Machine Tool Diagnostics Using Part-Sensor Correlation Learning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies for collecting and utilizing information from machine tools are inefficient, as the vast amount of data collected is not effectively utilized, and existing systems lack methods for comprehensive data collection and utilization.
Innovation Solution
An information processing system that collects data from machine tools using sensors and machine learning to correlate part information with sensed data, predicting the lifetime of parts and determining maintenance needs, thereby optimizing maintenance operations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If data is collected from multiple machine tools using various sensors, then the quantity and variety of collected information increases, but the effectiveness of utilization decreases
Solution Approach 1:
The patent segments the enormous collected data into meaningful groups by learning normal ranges for different machine tools and their components. Instead of treating all data uniformly, it divides data into normal and abnormal patterns, making the information manageable and actionable for maintenance decisions.
Solution Approach 2:
The patent introduces an information processing device as an intermediary between the machine tools and maintenance personnel. This device learns normal ranges from collected data and automatically compares actual data against these ranges, transforming raw data into diagnostic insights without requiring manual analysis of all collected information.
2Reliability
If comprehensive data from multiple machine tools is collected, then diagnostic capability improves, but system complexity increases
Solution Approach 1:
The patent performs preliminary learning of normal ranges before actual diagnostic operations. The information processing device first learns what constitutes normal operation for each machine tool by analyzing historical data, storing these normal ranges for future comparison. This preliminary action simplifies subsequent diagnostics by having reference standards ready.
Solution Approach 2:
The system enables self-diagnosis by allowing machine tools to automatically compare their own operational data against learned normal ranges. The information processing device autonomously determines whether data falls within normal ranges and generates appropriate diagnostic information, reducing the need for complex external analysis systems.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Provided is an information processing method capable of effectively utilizing a variety of types of information collected from a machine tool. The information processing method includes the step of communicating with a plurality of machine tools (S10). The plurality of machine tools each include a sensor to sense information about the machine tool as sensed information. The step of communicating (S10) includes the step of receiving as collected data from each of the plurality of machine tools part information about a part of the machine tool and the sensed information obtained by the sensor. The information processing method further includes the steps of: storing in a storage unit the collected data received from each of the plurality of machine tools (S20A, S20B); and based on the plurality of collected data stored in the storage unit, learning a correlation between part information of a machine tool and sensed information obtained by the sensor internal to that machine tool (S32).