Machine Learning Washing Prediction for Machine Tools
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for washing machine tools with a fixed frequency are inefficient, as they require manual decision-making based on chip accumulation, leading to unnecessary downtime and reduced productivity due to frequent washing and visual sensor checks.
Innovation Solution
A machine learning apparatus and washing prediction system that acquires and analyzes machining and washing conditions, along with state information, to generate a trained model that predicts optimal washing timing and frequency, eliminating the need for visual sensor confirmation and minimizing chip accumulation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If washing is performed with a fixed frequency based on empirical decisions, then chip accumulation is controlled, but productivity is reduced due to frequent washing and visual sensor checks
Solution Approach 1:
The patent replaces the mechanical/visual inspection system with an acoustic detection system. Acoustic sensors detect chip accumulation by listening to the sounds generated within the machine tool, eliminating the need for visual sensor checks and empirical washing schedules. This substitution enables continuous monitoring without stopping machining operations.
Solution Approach 2:
The system enables the machine tool to self-monitor its own chip accumulation status through acoustic sensors integrated within the machine tool structure. The self-diagnosis function allows the machine to automatically determine when washing is needed based on acoustic signals, eliminating the need for external visual inspection and empirical decision-making by operators.
2Measurement precision
If visual sensor checks are performed frequently to monitor chip accumulation, then chip accumulation is detected, but time is lost due to stopping machining operations
Solution Approach 1:
The acoustic monitoring system operates continuously during machining operations without requiring the machine tool to stop. The acoustic sensors continuously detect chip accumulation sounds, enabling real-time monitoring while machining continues uninterrupted, thus eliminating the downtime associated with periodic visual sensor checks.
Solution Approach 2:
The patent replaces the mechanical/visual inspection system with an acoustic detection system. Acoustic sensors detect chip accumulation by listening to the sounds generated within the machine tool, eliminating the need for visual sensor checks and empirical washing schedules. This substitution enables continuous monitoring without stopping machining operations.
3Productivity
If washing frequency is reduced to increase productivity, then productivity improves, but chip accumulation control becomes insufficient
Solution Approach 1:
The acoustic monitoring system provides continuous feedback on chip accumulation status through acoustic sensors that detect sounds generated by chips. This feedback mechanism allows the system to determine the optimal washing timing based on actual chip accumulation conditions rather than fixed schedules, enabling extended periods between washes while maintaining effective chip control.
Solution Approach 2:
The system performs preliminary detection of chip accumulation conditions using acoustic sensors before critical accumulation occurs. By detecting chip sounds early, the system can plan washing operations in advance based on actual conditions, allowing for optimized washing schedules that maximize productivity while maintaining effective chip control.
Data Source
AI summary
A machine learning device has an input data acquisition unit that acquires input data including an arbitrary machining condition for an arbitrary workpiece to be machined by an arbitrary machine tool, a cleaning condition for cleaning, by an arbitrary cleaning device, the inside of a machine by which the workpiece was machined under the machining condition and state information indicating a degree of dirtiness of the inside of the machine before the machining under the machining condition. The device also has a label acquisition unit that acquires, as label data, state information about the inside of the machine after machining under the machining condition and cleaning under the cleaning condition included in the input data, and a learning unit that executes supervised learning using the input data and the label data, so as to generate a learned model.