Defect Quality Attribute Classification for Faster Triage
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for classifying defects in industrial and autonomous systems are manual and do not consider quality attributes, leading to inefficiencies in defect triage and resource allocation, potentially resulting in system malfunctions and increased costs.
Innovation Solution
A computer-implemented method using a classification algorithm to determine quality attributes for defects, based on input data sets that include defect descriptions, allowing for efficient and reliable classification and prioritization of defects.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual defect classification by experts is used, then defect classification can be performed with project-specific criteria, but the process is time-consuming and inefficient
Solution Approach 1:
The patent replaces the manual mechanical classification process performed by experts with an automated machine learning classification system. The classifier automatically assigns quality attributes to defects based on input data, eliminating the need for manual expert review while maintaining classification accuracy through trained algorithms.
Solution Approach 2:
The defect classification system performs self-service by automatically classifying defects without requiring expert intervention. The machine learning model has been trained to independently evaluate defect data and assign appropriate quality attributes, enabling the system to serve itself in the classification task.
2Reliability
If manual defect classification is used, then quality attributes can be assessed, but resource allocation efficiency is reduced
Solution Approach 1:
The patent replaces manual quality attribute assessment with an automated machine learning classifier that rapidly evaluates defects and assigns quality attributes. This substitution maintains reliable quality assessment through consistent algorithmic application while dramatically improving resource allocation efficiency by processing defects at machine speed.
3Productivity
If automated classification is implemented, then defect triage efficiency improves, but the system requires training data and initial setup
Solution Approach 1:
The patent applies preliminary action by training the machine learning classifier on historical defect data before deployment. This pre-training phase establishes the classification model's capabilities, enabling efficient automated defect triage once the system is operational. The initial setup investment pays off through sustained high-speed classification performance.
4Ease of operation
If conventional defect classification methods are used, then simple categorization is achieved, but quality attribute determination is not performed
Solution Approach 1:
The patent applies universality by designing a classification system that performs multiple functions simultaneously. The machine learning classifier not only categorizes defects into standard groups but also determines quality attributes such as correctness, reliability, and performance impact. This multi-functional approach eliminates the need for separate quality assessment processes.
Data Source
AI summary
Provided is a computer-implemented method for determining at least one quality attribute for at least one defect of interest, including the steps: a. providing an input data set including the at least one defect of interest; b. determining the at least one quality attribute for the at least one defect of interest using a classification algorithm based on the input data set; and c. providing the determined at least one quality attribute and/or additional output information as output. Further, a computing unit and a computer program product are provided.