Odor Sensor Model Training With Adaptive Measurement Conditions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Odor sensors face challenges in achieving high detection accuracy due to varying measurement conditions such as temperature and humidity, requiring numerous variations in conditions to be prepared, which is impractical and difficult to determine the necessary number for sufficient accuracy.
Innovation Solution
A learning model generation support apparatus and method that acquires sensor data and condition data under specific measurement conditions, inputs this data into a machine learning engine to generate a learning model, and sets new measurement conditions based on predictive accuracy to enhance detection accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple variations of measurement conditions are prepared to enhance detection accuracy, then detection accuracy is improved, but device complexity and difficulty in determining necessary variations increase
Solution Approach 1:
The patent changes the parameters of measurement conditions systematically by using a machine learning model to identify which condition parameters (temperature, humidity, etc.) most affect detection accuracy. This allows focusing on critical parameter variations rather than preparing all possible condition variations, thereby improving detection accuracy while reducing the complexity of preparing multiple condition sets.
Solution Approach 2:
The patent implements feedback through machine learning models that analyze detection results under different measurement conditions and provide information about which conditions yield optimal accuracy. This feedback mechanism allows the system to adaptively determine necessary condition variations, reducing the need to pre-prepare extensive condition variations while maintaining high detection accuracy.
2Measurement precision
If numerous measurement condition variations are prepared, then detection accuracy under varying conditions is improved, but time and resources required for data collection increase
Solution Approach 1:
The patent applies preliminary action by using machine learning models to predict which measurement condition variations will be most beneficial for detection accuracy before actual data collection begins. This allows the system to prioritize data collection under specific conditions, significantly reducing the time required compared to collecting data under all possible condition variations.
Solution Approach 2:
The patent implements partial action by collecting data under a selected subset of measurement conditions determined by machine learning analysis, rather than collecting data under all possible conditions. This partial approach to data collection achieves sufficient detection accuracy while dramatically reducing the time and resources required.
3Adaptability or versatility
If measurement conditions are fixed, then device complexity is reduced, but adaptability to different odors and conditions decreases
Solution Approach 1:
The patent applies dynamics by implementing a machine learning-based system that dynamically adapts to different odor types and measurement conditions. Rather than using fixed measurement conditions, the system learns optimal conditions for each odor type, enabling high versatility while managing complexity through adaptive algorithms rather than extensive hardware configurations.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Supports the setting of optimal measurement conditions for odor sensors with non-fixed analysis targets, enabling the creation of highly accurate learning models for odor detection.
Implementation Method 1
A piezoresistive element is embedded in each bridge. In such a configuration, the circular portion deforms due to stress occurring in the sensitive membrane when a substance sticks to the sensitive membrane, leading to stress being applied to the bridges. As a result, the electrical resistance of the piezoresistive elements embedded in the bridges changes greatly
Implementation Method 2
An odor sensor detects a specific odor, by using a sensor element to detect an airborne chemical substance that produces the specific odor
Data Source
AI summary
A learning model generation support apparatus 10 is an apparatus for supporting generation of a learning model to be utilized in odor detection using an odor sensor that reacts to a plurality of types of odors. The learning model generation support apparatus 10 includes a data acquisition unit 11 that acquires sensor data output by the odor sensor under specific measurement conditions and condition data specifying the measurement conditions, and inputs, as training data, the acquired sensor data and condition data to a machine learning engine 31 that generates the learning model, and a condition setting unit 12 that acquires a predictive accuracy output by the machine learning engine in response to input of the training data, and sets new measurement conditions for when the odor sensor newly outputs sensor data as training data, based on the acquired predictive accuracy.