Information processing system and information processing method

Abstract

To enable estimation of information about an object based on actual odor features of the object.SOLUTION: An information processing system (100) is provided, comprising: an odor measurement device (20) including a sensor chamber (60) having multiple sensor elements (31A) provided therein, and a supply mechanism (40) capable of alternately supplying a first gas and a second gas into the sensor chamber; an acquisition unit (11) for acquiring a detection signal from each of the multiple sensor elements; an extraction unit (12) for extracting feature information indicative of odor features of an object; and an estimation unit (16) for estimating object information about the object.SELECTED DRAWING: Figure 11

Description

[Technical field]

[0001] The present invention relates to an information processing system and an information processing method capable of estimating information about an object based on the odor characteristics of the object. [Background technology]

[0002] Conventionally, a technique for measuring odors is known in which a specific gaseous chemical substance related to the odor is detected and measured from among a group of chemical substances that constitute the odor emitted by a measurement target. For example, in Patent Document 1, the type and amount of gas emitted from soil is measured by an odor sensor. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Patent Publication No. 2021-193894 Summary of the Invention [Problem to be solved by the invention]

[0004] Known odor-causing substances, which are chemical substances that make up odors, include, for example, ammonia, mercaptan, aldehyde, hydrogen sulfide, and amines. Conventional odor measurements use sensors such as an ammonia sensor or a hydrogen sulfide sensor that are designed to measure these odor-causing substances individually.

[0005] However, the odor actually emitted from the measurement target contains a mixture of many unspecified chemical substances. In conventional odor measurements, it is possible to measure the substances contained in the actual odor that are the measurement target of the sensor, but it is not possible to measure the substances contained in the actual odor that are not the measurement target of the sensor. Therefore, when the odor emitted from the measurement target contains a large number of unspecified chemical substances, it cannot be said that conventional odor measurements can fully measure the entire content of the actual odor.

[0006] There is a need to measure the actual odor of a measurement target, which may contain a mixture of an unspecified number of odor-causing substances, and to detect, distinguish, or recognize the odor as a specific odor. Therefore, an object of the present invention is to identify information about an odor-emitting target based on the target's actual odor. [Means for solving the problem]

[0007] An information processing system according to one aspect of the present invention includes at least one odor measuring device including a sensor chamber provided with multiple sensor elements and a supply mechanism capable of alternately supplying a first gas containing an odorant corresponding to the odor of a target and a second gas not containing an odorant corresponding to the odor of the target into the sensor chamber, an acquisition unit that acquires detection signals from each of the multiple sensor elements, an extraction unit that extracts characteristic information indicating characteristics of the odor of the target from the acquired detection signals, and an estimation unit that estimates target information regarding the target based on the extracted characteristic information. Equipped with.

[0008] An information processing method according to one embodiment of the present invention is a control method executed by one or more information processing devices, and includes an acquisition step of acquiring detection signals from each of the multiple sensor elements of at least one or more odor measuring devices having a sensor chamber provided with multiple sensor elements and a supply mechanism capable of alternately supplying a first gas containing an odorant corresponding to the odor of a target and a second gas not containing an odorant corresponding to the odor of the target into the sensor chamber, an extraction step of extracting characteristic information indicating characteristics of the odor of the target from the acquired detection signal, and an estimation step of estimating target information regarding the target based on the extracted characteristic information. Effect of the Invention

[0009] According to one aspect of the present invention, information about an object can be inferred based on the object's actual odor signature. [Brief description of the drawings]

[0010] [Figure 1] 1 is a schematic diagram illustrating an example of a configuration of an information processing system according to an embodiment of the present invention. [Diagram 2] FIG. 2 is a top view showing an example of the configuration of a sensor element. [Diagram 3] 3 is a cross-sectional view showing an example of the configuration of the sensor element shown in FIG. 2. [Figure 4] 1 is a schematic diagram showing an example of the configuration of an odor measurement device according to one embodiment of the present invention. [Diagram 5] FIG. 1 is a schematic diagram showing an example of an overview of an odor measuring device. [Figure 6] FIG. 2 is a top view showing an example of the configuration of a sensor element of the present invention. [Figure 7] FIG. 2 is a top view showing an example of the configuration of a sensor element of the present invention. [Figure 8] FIG. 2 is a perspective view showing an example of the configuration of a sensor element of the present invention. [Figure 9] FIG. 1 is a functional block diagram illustrating an example of a configuration of an information processing system. [Figure 10] 11 is a flowchart showing an example of a process flow in which the estimation device generates an estimation model. [Figure 11] FIG. 1 is a functional block diagram illustrating an example of a configuration of an information processing system. [Figure 12] 13 is a flowchart showing an example of a process flow in which the estimation device estimates target information. [Figure 13] 1 is a graph showing the change over time in detection intensity in measurements using a foam containing triethylamine (TEA) and a foam not containing TEA. [Figure 14] 14 is a graph showing the detection results when measurements were performed multiple times using the same subject as in the measurement shown in FIG. 13. [Figure 15] A diagram showing the correspondence between object information indicating whether an object is acceptable for use as a product, and feature information indicating the detection intensity of a detection signal in a measurement performed using the object. [Figure 16]FIG. 13 is a diagram showing feature quantities based on a sensory evaluation test and measurements performed using an odor measuring device for multiple subjects. [Figure 17] The diagram shows two objects and the features extracted as feature information from the detection signals obtained by measurements performed using each object. [Figure 18] The results of measurements taken on two objects using multiple sensor elements are shown. [Figure 19] 1 is a two-dimensional map showing the results of measurements taken before and after degradation of Additive 1 and Additive 2 as objects. [Figure 20] 13 shows a graph illustrating the relationship between detection signal strength and time when an odor substance in a specific space is measured using an odor sensor having multiple sensor elements. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0011] An embodiment of the present invention will be described below, but the present invention is not limited thereto. In addition, unless otherwise specified in this specification, "A to B" representing a numerical range means "A or more and B or less."

[0012] [Overview of Information Processing System 100] First, an overview of an information processing system 100 according to the present invention will be described with reference to FIG. 1. FIG. 1 is a schematic diagram showing an example of the configuration of an information processing system 100 according to an embodiment of the present invention. As shown in FIG. 1, the information processing system 100 includes an odor measurement device 20 capable of outputting a detection signal based on an odor substance corresponding to the target odor, and an estimation device 10, 10a capable of estimating target information based on the detection signal. Here, the "odor substance corresponding to the target odor" is an "odor substance constituting the target odor" and may include an "odor substance that characterizes the target odor". The information processing system 100 is a system that acquires a detection signal based on an odor substance corresponding to the odor of a target to be estimated, extracts feature information indicating the characteristics of the odor of the target from the detection signal, and estimates target information regarding the target from the feature information.

[0013] Here, characteristic information is, for example, the intensity of a detection signal related to an odor substance and the pattern of change in said intensity over time. Target information is information about a target, and may be, for example, evaluation information about the odor of a target that emits an odor substance, classification information that classifies the odor of the target based on a predetermined criterion, and status information that indicates the state of the target. The "predetermined criterion" may be, for example, the degree of similarity with a reference odor and the content (or content ratio) of a predetermined odor substance. The information processing system 100 can estimate the state of the target by using a detection signal based on the odor substance of the target.

[0014] For example, if the target is food, the target information may be the freshness of the food, the state of preservation, the progress of fermentation, the degree of heating, etc. If the target is a space used for a specific purpose (e.g., a bathroom, a toilet, a dining room, a kitchen, etc.), the target information may be the cleanliness of the space, the necessity for ventilation, etc. If the target is waste liquid and sewage, etc., the target information may be the type of purification treatment required before discharge and the degree of necessity, etc.

[0015] In this specification, "odor substance" means a substance that can be adsorbed to an odor substance receiving layer in a broad sense. Therefore, it also includes substances that are not generally considered to be the cause of odor. "Odor" often contains multiple odor substances that cause it, and there are also substances that are not recognized as odor substances or unknown odor substances. One embodiment of the present invention focuses on the fact that the amount of odor substance adsorbed to the odor substance receiving layer differs depending on the type of odor substance.

[0016] In addition, even when the term "odor substance" is simply used in this specification, it may mean a "collection of odor substances" that may contain multiple odor substances, rather than an individual odor substance.

[0017] In the information processing system 100, the odor measurement device 20 includes an odor sensor 30 capable of outputting a detection signal corresponding to an odor substance, and the odor sensor 30 includes a plurality of sensor elements 31A having a resin composition that reacts to an odor substance. First, the resin composition of each sensor element 31 of the odor sensor 30 will be described below. Then, the details of the information processing system 100 including the sensor element 31 having the resin composition, the odor sensor 30 to which the sensor element 31 is applied, and the odor measurement device 20 including the odor sensor 30 will be described.

[0018] [1. Resin composition] The resin composition according to one embodiment of the present invention is a resin composition for forming an odorant receiving layer 315, and contains a resin (A) and a conductive carbon material (C). The resin composition may further contain a surfactant (B). The resin (A), the surfactant (B) and the conductive carbon material (C) will be described later with specific examples.

[0019] In this specification, "odor substance receiving layer" means a layer that adsorbs the odor substance to be recognized. The odor substance receiving layer 315 is formed from the above-mentioned resin composition. The odor substance receiving layer 315 may be provided as a part of the sensor element 31 described below. The electrical resistance value of this odor substance receiving layer 315 changes depending on the adsorption of the odor substance, etc. In other words, the sensor element 31 is an odor detection device equipped with such an odor substance receiving layer 315, and the odor measurement method of the sensor element 31 is a chemiresistor type.

[0020] <Resin (A)> The resin (A) contained in the resin composition according to one embodiment of the present invention is not particularly limited, and may be a urethane resin, a polyalkylene oxide, an acrylic resin, a fluorine-containing resin, a vinyl polymer resin, a silicone resin, a polyamide resin, a polypropylene resin, paraffin wax, a polyester resin, or the like.

[0021] <Surfactant (B)> The resin composition according to one embodiment of the present invention may contain a surfactant (B) as described below. The surfactant (B) acts as a dispersant for the conductive carbon material (C) described below. The surfactant (B) can be appropriately selected from known surfactants within the range in which the above-mentioned action is exerted.

[0022] Examples of the surfactant (B) include anionic surfactants, cationic surfactants, amphoteric surfactants and nonionic surfactants.

[0023] Examples of the anionic surfactant include alkali metal salts of carboxylic acids having 10 to 24 carbon atoms, alkali metal salts of alkylsulfonic acids having 14 to 24 carbon atoms, and amine salts of polyether acid esters.

[0024] Examples of the carboxylic acid having 10 to 24 carbon atoms include decanoic acid, undecanoic acid, dodecanoic acid, tridecanoic acid, tetradecanoic acid, hexadecanoic acid, heptadecanoic acid, octadecanoic acid, pentadecanoic acid, nonadecanoic acid, icosanoic acid, henicosanoic acid, docosanoic acid, tricosanoic acid, and tetracosanoic acid.

[0025] Examples of the alkyl group contained in the alkylsulfonic acid having 14 to 24 carbon atoms include a tetradecyl group, a pentadecyl group, a hexadecyl group, a heptadecyl group, an octadecyl group, a nonadecyl group, an icosyl group, a henicosyl group, a docosyl group, a tricosyl group, and a tetracosyl group.

[0026] Examples of the alkali metal contained in the alkali metal salt include sodium and potassium.

[0027] Examples of the cationic surfactant include halide salts of quaternary ammonium having an alkyl group having 12 to 24 carbon atoms.

[0028] Examples of the quaternary ammonium having an alkyl group having 12 to 24 carbon atoms include tetrapropylammonium, tetrabutylammonium, tetrapentylammonium, tetrahexylammonium, dimethyldioctylammonium, didecyldimethylammonium, decyltrimethylammonium, dodecyltrimethylammonium, tridecyltrimethylammonium, hexadecyltrimethylammonium, methyltrioctylammonium, octyltrimethylammonium, tributylmethylammonium, octadecyltrimethylammonium, tetradecyltrimethylammonium, nonadecyltrimethylammonium, icosyltrimethylammonium, henicosyltrimethylammonium, heptadecyltrimethylammonium, and pentadecyltrimethylammonium.

[0029] The halide salts include, for example, fluoride salts, chloride salts, bromide salts, and iodide salts.

[0030] Examples of amphoteric surfactants include dimethyl(3-sulfopropyl)ammonium inner salts having an alkyl group with 10 to 22 carbon atoms, and N-alkyl-N,N-dimethylglycines having an alkyl group with 10 to 22 carbon atoms.

[0031] Examples of dimethyl(3-sulfopropyl)ammonium hydroxide inner salts having an alkyl group having 10 to 22 carbon atoms include decyldimethyl(3-sulfopropyl)ammonium hydroxide inner salt, undecyldimethyl(3-sulfopropyl)ammonium hydroxide inner salt, dodecyldimethyl(3-sulfopropyl)ammonium hydroxide inner salt, tridecyldimethyl(3-sulfopropyl)ammonium hydroxide inner salt, tetradecyldimethyl(3-sulfopropyl)ammonium hydroxide inner salt, pentadecyldimethyl(3-sulfopropyl)ammonium hydroxide inner salt, Examples of such ammonium hydroxide inner salt include hexadecyldimethyl(3-sulfopropyl)ammonium hydroxide inner salt, heptadecyldimethyl(3-sulfopropyl)ammonium hydroxide inner salt, octadecyldimethyl(3-sulfopropyl)ammonium hydroxide inner salt, nonadecyldimethyl(3-sulfopropyl)ammonium hydroxide inner salt, icosyldimethyl(3-sulfopropyl)ammonium hydroxide inner salt, henicosyldimethyl(3-sulfopropyl)ammonium hydroxide inner salt, and docosyldimethyl(3-sulfopropyl)ammonium hydroxide inner salt.

[0032] Examples of N-alkyl-N,N-dimethylglycines having an alkyl group having 10 to 22 carbon atoms include N-dodecyl-N,N-dimethylglycine and N-octadecyl-N,N-dimethylglycine.

[0033] Examples of nonionic surfactants include higher alcohol ethylene oxide adducts.

[0034] Examples of higher alcohols include 1-hexyl alcohol, 1-heptyl alcohol, 1-octyl alcohol, 1-nonyl alcohol, 1-decyl alcohol, 1-undecyl alcohol, 1-dodecyl alcohol, 1-tridecyl alcohol, 1-tetradecyl alcohol, 1-pentadecyl alcohol, 1-hexadecyl alcohol, 1-heptadecyl alcohol, and 1-octadecyl alcohol.

[0035] The number of moles of ethylene oxide added is preferably 5 to 50, more preferably 5 to 40, and even more preferably 5 to 30, from the viewpoint of odor discrimination performance.

[0036] The weight ratio of the resin (A) to the surfactant (B) [(A) / (B)] is preferably 1.0 to 50.0 from the viewpoint of odor discrimination performance.

[0037] The resin (A) and the surfactant (B) may or may not be compatible with each other.

[0038] From the viewpoint of dispersibility in the conductive carbon material (C), the surfactant (B) is preferably a nonionic surfactant. From the viewpoint of dispersibility in the conductive carbon material (C), the surfactant (B) preferably has at least one of an amide group, a primary amino group, a secondary amino group, and a tertiary amino group. From the viewpoint of dispersibility in the conductive carbon material (C), the surfactant (B) preferably has at least one of an oxyethylene chain, an oxypropylene chain, and a random structure or a block structure of oxyethylene-oxypropylene. The random structure of oxyethylene-oxypropylene is a chain structure in which both oxyethylene and oxypropylene are irregularly linked. The block structure of oxyethylene-oxypropylene is a chain structure in which an oxyethylene block formed by linking oxyethylenes and an oxypropylene block formed by linking oxypropylenes are linked.

[0039] <Conductive carbon material (C)> The resin composition according to one embodiment of the present invention contains a conductive carbon material (C). In this specification, the conductive carbon material (C) refers to a carbon material having a volume resistivity of 0.1 Ω·cm or less. The above-mentioned resin composition is in a state in which the conductive carbon material (C) is dispersed in the resin (A). The conductive carbon materials (C) come into contact with each other to form a conductive path, which gives the resin composition electrical conductivity.

[0040] Examples of the conductive carbon material (C) include carbon black, carbon nanotubes, and graphene.

[0041] Commercially available carbon black products include Ketjen Black EC (trade name manufactured by Akzo Co., Ltd., Netherlands), Ketjen Black EC-300J (trade name manufactured by Lion Specialty Chemicals Co., Ltd.), Ketjen Black EC-600JD (trade name manufactured by Lion Specialty Chemicals Co., Ltd.), Seast G116, 116 (trade names manufactured by Tokai Carbon Co., Ltd.), Niteron #10 (trade name manufactured by Nippon Steel Chemical Co., Ltd.), Denka Black (trade name manufactured by Denka Co., Ltd.), Toka Black (trade name manufactured by Tokai Carbon Co., Ltd.), and SUPER C-65 (trade name manufactured by MTI Corporation, USA).

[0042] Commercially available carbon nanotubes include VGCF-H (product name manufactured by Showa Denko KK).

[0043] Commercially available graphene products include those manufactured by Sigma-Aldrich.

[0044] The conductive carbon material (C) is preferably in the form of fibers or spheres.

[0045] In the case of a fibrous form, the fiber diameter is preferably 0.1 to 10 μm, more preferably 0.1 to 5 μm, and the fiber length is preferably 0.1 to 10 μm, more preferably 1 to 10 μm.

[0046] When it is spherical, the primary particle size is preferably 10 nm to 200 nm, and more preferably 20 nm to 150 nm.

[0047] In addition, the conductive carbon material preferably has a primary particle diameter of 100 nm or less from the viewpoint of conductivity in the resin composition and sensor sensitivity. The particle diameter of the conductive carbon material can be determined by a known method. For example, the particle diameter of the conductive carbon material can be measured by observing with a transmission electron microscope (TEM) and analyzing the image using an image processing device (for example, a digital microscope VHX-700F manufactured by Keyence). When the conductive carbon material is a known or commercially available product, the particle diameter may be a literature value or a catalog value.

[0048] The content of the conductive carbon material (C) is preferably 10 to 60% by weight, when the total amount of the resin (A) and the conductive carbon material (C) is taken as 100% by weight, from the viewpoint of the sensor element formed from the resin composition exhibiting sufficient conductivity as an odor sensor and sufficient sensitivity as the odor sensor. The content of the carbon black in the odorant receiving layer 315 may more preferably be 10 to 55% by weight, when the total amount of the resin (A) and the conductive carbon material (C) is taken as 100% by weight.

[0049] The resin composition may further contain other components other than the above-mentioned resin (A), surfactant (B) and conductive carbon material (C) within the range where the effects of the present invention can be obtained. Examples of other components include a solvent (D). The other components can be suitably used within the range where both the effects of the present invention and the effects of the other components can be obtained.

[0050] The solvent (D) can be added to the resin composition from the viewpoint of increasing the compatibility between the resin (A) and the conductive carbon material (C), increasing the dispersibility of the surfactant (B) in the resin composition, or increasing the coatability of the resin composition. Examples of the solvent (D) include N-methylpyrrolidone (hereinafter also referred to as NMP), propylene glycol monomethyl ether acetate, ethyl butyrate, butyl butyrate, ethyl acetate, N,N-dimethylformamide, N,N-dimethylacetamide, toluene, and xylene.

[0051] The content of the solvent (D) in the resin composition can be appropriately determined from the above viewpoints. For example, from the viewpoint of coatability, the content of the solvent (D) in the resin composition is preferably 100 to 10,000 parts by weight per 100 parts by weight of the total of the resin (A), the surfactant (B) and the conductive carbon material (C).

[0052] <Method of producing resin composition> A specific example of the method for producing a resin composition according to one embodiment of the present invention is as follows.

[0053] The resin composition is obtained as a slurry by mixing the resin (A) and the conductive carbon material (C) and kneading them uniformly with a stirrer. When mixing the resin (A) and the conductive carbon material (C), a surfactant (B) and a solvent (D) may be added as necessary. As the stirrer, for example, a rotation-revolution mixer is used. As the rotation-revolution mixer, for example, ARE-310 (trade name of Thinky Corporation) or HR003-04A / V (trade name of Samsung Industries Co., Ltd.) can be mentioned. The resin composition is obtained as a slurry having a desired porosity by mixing the resin (A) and the conductive carbon material (C) and kneading them with a stirrer at a rotation speed of, for example, 2000 rpm and a rotation time of about 10 to 60 minutes. When the solvent (D) is added, the solvent (D) is distilled off from the resin composition. The solvent (D) may be distilled off from the resin composition produced by uniform mixing, or may be distilled off from the coating film produced during the production of the sensor element described below.

[0054] [2. Sensor element 31] The above-mentioned resin composition exhibits different changes in electrical conductivity over time when odorant A is adsorbed to the resin composition than when odorant B, which is different from odorant A, is adsorbed to the resin composition. By utilizing this property, a sensor element 31 capable of detecting and identifying odorants can be realized.

[0055] The outline and effects of a sensor element 31 to which a resin composition according to one embodiment of the present invention is applied will be described below.

[0056] The sensor element 31 includes an odorant receiving layer 315 containing the above-mentioned resin composition, a first metal wiring 313A, and a second metal wiring 313B. In the following description, when there is no need to distinguish between the first metal wiring 313A and the second metal wiring 313B, they may be referred to as metal wiring 313.

[0057] Here, the first metal wiring 313A and the second metal wiring 313B will be described with reference to Fig. 2 and Fig. 3. Fig. 2 is a top view showing an example of the configuration of sensor element 31, and Fig. 3 is a cross-sectional view showing an example of the configuration of sensor element 31 shown in Fig. 2.

[0058] The first metal wiring 313A and the second metal wiring 313B are metal wirings that function as electrodes for measuring changes in the electrical conductivity of the odorant receiving layer 315 (i.e., the resin composition). That is, the first metal wiring 313A and the second metal wiring 313B are spaced apart from each other, and the odorant receiving layer 315 is in contact with at least a portion of the first metal wiring and at least a portion of the second metal wiring. In one example, the first metal wiring 313A and the second metal wiring 313B are metal wirings that are not in direct contact with each other, and may be metal wirings that are approximately parallel to each other, as shown in FIG. 2.

[0059] As shown in FIG. 2, the metal wiring 313 including the first metal wiring 313A and the second metal wiring 313B may be disposed on a substrate 311. The substrate 311 may be a substrate such as glass epoxy generally used for electronic circuits. The substrate 311 is not limited to glass epoxy, and may be a substrate of paper phenol, glass composite, polyimide, PET, glass ceramic, alumina, or aluminum. The metal wiring 313 may be a metal wiring such as copper or gold. The thickness of each of the first metal wiring 313A and the second metal wiring 313B as viewed from a direction perpendicular to the surface of the substrate is preferably 10 μm to 2 mm, and more preferably 10 μm to 1 mm. The height, i.e., thickness, of each of the first metal wiring 313A and the second metal wiring 313B as viewed from a direction parallel to the surface of the substrate is preferably 1 μm to 100 μm, and more preferably 10 μm to 50 μm. The distance between first metal wiring 313A and second metal wiring 313B is preferably 1 μm to 3 mm, and more preferably 1 μm to 1.5 mm. The length of metal wiring 313 is preferably 100 μm to 50 mm, and more preferably 500 μm to 30 mm.

[0060] Fig. 3 shows a cross section taken along the line AA in Fig. 2. The odorant receiving layer 315 may be in contact with at least a portion of the first metal wiring 313A and at least a portion of the second metal wiring 313B. The odorant receiving layer 315 may be arranged to fill the area between the first metal wiring 313A and the second metal wiring 313B, as shown in Figs. 2 and 3, for example.

[0061] When the electrical conductivity of the odorant receiving layer 315 (ie, the electrical conductivity of the sensor element 31) is low, it is desirable that the distance between the first metal wiring 313A and the second metal wiring 313B be a predetermined distance (eg, 500 μm) or less.

[0062] The sensor element 31 can detect and distinguish various odorants by applying a resin composition that exhibits different changes in electrical conductivity over time when odorant A is adsorbed and when odorant B, which is different from odorant A, is adsorbed. In the odor sensor 30 described later, a plurality of sensor elements 31 each having a substrate 311 provided with a configuration for detecting an odorant (metal wiring 313 and an odorant receiving layer 315) may be arranged. Each substrate 311 may be provided with a plurality of sets including an odorant receiving layer 315 of the same composition. When a plurality of sensor elements 31 are arranged, a constant voltage power supply and a voltmeter are provided for each sensor element 31. In the odor sensor 30, one configuration for detecting an odorant (metal wiring 313 and an odorant receiving layer 315) may be arranged on each substrate 311. Alternatively, in the odor sensor 30, a plurality of sets of a configuration for detecting an odorant (metal wiring 313 and an odorant receiving layer 315) may be arranged on one substrate 311. In the latter case, a constant voltage power supply and a voltmeter are connected to each of the sets provided on the substrate 311.

[0063] The composition of each of the multiple odorant receiving layers 315 included in the odor sensor 30 may be the same or different. When the odor sensor 30 includes odorant receiving layers 315 with the same composition, the same odorant can be detected in each of the multiple odorant receiving layers 315. When the odor sensor 30 includes odorant receiving layers 315 with different compositions, each of the multiple odorant receiving layers 315 responds differently to the odorant. In this way, by providing multiple sets of configurations for detecting odorants, the accuracy of odorant identification in the odor sensor 30 can be improved.

[0064] [3. Odor Sensor 30] The outline and effects of the odor sensor 30 to which the sensor element 31 is applied will be described below with reference to Fig. 4. Fig. 4 is a block diagram showing an example of the configuration of the odor sensor 30 to which the sensor element 31 is applied.

[0065] The odor sensor 30 includes a sensor element 31 that detects odor substances, a constant-voltage power supply 32 (power supply), and a voltmeter 33 (measuring device).

[0066] The first metal wiring 313A and the second metal wiring 313B of the sensor element 31 are connected by a lead wire W. An example in which a constant voltage power supply 32 and a voltmeter 33 are provided on the lead wire W is shown in FIG.

[0067] The constant voltage power supply 32 is a power supply for supplying power to the sensor element 31. The constant voltage power supply 32 supplies a constant voltage via lead wires to the sensor element 31. The voltage value supplied by the constant voltage power supply 32 is 0.5V to 10V, for example, 2.5V.

[0068] The voltmeter 33 measures the potential difference generated between the first metal wiring 313A and the second metal wiring 313B when a constant voltage supplied from the constant voltage power supply 32 is supplied to the odorant receiving layer 315.

[0069] In addition, the odor sensor 30 includes an amplifier (not shown) in front of the voltmeter 33 in the circuit for measuring odor substances, and the amplifier amplifies the acquired signal and supplies it to the voltmeter 33.

[0070] In addition, the odor sensor 30 is equipped with a reference circuit in addition to the circuit for measuring odorous substances, and the voltmeter 33 acquires the difference (potential difference) between the value acquired in the circuit for measuring odorous substances and the value acquired in the reference circuit as a voltage value.

[0071] The odor sensor 30 may include a constant current source (power supply) (not shown) instead of the constant voltage power supply 32, and an ammeter (measuring device) (not shown) instead of the voltmeter 33. In this case, the constant current source functions as a power supply for supplying power to the sensor element 31, and applies a constant current to the sensor element 31 via a lead wire. Meanwhile, the ammeter measures the value of the current flowing between the first metal wiring 313A and the second metal wiring 313B when a constant current is applied to the odorant receiving layer 315. Both the first metal wiring 313A and the second metal wiring 313B can function as electrodes. Hereinafter, when the metal wiring 313 functions as an electrode, it may be referred to as "electrode 313".

[0072] The odor sensor 30 outputs a measurement value indicating the change over time in the electrical conductivity of the sensor element 31 before and after an odor substance is adsorbed to the sensor element 31. This makes it possible to detect and identify various odor substances. In other words, the sensor element 31 is a chemiresistor type sensor.

[0073] 4. Information Processing System 100 The odor sensor 30 described above can output the change over time in the electrical conductivity of the sensor element 31 for each odor substance when various odor substances are adsorbed to the sensor element 31. By applying this odor sensor 30, it is possible to compare the change over time in the electrical conductivity of the sensor element 31 when odor substance A is adsorbed to the sensor element 31 with the change over time in the electrical conductivity of the sensor element 31 when odor substance B is adsorbed to the sensor element 31. Based on such a comparison result, it is possible to realize an information processing system 100 that can estimate target information regarding a target related to the odor substance adsorbed to the sensor element 31.

[0074] Furthermore, the information processing system 100 can estimate odor substances with high accuracy by using the estimation model 22 generated by machine learning. The estimation model 22 can be generated using learning data including a combination of measurement values ​​measured when each of a plurality of odor substances is adsorbed to at least one sensor element 31, and identification information unique to the odor substance to which the measurement value is applied.

[0075] The following describes the outline and effects of an information processing system 100 including an odor measuring device 20 using an odor sensor 30 and information processing devices 10, 10a. The odor measuring device 20 is a device capable of outputting a detection signal based on an odor substance adsorbed to a sensor element 31 using the above-mentioned resin composition from a change in electrical conductivity that occurs in the sensor element 31. The information processing devices 10, 10a are devices that extract feature information that indicates the features of the odor of a target based on the detection signal, and estimate target information related to the target based on the extracted feature information.

[0076] The odor measuring device 20 of this embodiment is separately provided with a sensor chamber 60 including a plurality of sensor elements 31A (hereinafter also referred to as a "sensor element group 31A"), and a target sample receiving section 50 (sample receiving section) into which a target sample containing an odorant is introduced and into which gas containing the odorant generated from the target sample is enclosed. In this embodiment, each sensor element 31 included in the sensor element group 31A will be simply referred to as a "sensor element 31".

[0077] The odor measuring device 20 of this embodiment employs a configuration in which the gas containing the odorant inside the target sample receiving section 50 is pushed toward the sensor chamber 60 by using a gas other than the gas containing the odorant. In this embodiment, the gas inside the target sample receiving section 50 when the target sample is introduced into the target sample receiving section 50 (i.e., the gas containing the odorant to be detected) is referred to as the first gas. Meanwhile, the gas for pushing the first gas toward the sensor chamber 60 is referred to as the second gas.

[0078] 1, the odor measurement device 20 includes an odor sensor 30, a supply mechanism 40, a target sample receiving unit 50, a sensor chamber 60, and a gas supply unit 80. The odor measurement device 20 may further include an adjustment unit 51. In the information processing system 100, the odor measurement device 20 is communicatively connected to the estimation device 10. The information processing system 100 may further include an estimation device 10a.

[0079] The supply mechanism 40 may replace the first gas in the sensor chamber 60 with the second gas for 0.01 to 10 seconds, and replace the second gas in the sensor chamber 60 with the first gas for 0.01 to 10 seconds. The flow rate when the supply mechanism 40 supplies the first gas and the second gas to the sensor chamber 60 may be set based on the time required to replace the gas in the sensor chamber 60 and the volume of the sensor chamber 60. For example, the flow rate when the supply mechanism 40 supplies the first gas and the second gas to the sensor chamber 60 may be set based on an equation that satisfies (time)=(volume of the sensor chamber 60) / (flow rate). This makes it possible to stably measure the odor substance contained in the first gas while suppressing vibrations that may occur in the sensor element 31 arranged in the sensor chamber 60.

[0080] The gas supply unit 80 is, for example, a pump, and operates under the control of the supply control unit 41 to supply the second gas to the storage space of the target sample receiving unit 50 or the sensor chamber 60. For example, the gas supply unit 80 is connected to the first port 501 of the target sample receiving unit 50, and sends the first gas from the target sample receiving unit 50 toward the sensor chamber 60 by sending the second gas into the inside of the target sample receiving unit 50.

[0081] Gas supply unit 80 may also function as supply control unit 41. Fig. 1 shows, as an example, an example in which gas flows through gas supply unit 80, target sample receiving unit 50, and sensor chamber 60 in that order. Gas supply unit 80, target sample receiving unit 50, and sensor chamber 60 are connected by pipes 91 to 95, respectively.

[0082] The sensor chamber 60 may have one flow path, and all of the sensor elements 31 may be disposed on the flow path. Alternatively, the sensor chamber 60 may have multiple flow paths, and some of the sensor elements 31 may be disposed in each of the multiple flow paths. When the sensor chamber 60 has multiple flow paths, each of the multiple flow paths may be a flow path branched off from one flow path.

[0083] [Supply mechanism 40] The supply mechanism 40 includes a supply control unit 41, a relay 42, and valves 81 to 83, and supplies the second gas to the storage space of the target sample receiving unit 50, thereby pushing the first gas in the storage space from the storage space toward the sensor chamber 60 at a desired flow rate. The supply control unit 41 is, for example, a mass flow controller provided on the first port 501 side of the target sample receiving unit 50, more specifically, between the valve 81 and the gas supply unit 80, and controls the supply of gas from the gas supply unit 80 to the storage space and the sensor chamber 60. By supplying the second gas to the storage space of the target sample receiving unit 50, the first gas in the storage space is pushed from the storage space toward the sensor chamber 60 at a desired flow rate. The relay 42 controls the opening and closing of the valves 81 to 83 provided in the odor measurement device 20, and adjusts the start and stop of the supply of gas in the odor measurement device 20.

[0084] Valves 81 to 83 are, for example, solenoid valves that open and close under the control of relay 42 to control the flow of gas in odor measuring device 20, such as by adjusting the start and stop of gas supply from gas supply unit 80. As shown in FIG. 1, valve 81 is provided between pipes 91 and 92 that connect gas supply unit 80 and target sample receiving unit 50. Valve 82 is provided in pipe 95 that connects gas supply unit 80 and sensor chamber 60. Valve 83 is provided in pipe 93 that connects gas supply unit 80 and target sample receiving unit 50 with sensor chamber 60.

[0085] In this way, the gas supply unit 80 pushes the gas from the first opening 501 side of the target sample receiving section 50 to send the first gas into the sensor chamber 60, so the pressure in the sensor chamber 60 is positive. For this reason, the odor measurement device 20 can obtain stable measurement results. Furthermore, the odor measurement device 20 can send the second gas into the accommodation space by opening and closing the valve 81. Furthermore, the odor measurement device 20 can send the first gas from the target sample receiving section 50 toward the sensor chamber 60 at any timing by opening and closing the valve 83. According to this, when the odor measurement device 20 repeatedly measures the odor substance contained in the first gas using the sensor element group 31A, it can improve the reproducibility of the shape of the waveform output by each sensor element 31. Furthermore, the odor measurement device 20 can directly supply the second gas to the sensor chamber 60 by opening and closing the valve 82. This allows the first gas to be quickly removed from the sensor chamber 60, and preparations for the next measurement can be made.

[0086] The second gas may be an inert gas or air. Examples of the inert gas include argon and nitrogen. When the second gas is an inert gas, the gas supply unit 80 may be a gas cylinder.

[0087] Furthermore, when the second gas is air, the gas supply unit 80 may be a pump. In this case, in order to remove components that react with the first gas contained in the target sample receiving unit 50, the odor measuring device 20 may be provided with, for example, an activated carbon filter, a dehumidifier, a silica gel column, or a dust filter on the first port 501 side of the target sample receiving unit 50.

[0088] 1 shows a configuration in which the first gas and the second gas leaving the inside of the target sample receiving part 50 pass through the tube 93 and the sensor chamber 60, but the configuration is not limited to this. Since the target sample receiving part 50 of the odor measurement device 20 has a larger volume than the sensor chamber 60, it is not necessary to send the entire amount of the first gas inside the target sample receiving part 50 to the sensor chamber 60 during measurement. Furthermore, when the inside of the target sample receiving part 50 is purged with the second gas after the measurement, it is not necessary for the target sample receiving part 50 and the sensor chamber 60 to be connected. Therefore, the odor measurement device 20 may be configured such that a valve (not shown) is arranged in the tube 93 so that the first gas and the second gas leaving the inside of the target sample receiving part 50 can be exhausted without passing through the sensor chamber 60.

[0089] [Target sample receiving section 50] The target sample receiving section 50 can receive a target sample containing an odorant and hold a first gas. The target sample receiving section 50 has a first port 501 through which the second gas entering the inside passes, and a second port 502 through which the first gas and the second gas leaving the inside can pass. In FIG. 1, the first port 501 is installed at the upper part of the target sample receiving section 50 and the second port 502 is installed at the lower part of the target sample receiving section 50, but is not limited to this. For example, the positions of the first port 501 and the second port 502 can be appropriately set depending on the type and combination of odor components contained in the first gas. For example, the positions of the first port 501 and the second port 502 may be changed depending on whether the weight per unit volume of the odor components contained in the first gas (i.e., specific gravity) is heavier or lighter than that of the second gas. Furthermore, the target sample receiving section 50 may include an airflow generating fan 35 therein, as in FIG.

[0090] The target sample receiving section 50 includes a sample inlet 503 for receiving a liquid or solid target sample. The target sample receiving section 50 may include a placement section (not shown) for placing the target sample. When the target sample is liquid, the placement section may be a cup for holding the liquid, and when the target sample is solid, the placement section may be a petri dish on which the solid is placed. The target sample may be introduced into the target sample receiving section 50 in a gaseous state as a first gas from the sample inlet 503. In this way, since the target sample receiving section 50 can receive a liquid or solid target sample, it is possible to adjust the concentration of the odorant in the first gas. For example, even if the odorant is the same, it is easy to adjust the concentration of the odorant in the first gas.

[0091] A material inactive to odorants may be disposed on the inner surface of the target sample receiving portion 50. A material inactive to odorants is a material that does not significantly change the concentration of each odorant contained in the gas sent to the sensor chamber 60. For example, a material inactive to odorants is a material in which odorants are unlikely to be adsorbed or dissolved. Examples of materials inactive to odorants include glass, metal, and resin. When metal is used, stainless steel (SUS) is preferable, and when resin is used, fluorine-based resin, polypropylene (PP), polyethylene (PE), ABS resin, and polyethylene terephthalate (PET) are preferable.

[0092] If the inner surface of target sample receiving portion 50 is made of a material that adsorbs odorous substances contained in the first gas, the odorous substances may be adsorbed to various portions, which may affect subsequent measurements.

[0093] In this way, the inner surface of the target sample receiving portion 50 is made of a material that is inactive against odorants, thereby reducing the risk that the material of the inner surface will react with the odorants contained in the first gas, or that the odorants will be adsorbed onto the inner surface. This reduces the risk that the odorants contained in the first gas supplied to the sensor chamber 60 will change while contained in the target sample receiving portion 50, or that the concentration of the odorants will be diluted.

[0094] Even if the target sample is a liquid or solid, the odor measurement device 20 is provided with the target sample receiving section 50, thereby making the concentration of the first gas uniform in the target sample receiving section 50 before the first gas is sent to the sensor chamber 60. Furthermore, by providing the target sample receiving section 50, the odor measurement device 20 is able to push the first gas into the sensor chamber 60 at a constant flow rate. As a result, even when measurements are repeated, the odor measurement device 20 can send the first gas to the sensor chamber 60 under the same conditions each time, thereby enabling repeated and stable measurements to be performed.

[0095] The volume of the target sample receiving section 50 is preferably 1 to 200 times the volume of the sensor chamber 60. In particular, the volume of the target sample receiving section 50 is preferably larger than the volume of the sensor chamber 60. The volume of the target sample receiving section 50 is more preferably 2 or more times the volume of the sensor chamber 60, and more preferably 4 or more times. In addition, the volume of the target sample receiving section 50 is preferably 100 or less times the volume of the sensor chamber 60, and more preferably 60 or less times. By having the volume of the target sample receiving section 50 be 1 or more times the volume of the sensor chamber 60, the concentration of the odor substance in the sensor chamber 60 is appropriately adjusted, and the measurement result by the sensor provided in the sensor chamber 60 is stably output. In addition, by having the volume of the target sample receiving section 50 be 200 or less times the volume of the sensor chamber 60, it is easy to adjust the temperature and humidity in the target sample receiving section 50, so that the measurement result by the sensor is stably output and the size of the odor measurement device 20 can be kept compact.

[0096] If the volume of the target sample receiving section 50 is less than one time the volume of the sensor chamber 60, the odor substances generated in the target sample receiving section 50 may be diluted in the sensor chamber 60, and the measurement sensitivity of the sensor may decrease. If the volume of the target sample receiving section 50 is more than 200 times the volume of the sensor chamber 60, the volume of the target sample receiving section 50 is too large, so the uniformity of the concentration, temperature, and humidity of the first gas may decrease, and when measurements are repeated, the first gas may not be sent to the sensor chamber 60 under the same conditions. Furthermore, the overall size of the odor measurement device 20 may become large.

[0097] FIG. 1 shows an example in which the volume of the target sample receiving portion 50 is eight times the volume of the sensor chamber 60 .

[0098] For example, if the inner surface of the tube 93 is made of a material that adsorbs odorous substances contained in the first gas, the odorous substances may be adsorbed to various parts, which may affect the subsequent measurement. Therefore, it is preferable that a material that is inactive to odorous substances is arranged on the inner surface of the tube 93 that guides the first gas from the target sample receiving section 50 to the sensor chamber 60, similar to the inner surface of the target sample receiving section 50. Examples of materials that are inactive to odorous substances include glass, metal, and resin. When metal is used, stainless steel (SUS) is preferable, and when resin is used, fluorine-based resin, polypropylene (PP), polyethylene (PE), ABS resin, and polyethylene terephthalate (PET) are preferable.

[0099] The target sample receiving section 50 may be configured to be detachable from the tubes 92 and 93. In this manner, since the target sample receiving section 50 is detachable, when the previous measurement is completed and the next measurement is to be performed, a new target sample receiving section 50 can be attached without purging the inside of the target sample receiving section 50. This allows the odor measuring device 20 to perform multiple measurements in a short period of time.

[0100] Furthermore, since the target sample receiving section 50 is detachable, the target sample receiving section 50 into which the target sample has been introduced can be maintained at a desired temperature using a heat retention chamber separate from the odor measurement device 20. This allows the odor measurement device 20 to adjust the temperature of the target sample receiving section 50, even if the odor measurement device 20 cannot be provided with an adjustment section 51, which will be described later. As the heat retention chamber, for example, a water bath, a dry bath, a heat jacket, a silicon heater, a forward air dryer, a thermo-hygrostat, a sprayer, or the like may be used.

[0101] [Adjustment section 51] The adjustment unit 51 is a temperature adjustment mechanism capable of adjusting the temperature of the accommodation space. The adjustment unit 51 adjusts at least one of the temperature and humidity of the first gas contained in the target sample receiving unit 50. When the adjustment unit 51 adjusts the temperature, the adjustment unit 51 is, for example, a heater or a cooler. In this case, the adjustment unit 51 may be configured to cover the entire target sample receiving unit 50. When the adjustment unit 51 adjusts the humidity, the adjustment unit 51 is, for example, a humidifier or a dehumidifier. The adjustment unit 51 may adjust at least one of the temperature and humidity for each type of first gas, or may change at least one of the temperature and humidity at predetermined time intervals during measurement of the same first gas. As the adjustment unit 51, for example, a water bath, a dry bath, a heat jacket, a silicon heater, a wind dryer, a thermohygrostat, a sprayer, or the like capable of changing at least one of the temperature and humidity may be used.

[0102] By the adjustment section 51 adjusting at least one of the temperature and humidity of the first gas in the target sample receiving section 50, the odor measuring device 20 can send the first gas to the sensor chamber 60 using conditions according to, for example, the type of the first gas (weight of the gas, volatility, etc.). Furthermore, this allows the odor measuring device 20 to send the first gas at a stable concentration to the sensor chamber 60, improving the measurement accuracy.

[0103] [Sensor Chamber 60] The sensor chamber 60 is a space for storing the sensor element 31 for measuring odor substances. The sensor chamber 60 is connected to the second port 502 of the target sample receiving portion 50. Specifically, the sensor chamber 60 includes a gas supply port 601 and a gas exhaust port 602, and the second port 502 of the target sample receiving portion 50 is connected to the gas supply port 601.

[0104] The sensor chamber 60 includes a plurality of sensor elements 31A capable of outputting a measurement result according to an odor substance contained in the first gas. The plurality of sensor elements 31A may be sensor elements 31 each having a different resin composition as a substance receiving layer. That is, the plurality of sensor elements 31A may have different sensitivities and specificities to the odor substance. The sensor chamber 60 in FIG. 1 includes, as an example, a sensor element 31 and a sensor element 31b, but is not limited thereto. The sensor chamber 60 in FIG. 1 also includes a sensor element 31c having a resin composition different from the sensor elements 31 and 31b as a substance receiving layer. The sensor element 31 and the sensor element 31b are capable of outputting a measurement result according to the same odor substance contained in the first gas, but the measurement results output by each of them are different. The measurement result according to the odor substance is, for example, a measurement result according to the concentration of the odor substance. In the following description, when there is no particular distinction between the sensor elements 31, 31b, 31c and the sensor element 31d described later, they are collectively referred to as "sensor element 31".

[0105] The sensor elements 31 each having a different resin composition as a substance receiving layer may be installed in any combination and arrangement in the sensor chamber 60. Furthermore, the sensor chamber 60 may be installed with a plurality of sensor elements 31 each having the same resin composition as a substance receiving layer.

[0106] Here, the sensor element 31 and the sensor element 31b may be sensor elements capable of outputting measurement results corresponding to different odor substances. For example, the sensor chamber 60 may be provided with a sensor element 31 capable of outputting measurement results corresponding to an odor substance contained in a first gas, and a sensor element 31b capable of outputting measurement results corresponding to a second odor substance different from the odor substance contained in the first gas. For example, the odor sensor 30 may include sensor elements 31 and 31b having different resin compositions used in the odor substance receiving layer 315.

[0107] By providing multiple sensor elements 31 in which resin compositions with different odorant adsorption properties are used in the odorant receiving layer 315, the information processing system 100 can simultaneously perform estimations for multiple odorants. Note that as a sensor element 31 according to one embodiment of the present invention, a sensor element 31 that does not contain a surfactant (B) in the odorant receiving layer 315 and a sensor element 31 that contains a surfactant (B) in the odorant receiving layer 315 may be used in combination.

[0108] In addition, by using the odor measuring device 20, it is possible to obtain a first change pattern indicating a change in the electrical conductivity of the sensor element 31 and a second change pattern indicating a change in the electrical conductivity of the sensor element 31b for each known odor substance. The estimation model 22 may be generated by machine learning using both the first change pattern and the second change pattern. The information processing system 100 estimates the odor substance using the estimation model 22 generated in this way, and therefore can more precisely identify each odor substance.

[0109] 1 includes a plurality of sensor elements 31A arranged in a 4×4 array, as an example. The number of sensor elements 31A and the manner in which the sensor elements 31A are arranged are not limited. The total number of sensor elements 31A included in the sensor chamber 60 is also not particularly limited, and may be, for example, 2, 8, 16, or 64. The total number of sensor elements 31A may be 8 or more and 16 or less.

[0110] The material of the inner surface of the sensor chamber 60 is preferably a material that is inactive to odorants, similar to the target sample receiving section 50. Examples of materials that are inactive to odorants include glass, metal, and resin. When metal is used, stainless steel (SUS) is preferable, and when resin is used, fluororesin, polypropylene (PP), polyethylene (PE), ABS resin, and polyethylene terephthalate (PET) are preferable. When the material of the inner surface of the sensor chamber 60 is a material that adsorbs odorants contained in the first gas, the odorants adsorbed in the sensor chamber may reduce the amount of change in output from the sensor element 31 in a subsequent measurement, and the odor measuring device 20 may not be able to perform an accurate measurement.

[0111] The sensor chamber 60 is detachably attached to the odor measurement device 20. Here, the odor measurement device 20 may include a first housing 70 that covers at least a part of the supply mechanism 40. Here, at least a part of the supply mechanism 40 may include, for example, a target sample receiving section 50. By covering the target sample receiving section 50 with the first housing 70, it may be possible to easily adjust the temperature of the storage space.

[0112] FIG. 5 is a schematic diagram showing an example of an overall view of the odor measurement device 20. Alternatively, as shown in FIG. 5, the odor measurement device 20 may include at least a part of the supply mechanism 40 and a second housing 71 that covers the sensor chamber 60. In FIG. 5, the supply mechanism 40 is omitted from illustration for simplification. For example, by covering the sensor chamber 60 with the second housing 72, the temperature of the space in the sensor chamber 60 can be stabilized, and the reliability of the measurement result can be improved. The second housing 71 may include an openable and closable lid 72. When the second housing 71 is in an open state, the sensor chamber 60 may be configured to be detachably attached to the odor measurement device 20. This allows the sensor chamber 60 to be replaced without removing or moving the odor measurement device 20.

[0113] [Multiple sensor elements 31A (sensor element group 31A)] The sensor elements 31A may include a thin film. As an example, the odorant receiving layer 315 in FIG. 2 and FIG. 3 is a thin film.

[0114] As an embodiment for sending the first gas containing an odorant into the sensor chamber 60, for example, a vacuum pump may be installed on the gas exhaust port 602 side of the sensor chamber 60, and the odorant may be sent into the sensor chamber 60 from the gas supply port 601 side of the sensor chamber 60 by drawing the gas using the vacuum pump. However, if the sensor elements 31 and 31b have a thin film, the thin film may expand if the sensor chamber 60 is under negative pressure, and the sensor elements 31 and 31b may not be able to output a stable measurement result. In contrast, if the gas supply unit 80 pushes the gas from the sensor chamber 60 and the first port 501 side of the target sample receiving unit 50 to send the first gas into the sensor chamber 60, the pressure in the sensor chamber 60 is positive. Therefore, it is preferable that the odor measurement device 20 according to this embodiment adopts a configuration in which the gas supply unit 80 pushes the gas from the sensor chamber 60 and the first port 501 side of the target sample receiving unit 50 to send the first gas into the sensor chamber 60. This allows the odor measuring device 20 to obtain stable measurement results even if multiple sensor elements 31A are equipped with thin films.

[0115] The thin film of the plurality of sensor elements 31A may contain a conductive carbon material, a resin composition, and a surfactant.

[0116] In addition, each of the multiple sensor elements 31A has an odorant receiving layer 315 containing a resin composition including resin and filler, and the resin compositions contained in the odorant receiving layer 315 of each of the multiple sensor elements 31A may be different from each other.

[0117] <Configuration example of sensor element 31c> Hereinafter, an example of the configuration of one sensor element 31c included in the sensor element group 31A will be described with reference to Figures 6 to 8. In Figures 6 to 8, the dotted and hatched areas indicate the odorant receiving layers 315c to 315d. In one sensor element 31c included in the sensor element group 31A, the surface roughness (Sa) of the odorant receiving layers 315c to j between the opposing electrodes is 0.5 μm to 5 μm, more preferably 0.5 μm to 3 μm. In Figures 6 to 8, the hatched areas are the areas of the odorant receiving layers 315c to 315d where the surface roughness is measured. In the following description, when there is no particular distinction between the odorant receiving layers 315c and 315d, they will be collectively referred to as the "odorant receiving layer 315."

[0118] FIG. 6 is a top view showing an example of the configuration of one sensor element 31c included in the sensor element group 31A. The sensor element 31c includes an electrode 313 (first metal wiring 313C, second metal wiring 313D) arranged on a substrate 311, and a circular odorant receiving layer 315c formed on the electrode 313. As shown in FIG. 6, the first metal wiring 313C and the second metal wiring 313D are arranged in parallel straight lines. Specifically, the first metal wiring 313C includes metal wiring 313a and metal wiring 313b arranged in a T-shape so as to be perpendicular to each other. The second metal wiring 313D includes metal wiring 313c and metal wiring 313d arranged in a T-shape so as to be perpendicular to each other. Furthermore, the metal wiring 313a and the metal wiring 313c are arranged so as to be located on parallel lines. The diameter R of the odorant receiving layer 315c is 0.2 mm or more and 5 mm or less. In FIG. 6, the shape of the odorant receiving layer 315c of the sensor element 31c is an ellipse, but is not limited thereto. When the shape of the odorant receiving layer 315c is an ellipse, the average of the minor axis and the major axis may be 0.2 mm or more and 5 mm or less. The shape of the odorant receiving layer 315c may also be a perfect circle.

[0119] 7 is a top view showing an example of the configuration of one sensor element 31d included in the sensor element group 31A. The sensor element 31d comprises an electrode 313 (first metal wiring 313C, second metal wiring 313D) arranged on the substrate 311, and a strip-shaped odorant receiving layer 315d formed on the electrode 313. The length of the width in the short direction of the odorant receiving layer 315d is 0.2 mm or more and 5 mm or less.

[0120] 8 is a perspective view showing an example of the configuration of one sensor element 31c included in the sensor element group 31A. As shown in FIG. 8, in the sensor element 31c, the first metal wiring 313C and the second metal wiring 313D are connected to pins 316 at the ends where the first metal wiring 313C and the second metal wiring 313D do not face each other. The pins 316 are conductive members for electrically connecting the first metal wiring 313C and the second metal wiring 313D to other members of the odor sensor 30. Although not shown, the sensor elements 31c and 31d shown in FIG. 6 and FIG. 7 also include the pins 316 shown in FIG. 8.

[0121] The electrodes 313 of the sensor elements 31 included in the sensor element group 31A each have a first metal wiring 313C and a second metal wiring 313D, and the first metal wiring 313C and the second metal wiring 313D may be arranged in the form of parallel lines, parallel curves, a comb shape, or concentric circles. Regardless of which shape is adopted, the first metal wiring 313C and the second metal wiring 313D are preferably arranged in line symmetry or point symmetry with each other. By arranging the first metal wiring 313C and the second metal wiring 313D in this way, the odor measuring device 20 can measure odor substances contained in the gas with high accuracy.

[0122] The sensor element 31c in Fig. 6 has a first metal wiring 313C and a second metal wiring 313D. As an example, the first metal wiring 313C is composed of a metal wiring 313a and a metal wiring 313b, and the two metal wirings are arranged in a T-shape so as to be perpendicular to each other. The second metal wiring 313B is also composed of two metal wirings 313c and 313d like the first metal wiring 313C, and is configured so as to be arranged in a T-shape so as to be perpendicular to each other. The first metal wiring 313C and the second metal wiring 313D are arranged in parallel lines so that the metal wiring 313a and the metal wiring 313c face each other.

[0123] The first metal wiring 313C and the second metal wiring 313D are arranged in a T-shape, so that the first metal wiring 313C and the second metal wiring 313D can be arranged at a suitable distance, and the resistance value of the electrode can be stabilized. For example, if the electrodes are arranged in a comb shape, the distance between the electrodes becomes short, and the resistance value of the electrode may become too small. In addition, since the first metal wiring 313C and the second metal wiring 313D are arranged in a T-shape, when manufacturing the sensor element, in the process of applying a slurry to the substrate to form the odorant receiving layer 315, there is no uneven part of the electrode that may hinder the slurry from spreading in the area where the slurry spreads, so the slurry spreads easily. In addition, since the slurry spreads easily, there is an effect that the thickness of the odorant receiving layer 315 after drying becomes constant.

[0124] [Estimation device 10] The estimation device 10 is a device that estimates an odor substance detected by the odor sensor 30. The estimation device 10 is, for example, a computer, and includes a CPU and a memory (not shown). The estimation device 10 is communicably connected to the odor measurement device 20. Specifically, the estimation device 10 performs estimation of an odor substance corresponding to a target odor by analyzing a detection signal acquired from the odor sensor 30. When the sensor chamber 60 further includes a sensor element 31c in which a resin composition different from the sensor element 31 and the sensor element 31b is used in the substance receiving layer 315, the estimation device 10 may further acquire and analyze a measurement value measured by a voltmeter by supplying a constant voltage to the sensor element 31c. In addition, the estimation device 10 may display the measurement value itself based on the detection signal, a waveform in which the measurement value is plotted, and an estimation result of an unknown odor substance based on an estimation model. In addition, the estimation device 10 may display a numerical value and a graph showing a change in the presence ratio of each odor substance for a gas containing multiple odor substances. The estimation device 10 may have a function of generating an estimation model 22 used to estimate an odor substance, but is not limited to this configuration. For example, the estimation device 10 may be configured to use the estimation model 22 generated by a computer other than the estimation device 10.

[0125] <Generation of Estimation Model 22> Next, the configuration of the information processing system 100 that performs the process of generating the estimation model 22 used to estimate odor substances, and the process of generating the estimation model 22 will be described with reference to FIGS. 9 and 10. FIG.

[0126] The estimation model 22 is generated by machine learning using learning data including a combination of measurement values ​​measured by the voltmeter 33 when each of a plurality of odor substances is adsorbed to at least one sensor element 31, and identification information unique to the odor substance that gave the measurement value. Here, the identification information unique to the odor substance may be, for example, the name, CAS number, and chemical formula of the odor substance.

[0127] (Configuration of the Estimation Device 10 (Generation of Estimation Model 22)) Fig. 9 is a functional block diagram showing an example of the configuration of the information processing system 100. For ease of explanation, the same reference numerals are given to components having the same functions as those described in Fig. 1, and the description thereof will not be repeated.

[0128] As shown in FIG. 9, the estimation device 10 includes an input unit 15, a control unit 1, and a storage unit 2.

[0129] The input unit 15 is for accepting various input operations from a user, and may be, for example, a keyboard, a mouse, a touch panel, or the like.

[0130] The control unit 1 includes a measurement value acquiring unit 11 (acquiring unit), an extracting unit 12, a learning control unit 13, and an estimation model generating unit .

[0131] The measurement value acquiring unit 11 acquires a detection signal from each of the multiple sensor elements 31A. Specifically, the measurement value acquiring unit 11 acquires a measurement value from a voltmeter 33. The measurement value acquiring unit 11 also uses the acquired measurement value to calculate a value indicating the electrical conductivity of the sensor element 31 (e.g., resistance value, impedance, etc.). The measurement value acquiring unit 11 may acquire the measurement value from the voltmeter 33 at a predetermined time interval (e.g., 0.1 second interval).

[0132] The extraction unit 12 extracts characteristic information indicating the characteristics of the target odor from the detection signal acquired by the measurement value acquisition unit 11.

[0133] The characteristic information includes at least one of the intensity of the detection signal output from each of the multiple sensor elements 31A during a first period in which the first gas is supplied into the sensor chamber 60 and the pattern of change over time of the intensity. The characteristic information also includes at least one of the intensity of the detection signal output from each of the multiple sensor elements 31A during a second period in which the first gas supplied into the sensor chamber 60 is pushed out of the sensor chamber 60 by the second gas. When an odor substance acts on the sensor element 31 (adsorption / desorption) over time, the electrical characteristic (resistance) of the sensor element 31 changes. The extraction unit 12 extracts a waveform based on the temporal change in the electrical characteristic, or a characteristic numerical value included in the waveform, as the characteristic information.

[0134] The extraction unit 12 extracts, for example, a change over time in the electrical conductivity of at least one sensor element 31. The extraction unit 12 uses the resistance value calculated by the measurement value acquisition unit 11 to calculate a value indicating the amount of change in the electrical conductivity of the sensor element 31 due to the adsorption of an odorant. The extraction unit 12 generates data indicating a change pattern indicating the time change in the calculated amount of change in electrical conductivity. If the generated change pattern is the same as the change pattern when a known odorant is detected, the extraction unit 12 may associate the generated change pattern with identification information unique to the known odorant and store it in the change pattern database 21 (learning data).

[0135] The learning control unit 13 reads out the change pattern database 21 from the storage unit 2 and controls the generation of the estimation model 22 by machine learning. Here, the change pattern database 21 is a database that includes combinations of measurement values ​​measured when multiple odor substances are adsorbed to the sensor element 31 and identification information unique to the known odor substances that gave the measurement values. The learning control unit 13 inputs the change pattern read out from the change pattern database 21 to the estimation model generation unit 14. In addition, the learning control unit 13 compares the odor substance corresponding to the change pattern input to the estimation model generation unit 14 with the estimation result output from the estimation model generation unit 14, and outputs a correction instruction according to the comparison result to the estimation model generation unit 14.

[0136] The estimation model generation unit 14 generates the estimation model 22 by a machine learning algorithm using the change patterns stored in the change pattern database 21. The estimation model generation unit 14 may be configured to generate the estimation model 22 by using a known supervised machine learning algorithm. Examples of machine learning algorithms that can be applied to the estimation model generation unit 14 include the k-nearest neighbor method, logistic regression, support vector machine, random forest, and neural network.

[0137] Feature extraction method When generating an estimation model by machine learning, the data on which the model is based may be the measurement value itself, or may be data extracted from the feature of the measurement value, such as a statistical quantity of the measurement value, a differential integral value, a peak detection value, or an autocorrelation value. Examples of the statistical quantity include an average value, a variance value, a maximum value, a minimum value, a difference between a maximum value and a minimum value, and a standard deviation. Examples of the differential integral include a slope and an area. Examples of the peak detection value include the number of peaks and the height. Examples of the autocorrelation value include a step difference.

[0138] Pretreatment method When generating an estimation model by machine learning, the base data may be used for machine learning as it is, or the base data may be pre-processed as necessary. When pre-processing is performed, the pre-processing may be performed before feature extraction, after feature extraction, or both before and after feature extraction. The pre-processing may be performed by a known method, and may include correction, noise removal, standardization, data conversion, smoothing, data expansion, and other methods. Correction may be performed based on the measurement results of multiple elements, commercially available sensors (for example, temperature sensors, humidity sensors, and the like), and standard gases. For example, correction may be performed by output ratio calculation, independent component analysis (ICA), statistical calculation, integration, addition, subtraction, division, and other methods. Examples of noise removal include removal of outliers and white noise. Examples of standardization include normalization and regularization. Examples of data transformation include trend removal, frequency conversion, logarithmic conversion, and the like. Examples of smoothing include moving average, difference, and the like. Examples of data augmentation include adding the same sample data (for example, assuming a normal distribution) and adding new sample data (for example, adding using a vector mixture ratio).

[0139] Machine learning algorithms Machine learning algorithms applicable to the estimation model generation unit 14 include regression analysis, classification, tree, time series analysis, neural network, clustering, etc. Examples of regression analysis include logistic regression, Lasso, elastic net, support vector regression (SVR), linear, Ridge, ensemble regression, etc. Examples of classification include k-nearest neighbor method, support vector classification (SVC), Naibe Bayes classifier, stochastic gradient descent (SGD), kernel approximation, etc. Examples of trees include decision trees, regression trees, random forests, boosting (lightGBM, XGboost), stacking, etc. Examples of time series include AR, MA, ARIMA, state space, etc. Examples of neural networks include multi-layer perceptrons (MLPs), convolutional neural networks (CNNs), recurrent neural networks (RNNs), residual neural networks (ResNets), transformers, graph neural networks (GNNs), etc. Examples of clustering include Gaussian mixture models (GMMs), k-means methods (kmeas), minikmeans, variational Gaussian mixture models (VBGMMs), kernel approximations, etc.

[0140] (Process for generating estimation model 22) The process of generating the estimation model 22 using the odor measurement device 20 will be described below with reference to Fig. 10. Fig. 10 is a flow chart showing an example of the process flow in which the estimation device 10 of the odor measurement device 20 generates the estimation model 22. The estimation model 22 is generated by machine learning using learning data including a combination of measurement values ​​measured by the voltmeter 33 when each of a plurality of odor substances is adsorbed to at least one sensor element 31, and identification information unique to the odor substance that gave the measurement value. Here, the identification information unique to the odor substance may be, for example, the name, CAS number, and chemical formula of the odor substance.

[0141] First, the measurement value acquiring unit 11 acquires the voltage value V0 measured by the odor sensor 30 before the odor substance is introduced into the target sample receiving unit 50, and calculates the resistance value R0. The resistance value R0 is preferably 500 to 3000 Ω, more preferably 800 to 2800 Ω, and most preferably 1000 to 2500 Ω. Then, the odor substance is introduced into the target sample receiving unit 50 (step S1).

[0142] Meanwhile, the input unit 15 receives an input of the name of the known odor substance introduced into the target sample receiving unit 50 (step S2). The process of step S2 may be performed before step S1.

[0143] Next, the measurement value acquiring unit 11 acquires detection signals from each of the plurality of sensor elements 31A in the process before and after the odor substance is adsorbed and desorbed to the sensor element 31 (step S3: acquiring step).

[0144] Next, the extraction unit 12 extracts characteristic information indicating characteristics of the target odor from the detection signal acquired by the measurement acquisition unit 11 (step S4: extraction step). For example, the extraction unit 12 acquires data on the amount of change (ΔV) in the voltage value V (waveform or pattern of change over time). The extraction unit 12 associates the data on the amount of change (ΔV) in the voltage value V (waveform or pattern of change over time) before and after the odor substance adsorption / desorption with the input known target information, and stores the data in the change pattern database.

[0145] If no change pattern has been stored for a given type of existing odor substance (NO in step S5), that is, if there is still insufficient data to use for machine learning, the process returns to step S1.

[0146] If a change pattern is stored for a predetermined type of existing odor substance (YES in step S5), the learning control unit 13 reads out the change pattern for the known odor substance stored in the change pattern database 21 and inputs it to the estimation model generation unit 14. The estimation model generation unit 14 generates an estimation model 22 by machine learning based on the time-dependent change pattern (or feature amount extracted from the change pattern) stored in the change pattern database 21 (step S6).

[0147] The estimation model generating unit 14 stores the estimation model 22 generated by a predetermined machine learning in the storage unit 2 (step S7).

[0148] 9 and 10, the estimation device 10 generates the estimation model 22, but is not limited to this. For example, the estimation device 10 may provide an external computer different from the estimation device 10 and having the same functions as the learning control unit 13 and the estimation model generation unit 14 with the same data as the change pattern database 21 to generate the estimation model 22.

[0149] <Identification of odor substances> Next, the configuration of the information processing system 100a that estimates target information using the estimation model 22 and the estimation process will be described with reference to FIGS.

[0150] (Configuration of Estimation Device 10a (Execution of Estimation Process)) Fig. 11 is a functional block diagram showing an example of the configuration of the information processing system 100a. For convenience of explanation, the same reference numerals are given to components having the same functions as those described in Fig. 1 and Fig. 9, and the description thereof will not be repeated.

[0151] As shown in Fig. 11, the estimation device 10a includes a control unit 1a, a storage unit 2a, and an output unit 18. Here, Fig. 11 shows a configuration example in which the estimation device 10 shown in Fig. 9 is used for odor substance estimation processing. That is, the estimation device 10 shown in Fig. 9 and the estimation device 10a shown in Fig. 11 may be computers having the same hardware configuration.

[0152] The output unit 18 is for presenting the estimation result to the user, and may be, for example, a display, a speaker, a lamp, or the like.

[0153] The control unit 1 a includes a measurement value acquiring unit 11 (acquiring unit), an extracting unit 12, an estimating unit 16, and an output control unit 17.

[0154] The estimation unit 16 uses the estimation model 22 to estimate the odor substance detected by the odor sensor 30. The estimation unit 16 also estimates target information related to the target based on the feature information extracted by the extraction unit 12 and / or the estimation result of the odor substance.

[0155] The target information is at least one of evaluation information about the target's odor, classification information that classifies the target's odor based on a predetermined criterion, and status information that indicates the target's status. The estimation unit 16 uses a learning model obtained by extracting features based on waveform data acquired from targets at multiple levels, for example, to determine which target status in past data the input unknown target status is closest to.

[0156] The output control unit 17 controls the output unit 18 to output the estimation result. Here, the estimation result may be at least one of an estimation result of an odor substance and an estimation result of target information.

[0157] (Estimation process) Hereinafter, specific processes performed by each part of the control unit 1a will be described with reference to Fig. 12. Fig. 12 is a flowchart showing an example of the process flow in which the estimation device 10a estimates target information.

[0158] First, the measurement value acquiring unit 11 acquires the voltage value V0 measured by the odor sensor 30 before the odor substance is introduced into the target sample receiving unit 50, and calculates the resistance value R0. After that, an unknown odor substance (regardless of its properties) is introduced into the target sample receiving unit 50 (step S11).

[0159] Next, the measurement value acquisition unit 11 acquires detection signals in which each of the multiple sensor elements 31 detects an odor substance corresponding to an odor emitted from an object having unknown object information to the sensor element 31 (step S12: acquisition step).

[0160] The extraction unit 12 extracts characteristic information from the detection signal acquired by the measurement acquisition unit 11 (step S13: extraction step). For example, the extraction unit 12 extracts data (waveform or time-dependent change pattern) of the change in voltage value V (ΔV) before and after the odor substance adsorption / desorption as the characteristic information.

[0161] Next, the estimation unit 16 uses the estimation model 22 to estimate an odor substance corresponding to the odor of the target based on the extracted characteristic information, and estimates target information based on the extracted characteristic information (step S14: estimation step). For example, the estimation unit 16 estimates target information indicating the state of the target from a time-dependent change pattern of the voltage value V (or a feature amount extracted from the change pattern).

[0162] The output control unit 17 controls the output unit to output the estimation result (step S15).

[0163] In the above embodiment, the estimation device 10 that generates the estimation model 22 and the estimation device 10a that estimates odor substances using the estimation model 22 have been described. Note that the estimation device 10 and the estimation device 10a may be separate devices or may be a single device.

[0164] (Example of data analysis) The following describes a data analysis method provided by the information processing system 100. The information processing system 100 according to the present disclosure has the above-mentioned configuration and can provide analysis results in the form of superposition of waveforms showing characteristics of a target, principal component analysis of feature amounts extracted from the waveforms, linear discriminant analysis, bar graphs, radar charts, and the like.

[0165] The information processing system 100 can be used, for example, for product quality control, product PR, and target monitoring.

[0166] For example, the information processing system 100 can support the development of food ingredients as a quality control. As an example, the information processing system 100 can create a library showing characteristic information and target information of products shipped in the past, and compare the characteristic information and target information of a newly manufactured product with the information included in the library.

[0167] Furthermore, the results of the estimation performed by the information processing system 100 can be used for product PR. For example, the characteristics of a product can be compared with those of other companies' products based on the information estimated using the information processing system 100. Furthermore, the trends of product characteristics can be identified based on the estimation results of the information processing system 100 and used for matching with customer preferences.

[0168] The results of the estimation performed by the information processing system 100 can be used for monitoring the change over time in the degree of food deterioration, extending the shelf life of food, visualizing the state of food, etc. The results of the estimation can also be used to understand the fermentation state and optimize the conditions.

[0169] For example, a learning model can be generated using target information indicating the deterioration state of a certain food and feature information extracted from a detection signal obtained using the food, and the deterioration degree of the food can be estimated by performing measurements using food whose condition is unknown. In addition, by performing measurements multiple times, the deterioration degree of the food can be monitored over time.

[0170] Furthermore, based on the target information indicating the fermentation state and fermentation conditions of a certain food and the feature information extracted from the detection signal obtained using the food, it is possible to estimate the degree of fermentation of a food whose state is unknown and to estimate the optimal fermentation conditions for the food.

[0171] For example, the information processing system 100 can be used to determine the odor of urethane resin contained in a target. FIG. 13 is a graph showing the change over time in detection intensity in a measurement using a foam containing amines (urethane foam) as a target and in a measurement using a foam not containing amines. FIG. 14 is a graph showing the detection results when a plurality of measurements are performed using the same target as in the measurement shown in FIG. 13. In FIG. 14, the peak top values ​​in each measurement are plotted. In FIG. 13 and FIG. 14, the symbol A indicates the result based on the measurement using a foam containing amines, and the symbol B indicates the result based on the measurement using a foam not containing amines. In FIG. 14, the symbol C indicates the result based on the measurement performed without using a target.

[0172] As shown in Figure 13, a comparison of the detection intensity when using a foam containing amines and a foam not containing amines confirmed that the detection intensity tended to be higher when using a foam containing amines. This tendency was consistent with the sensation when sniffing with the nose. In addition, as shown in Figure 14, even when the same subject was used for five consecutive measurements, the above order (with amines > without) was confirmed to remain unchanged.

[0173] In this way, by using the information processing system 100, it is possible to estimate whether the odor of a target is closer to that of a foam containing amines or that of a foam not containing amines, without relying on human nasal sensation.

[0174] Furthermore, the information processing system 100 can generate a learning model that has been learned to link the detection intensity of a measured signal with the result of a sensory evaluation test conducted in advance, and perform linear discriminant analysis based on the learning model. FIG. 15 is a diagram showing the correspondence between object information indicating whether or not an object is acceptable as a product, and feature information indicating the detection intensity of a detection signal in a measurement conducted using the object. In FIG. 15, "●" indicates data plotted for an object having object information indicating that the object is acceptable as a product. Also, in FIG. 15, "×" indicates data plotted for an object having object information indicating that the object is problematic as a product. The feature information of the object indicated by "●" and "×" is identified by a sensory evaluation test conducted in advance.

[0175] For example, the information processing system 100 generates a learning model that has been trained based on feature information indicating detection strength and target information indicating the state of a product, as shown in Fig. 15. By using this learning model and performing linear discriminant analysis, it is possible to determine whether a target with unknown target information (target indicated by the symbol "★" in Fig. 15) is acceptable as a product, based on the feature information of the target.

[0176] In addition, the information processing system 100 can generate a learning model that has been trained to link the detection intensity of the measured signal with the results of a sensory evaluation test conducted in advance, and perform principal component analysis based on the learning model.

[0177] FIG. 16 is a diagram showing feature quantities based on a sensory evaluation test and measurements performed using the odor measuring device 20 for multiple objects. For example, the information processing system 100 generates a learning model that performs learning for multiple objects based on feature information and object information indicating the state of the product. The information processing system 100 uses the learning model to perform principal component analysis, thereby determining which object whose object information is unknown (objects indicated by the symbol "★" in FIG. 16) is similar in characteristics to which object that has been previously evaluated. For example, as shown in FIG. 16, an object plotted in a position close to the position on the graph of the object to be determined can be considered to show relatively similar object information.

[0178] Fig. 17 shows two objects and feature amounts extracted as feature information from detection signals obtained by measurements performed using each object. As shown in Fig. 17, by using the information processing system 100, it is possible to allow a user to understand that the states of a plurality of objects are different from each other, for example, that the state of the object shown by symbol E is different from that of the object shown by symbol D. For example, if the greater the feature amount shown in Fig. 17, the stronger the odor, the user can understand that the odor of the object shown by symbol E is more suppressed than that of the object shown by symbol D.

[0179] Fig. 18 shows the results of measurements of two objects using multiple sensor elements 31A. In Fig. 18, the vertical axis indicates the intensity of the detection signal detected by each of the multiple sensor elements 31A, and the numbers on the horizontal axis indicate the identification numbers of each of the multiple sensor elements 31A (detection elements) that detected the signal.

[0180] As shown in Fig. 18, by using the information processing system 100, it is possible to compare object information indicating the states of multiple objects and identify the tendency of each object. For example, if the vertical axis in Fig. 18 indicates the odor intensity as the object information of the object, it can be seen that the object with code F and the object with code G have similar odors, and that the object with code G generally emits a stronger odor than the object with code F.

[0181] Furthermore, the information processing system 100 can identify the tendency of the deterioration of the object. FIG. 19 is a two-dimensional map showing the results of measurements of food additives, additive 1 and additive 2, before and after deterioration as objects. The deterioration of an object refers to the deterioration of taste and flavor due to, for example, light irradiation or the passage of time. In FIG. 19, the greater the distance between the plots on the map, the greater the qualitative difference in the odor of the object. Furthermore, the deterioration degree refers to the degree of change in an object from before deterioration to after deterioration.

[0182] It has been known that additive 1 deteriorates more than additive 2. As shown in Fig. 19, when measurements are performed using the information processing system 100, results are obtained that indicate that additive 1 deteriorates more than additive 2, similar to the previously known tendency.

[0183] 20 shows a graph illustrating the relationship between the intensity of a detection signal and time when an odor substance in a specific space is measured using an odor sensor 30 including multiple sensor elements 31A. Here, the space to be measured may be a place where a hygienic environment should be maintained, such as a toilet room.

[0184] From the results shown in FIG. 20, it can be seen that one graph (symbol X in FIG. 20) shows abnormal fluctuations compared to the other graphs. In this way, by using the information processing system 100 to monitor fluctuations in measurement values ​​over time, it is possible to identify sensor elements 31 that cannot perform normal measurement among the multiple sensor elements 31A used in the measurement. Furthermore, by excluding results showing abnormal fluctuations from these results and performing analysis, more accurate results can be obtained. By obtaining more accurate measurement results, it is possible to more accurately grasp the odor situation in places where a hygienic environment should be maintained.

[0185] <Example of software implementation> The control block (particularly the control unit 1) of the estimation device 10, 10a may be realized by a logic circuit (hardware) formed on an integrated circuit (IC chip) or the like, or may be realized by software.

[0186] In the latter case, the estimation device 10, 10a includes a computer that executes instructions of a program, which is software that realizes each function. The computer includes, for example, one or more processors, and a computer-readable recording medium that stores the program. The object of the present invention is achieved by the processor reading the program from the recording medium and executing it in the computer. The processor may be, for example, a CPU (Central Processing Unit). The recording medium may be a "non-transient tangible medium", such as a ROM (Read Only Memory), a tape, a disk, a card, a semiconductor memory, a programmable logic circuit, or the like. The device may further include a RAM (Random Access Memory) that expands the program. The program may be supplied to the computer via any transmission medium (such as a communication network or a broadcast wave) that can transmit the program. Note that one aspect of the present invention may also be realized in the form of a data signal embedded in a carrier wave, in which the program is embodied by electronic transmission.

[0187] 〔summary〕 The information processing system according to aspect 1 of the present disclosure comprises at least one odor measuring device having a sensor chamber provided with a plurality of sensor elements and a supply mechanism capable of alternately supplying a first gas containing an odorant corresponding to the odor of a target and a second gas not containing an odorant corresponding to the odor of the target into the sensor chamber, an acquisition unit that acquires detection signals from each of the plurality of sensor elements, an extraction unit that extracts characteristic information indicating characteristics of the odor of the target from the acquired detection signals, and an estimation unit that estimates target information regarding the target based on the extracted characteristic information.

[0188] The information processing system according to aspect 2 of the present disclosure is, in the above-mentioned aspect 1, further comprising a sample receiving section having a storage space capable of accommodating the object and holding the first gas, and the supply mechanism may supply the second gas to the storage space, thereby pushing the first gas within the storage space out of the storage space toward the sensor chamber.

[0189] The information processing system according to Aspect 3 of the present disclosure is in Aspect 2 above, and may further include a temperature adjustment mechanism capable of adjusting the temperature of the storage space.

[0190] An information processing system according to aspect 4 of the present disclosure, in aspect 2 or 3 above, may be such that the supply mechanism replaces the first gas in the sensor chamber with the second gas in at least 0.01 seconds and not more than 10 seconds, and replaces the second gas in the sensor chamber with the first gas in at least 0.01 seconds and not more than 10 seconds.

[0191] In an information processing system according to aspect 5 of the present disclosure, in any of aspects 1 to 4 above, the characteristic information may include: (1) at least one of the intensity and the time change pattern of the intensity of the detection signal output from each of the multiple sensor elements during a first period in which the first gas is supplied into the sensor chamber, and (2) at least one of the intensity and the time change pattern of the intensity of the detection signal output from each of the multiple sensor elements during a second period in which the first gas supplied into the sensor chamber is pushed out of the sensor chamber by the second gas.

[0192] In an information processing system according to aspect 6 of the present disclosure, in any of aspects 1 to 5 above, the target information may be at least one of evaluation information about the odor of the target, classification information that classifies the odor of the target based on predetermined criteria, and status information that indicates the status of the target.

[0193] In the information processing system according to a sixth aspect of the present disclosure, in any one of the first to sixth aspects, the sensor chamber may be detachably attached to the odor measuring device.

[0194] An information processing system according to an eighth aspect of the present disclosure is based on the seventh aspect, and the odor measuring device may include a first housing that covers at least a part of the supply mechanism.

[0195] An information processing system according to aspect 9 of the present disclosure, in aspect 7 or 8 above, is such that the odor measuring device includes a second housing that covers at least a portion of the supply mechanism and the sensor chamber, the second housing being openable and closable, and when the second housing is in the open state, the sensor chamber may be removably attached to the odor measuring device.

[0196] An information processing system according to a tenth aspect of the present disclosure is in any one of the first to ninth aspects, wherein the plurality of sensor elements are chemiresistor-type sensors.

[0197] An information processing system according to aspect 11 of the present disclosure, in any one of aspects 1 to 10 above, wherein each of the plurality of sensor elements has an odorant receiving layer including a resin composition including a resin and a filler, and the resin compositions contained in the odorant receiving layer of each of the plurality of sensor elements may be different from each other.

[0198] An information processing system according to a twelfth aspect of the present disclosure is any one of the first to eleventh aspects, in which the second gas may be an inert gas or air.

[0199] An information processing method according to aspect 13 of the present disclosure is a control method executed by one or more information processing devices, and includes an acquisition step of acquiring detection signals from each of the multiple sensor elements of at least one or more odor measuring devices having a sensor chamber provided with multiple sensor elements and a supply mechanism capable of alternately supplying a first gas containing an odorant corresponding to the odor of a target and a second gas not containing an odorant corresponding to the odor of the target into the sensor chamber, an extraction step of extracting characteristic information indicating characteristics of the odor of the target from the acquired detection signals, and an estimation step of estimating target information regarding the target based on the extracted characteristic information.

[0200] A program according to aspect 14 of the present disclosure is a program for causing a computer to function as an information processing system according to any one of aspects 1 to 11 above, and is a program for causing a computer to function as the acquisition unit, the extraction unit, and the estimation unit. [Industrial Applicability]

[0201] The present invention is useful as an odor identification sensor for medical use, gas detection, agriculture, and other industrial and daily uses. For example, farmers can use the odor identification sensor to determine the maturity of fragrant crops and manage the optimal harvest timing. In addition, the odor identification sensor can be used to digitize the odors of products such as food and cosmetics, helping to improve the efficiency of product development and stabilize quality. [Explanation of symbols]

[0202] 10, 10a Estimation device 11 Measurement value acquisition unit (acquisition unit) 12 Extraction part 16 Estimation part 30 Odor Sensor 31A Multiple sensor elements (sensor element group) 31, 31b, 31c Sensor element 32 Constant voltage power supply (power supply) 33 Voltmeter (measuring instrument) 20 Odor measuring device 40 Supply mechanism 100, 100a Information Processing Systems 313A, 313C 1st metal wiring 313B, 313D 2nd metal wiring 315, 315c, 315d Odorant receptor layer 50 Target sample receiving section (sample receiving section) 51 Adjustment part 60 Sensor Chamber 80 Gas supply section 91, 92, 93, 94 Body 501 1st mouth 502 2nd mouth 503 Sample inlet

Claims

1. At least one odor measuring device including a sensor chamber provided with a plurality of sensor elements and a supply mechanism capable of alternately supplying a first gas containing an odorant corresponding to a target odor and a second gas not containing an odorant corresponding to the target odor into the sensor chamber; an acquisition unit that acquires a detection signal from each of the plurality of sensor elements; an extraction unit that extracts characteristic information indicating characteristics of the target odor from the acquired detection signal; an estimation unit that estimates object information related to the object based on the extracted feature information; An information processing system comprising:

2. a sample receiving section having a storage space capable of storing the target and holding the first gas; the supply mechanism supplies the second gas to the storage space, thereby pushing the first gas in the storage space out of the storage space toward the sensor chamber. The information processing system according to claim 1 .

3. Further comprising a temperature adjustment mechanism capable of adjusting the temperature of the storage space. The information processing system according to claim 2 .

4. The supply mechanism includes: replacing the first gas in the sensor chamber with the second gas for 0.01 seconds or more and 10 seconds or less; replacing the second gas in the sensor chamber with the first gas for 0.01 seconds or more and 10 seconds or less; The information processing system according to claim 2 .

5. The characteristic information includes: (1) at least one of an intensity of a detection signal output from each of the plurality of sensor elements during a first period in which the first gas is supplied into the sensor chamber and a pattern of change in the intensity over time; and (2) at least one of an intensity of a detection signal output from each of the plurality of sensor elements during a second period in which the first gas supplied into the sensor chamber is pushed out of the sensor chamber by the second gas. The information processing system according to claim 1 .

6. The target information is at least one of evaluation information about the odor of the target, classification information that classifies the odor of the target based on a predetermined criterion, and status information that indicates the status of the target. The information processing system according to claim 1 .

7. The sensor chamber is detachably attached to the odor measuring device. The information processing system according to claim 1 .

8. The odor measurement device includes a first housing that covers at least a portion of the supply mechanism. The information processing system according to claim 7.

9. The odor measurement device includes a second housing that covers at least a part of the supply mechanism and the sensor chamber, The second housing is openable and closable, and when the second housing is in an open state, the sensor chamber can be detachably attached to the odor measuring device. The information processing system according to claim 7.

10. The plurality of sensor elements are chemiresistor type sensors. The information processing system according to claim 1 .

11. Each of the plurality of sensor elements includes an odorant receiving layer including a resin composition including a resin and a filler; The resin compositions contained in the odorant receiving layers of the plurality of sensor elements are different from each other. The information processing system according to claim 1 .

12. The second gas is an inert gas or air. The information processing system according to claim 1 .

13. A control method executed by one or more information processing devices, comprising: An acquisition step of acquiring a detection signal from each of the plurality of sensor elements of at least one or more odor measuring devices comprising a sensor chamber provided with a plurality of sensor elements and a supply mechanism capable of alternately supplying a first gas containing an odorant corresponding to a target odor and a second gas not containing an odorant corresponding to the target odor into the sensor chamber; An extraction step of extracting characteristic information indicating characteristics of the target odor from the acquired detection signal; an estimation step of estimating object information related to the object based on the extracted feature information; An information processing method comprising:

14. A program for causing a computer to function as the information processing system according to claim 1 , the program causing a computer to function as the acquisition unit, the extraction unit, and the estimation unit.

Images