Vehicle Fault Diagnosis Using Feature-Master Similarity Detection

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current vehicle fault diagnosis systems rely on specialized sensors to detect unusual noises and odors, which not all vehicles are equipped with, and existing predictive systems like Patent Literature 1 can only predict age-related faults, failing to detect daily life faults.

Innovation Solution

A fault diagnosis device and system that receives input of fault and observation data from vehicles, generates a feature master associating fault content with extracted features, and transmits this to target vehicles for similarity-based fault detection without requiring additional sensors.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If specialized sensors are installed to detect unusual noises and odors, then fault detection capability is improved, but device complexity and cost increase

Engineering Contradiction:

Improvefault detection capabilityVSAvoidsensor installation requirements

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The patent creates a virtual model (feature master) that copies the characteristic patterns of fault conditions from historical data. This virtual model is then used to detect faults by comparing current observation data against the stored feature master, eliminating the need for specialized physical sensors while maintaining fault detection capability

Inventive Principle:

Principle #26Copying

Solution Approach 2:

The patent replaces the mechanical sensor-based detection system with an information-processing system that uses data analysis and pattern recognition. Instead of physical sensors detecting noises and odors, the system uses observation data from existing vehicle systems processed through feature extraction and similarity comparison algorithms

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Reliability

If fault prediction is based on age-related deterioration patterns, then long-term fault prediction is improved, but detection of daily life faults deteriorates

Engineering Contradiction:

Improveage-related fault predictionVSAvoiddaily life fault detection

Core Design Contradiction:

ReliabilityVSAdaptability or versatility

Solution Approach 1:

The patent makes the fault prediction system dynamic by adapting the feature master based on accumulated observation data. The system can adjust to different fault patterns over time, allowing it to handle both age-related deterioration and sudden daily life faults by updating its understanding of normal and abnormal patterns

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The patent changes the approach from fixed age-based prediction to flexible parameter-based prediction. By extracting multiple features from observation data and comparing them against the feature master, the system can detect various types of faults regardless of vehicle age, adapting to different fault scenarios through parameter comparison rather than predetermined age thresholds

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS12190658B2Fault diagnosis device, fault diagnosis system, fault diagnosis method, and fault diagnosis program

Publication Date: 2025.01.07 NEC CORP

AI summary

A fault diagnosis device 80 includes an input unit 81 and a generation unit 82. The input unit 81 receives input of fault data obtained from a vehicle when a fault of the vehicle occurs and observation data observed in time series by each device of the vehicle until immediately before the fault occurs. The generation unit 82 generates a feature master that associates a content of the fault indicated by the fault data with features extracted from the corresponding observation data.