Digital olfactory data service platform

Abstract

The present invention provides a digital olfactory data service platform comprising: a digital olfactory data collection system including a multi-channel olfactory sensor array provided to be able to collect digital olfactory data by using various types of odor collection samples collected in a plurality of fields, and a sensing management module provided to be able to manage the odor collection samples and the digital olfactory data; a digital olfactory AI data service system including a digital olfactory data lake, a digital olfactory data service management module, and a digital olfactory data management module, storing olfactory sample-related metadata, digital olfactory data, digital olfactory recognition data, and sensor management data, providing various data management functions stored in the digital olfactory data lake, and providing functions such as product management and subscription management for data services; and an olfactory recognition AI deep learning system including an olfactory recognition deep learning model for performing AI deep learning on the digital olfactory sensor data.

Description

Digital Olfactory Data Service Platform

[0001] The present invention relates to a digital olfactory data service platform, and more specifically, to a digital olfactory data service platform having a digital olfactory data distribution function, which enables the provision or sale of digital olfactory data to companies, organizations, or institutions that intend to develop olfactory AI technology or services.

[0002] Existing odor recognition sensors are being developed either by being composed of sensors in the form of electronic noses that react to VOCs (volatile organic compounds) of specific molecules or substances to be detected, or by mimicking the functions of olfactory cells through convergence with biotechnology.

[0003] Furthermore, existing odor recognition methods react only to specific substances, so while they offer high accuracy in detecting those substances in a short time, most are still in the early stages of technological development, with the exception of certain toxic gases.

[0004] In addition, existing dedicated sensors had problems such as short lifespans, making them unsuitable for generating and utilizing various odor data content.

[0005] Registered Patent No. 10-2253148 (registered on May 11, 2021) discloses an olfactory detection method for measuring odors.

[0006] The olfactory detection method for measuring an odor according to the above-mentioned registered patent detects an odor around an electronic device by one or more of a plurality of odor sensors of an electronic device, obtains a first odor count coefficient for the odor sensor that detected the odor among the plurality of odor sensors by a processor of the electronic device, and obtains a change amount of the first odor count coefficient by subtracting the obtained first odor count coefficient from a second odor count coefficient that was stored in advance before detecting the odor by the processor.

[0007] And the aforementioned previously stored second odor count coefficient is stored in a first table of memory of the electronic device, and the processor obtains a temperature count value from at least one monitor sensor, estimates the current temperature around the electronic device based on the temperature count value obtained by the processor, and obtains a first temperature count value for an odor sensor corresponding to the current temperature using the temperature count value obtained by the processor.

[0008] Additionally, a second temperature count value for an odor sensor corresponding to a reference temperature is subtracted from a first temperature count value obtained by the processor to obtain a second odor count coefficient change amount, the second temperature count value is stored in a second table of the memory, and the processor subtracts the second odor count coefficient change amount from the first odor count coefficient change amount to identify the level of a detected odor around the electronic device, and the identified level of the odor is displayed on the display of the electronic device.

[0009] A conventional olfactory detection method for measuring odors configured in this way can reduce costs and reduce the size of the olfactory detection device by compensating for the temperature of the odor in the olfactory detection device to measure the odor, thereby eliminating the need for a separate device for measuring temperature.

[0010] However, conventional olfactory detection methods that have the above-mentioned effects are merely at the level of measuring odor by compensating for temperature in the olfactory detection device, and thus have the problem of being unsuitable for generating and utilizing various odor data content.

[0011] Due to these problems, it was previously impossible to provide or sell odor data.

[0012] (Prior Art Literature)

[0013] (Patent Literature)

[0014] (Patent Document 1) Olfactory sensing method for measuring odors according to Korean Registered Patent Publication No. 10-2253148 (Registered May 11, 2021).

[0015] (Patent Document 2) Composite for electronic olfactory sensor and method for manufacturing the same according to Korean Registered Patent Publication No. 10-0477799 (Registered March 10, 2005)

[0016] (Patent Document 3) Electronic nose system and method for gas classification according to Korean Registered Patent Publication No. 10-1852074 (Registered April 19, 2018)

[0017] (Patent Document 4) Gas collection unit, gas analysis device including the same, and gas analysis method of Korean Registered Patent No. 10-2264407 (registered June 8, 2021).

[0018] (Patent Document 5) Cell-based olfactory sensor that increases olfactory response and method for increasing olfactory response using the same according to Korean Registered Patent Publication No. 10-1129474 (Registered March 16, 2012).

[0019] (Patent Document 6) Method for manufacturing a biosensor platform with a pattern deposited olfactory receptor protein according to Korean Registered Patent Publication No. 10-1267238 (Registered May 20, 2013), biosensor platform manufactured thereby, field-effect transistor including the same, and biosensor

[0020] The present invention was created to solve the aforementioned problems, and its purpose is to provide a digital olfactory data service platform equipped with a digital olfactory data distribution function, which enables the provision or sale of digital olfactory data to companies, organizations, or institutions that intend to develop olfactory AI technology or services.

[0021] The digital olfactory data service platform of the present invention for achieving the above-mentioned purpose is,

[0022] A digital olfactory data collection system comprising a multi-channel olfactory sensor array configured to collect digital olfactory data using various types of odor collection samples collected from multiple fields, and a sensing management module configured to manage the odor collection samples and the digital olfactory data;

[0023] A digital olfactory AI data service system comprising a digital olfactory data lake, a digital olfactory data service management module, and a digital olfactory data management module, which stores metadata related to olfactory samples, digital olfactory data, digital olfactory recognition data, and sensor management data, provides various data management functions stored in the digital olfactory data lake, and provides functions such as product management and subscription management for data services;

[0024] The system is characterized by including an olfactory recognition AI deep learning system equipped with an olfactory recognition deep learning model that performs AI deep learning on the digital olfactory sensor data.

[0025] In the present invention, the various types of odor collection samples include target gases in the fields of food, living organisms, natural objects, and artificial objects, and olfactory data generated by mixing the target gas with an environmental gas (a substance other than the target substance or air in a specific space) to recognize the characteristics of the target gas.

[0026] In the present invention, the multi-channel olfactory sensor array comprises multiple types of sensors arranged in a stacked structure, allowing for the acquisition of various data based on sensor locations.

[0027] In the present invention, the multi-channel olfactory sensor array is configured to deliver input gas from an air distributor to each chamber under the same conditions by applying a multi-chamber method.

[0028] In the present invention, the digital olfactory data lake comprises: a raw data processing unit that stores and processes a plurality of digital olfactory sensor raw data generated through the multi-channel olfactory sensor array; a sensor data processing unit that stores and processes digital olfactory sensor data generated by preprocessing the digital olfactory sensor raw data; and a recognition data processing unit that stores and processes digital olfactory recognition data generated through AI deep learning of the digital olfactory sensor data.

[0029] In the present invention, the raw data of the digital olfactory sensor is sensor data transmitted from the olfactory sensor array, and tasks such as deep learning training and recognition are performed based on the data prior to the preprocessing stage.

[0030] In the present invention, the digital olfactory sensor data is data generated when the digital olfactory sensor raw data is preprocessed, and the digital olfactory recognition data is olfactory recognition data recognized through olfactory recognition AI deep learning using the digital olfactory sensor raw data or the digital olfactory sensor data.

[0031] In the present invention, the digital olfactory data lake comprises a captured sample information storage unit, a sensor array information storage unit, and a preprocessing information storage unit, wherein the captured sample information storage unit stores basic information regarding odor capture data that recognizes a sensor as metadata for the captured sample, the sensor array information storage unit stores various setting values ​​of the sensor array as metadata for the sensor array used when recognizing each captured sample as a sensor, and the preprocessing information storage unit stores data regarding which setting value the preprocessing was performed on in the case of olfactory sensor data on which a preprocessing process was performed.

[0032] In the present invention, the digital olfactory data service management module is provided to supply or service digital olfactory data and is composed of functions related to actual services for providing digital olfactory data.

[0033] In the present invention, the digital olfactory data management module is provided to manage digital olfactory data, manages various data to be stored in the digital olfactory data lake, stores olfactory data in the digital olfactory data lake, and provides related functions in the form of an API so as to provide necessary data to the olfactory data service management module.

[0034] According to an embodiment of the present invention, by providing olfactory data, which is one of the biggest obstacles to the advancement of olfactory AI technology, a foundation for developing olfactory AI technology can be established, and by providing an integrated solution for all procedures related to olfactory recognition, such as odor sample collection, odor sample sensing, digital olfactory data generation, and olfactory recognition AI, it can contribute to the development of olfactory AI learning technology.

[0035] Furthermore, the development of olfactory recognition AI technology enables the advancement of sensory technologies in the field of olfaction, thereby contributing to the development of multi-sensory technologies based on the five senses in a true sense. Additionally, by providing olfactory data to contribute to the advancement of various olfactory AI industries, it is possible to secure a technological lead in the globally nascent olfactory AI sector.

[0036] Furthermore, by contributing to the development of various olfactory AI-based services, the scale of the olfactory sensing industry can be expanded, and since olfactory AI can be integrated with other industries such as VR, the scale of the ICT convergence industry can be expanded.

[0037] Furthermore, the development of new sensory intelligence enables the advancement of various olfactory services, thereby contributing to the improvement of human Quality of Life (QOL); olfactory AI can be applied to healthcare fields such as the diagnosis of diseases like lung cancer, the treatment of olfactory disorders, and the alleviation of dementia symptoms, thus contributing to the improvement of public health; and olfactory AI can be applied to hazardous fields such as explosive detection, thereby contributing to the enhancement of public safety.

[0038] FIG. 1 is a block diagram schematically showing the overall configuration of a digital olfactory data service platform according to the present invention.

[0039] FIG. 2 is a detailed configuration diagram of the digital olfactory AI data service system of FIG. 1.

[0040] FIG. 3 is a conceptual diagram of the sensor array structure configured in FIG. 1.

[0041] FIG. 4 is a multi-chamber sensor array with a stacked structure.

[0042] Hereinafter, preferred embodiments according to the present invention will be described in detail with reference to the attached drawings.

[0043] FIG. 1 shows a block diagram schematically illustrating the overall configuration of a digital olfactory data service platform according to the present invention.

[0044] And Figure 2 shows a detailed configuration diagram of the digital olfactory AI data service system of Figure 1.

[0045] In addition, Fig. 3 shows a conceptual diagram of the sensor array structure configured in Fig. 1, and Fig. 4 shows a multi-chamber sensor array with a stacked structure.

[0046] Referring to FIGS. 1 to 4, the digital olfactory data service platform according to the present invention comprises a digital olfactory data collection system (100), a digital olfactory AI data service system (200), and an olfactory recognition AI deep learning system (300).

[0047] Specifically, the digital olfactory data collection system (100) comprises a multi-channel olfactory sensor array (110) configured to collect digital olfactory data using various types of odor collection samples collected from multiple fields, and a sensing management module (120) configured to manage odor collection samples (10) and digital olfactory data.

[0048] And the above-mentioned various types of odor collection samples (10) include food, living organisms, natural objects, artificial objects, and target olfactory data.

[0049] And the above-mentioned multi-type odor collection sample (10) is composed of a target gas in fields such as food, living organisms, natural objects, and artificial objects, and an environmental gas (a substance other than the target substance or air in a specific space) that is mixed with the target gas to infer the characteristics of the target gas in various environments.

[0050] In addition, the multi-channel olfactory sensor array (110) has multiple types of sensors arranged as shown in FIG. 3 and can be expanded into a stacked structure to obtain various data according to the sensor positions.

[0051] This multi-channel olfactory sensor array (110) is configured to apply a multi-chamber method, as shown in FIG. 4, to deliver input gas from an air distributor to each chamber under the same conditions.

[0052] And the above digital olfactory AI data service system (200) is equipped with a digital olfactory data lake (200a), a digital olfactory data service management module (200b), and a digital olfactory data management module (200c).

[0053] The digital olfactory data lake (200a) is equipped with a raw data processing unit (210) that stores and processes a plurality of digital olfactory sensor raw data generated through a multi-channel olfactory sensor array, a sensor data processing unit (220) that stores and processes digital olfactory sensor data generated by preprocessing the digital olfactory sensor raw data, and a recognition data processing unit (230) that stores and processes data including a digital olfactory recognition deep learning model generated through AI (Artificial Intelligence) deep learning of the digital olfactory sensor data, the type of the recognized substance, and information related to the model generation.

[0054] In addition, the raw data of the digital olfactory sensor is sensor data transmitted from an olfactory sensor array, which is data prior to preprocessing, and tasks such as deep learning training and recognition are performed based on this data.

[0055] In addition, the above digital olfactory sensor data is data generated when the raw digital olfactory sensor data is preprocessed, and the decision to preprocess is made based on the purpose and nature of the olfactory data.

[0056] In addition, the digital olfactory recognition data is an olfactory recognition model and metadata recognized through olfactory recognition AI deep learning using the digital olfactory sensor raw data or the digital olfactory sensor data, and various detailed setting values ​​for the recognition data are stored.

[0057] And the digital olfactory data lake (200a) is equipped with a collection sample information storage unit (240), a sensor array information storage unit (250), and a preprocessing information storage unit (260).

[0058] The above-mentioned captured sample information storage unit (240) stores basic information about odor capture data and environmental gases that recognize the sensor as metadata for the captured sample.

[0059] In addition, the sensor array information storage unit (250) stores various setting values ​​of the sensor array as metadata for the sensor array used when recognizing each captured sample as a sensor, and allows the sensor data recognized by the various setting values ​​to be compared so that the optimal sensor setting value can be found.

[0060] And the above preprocessing information storage unit (260) stores data on what setting value the preprocessing was performed on in the case of olfactory sensor data for which the preprocessing process was performed, and can be used to identify basic information on the necessity of preprocessing.

[0061] In addition, the digital olfactory data service management module (200b) is provided to supply or service digital olfactory data, and is configured with functions related to actual services to provide digital olfactory data to companies, organizations, or institutions that require digital olfactory data, as shown in FIG. 1.

[0062] And the digital olfactory data management module (200c) is provided to manage digital olfactory data and has the function of managing various data to be stored in the digital olfactory data lake, storing olfactory data in the digital olfactory data lake (200a), and providing related functions in the form of an API so that the olfactory data service management module (200b) can provide the necessary data.

[0063] In addition, the above-mentioned olfactory recognition AI deep learning system (300) is equipped with an olfactory recognition deep learning model (310) that performs AI deep learning on digital olfactory sensor data.

[0064] As described above, the digital olfactory data service platform according to the present invention can provide and sell digital olfactory data to companies, organizations, and institutions that wish to develop olfactory AI technology or services by implementing a digital olfactory data collection system (100) based on a multi-channel olfactory sensor array, a deep learning-based olfactory recognition AI deep learning system (300), and a digital olfactory AI data service system (200) having a digital olfactory data distribution function, so as to collect multiple digital olfactory data for olfactory samples in various fields such as nature, environment, food, and organisms, including target olfactory data that is economically valuable and can be sold immediately.

[0065] And the multi-channel sensor array (110) of the digital olfactory data collection system (100) is a sensor array composed of 40 or more sensors of 3 or more types, and the sensing management module (120) performs sensor array management and sensing control functions.

[0066] Through this digital olfactory data collection system (100), olfactory samples and digital olfactory data can be obtained, and digital olfactory data including target olfactory data that is economically valuable and can be sold immediately can be generated.

[0067] In addition, the above-mentioned olfactory recognition AI deep learning system (300) implements a deep learning model based on CNN or ConvNet (Convolutional neural network).

[0068] And the digital olfactory AI data service system (200) is equipped with a digital olfactory data lake (200a) to store metadata related to olfactory samples, digital olfactory data, digital olfactory recognition data, sensor management data, etc., and is equipped with a digital olfactory data management module (200c) to provide various data management functions stored in the digital olfactory data lake (200a), and is equipped with a digital olfactory data service management module (200b) to provide functions such as product management and subscription management for data services.

[0069] As such, the digital olfactory data service platform according to the present invention can implement a multi-channel sensor array to identify various characteristics such as physical, electrical, and chemical properties of odors using multi-channel olfactory sensor array technology, and can automate the reflection of various generation conditions to generate large volumes of digital olfactory data using olfactory sensing management technology.

[0070] In addition, the digital olfactory data service platform according to the present invention can implement olfactory recognition AI technology and deep learning optimization technology using CNN-based deep learning technology, and can build an integrated service platform for providing digital olfactory data and olfactory recognition AI technology.

[0071] As described above, the present invention has been explained with reference to an embodiment illustrated in the drawings, but this is merely illustrative, and those skilled in the art will understand that various modifications and equivalent embodiments are possible therefrom.

[0072] Therefore, the true scope of protection of the present invention must be determined solely by the appended claims.

[0073] (Explanation of symbols)

[0074] 10: Odor collection sample

[0075] 100: Digital Olfactory Data Collection System

[0076] 110: Multichannel olfactory sensor array

[0077] 120: Sensing Management Module

[0078] 200: Digital Olfactory AI Data Service System

[0079] 200a: Digital Olfactory Data Lake

[0080] 200b: Digital Olfactory Data Service Management Module

[0081] 200c: Digital Olfactory Data Management Module

[0082] 210: Raw Data Processing Unit

[0083] 220: Sensor data processing unit

[0084] 230: Recognition data processing unit

[0085] 240: Captured sample information storage unit

[0086] 250: Sensor array information storage unit

[0087] 260: Preprocessing Information Storage Unit

[0088] 300: Olfactory Recognition AI Deep Learning System

[0089] 310: Deep learning model for olfactory recognition

Claims

1. A digital olfactory data collection system comprising a multi-channel olfactory sensor array configured to collect digital olfactory data using various types of odor collection samples collected from multiple fields, and a sensing management module configured to manage the odor collection samples and the digital olfactory data; A digital olfactory AI data service system comprising a digital olfactory data lake, a digital olfactory data service management module, and a digital olfactory data management module, which stores metadata related to olfactory samples, digital olfactory data, digital olfactory recognition data, and sensor management data, provides various data management functions stored in the digital olfactory data lake, and provides functions such as product management and subscription management for data services; A digital olfactory data service platform characterized by including an olfactory recognition AI deep learning system equipped with an olfactory recognition deep learning model that performs AI deep learning on the digital olfactory sensor data.

2. In Paragraph 1, The above-mentioned multi-type odor-capturing samples include target gases selected from the fields of food, living organisms, natural objects, and artificial objects; and Olfactory data generated by mixing a target gas and an environmental gas to recognize the characteristics of the above-mentioned target gas; A digital olfactory data service platform characterized by including 3. In Paragraph 1, A digital olfactory data service platform characterized by the above-mentioned multi-channel olfactory sensor array having multiple types of sensors arranged and expanded into a stacked structure to obtain various data according to the sensor positions.

4. In Paragraph 1, A digital olfactory data service platform characterized by the above-mentioned multi-channel olfactory sensor array being configured to deliver input gas from an air distributor to each chamber under the same conditions by applying a multi-chamber method.

5. In Paragraph 1, The above digital olfactory data lake is, A raw data processing unit that stores and processes multiple digital olfactory sensor raw data generated through the above-mentioned multi-channel olfactory sensor array; A sensor data processing unit that stores and processes digital olfactory sensor data generated by preprocessing the above-mentioned digital olfactory sensor raw data; A digital olfactory data service platform characterized by including: a recognition data processing unit that stores and processes digital olfactory recognition data generated through AI deep learning of the above digital olfactory sensor data.

6. In Paragraph 5, A digital olfactory data service platform characterized by performing tasks such as deep learning training and recognition based on data prior to preprocessing, wherein the raw data of the digital olfactory sensor is sensor data transmitted from the olfactory sensor array.

7. In Paragraph 5, A digital olfactory data service platform characterized in that the above digital olfactory sensor data is data generated when the above digital olfactory sensor raw data is preprocessed.

8. In Paragraph 5, A digital olfactory data service platform characterized in that the digital olfactory recognition data is olfactory recognition data recognized through olfactory recognition AI deep learning using the digital olfactory sensor raw data or the digital olfactory sensor data.

9. In Paragraph 1, The above digital olfactory data lake includes a captured sample information storage unit, a sensor array information storage unit, and a preprocessing information storage unit, wherein The above-mentioned captured sample information storage unit stores basic information about odor capture data that recognizes a sensor as metadata for the captured sample, and The sensor array information storage unit stores various setting values ​​of the sensor array as metadata for the sensor array used when recognizing each captured sample as a sensor, and A digital olfactory data service platform characterized by the above-mentioned preprocessing information storage unit storing data regarding what setting value the preprocessing was performed on in the case of olfactory sensor data on which a preprocessing process was performed.

10. In Paragraph 1, A digital olfactory data service platform characterized by the above-mentioned digital olfactory data service management module being equipped to supply or service digital olfactory data and composed of functions related to actual services for providing digital olfactory data.

11. In Paragraph 1, A digital olfactory data service platform characterized by the above-mentioned digital olfactory data management module being provided to manage digital olfactory data, managing various data to be stored in the digital olfactory data lake, storing olfactory data in the digital olfactory data lake, and providing related functions in the form of an API so as to provide necessary data to the olfactory data service management module.

Images