Lifecycle Quality Monitoring With AI Claim-to-FMEA Mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods lack a comprehensive system for monitoring and analyzing quality problems throughout a product's lifecycle, particularly after the warranty period, and fail to accurately translate customer claims into Failure Mode and Effect Analysis (FMEA) categories.
Innovation Solution
A method and system that processes claims in natural language, translates them into a predetermined language, removes irrelevant words, calculates TF_IDF indices, associates keywords with FMEA categories using a probability map, and integrates data from official and unofficial channels to monitor and predict quality issues across the lifecycle.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If NLP techniques are used to extract keywords from claims, then the automatic processing capability is improved, but the accuracy of translating claims into FMEA categories deteriorates because the extracted keywords do not correspond to the categories used by technicians during design
Solution Approach 1:
The patent introduces an intermediary translation layer that maps NLP-extracted keywords to FMEA categories through a predefined mapping relationship. This intermediary mechanism bridges the gap between automated keyword extraction and structured FMEA categorization, allowing automatic processing while maintaining translation accuracy through the mediating mapping layer.
Solution Approach 2:
The patent replaces the manual mechanical process of translating claims to FMEA categories with an automated computational system. By substituting human expert translation with algorithmic processing based on keyword-FMEA mapping, the system achieves both automation and acceptable accuracy through systematic rule-based transformation.
2Ease of manufacture
If separate systems are used for monitoring products under warranty and after warranty, then the specialization for each stage is improved, but the overall efficiency of lifecycle monitoring deteriorates due to lack of integration
Solution Approach 1:
The patent merges previously separate warranty monitoring and post-warranty monitoring systems into a unified lifecycle monitoring platform. By combining data sources, processing capabilities, and analysis functions into a single integrated system, the patent achieves comprehensive lifecycle coverage while maintaining the specialized handling of different product stages through modular architecture.
Solution Approach 2:
The patent creates a universal monitoring system that handles multiple product lifecycle stages (under warranty and after warranty) through a single platform. This multi-functional system performs diverse tasks including claim processing, failure analysis, and predictive modeling across different product phases, improving overall efficiency while preserving stage-specific capabilities.
3Reliability
If comprehensive data collection from multiple channels is implemented, then the coverage of quality monitoring is improved, but the system complexity increases making integration and analysis more difficult
Solution Approach 1:
The patent segments the complex data collection system into distinct modular components, each responsible for specific data sources (official channels, unofficial channels, social media, etc.). This segmentation allows comprehensive data coverage while managing complexity through independent, manageable modules that can be developed, maintained, and scaled separately.
Solution Approach 2:
The patent introduces intermediary data processing layers that standardize and normalize data from diverse channels before integration. These intermediary components act as adapters and translators, converting various data formats and structures into a unified schema, thereby enabling comprehensive monitoring while reducing integration complexity through standardized interfaces.
Data Source
AI summary
Application, in a single software instrument, of AI and advanced statistics methods for monitoring and analyzing quality problems over the whole lifecycle of a product is provided. A method based on web scraping techniques and AI for monitoring the product out of the warranty period is also provided.