IoT Measurement Data Aggregation for Cross-Base Quality Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current measurement systems face challenges in data aggregation, immediacy of data analysis, consolidation of data across manufacturing bases, and automation of data analysis, leading to increased costs and delayed responses to quality issues.
Innovation Solution
A measurement solution service system that integrates IoT technology with AI, using an IoT relay device to collect and standardize measurement data, and a computing system for aggregation, analysis, and display processing, employing machine learning and deep learning for predictive analytics and optimal condition computation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If measurement data is collected and aggregated manually from multiple measurement sources, then data aggregation can be performed, but operating costs increase due to human labor requirements
Solution Approach 1:
The system enables automated data collection and aggregation through the measurement data collecting unit that automatically gathers data from multiple measurement sources without human intervention. The aggregation is performed automatically by the aggregate processing unit in the computing system, eliminating the need for manual human labor in data collection and aggregation processes.
2Loss of time
If data aggregation is performed without automated processing, then system complexity is reduced, but response time to quality issues is delayed
Solution Approach 1:
The system performs preliminary actions by continuously and automatically collecting measurement data from multiple sources and pre-aggregating it in real-time before quality issues manifest. The analyzing unit is ready to immediately analyze the pre-collected and pre-aggregated data, enabling rapid response to quality deviations without waiting for manual data collection.
3Extent of automation
If data analysis is automated at each manufacturing base individually, then analysis can be performed locally, but costs increase due to requiring separate servers and dedicated software at each base
Solution Approach 1:
The system merges the data analysis functionality into a centralized computing system that receives measurement data from multiple manufacturing bases through the measurement data collecting unit. The aggregate processing unit consolidates data from various bases and performs unified analysis, eliminating the need for separate servers and dedicated software at each individual base while maintaining automated analysis capabilities.
4Productivity
If measurement data is not standardized before aggregation, then data collection is simpler, but data consolidation across manufacturing bases becomes difficult and requires many man-hours
Solution Approach 1:
The measurement data collecting unit performs preliminary standardization of measurement data before it reaches the aggregation stage. By standardizing data formats, structures, and protocols at the collection point, the system prepares data for efficient consolidation across multiple manufacturing bases, eliminating the need for time-consuming manual standardization efforts during later consolidation phases.
Data Source
AI summary
A computing system is configured to analyze both measurement data and indicator data as big data aggregated in measurement database and indicator database by deep learning for each lot of a part or for each lot of a finished product and a part pre-associated with each other, and also for each consolidation target between bases subordinate to the same start point corresponding to identification information that specifies a business user of the computing system. Analysis target layers by the deep learning are a three-layer serial hierarchical structure containing a production condition layer and an environment condition layer as a start point for analysis of a part layer, or a four-layer serial hierarchical structure containing a part layer, a production condition layer, and an environment condition layer as a start point for analysis of a finished product layer.