Abnormal Sound Specifying Device Using AI Range Overlap Verification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for specifying abnormal sounds from vehicle recordings rely on human intervention, which can lead to inaccuracies, including the specification of sounds that do not occur in reality, and lack the precision needed for reliable output.
Innovation Solution
An abnormal sound specifying device utilizing artificial intelligence, equipped with a learned model that performs frequency-time data analysis and determines the type of abnormal sound by comparing designated and basis ranges, preventing incorrect outputs by ensuring overlap between these ranges.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If artificial intelligence is used to specify abnormal sound types from frequency-time data, then automation and productivity are improved, but measurement precision deteriorates due to inaccurate specifications including sounds that do not occur in reality
Solution Approach 1:
The system generates explanation information that provides feedback on the basis for AI's abnormal sound specifications. This feedback mechanism allows verification of whether the specified abnormal sound types are reasonable by comparing the basis range (where sound energy is concentrated) with the designated range (user-specified range of interest), thereby improving measurement precision while maintaining automation.
Solution Approach 2:
The patent introduces an intermediary verification process between the AI specification and final output. The explanation information acts as a mediator that bridges the AI's automated specification and the user's understanding, allowing indirect verification through comparison of basis and designated ranges without requiring direct human intervention in every case.
2Productivity
If the AI model specifies abnormal sound types directly from frequency-time data, then productivity is improved, but reliability deteriorates due to specification of abnormal sounds that do not occur in reality
Solution Approach 1:
The system performs preliminary action by generating explanation information that identifies the basis range before final specification is made. This preliminary step allows the system to prepare verification data (the basis where sound energy is concentrated) that can be used to check the reliability of the AI's specification, thereby improving reliability without significantly reducing productivity.
Solution Approach 2:
The explanation information provides feedback on the AI's specification by showing the basis range. This feedback loop allows the system to verify whether the specified abnormal sound type is reliable by checking if the basis range overlaps with the designated range, thereby improving reliability while maintaining automated productivity.
3Measurement precision
If human users specify abnormal sound types based on Fast Fourier Transform analysis, then measurement precision is maintained, but productivity and ease of operation deteriorate due to manual intervention requirements
Solution Approach 1:
The AI model performs self-service by automatically specifying abnormal sound types and generating explanation information. This self-service capability maintains measurement precision through the built-in verification mechanism (comparing basis range with designated range) while dramatically improving productivity by eliminating manual Fast Fourier Transform analysis and user specification requirements.
Solution Approach 2:
The patent substitutes the mechanical process of manual Fast Fourier Transform analysis and human judgment with an automated AI system. The AI model, equipped with explanation generation capability, replaces the mechanical human operation while maintaining or improving measurement precision through automated verification, thereby significantly improving productivity.
4Reliability
If verification processes are added to check AI specifications, then reliability is improved, but device complexity increases
Solution Approach 1:
The AI model serves multiple functions: it specifies abnormal sound types and simultaneously generates explanation information that serves as verification data. This multi-functionality approach improves reliability through built-in verification while avoiding additional separate verification devices or complex external systems, thereby minimizing device complexity.
Solution Approach 2:
The verification function is merged with the AI specification process. The explanation information generation is combined with the abnormal sound type specification in a single integrated process, eliminating the need for separate verification systems and reducing overall device complexity while maintaining improved reliability.
Data Source
AI summary
An abnormal sound specifying device may include an arithmetic device configured to access a learned model of artificial intelligence and an output device. The arithmetic device may perform: specifying frequency-time data of sound recorded at a vehicle; causing the learned model to specify a type of abnormal sound included in the sound based on the frequency-time data and causing the learnt model to specify a basis range from the frequency-time data, the basis range indicating a frequency range and a time range that are used to specify the type of the abnormal sound; designating a designated range indicating frequency and time ranges; and determining whether to cause the output device to output the type of the abnormal sound in a determination process, the determination process including, as a determination element, at least a determination on whether the basis range and the designated range overlap each other.