simulation results.
Solution Approach 2:
The patent focuses analysis on key process parameters (KPPs) rather than all possible process variables. By identifying and concentrating on the most critical parameters that significantly impact manufacturing outcomes, the system reduces data processing time while maintaining simulation accuracy through targeted analysis of only the most influential parameters.
3Manufacturing precision
If conventional data processing methods are used to improve simulation accuracy, then manufacturing precision is improved, but versatility deteriorates
Solution Approach 1:
The patent develops a universal data analysis framework with standardized data structures, collection methods, and analytical approaches that can be applied across different manufacturing facilities, process types, and data sources. This versatile framework maintains simulation accuracy while enabling the system to handle diverse manufacturing data from various facilities without requiring facility-specific customizations.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The present invention realizes highly versatile data processing for data collected during a manufacturing process. This data processing device has: a calculation means for collecting data groups associated with a prescribed step in a process, and calculating effects in the prescribed step for the data groups; a dividing means for dividing a feature space such that the distribution, in the feature space, of the data groups associated with the prescribed step results in classification by the calculated effects; and an output means for outputting identification data for identifying regions obtained by dividing the feature space.