JP2022114254A - Information processing device, smell measurement system, and program - Google Patents

Abstract

To make it possible to improve the accuracy of smell identification and the accuracy of smell level analysis when measuring a smell type in accordance with a user's request.SOLUTION: An information processing device includes: a collection unit 33 for collecting, for a plurality of users, a detection result from a smell sensor 10 capable of detecting a plurality of smells and smell type information which is information on a smell type; a learning unit 34 for learning a relationship between the detection results and the smell type information collected by the collection unit 33 for the plurality of users; a transmitting/receiving unit 31 that acquires information on a smell that a user wants to measure from among the plurality of smells; and an analyzing unit 35 for analyzing a smell level that the user wants to measure on the basis of the relationship learned by the learning unit 34.SELECTED DRAWING: Figure 3

Description

The present invention relates to a technique for analyzing types of odors. More specifically, the present invention relates to technology for analyzing types of odors based on detection results obtained by odor sensors.

In recent years, the development of odor sensors has been remarkable. For example, an odor sensor has been proposed that has a plurality of sensor elements in one device and can identify many odors. For example, sensing elements are coated with sensitive films that adsorb individual odor components, and odors are identified from patterns of odor molecules detected by a plurality of sensing elements.

Patent Literature 1 discloses an odor detection device. This odor detection device includes a plurality of odor sensors having different odor response characteristics. In addition, based on the output values of multiple odor sensors, the odor components contained in the gas to be measured and their concentrations are specified, and based on the specified odor components and their concentrations, the type of odor is discriminated and determined. Outputs the type of odor.
Patent Literature 2 discloses a biological gas detection device. This biological gas detection device acquires the output value in the reference atmosphere of a semiconductor gas sensor sensitive to the interference gas and the desired gas contained in the biological gas as the reference output value. Then, the concentration of the desired gas is obtained based on the first output value of the semiconductor gas sensor during measurement of the biogas, and the second output value of the semiconductor gas sensor in the atmosphere before the measurement of the biogas is obtained; The concentration of the desired gas is corrected based on the reference output value, and information corresponding to the corrected corrected desired gas concentration is output.

Japanese Unexamined Patent Application Publication No. 2007-7006 JP 2018-155509 A

However, the conventional technology has limitations in identifying types of odors, and the identification accuracy is insufficient.
SUMMARY OF THE INVENTION It is an object of the present invention to improve the odor identification accuracy and the odor level analysis accuracy when measuring the type of odor according to the user's request.

Thus, according to the present invention, the collecting unit collects the detection results of the odor sensors capable of detecting a plurality of odors and the odor type information, which is information about the types of odors, for a plurality of users; A learning unit that learns the relationship between detection results and odor type information for multiple users, an acquisition unit that acquires information on the odor that one user wants to measure from among multiple odors, and the relationship learned by the learning unit. and an analysis unit that analyzes the level of odor that one user wants to measure based on the above. In this case, it is possible to improve the accuracy of odor identification and the analysis accuracy of the odor level when measuring the type of odor according to the user's request.

Here, the analysis unit can correct the detection result based on the temperature and/or humidity of the location where the odor that one user wants to measure is measured. In this case, the detection result will be more accurate.
Then, the analysis unit can analyze the odor level that the one user wants to measure, taking into account the user information, which is information about the one user. In this case, the odor type information becomes more accurate.
User information can also include information about a user's preferences. In this case, noise can be removed from the detection result, and the type of odor can be predicted.
Furthermore, the analysis unit can further create advice for the one user to deodorize the odor that the one user wants to measure. In this case, the user can obtain specific means for deodorizing the odor.
Also, the advice can guide a product that deodorizes the odor that one user wants to measure. In this case, user convenience is improved.
Furthermore, the collection unit can collect evaluations of results analyzed by the analysis unit from one user. In this case, it is possible to obtain user's feedback on the relationship between the detection result and the odor type information.
Still further, the learner can modify the relationship based on the evaluation collected by the collector. In this case, a more accurate learning model can be constructed.
Then, the learning unit can learn the evaluations collected by the collection unit above a predetermined value, and can not learn the evaluations less than the predetermined value. In this case, highly accurate learning data can be learned.
Further, the odor sensor has a plurality of sensors that output different detection results depending on the type of odor, and the collection unit can acquire the detection results of the plurality of sensors. In this case, the types of odors that can be detected are increased.

Furthermore, according to the present invention, an odor sensor capable of detecting a plurality of odors, an information processing device for obtaining the level of an odor that one user wants to measure from among the plurality of odors, and a display device for displaying the odor level. and an information processing apparatus comprising: a collection unit for collecting, for a plurality of users, results of detection by an odor sensor and odor type information, which is information regarding types of odors; A learning unit that learns the relationship between the detection result and the odor type information; and an analysis unit that analyzes the level of odor that one user wants to measure. It is possible to provide an odor measurement system capable of improving the accuracy of odor identification and the analysis accuracy of the odor level when measuring the type of odor according to the user's request.
Here, the display device can associate the odor type information, the odor level indicated by the odor type information, and the advice to deodorize the odor indicated by the odor type information, and display them to one user. In this case, user convenience is improved.

Then, according to the present invention, the computer has a collection function for collecting, for a plurality of users, the results of detection by odor sensors capable of detecting a plurality of odors, and odor type information, which is information relating to the types of odors, and a collection function. A learning function that learns the relationship between the detection results and odor type information collected by multiple users, and an acquisition function and a learning function that acquires information on the odor that one user wants to measure from among multiple odors. A program is provided for realizing an analysis function for analyzing the odor level that one user wants to measure based on the learned relationship. A computer can realize a function capable of improving the accuracy of odor identification and the analysis accuracy of the odor level when measuring the type of odor according to the user's request.

According to the present invention, it is possible to improve the accuracy of odor identification and the analysis accuracy of the odor level when measuring the type of odor according to the user's request.

1 is a diagram showing a configuration example of an odor measurement system according to the present embodiment; FIG. 1 is a schematic diagram showing the structure of an odor sensor; FIG. 1 is a block diagram showing a functional configuration example of an odor measurement system according to the present embodiment; FIG. FIG. 10 is a diagram showing an example of odor identification results; 4 is a flow chart explaining the operation of the odor measurement system. 6 is a diagram showing an example of a list presented to the user in step 105 of FIG. 5; FIG. 6 is a diagram showing a method of performing correction in step 111 of FIG. 5; FIG. (a) to (e) are diagrams showing analysis results of odors displayed on the terminal device. It is a figure which shows the structural example of the smell measurement system in a modification. FIG. 11 is a block diagram showing an example of functional configuration of an odor measurement system in a modified example;

BEST MODE FOR CARRYING OUT THE INVENTION Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

<Explanation of entire odor measurement system 1>
FIG. 1 is a diagram showing a configuration example of an odor measurement system 1 according to this embodiment.
As illustrated, the odor measurement system 1 of the present embodiment includes odor sensors 10a, 10b, and 10c for detecting odors, detection results of the odor sensors 10a, 10b, and 10c, analysis results of the odors, evaluation of the analysis results, and the like. Terminal devices 20 a , 20 b , and 20 c for transmitting/receiving and displaying odor analysis results, and a management server 30 , which is an example of an information processing device, for managing the entire odor measurement system 1 . Although there are three odor sensors 10a, 10b, and 10c and three terminal devices 20a, 20b, and 20c here, any number of them may be used as long as they are plural. User A owns the odor sensor 10a and the terminal device 20a. Similarly, user B owns odor sensor 10b and terminal device 20b, and user C owns odor sensor 10c and terminal device 20c. User A, user B, and user C are examples of multiple users, and user A, user B, and user C are examples of one user. Also, here, the number of users is three, but any number of users may be used. Further, hereinafter, when the odor sensor 10a, the odor sensor 10b, and the odor sensor 10c are not distinguished from each other, they may simply be referred to as "the odor sensor 10". Furthermore, when the terminal device 20a, the terminal device 20b, and the terminal device 20c are not distinguished from each other, they may simply be referred to as the "terminal device 20." In the following description, users A, B, and C may simply be referred to as "users" when they are not distinguished from each other.

The odor sensor 10 and the terminal device 20 are connected by a wired communication line or a wireless communication line, and the information on the odor acquired by the odor sensor 10 is transmitted to the terminal device 20 . Wired communication lines include, for example, wired LAN (Local Area Network) lines, USB (Universal Serial Bus) lines, optical communication lines, and public telephone lines. As wireless communication lines, mobile phone lines, PHS (Personal Handy-phone System) lines, Wi-Fi (Wireless Fidelity, registered trademark), Bluetooth (registered trademark), ZigBee (registered trademark), UWB (UltraWideband), etc. can be used. Also, the Internet or the like may be used together.
Also, the terminal device 20 and the management server 30 are connected via a network 40 . The network 40 is communication means used for information communication between the terminal device 20 and the management server 30, and is, for example, a LAN or the Internet.

The odor sensor 10 is a sensor device capable of detecting multiple odors. The odor sensor 10 identifies odors by detecting odor molecules that are sources of odors.
FIG. 2 is a schematic diagram showing the configuration of the odor sensor 10. As shown in FIG.
The illustrated odor sensor 10 includes a crystal oscillator 110 and a sensitive film 120 .
A crystal resonator 110 includes a crystal piece 111 and a pair of electrodes 112a and 112b.
The crystal piece 111 is, for example, a thin piece of crystal cut by an AT cut. The crystal piece 111 is sandwiched between a pair of electrodes 112a and 112b.
A pair of electrodes 112a and 112b are connected to an oscillating circuit Hc, and when an AC voltage is applied by the oscillating circuit Hc, resonance occurs in the crystal blank 111 at a predetermined frequency. This frequency is a resonance frequency, and is determined by, for example, the cut surface and thickness of the crystal blank 111, the elastic modulus of the crystal, and the like.

The sensitive film 120 adsorbs and desorbs chemical components contained in the air that cause odors. In this case, the sensitive film 120 adsorbs and desorbs odor molecules as chemical components that cause odors. Then, when the odor molecules are adsorbed, the resonance frequency of the crystal piece 111 changes. This can also be said to change the vibration mode of the crystal blank 111 . By detecting the amount of change in the resonance frequency with the controller S, the presence or absence and amount of odor molecules adsorbed to the sensitive film 120 can be determined.

As the sensitive film 120, one having selectivity for adsorbed odor molecules can be used. Thus, it is possible to create an odor sensor 10 that responds to a specific odor caused by this odor molecule. For example, by using the sensitive film 120 that selectively adsorbs ethanol, the odor sensor 10 that can detect the odor of alcoholic beverages can be produced. By arranging a plurality of odor sensors 10 having sensitive films 120 with various selectivities, various odor molecules can be adsorbed and various odors can be detected. Also, a plurality of sensitive films 120 may be provided in one odor sensor 10 . In this case, it can be considered that the odor sensor 10 has a plurality of sensors that output different detection results depending on the type of odor.
The sensitive film 120 is a thin film because it is required that the vibration of the crystal resonator 110 is hardly disturbed. In addition, a barrier layer may be further provided on the sensitive film 120 in order to prevent adsorption of molecules other than specific odor molecules.

After the odor molecules are adsorbed on the sensitive film 120, when dry air is supplied, the odor molecules are desorbed from the adsorbed state, and the resonance frequency returns to its original value. In other words, the odor can be continuously detected by alternately flowing the air containing the odor molecules and the dry air.

The odor sensor 10 may be a property purchased by the user, or may be provided by loan, for example. Also, on the condition of this loan, it may be possible to receive a service of analyzing odors, which will be described later. The lending of the odor sensor 10 may be free or charged. Also, the user may be charged when receiving the service of analyzing the odor.

The terminal device 20 is an example of a display device that displays an odor level. The terminal device 20 notifies the user of the odor analysis results sent from the management server 30 . Although the details will be described later, the terminal device 20 notifies the user of the odor analysis result by displaying the identified odor type and level. Alternatively, the user may be notified by emitting a sound such as voice or warning sound.
The terminal device 20 is, for example, a computer device such as a smart phone, tablet, mobile phone, desktop computer, or mobile computer.

The management server 30 is a server computer that manages the entire odor measurement system 1 . Although the details will be described later, for example, the management server 30 analyzes the odor based on the detection result of the odor sensor 10 .

The terminal device 20 and the management server 30 are provided with a CPU (Central Processing Unit), which is computing means, and a main memory, which is storage means. Here, the CPU executes various types of software such as an OS (basic software) and applications (application software). The main memory is a storage area for storing various software and data used for executing the software. Further, the terminal device 20 and the management server 30 include a communication interface (hereinafter referred to as "communication I/F") for communicating with the outside, a display mechanism including a video memory and a display, input buttons, An input mechanism such as a touch panel and a keyboard is provided. The terminal device 20 and the management server 30 also have a storage as an auxiliary storage device. The storage is, for example, an HDD (HardDiskDrive) or an SSD (SolidStateDrive). Furthermore, the terminal device 20 and the management server 30 can also be equipped with a speaker that emits sound.

Next, the detailed functional configuration and operation of the odor measurement system 1 of this embodiment will be described.
<Description of the functional configuration of the odor measurement system 1>
FIG. 3 is a block diagram showing a functional configuration example of the odor measurement system 1 according to this embodiment.
Here, among the various functions of the odor measurement system 1, those related to the present embodiment are selected and illustrated.

The odor sensor 10 includes a transmission/reception unit 11 that transmits information to the outside, an odor detection unit 12 that detects odors, an odor information generation unit 13 that generates detection results of odors, and a temperature/humidity measurement unit that measures temperature and humidity. 14.

The transmitting/receiving unit 11 transmits output information including the odor detection result to the terminal device 20 . The transmitting/receiving unit 11 corresponds to the control device S, for example.
The odor detection unit 12 is a functional unit that detects odor molecules. That is, it corresponds to the crystal oscillator 110, the sensitive film 120, the oscillation circuit Hc, and the control device S described above. The odor detection unit 12 detects the resonance frequency of the crystal oscillator 110 .

The odor information creating unit 13 creates output information including the odor detection result. In this case, the odor detection result is, for example, vibration information created based on the resonance frequency detected by the odor sensor 10 . The vibration information may be information representing a resonance frequency, or may be information representing a difference in resonance frequency between when odor molecules are detected and when odor molecules are not detected. Further, the odor detection result may be information processed based on vibration information. For example, a state in which odor molecules are not detected is 0, and a state in which odor molecules are detected up to the upper limit of the detection limit is 99, and the odor detection results may be expressed as numerical values in 100 steps.
Also, the output information may include other information in addition to the odor detection result. For example, it may include the acquisition time of the odor, the unique ID of the odor sensor 10 that acquired the odor, and the like. The output information also includes information on the temperature and humidity of the location to be measured, which are measured by the temperature/humidity measuring unit 14 .
The odor information creating unit 13 corresponds to the control device S, for example.

The temperature/humidity measurement unit 14 measures the temperature and humidity around the odor sensor 10 . That is, the odor sensor 10 can measure the temperature and humidity of the location where the odor that the user wants to measure is measured by the temperature/humidity measurement unit 14 when the odor is measured. The temperature/humidity measuring unit 14 is composed of a thermometer and a hygrometer.

The terminal device 20 includes a reception unit 21 that receives user instructions, an odor information acquisition unit 22 that acquires output information including odor detection results from the odor sensor 10, and a server communication that acquires odor analysis results from the management server 30. It includes a unit 23, an output unit 24 that creates and outputs measurement results to the user, and a storage unit 25 that stores odor detection results and the like.

The receiving unit 21 receives, as an instruction from the user, information on the odor that the user wants to measure. The user selects an odor to be measured from a list of odors measurable by the odor sensor 10, which will be described later in detail.
The odor information acquisition unit 22 acquires output information from the odor sensor 10 as an odor detection result.

The server communication unit 23 transmits sensor identification information for identifying the odor sensor 10 from the odor sensor 10 . The sensor identification information is, for example, the manufacturer and model number of the odor sensor 10 . Alternatively, a unique ID of the odor sensor 10 may be used. Furthermore, it may be a unique ID of the sensor chip mounted in the odor sensor 10 .
In addition, the server communication unit 23 associates the output information including the odor detection result with the user information, which is information about the user, and transmits the output information to the management server 30 . The output information includes information on the temperature and humidity of the location where the measurement was performed, as described above. User information includes, for example, the user's sex, nationality, age, place of residence, and user preferences. Among these, the user's preference corresponds to, for example, whether or not he/she smokes or drinks alcohol, and foods that he/she eats often.
The server communication unit 23 also receives from the management server 30 a list of odors measurable by the odor sensor 10 , which is created based on sensor identification information for identifying the odor sensor 10 .
Further, the server communication unit 23 receives the odor analysis result from the management server 30 . The analysis results will be detailed later.

The output unit 24 provides the user with the list of odors and analysis results of the odors acquired by the server communication unit 23 from the management server 30 . Here, a list of odors and analysis results of odors are provided to the user by displaying them or emitting sounds such as voice.

The storage unit 25 stores odor information that the user wants to measure, sensor identification information, output information, odor analysis results, and the like.

The reception unit 21 corresponds to, for example, an input mechanism. Also, the odor information acquisition unit 22 and the server communication unit 23 correspond to the CPU and the communication I/F, for example. Furthermore, the output unit 24 corresponds to, for example, a display mechanism, and the storage unit 25 corresponds to, for example, storage.

The management server 30 includes a transmission/reception unit 31 that transmits and receives information to and from the outside, a creation unit 32 that creates a list of odors that can be detected by the odor sensor 10, a collection unit 33 that collects detection results from a plurality of users, A learning unit 34 for learning information collected by the collecting unit 33, an analysis unit 35 for analyzing odors, and a storage unit 36 for storing various information are provided.

The transmitter/receiver 31 receives sensor identification information, output information, and user information from the terminal device 20 .
The transmitting/receiving unit 31 is included in the output information and functions as an acquisition unit that acquires information on an odor that one user wants to measure from among multiple odors. In this case, the transmitting/receiving unit 31 acquires, for example, odor information that each of user A, user B, and user C in FIG. 1 wants to measure. The odor information that the user wants to measure is received from the user by the receiving unit 21 of the terminal device 20 .

The transmitter/receiver 31 also transmits a list of odors measurable by the odor sensor 10 to the terminal device 20 based on the sensor identification information included in the output information.
Further, the transmitting/receiving unit 31 transmits the analysis result of the odor by the analyzing unit 35 to the terminal device 20 .

Based on the sensor identification information of the odor sensor 10 , the creation unit 32 creates a list of odors measurable by the odor sensor 10 for the terminal device 20 . The relationship between the odor sensor 10 and measurable odors is stored in the storage unit 36, and the creation unit 32 acquires the types of odors measurable by the odor sensor 10 by referring to the storage unit 36, and creates a list. to create

The collection unit 33 collects detection results from the odor sensor 10 and odor type information, which is information on the types of odors, for a plurality of users. In this case, the "information about the type of smell" is information that is analyzed by the analysis unit 35 and indicates what kind of smell it is. Although details will be described later, the odor type information is information indicating, for example, the type of odor, such as the smell of black coffee, the smell of wine, the smell of soy sauce, the smell of aging, and the like. Also, if the odor sensor 10 has a plurality of sensitive films 120 that output different detection results depending on the type of odor, the collection unit 33 acquires the detection results from these.

The learning unit 34 learns the relationship between the detection results and the odor type information for multiple users collected by the collection unit 33 . That is, the learning unit 34 uses the odor detection results and the odor type information as input data for learning, and machine-learns the relationship between the odor detection results and the odor type information. This input data is also called learning data or teacher data. Then, the learning unit 34 creates a learning model that learns about the relationship between the odor detection result and the odor type information. This learning model is stored in the storage unit 36 .

The analysis unit 35 acquires odor identification results based on differences in reactions with various odor molecules. The odor identification result is odor information that can be identified by the odor sensor 10, and is obtained in advance by performing a test or the like. In other words, it can be said that the odor identification result is a history of odor information identified by the odor sensor 10 in the past.

FIG. 4 is a diagram showing an example of odor identification results.
FIG. 4 shows the result of two-dimensional mapping of the detection result by the odor sensor 10 by principal component analysis as the odor identification result by the odor sensor 10 . Here, the horizontal axis is the first principal component and the vertical axis is the second principal component. In the figure, they are shown as PC (Principal Component) 1 and PC2, respectively.

FIG. 4 shows the results of principal component analysis of the detection results of the odor sensor 10 for the respective odors of pure water, black coffee, wine, and soy sauce. Here, as the odor sensor 10, nose@MEMS manufactured by I-PEX Co., Ltd. was used. This odor sensor 10 has 180 kinds of sensitive films 120 and can identify various odors.
As shown in FIG. 4, pure water, black coffee, wine, and soy sauce have different odor identification results and can be distinguished from each other. In other words, it is possible to distinguish between various odors based on the odor identification result. In addition to the case where the odor can be detected using one sensitive film 120, the odor can also be distinguished by measuring the odor using a plurality of sensitive films 120 and performing the analysis as shown in FIG. included. Therefore, in this case, it is necessary to preliminarily determine items of odors that can be distinguished by conducting tests in advance and assuming that they can be measured according to the characteristics of the odor sensor 10 .

The analysis unit 35 also identifies the type of odor detected by the odor sensor 10 . For example, based on the odor identification result shown in FIG. 4, the odor is identified as pure water, black coffee, wine, or soy sauce.
FIG. 4 shows the case of specifying the smell of pure water, black coffee, wine, and soy sauce, but the present invention is not limited to this. For example, it is possible to identify aging odors, differences in fabric softener fragrances depending on brands, differences in perfume fragrances depending on brands, differences in beer odors depending on brands, differences in odors depending on the freshness of meat, and the like. The identified odor becomes odor type information, which is information about the type of odor.

Furthermore, the analysis unit 35 analyzes the odor level that one user wants to measure based on the relationship between the detection results and odor type information for multiple users learned by the learning unit 34 . Then, the analysis unit 35 outputs the analysis result of the odor.
At this time, the analysis unit 35 refers to the learning model created by the learning unit 34 and stored in the storage unit 36 . Then, using this learning model, the level of the odor that the user wants to measure and that is specified by the analysis unit 35 is analyzed. This "level" is the degree of odor measured by one user among a plurality of users. The level is represented by, for example, a numerical value, a symbol, or the like. Numerical values include, for example, a five-level score from 1 to 5 and a degree of satisfaction indicated by 0% to 100%. Symbols include, for example, symbols such as "⊚", "∘", "Δ", and "x", and symbols such as A, B, C, and D. The odor analysis result can be considered as output data output using the learning model.

The storage unit 36 stores output information, user information, odor type information, odor identification results, learning models, sensor identification information, and the like.

The transmission/reception unit 31 corresponds to, for example, a CPU and a communication I/F. Also, the creating unit 32, the collecting unit 33, the learning unit 34, and the analyzing unit 35 correspond to, for example, a CPU. Furthermore, the memory|storage part 36 respond|corresponds to a storage, for example.

<Description of the operation of the odor measurement system 1>
Next, an example of actual operation of the odor measurement system 1 will be described.
FIG. 5 is a flow chart explaining the operation of the odor measurement system 1. FIG.
First, the user activates application software such as an odor measurement application that operates on the terminal device 20 . Then, the user operates the terminal device 20 to input user information such as the user's gender, nationality, age, place of residence, and user's preferences (step 101).
Next, the server communication unit 23 of the terminal device 20 transmits the sensor identification information of the odor sensor 10 to the management server 30 . In the management server 30, the transmitter/receiver 31 receives the sensor identification information (step 102). The sensor identification information is, for example, the manufacturer and model number of the odor sensor 10 as described above, and may be acquired from the odor sensor 10 or registered in the terminal device 20 in advance.

In the management server 30, the creation unit 32 creates a list of odors measurable by the odor sensor 10 based on the sensor identification information of the odor sensor 10 (step 103). Then, the transmitting/receiving section 31 transmits this list to the terminal device 20 . In the terminal device 20, the server communication section 23 receives the list (step 104).
Furthermore, the output unit 24 presents this list to the user (step 105). The user who sees this list selects an odor to be measured from this list, and the reception unit 21 receives the user's instruction (step 106).

FIG. 6 is a diagram showing an example of the list presented to the user in step 105 of FIG.
Here, the case where the list L1 is displayed on the display mechanism 20A of the terminal device 20 is shown. In this list L1, each odor mode includes (A) body odor check mode, (B) bad breath check mode, (C) room odor check mode, (D) clothes odor transfer check mode, and (E) room drying odor check mode. can be selected, indicating that each odor can be identified. Also shown here is the case where the (A) body odor check mode is selected. In this case, for example, body odor, bad breath, room odor, clothing odor, room drying odor, or the presence or absence of these odors can be detected. Detectable odors include those that can be detected by measuring odors using a plurality of sensitive films 120 and performing analysis as shown in FIG. 4, as described above with reference to FIG. That is, some odors, such as ethanol, are composed of one odor component and can be directly detected by one sensitive film 120, but black coffee, wine, soy sauce, body odor, bad breath, room odor, clothing odor, Odors such as room drying odors are complex odors composed of a plurality of odor components. Such an odor can be detected by analyzing the values detected by the plurality of sensitive films 120 as shown in FIG. In this case, the output unit 24 further analyzes the odors detected by the odor sensor 10 as the odors to be measured by the user, thereby providing a plurality of types of odors determined to be detectable according to the characteristics of the odor sensor 10. It can also be said that the user is made to select from among
When the decision button Bt1 is pressed after the selection, the reception unit 21 receives the selection of the odor to be measured as the user's instruction. It should be noted that the measurement method for measuring the odor selected at this time may be displayed.

Returning to FIG. 5, the user then uses the odor sensor 10 to measure the selected odor to be measured. At this time, the detector 12 of the odor sensor 10 detects odor components in the air (step 107). Then, the odor information creating unit 13 creates output information including the odor detection result (step 108). The output information is sent to the terminal device 20 via the transmitting/receiving section 11, and received by the odor information acquisition section 22 of the terminal device 20 (step 109).

In the terminal device 20 , the server communication unit 23 associates the output information and the user information and transmits them to the management server 30 . In the management server 30, the transmission/reception section 31 receives the output information and the user information (step 110).
In the management server 30, the analysis unit 35 corrects the odor detection result included in the output information based on at least one of temperature and humidity (step 111). That is, the odor detection result of the odor sensor 10 changes according to temperature and humidity. Therefore, correcting this makes it possible to specify the type of odor more accurately.

FIG. 7 is a diagram showing a method of performing correction in step 111 of FIG.
Here, a case where correction is performed using a LUT (Look Up Table) is shown. Then, correction values corresponding to temperature and humidity are shown. For example, "+1" indicates that the detection result is incremented by one. Also, for example, "-1" indicates that the detection result is decremented by one. "0" indicates that the detection result is not corrected. If the temperature and humidity are not among these, the correction values can be determined by interpolation.

Returning to FIG. 5 again, next, the analysis unit 35 uses the corrected detection result to analyze the odor (step 112). At this time, the analysis unit 35 uses the learning model stored in the storage unit 36 to analyze the odor. The analysis unit 35 identifies the type of smell as the analysis of the smell. The analysis unit 35 also analyzes the level of the odor as the odor analysis. The odor level is obtained, for example, as a numerical value or a symbol, as described above.

Furthermore, the analysis unit 35 creates a diagnostic result for the user who performed the measurement (step 113). This diagnostic result is created in the form of sentences and graphs. When the diagnosis result is created as a sentence, the contents are the type of smell and advice to the user. Advice for the user includes, for example, advice for deodorizing odors. Moreover, when the diagnostic result is created as a graph, the content is a bar graph, a scatter diagram, or the like in comparison with other users.

Also, at this time, the analysis unit 35 analyzes the level of the odor, taking user information such as user's preference into account. For example, if the user is a smoker as a user's preference, it is expected that the odor component of cigarettes will become noise in identifying the type of odor. Further, if the user prefers to eat food with a strong smell, it is expected that the smell of the food will become noise. Also, depending on the user's age, it is expected that aging odors will become noise. Therefore, the analysis unit 35 can perform processing for removing such noise according to the user's preference.

Also, the analysis unit 35 can estimate the type of smell from the user's preference. For example, if the user prefers gyoza dumplings, it is expected that the smell of garlic will be present. Also, if the user is an elderly male, it is expected that the odor of aging will be present as a smell. The analysis unit 35 may specify the type of odor based on the odor expected to exist.
As described above, by adding user information, it is possible to improve the accuracy of analysis when analyzing odors.

The analysis unit 35 then transmits the analysis result to the terminal device 20 via the transmission/reception unit 31 . In the terminal device 20, the server communication section 23 receives the analysis result (step 114).
In the terminal device 20, the output unit 24 displays the odor analysis results to the user (step 115).

8A to 8E are diagrams showing analysis results of odors displayed on the terminal device 20. FIG.
In this case, the odor type information, which is information about the type of odor, the level of the odor indicated by the odor type information, and the advice to deodorize the odor indicated by the odor type information are associated and displayed to the user.

FIG. 8(a) is a diagram showing analysis results when "(A) body odor check mode" in FIG. 6 is selected. This analysis result consists of a message Me1 of "(A) aged odor level", "20/100", and "XX body soap is recommended." “(A) Age-old odor level” is odor type information, and indicates that an age-old odor is detected when “(A) Body odor check mode” in FIG. 6 is selected. Also, "20/100" is a score that is the level of the odor that the user wants to measure. In this case, the odor level was 20 points, with 0 being the lowest level and 100 being the highest level. Further, "XX body soap is recommended" is advice to the user, and is advice for deodorizing aging odor.

As shown here, the odor type information is the odor related to the odor that the user wants to measure, but the two do not necessarily match. That is, as shown in FIG. 6, when the "(A) body odor check mode" is selected, the odor analysis result is "(A) aged odor level" as shown in FIG. 8(a). is the case. For example, an analysis result such as "(A) sweat odor level" may be output instead of "(A) aging odor level".

FIG. 8(b) is a diagram showing the analysis result when "(B) halitosis check mode" in FIG. 6 is selected. This analysis result consists of a message Me2 of "(B) garlic odor level", "60/100", and "breath freshener XX is recommended." "(B) Garlic odor level" is odor type information, and indicates that garlic odor was detected when "(B) Halitosis check mode" in FIG. 6 was selected. Also, "60/100" is the same score as in FIG. 8(a), indicating that this was 60 points. Furthermore, "Breathing freshener XX is recommended." is advice for deodorizing garlic odor.

FIG. 8(c) is a diagram showing the analysis result when the "(C) room odor check mode" in FIG. 6 is selected. This analysis result consists of a message Me3 of "(C) Tobacco odor level", "40/100", and "Would you like to consign the work to △△ cleaning?". "(C) Cigarette odor level" indicates that cigarette odor was detected when "(C) Room odor check mode" in FIG. 6 was selected. Also, "40/100" is the same score as in FIG. 8(a), indicating that this was 40 points. Furthermore, "Would you like to entrust the work to △△ cleaning?" is advice for removing the smell of cigarettes.

FIG. 8(d) is a diagram showing analysis results when "(D) check mode for smell transfer to clothes" in FIG. 6 is selected. This analysis result consists of a message Me4 of "(D) Roasted meat smell level", "50/100", and "Let's use □□ spray." "(D) Grilled meat odor level" is odor type information, and indicates that a grilled meat odor is detected when the "(D) clothing smell transfer check mode" in FIG. 6 is selected. Also, "50/100" is the same score as in FIG. Furthermore, "Let's use □□ spray" is advice for removing the smell of grilled meat.

FIG. 8(e) is a diagram showing the analysis results when the "(E) room drying smell check mode" in FIG. 6 is selected. This analysis result consists of a message Me5 of "(E) sweat odor level", "20/100", and "recommend new product ◇◇ detergent". "(E) sweat odor level" is odor type information, and indicates that the sweat odor is detected when the "(E) room drying odor check mode" in FIG. 6 is selected. Also, "20/100" is the same score as in FIG. 8(a), indicating that this was 20 points. Furthermore, "New product ◇◇ detergent is recommended." is advice for deodorizing sweat odor.

After viewing the analysis result, the user inputs an evaluation for the analysis result (step 116). This evaluation is an evaluation of the degree of matching felt by the user with respect to the odor analysis results. In other words, if the user thinks that the analysis result of the odor matches the actual odor more closely, the evaluation will be higher. On the other hand, if it is considered that the smell does not match the actual smell, the evaluation will be low. This evaluation can be input, for example, on a scale of 1 to 5. In this case, "1 point" is the lowest evaluation, and "5 points" is the highest evaluation.

The user's evaluation is received by the reception unit 21, transmitted to the management server 30 by the server communication unit 23, and received by the transmission/reception unit 31 of the management server 30 (step 117).
Then, the collection unit 33 of the management server 30 collects evaluations of the analysis results. Specifically, when the user's evaluation is high, the collection unit 33 collects the relationship between the detection result of the odor sensor 10 and the odor type information, and the learning unit 34 learns these relationships (step 118). . In this case, the learning unit 34 learns when the user's evaluation is high, for example, when the user's evaluation is 4 points and 5 points. In this case, the learning unit 34 excludes cases where the user's evaluation is 1 to 3 points from the learning target. In other words, the learning unit 34 learns for those whose evaluation is greater than the predetermined value, and does not learn about those whose evaluation is less than the predetermined value. In other words, the learning unit 34 selects and learns the one that is highly evaluated by the user and considered to have a more correct relationship between the detection result of the odor sensor 10 and the odor type information.

Further, the learning unit 34 may weight the relationship between the detection result of the odor sensor 10 and the odor type information based on the user's evaluation. That is, the learning unit 34 learns by increasing the weight as the user's evaluation is higher, and decreasing the weight as the user's evaluation is lower.
The learning unit 34 receives feedback from the user by receiving an evaluation of the result analyzed by the analysis unit 35, and corrects the learning model. In this way, the learning unit 34 can improve the learning model by correcting the relationship between the detection result of the odor sensor 10 and the odor type information based on the user's evaluation.

In the form detailed above, the management server 30 learns from a plurality of users the relationship between the detection result of the odor sensor 10 and the odor type information. Then, the management server 30 analyzes the type of odor that one user wants to measure based on the learned relationship. As a result, the management server 30 can improve the accuracy of identifying the types of odors. Also, the management server 30 can improve the analysis accuracy of the odor level.
In addition, the management server 30 can provide more accurate results for the level of odor that one user wants to measure by improving the accuracy of identifying the type of odor and improving the accuracy of analyzing the level, thereby improving the user's convenience. more improved.
Furthermore, the management server 30 corrects the detection result of the odor sensor based on information on the temperature and humidity of the location where the odor that one user wants to measure can be corrected, thereby further improving the accuracy of identifying the type of odor. .
Furthermore, the management server 30 can further acquire an evaluation of the odor level from one user, thereby improving the learning model and further improving the accuracy of the relationship between the detection result of the odor sensor and the odor type information. The management server 30 can further improve the accuracy of identifying the type of odor. In addition, the management server 30 can further improve the analysis accuracy of the odor level.

<Modification>
Next, a modified example of the odor measurement system 1 will be described. In a modified example, the product information providing server 50 to be described below provides product information to the user based on the analysis result by the management server 30 .

FIG. 9 is a diagram showing a configuration example of the odor measurement system 1 in the modified example.
The illustrated odor measurement system 1 differs from the odor measurement system 1 shown in FIG. 1 in that a product information providing server 50 is added, and the rest is the same.
FIG. 10 is a block diagram showing a functional configuration example of the odor measurement system 1 in the modified example.
As illustrated, the functional configuration of the odor measuring system 1 differs from that of the odor measuring system 1 shown in FIG. 3 in that the functional configuration of the product information providing server 50 is added, and the rest is the same.
Therefore, the product information providing server 50 will be mainly described below, and descriptions of the same parts as in the above-described embodiment will be omitted.

The product information providing server 50 is a server device that guides products for deodorizing odors. This odor is the odor that the user wants to measure using the odor sensor 10 and is the odor analyzed by the management server 30 . The management server 30 provides the product information providing server 50 with user information and odor analysis results. Then, the product information providing server 50 guides the product for deodorizing the odor based on the provided information. The product information providing server 50 is, for example, a server device owned or rented by a manufacturer of deodorant products. Note that the number of product information providing servers 50 is not limited to one, and may be plural. For example, the product information providing server 50 may be installed for each manufacturer that handles deodorant products.

As shown in FIG. 10, the product information providing server 50 includes a transmitting/receiving unit 51 for transmitting/receiving information to/from the outside, a product determining unit 52 for determining the product to be introduced, a managing unit 53 for managing the price of the product, a user and a storage unit 54 for storing information and products provided to the user.

The transmitter/receiver 51 communicates with the terminal device 20 and the management server 30 . In this case, the transmission/reception unit 51 receives the user information and the odor analysis result from the management server 30 . In addition, the terminal device 20 is informed of the deodorizing product.
The product determination unit 52 determines a product for deodorizing the odor based on the odor analysis result received from the management server 30 . This product is, for example, an existing product. Specific examples include products such as deodorant sprays, stationary deodorants, laundry detergents, breath fresheners, toothpastes, toothbrushes, and interdental brushes.
However, the product is not limited to this, and may be an odor countermeasure product customized for each user. In other words, this product is made by adjusting the ingredients to deodorize for each user.

The management unit 53 provides payment information to the user and manages payment by the user when the user places an order for the product.

The storage unit 54 associates and manages user information and product order information. This makes it possible to manage the history of product orders made by each user.

According to the modified example, the user can receive guidance on more appropriate products for deodorizing odors from manufacturers of deodorizing products. In addition, the manufacturer can obtain an opportunity to sell the product to users who need deodorant.

<Explanation of the program>
Here, the processing performed by the management server 30 according to the present embodiment described above is prepared as a program such as application software, for example. This processing is realized through the cooperation of software and hardware resources. That is, the CPU inside the computer provided in the management server 30 executes the programs for realizing the functions described above, thereby realizing the functions.

Therefore, in the present embodiment, the processing performed by the management server 30 is to store the detection results of the odor sensor 10 capable of detecting a plurality of odors and the odor type information, which is information on the types of odors, in a computer for a plurality of users. a learning function that learns the relationship between the detection results of multiple users collected by the collection function and the odor type information; and an analysis function for analyzing the odor level that one user wants to measure based on the relationship learned by the learning function.

In the above example, the odor sensor 10 and the terminal device 20 are described as separate devices, but the terminal device 20 containing the odor sensor 10 may be used. Also, the management server 30 and the product information providing server 50 have been described as separate devices, but they can also be integrated as one device.
Furthermore, in the above example, the management server 30 analyzes the odor based on the detection results measured by the odor sensor 10 , but the processing performed by the management server 30 may be performed by the terminal device 20 . In this case, the management server 30 becomes unnecessary.

Also, the odor sensor 10 is not limited to that shown in FIG. For example, instead of the crystal oscillator 110, a CMOS (Complementary Metal Oxide Semiconductor) sensor is used, and a CMOS sensor and a sensitive film are combined. Instead of the crystal oscillator 110, a piezoelectric element ( (Piezo element) is used, and a piezoelectric element and a sensitive film are combined, and a surface stress sensor detects changes in surface stress when odor molecules are adsorbed to the sensitive film. In addition, there are other methods that use concentration of odor molecules by molecular nanowires to detect odor molecules with chemoresistors and FETs (field effect transistors), and biological membranes that mimic human olfactory receptors. Then, a camera captures the change when odor molecules are adsorbed on the biological membrane, and MOS (Metal Oxide Semiconductor) detects odors by using the change in resistance value that occurs when odor molecules come into contact. It may also be a semiconductor sensor, or a combination of a semiconductor sensor and an air quality sensor.

For example, in the above example, the nose@MEMS manufactured by I-PEX Co., Ltd. is used as the odor sensor 10. An odor sensor XP-329IIIR manufactured by New Cosmos Electric Co., Ltd., an MSS olfactory IoT sensor manufactured by the National Institute for Materials Science, and the like can also be used.

Although the present embodiment has been described above, the technical scope of the present invention is not limited to the range described in the above embodiment. It is clear from the scope of claims that various modifications and improvements to the above embodiment are also included in the technical scope of the present invention.

DESCRIPTION OF SYMBOLS 1... Odor measurement system 10... Odor sensor 20... Terminal device 30... Management server 31... Transmission/reception part 32... Creation part 33... Collection part 34... Learning part 35... Analysis part 36... Storage part , 50... Information providing server, 110... Crystal oscillator, 120... Sensitive film, Hc... Oscillation circuit, S... Control device

Claims (13)

a collection unit that collects detection results from odor sensors capable of detecting a plurality of odors and odor type information that is information on the types of odors for a plurality of users;
a learning unit that learns the relationship between the detection results and the odor type information for a plurality of users collected by the collecting unit;
an acquisition unit that acquires information on an odor that one user wants to measure from among a plurality of odors;
an analysis unit that analyzes the odor level that the one user wants to measure based on the relationship learned by the learning unit;
Information processing device. 2. The information processing apparatus according to claim 1, wherein the analysis unit corrects the detection result based on the temperature and/or humidity of a location where the smell desired by the one user is measured. 2. The information processing apparatus according to claim 1, wherein the analysis unit analyzes the odor level that the one user wants to measure by adding user information that is information about the one user. 4. The information processing apparatus according to claim 3, wherein said user information includes information about preferences of said one user. 2. The information processing apparatus according to claim 1, wherein the analysis unit further creates advice for the one user to deodorize the odor that the one user wants to measure. 6. The information processing apparatus according to claim 5, wherein the advice guides a product that deodorizes the smell that the one user wants to measure. The information processing apparatus according to claim 1, wherein the collection unit collects evaluations of results analyzed by the analysis unit from the one user. The information processing apparatus according to claim 7, wherein the learning unit corrects the relationship based on the evaluations collected by the collecting unit. 8. The information processing apparatus according to claim 7, wherein the learning unit learns for those whose evaluations collected by the collecting unit are above a predetermined value, and does not learn about those whose evaluations are less than a predetermined value. The odor sensor has a plurality of sensors that output different detection results depending on the type of odor,
The information processing apparatus according to claim 1, wherein the collection unit acquires the detection results from the plurality of sensors. an odor sensor capable of detecting a plurality of odors;
an information processing device that obtains the level of an odor that one user wants to measure from among a plurality of odors;
a display device for displaying the level of odor;
with
The information processing device is
a collection unit that collects detection results from the odor sensor and odor type information, which is information on types of odors, for a plurality of users;
a learning unit that learns the relationship between the detection results and the odor type information for a plurality of users collected by the collecting unit;
an acquisition unit that acquires information on an odor that the one user wants to measure from among a plurality of odors;
an analysis unit that analyzes the odor level that the one user wants to measure based on the relationship learned by the learning unit;
odor measurement system. 12. The odor according to claim 11, wherein the display device associates the odor type information, the level of the odor indicated by the odor type information, and advice to deodorize the odor indicated by the odor type information, and displays the information to the one user. measurement system. to the computer,
a collection function that collects, for multiple users, the results of detection by an odor sensor capable of detecting multiple odors, and odor type information that is information on the types of odors;
a learning function for learning the relationship between the detection result and the odor type information for a plurality of users collected by the collection function;
an acquisition function for acquiring information on the odor that one user wants to measure from among a plurality of odors; and an analysis function for analyzing the level of the odor that the one user wants to measure based on the relationships learned by the learning function. ,
A program to realize

Images