Machine Learning FMEA System for Automated Risk Calculation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current Failure Mode and Effect Analysis (FMEA) systems require significant time and are prone to human error due to the manual calculation of severity, frequency, and detectivity, which hampers their efficiency and accuracy in safety analysis.
Innovation Solution
A machine learning-based FMEA system that includes units for calculating failure severity, frequency, and detectivity, utilizing frameworks like open-source neural networks, Keras, or Tensorflow, to automate these calculations and provide users with automated analysis and evaluation information, including new failure modes and risk scenarios.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual calculation methods are used for FMEA analysis, then human operators can perform the analysis with existing tools, but the process requires considerable time and is prone to human error
Solution Approach 1:
The patent replaces manual mechanical calculation methods with an automated computing system that uses machine learning models. The system automatically calculates severity, frequency, and detectivity values by processing safety analysis information through trained algorithms, eliminating human operators from the calculation process. This substitution resolves the contradiction by providing both high accuracy (through consistent algorithmic processing) and time efficiency (through automated parallel processing).
Solution Approach 2:
The system enables self-service by automatically performing FMEA calculations without requiring human intervention in the computation process. The machine learning models independently process input data and generate results, allowing the system to serve itself in completing the analytical task. This approach simultaneously improves accuracy by eliminating human error and reduces time loss by removing manual processing steps.
2Productivity
If manual FMEA calculation processes are used, then existing tools can be utilized, but human error significantly impacts the accuracy of severity, frequency, and detectivity calculations
Solution Approach 1:
The patent replaces human manual calculation with automated machine learning-based computation. The system uses trained models to consistently calculate severity, frequency, and detectivity parameters without human intervention, eliminating variability and error associated with manual processes. This resolves the contradiction by achieving both high productivity (through automation) and high measurement precision (through consistent algorithmic execution).
Solution Approach 2:
The system incorporates feedback mechanisms where the machine learning models are trained on historical data and continuously improve their calculations. The automated system processes results and can refine its algorithms based on outcomes, ensuring both speed and precision are maintained and improved over time. This feedback loop resolves the contradiction by enabling rapid yet precise calculations that improve with use.
3Measurement precision
If automated machine learning systems are implemented for FMEA, then calculation speed and accuracy are improved, but system complexity increases
Solution Approach 1:
The patent introduces machine learning models as intermediary components between input safety analysis information and output failure parameters. These models act as intelligent mediators that automatically process complex calculations, shielding users from the underlying system complexity while delivering high-precision results. This resolves the contradiction by maintaining measurement precision through sophisticated algorithms while managing device complexity through modular architecture and automated processing.
Data Source
AI summary
A failure mode and effect analysis system according to the present invention may include: a failure severity calculation unit for calculating a failure severity by using machine learning on the basis of safety analysis information; a failure frequency calculation unit for calculating a failure frequency by using machine learning on the basis of safety analysis information; and a failure detectivity calculation unit for calculating a failure detectivity by using machine learning on the basis of safety analysis information.