Machine Tool State Determination Using Tuned Feature Distributions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing machine tool operating state determination systems struggle to accurately assess the state of facilities under varying conditions such as different installation locations, processing types, and work materials, as pre-defined determination criteria may not account for these variations, leading to inaccurate state identification.
Innovation Solution
A facility management system that includes a sensor edge device and a server device, which utilize sensor data, feature quantity tuning, and processing algorithms to dynamically adjust and refine the feature quantity distribution for accurate state determination, enabling real-time identification of operating states, abnormality detection, and failure sign detection across diverse conditions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If pre-defined determination criteria are used for each machine tool model, then the system is simple to operate, but determination accuracy deteriorates under varying conditions
Solution Approach 1:
The patent implements dynamic determination criteria that automatically adapt to different operating conditions (installation location, processing type, work material) rather than using static pre-defined criteria. The system learns and adjusts criteria based on actual sensor data from each specific facility configuration, resolving the contradiction between operational simplicity and determination accuracy.
Solution Approach 2:
The system changes the parameters of determination criteria based on detected operating conditions. When conditions such as installation location or work material change, the system modifies the determination criteria parameters accordingly, maintaining high accuracy across diverse conditions while keeping the interface simple for users.
2Measurement precision
If determination criteria are prepared for each combination of conditions (model, installation location, processing type, work material), then determination accuracy improves, but device complexity increases
Solution Approach 1:
The patent creates a universal determination criteria system that can handle multiple conditions and facility types through a single integrated approach. Rather than maintaining separate criteria for each combination, the system uses a unified framework that adapts to different scenarios, reducing complexity while maintaining comprehensive coverage.
Solution Approach 2:
The system performs self-service by automatically generating and adjusting determination criteria based on sensor data from the actual facility configuration. Rather than requiring manual preparation of criteria for each condition combination, the system autonomously learns and configures appropriate criteria, dramatically reducing device complexity.
3Measurement precision
If manual preparation of determination criteria for each condition combination is performed, then determination accuracy improves, but loss of time increases
Solution Approach 1:
The system performs preliminary action by automatically preparing determination criteria in advance based on facility configuration data, eliminating the need for manual preparation. The criteria are generated beforehand through automated learning processes, ensuring both high accuracy and time efficiency.
Solution Approach 2:
The system serves itself by automatically generating determination criteria without human intervention. The automated system learns from sensor data and configures appropriate criteria based on the specific facility conditions, eliminating time-consuming manual preparation while maintaining high determination accuracy.
Data Source
Figure 1
Figure 2(A)~2(B)
Figure 3
AI summary
To accurately determine a state of a facility under various conditions. A facility state determination device includes a catalog storage unit that stores a catalog having versatility; a catalog control unit that specifies the catalog corresponding to a target machine tool to be subjected to state determination based on machine tool information including at least an item of facility type and acquires the catalog from the catalog storage unit; a feature quantity extraction unit that extracts a feature quantity from sensor data detected from the target machine tool; a sensor data processing unit that performs state determination of the target machine tool based on the feature quantity distribution included in the acquired catalog and the feature quantity extracted from the sensor data; a feature quantity tuning unit that performs tuning of the feature quantity distribution by mapping the extracted feature quantity to the feature quantity distribution; and a catalog updating unit that updates the catalog of the catalog storage unit based on the feature quantity distribution after tuning.