Multi-Axis Machine Monitoring With Interpretable Fault Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing multi-axis machines, such as robots, face challenges in accurately identifying and preventing errors due to the lack of interpretability in machine learning algorithms, leading to inefficiencies and potential damage from undetected faults.
Innovation Solution
A method using a convolutional neural network with specific layer configurations to analyze data time series, providing interpretable results by calculating channel contributions to classification values, allowing for early detection and prevention of errors through probability distributions and warning signals.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If machine learning algorithms are used for error detection in multi-axis machines, then detection accuracy is improved, but interpretability deteriorates
Solution Approach 1:
The patent introduces an intermediary explanation layer that mediates between the black-box machine learning model and the user. This layer generates human-readable explanations that translate the model's internal decisions into understandable formats, allowing users to comprehend why errors were detected without sacrificing the model's predictive accuracy.
Solution Approach 2:
The patent segments the error detection system into distinct components: the machine learning model for prediction and an explanation module for interpretability. By dividing the system, the model can operate as a black-box for accurate predictions while the explanation module separately provides transparent reasoning, resolving the contradiction between accuracy and interpretability.
2Reliability
If complex machine learning models are deployed, then error detection capability is improved, but system complexity increases
Solution Approach 1:
The explanation module serves as an intermediary that simplifies the interaction with complex machine learning models. It translates complex model outputs into straightforward explanations, allowing the system to leverage powerful complex models while presenting a simplified interface to users and stakeholders.
3Device complexity
If traditional error detection methods are used, then system simplicity is maintained, but error detection accuracy deteriorates
Solution Approach 1:
The patent replaces traditional mechanical or rule-based error detection systems with machine learning-based detection. This substitution enables the system to achieve superior detection accuracy by leveraging patterns and insights that would be difficult to encode using traditional methods, while the explanation module maintains clarity.
4Productivity
If machine learning algorithms without explanation are used, then detection speed is improved, but user trust deteriorates
Solution Approach 1:
The explanation module acts as an intermediary that preserves detection speed while building user trust. It provides rapid, automated explanations that accompany detection results, allowing users to quickly understand and act on detections without manual analysis, thereby maintaining speed while enhancing trust through transparency.
Data Source
AI summary
A method for assessing and/or monitoring a process and/or a multi-axis machine includes recording at least one data time series, wherein the at least one data time series includes at least one channel describing at least one parameter of the process and/or of the multi-axis machine, and wherein the data time series is caused by the process. An interpretable result is determined by a machine learning algorithm based on the at least one data time series, wherein the result describes a classification value of a state in the process and/or of a state of the multi-axis machine. A warning is output when determining the result if the classification value of the state in the process and/or of the state of the multi-axis machine is assigned to a value of an error class that is in a warning range or corresponds to a warning range, and an all-clear signal is output if the classification value of the state in the process and/or of the state of the multi-axis machine is assigned to a value of an error class that is in an all-clear range or corresponds to an all-clear range.