Milling Machine Anomaly Monitoring With In-Operation Model Training
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional milling machines often fail to detect errors or failures until they occur, leading to significant delays in manufacturing and potential rejection of workpieces, as well as requiring immediate maintenance that may not be readily available, causing temporary production line standstills.
Innovation Solution
A method and device using an untrained machine learning model to monitor milling machines by analyzing time series data of rotational speed and supply current, allowing for early prediction of anomalies and proactive maintenance measures, independent of machine type or application.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional monitoring methods are used, then the system is simple, but errors are only detected after they occur leading to production delays
Solution Approach 1:
The patent applies preliminary action by training the machine learning model with historical data before actual monitoring begins, enabling the system to predict failures before they occur. The model learns normal operation patterns and can identify deviations that indicate upcoming failures, allowing maintenance to be scheduled proactively rather than reactively.
Solution Approach 2:
The system implements continuous feedback by constantly monitoring operating parameters and comparing them against the trained model's expectations. When anomalies are detected, the system provides feedback signals that trigger alerts or automated responses, creating a closed-loop monitoring system that continuously improves detection accuracy.
2Adaptability or versatility
If machine learning models are trained offline before deployment, then the model structure is fixed, but it cannot adapt to changes in milling machine operation modes
Solution Approach 1:
The patent implements dynamics by enabling the machine learning model to adapt its parameters continuously during operation based on incoming data. The model transitions from a static, pre-trained state to a dynamic system that learns and adjusts its internal parameters in real-time, allowing it to accommodate changes in operation modes, workpiece types, and machine conditions.
Solution Approach 2:
The system applies self-service by enabling the monitoring system to automatically train and retrain itself using data from the milling machine's operation. The model serves its own training needs by consuming operational data and autonomously updating its parameters without requiring external intervention or manual reconfiguration.
3Measurement precision
If comprehensive time series data is collected, then anomaly detection accuracy improves, but data processing requirements increase
Solution Approach 1:
The patent applies extraction by selecting and focusing on the most relevant operating parameters for anomaly detection rather than processing all available data equally. The system extracts key features from the time series data that are most indicative of anomalies, reducing the computational burden while maintaining detection accuracy.
Solution Approach 2:
The system implements parameter changes by dynamically adjusting which operating parameters are monitored and how they are processed based on the current operational context. The model can change its focus between different parameters depending on which are most relevant for detecting anomalies in the current operation mode, optimizing computational resource usage.
Data Source
AI summary
A method of monitoring a milling machine includes deploying an untrained machine learning model for determining one or more anomalies in time series data. During operation of the milling machine, first time series data representing a rotational speed of a milling head of the milling machine and at least one further operating parameter of the milling machine are obtained by the untrained machine learning model. The untrained machine learning model is trained, during operation of the milling machine, based on the obtained first time series data. Second time series data representing the rotational speed of the milling head of the milling machine and the further operating parameter are obtained by the trained machine learning model during operation of the milling machine. One or more anomalies in the second time series data are determined by the trained machine learning model during operation of the milling machine.