Health information estimation system and information estimation system

Abstract

To appropriately estimate information relating to human health.SOLUTION: A health information estimation system according to an embodiment includes: data acquisition means for acquiring data detected by a sensor disposed in a water-related space; data structure determination means for determining a data structure on the basis of a data array configured by associating the data acquired by the data acquisition means on the basis of date and time information; and health information estimation means for estimating health information of a user on the basis of the data array, the health information estimation means selecting, from among a plurality of health information estimation models, a model corresponding to the data structure and estimating a health condition of the user.SELECTED DRAWING: Figure 2

Description

[Technical Field]

[0001] The disclosed embodiments relate to a health information estimation system and an information estimation system. [Background technology]

[0002] Conventionally, technologies have been provided for collecting data by detecting sensors in wet spaces such as toilets. For example, a toilet seat device equipped with a gas sensor that can detect fecal gas (such as farts) emitted when a toilet user (hereinafter also referred to as "user") defecates has been provided (see, for example, Patent Documents 1 and 2). Also, a system has been provided that determines health information of a user based on excrement data acquired by a sensor installed in the toilet (see, for example, Patent Document 3). [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Laid-Open No. 2005-315836 [Patent Document 2] Japanese Patent Application Laid-Open No. 2016-145806 [Patent Document 3] U.S. Patent No. 11,881,313 Summary of the Invention [Problem to be solved by the invention]

[0004] However, the above-described conventional technology has room for improvement. For example, while the above-described conventional technology can estimate the health of a person, such as a user, using data acquired in a water-related space such as a toilet, it may be difficult to appropriately estimate information about a person's health when there is variation in the acquired data or when the data used for processing fluctuates. Furthermore, there is room for improvement in terms of how to estimate information about a person's health when there is a variety of data that can be acquired. Therefore, it is desirable to appropriately estimate information about a person's health depending on, for example, the data used for processing.

[0005] In view of the above, the challenge is to appropriately estimate information related to a person's health.

[0006] The disclosed embodiments aim to provide a health information estimation system and an information estimation system that can appropriately estimate information related to a person's health. [Means for solving the problem]

[0007] A health information estimation system according to one aspect of the embodiment comprises a data acquisition means for acquiring data detected by sensors placed in a water-related space, a data structure determination means for determining a data structure based on a data array constructed by associating the data acquired by the data acquisition means based on date and time information, and a health information estimation means for estimating a user's health information based on the data array, wherein the health information estimation means selects a model corresponding to the data structure from a plurality of health information estimation models to estimate the health condition of the user.

[0008] According to one aspect of the embodiment, a health information estimation system estimates a user's health condition based on acquired data by selecting from a plurality of health information estimation models a model corresponding to a data structure determined based on a data array of data detected by sensors installed in a wet space. For example, the health information estimation system can provide health information personalized to the user's condition by selecting an optimal health information estimation model from a plurality of health information estimation models based on the structure of data acquired from sensors installed in the wet space. This allows the health information estimation system to appropriately estimate information about a person's health.

[0009] A health information estimation system according to one aspect of the embodiment comprises a first data acquisition means for acquiring data detected by sensors placed in a water-related space, a second data acquisition means for acquiring external data, a data structure determination means for determining a data structure based on a data array formed by associating data acquired by at least one of the first data acquisition means and the second data acquisition means based on date and time information, and a health information estimation means for estimating a user's health information based on the data array, wherein the health information estimation means selects a model corresponding to the data structure from a plurality of health information estimation models to estimate the health condition of the user.

[0010] According to one aspect of the embodiment, a health information estimation system estimates a user's health condition based on acquired data by selecting from among multiple health information estimation models a model corresponding to a data structure determined based on the data array of data detected by sensors installed in a wet space or data acquired externally. For example, the health information estimation system can provide health information personalized to the user's condition by selecting an optimal health information estimation model from among multiple health information estimation models based on the structure of data acquired from sensors installed in a wet space. This allows the health information estimation system to appropriately estimate information about a person's health.

[0011] A health information estimation system according to one aspect of the embodiment includes a first data acquisition means for acquiring data detected by a sensor placed in a toilet space, a second data acquisition means for acquiring external data, a data structure determination means for determining a data structure based on a data array formed by associating data acquired by at least one of the first data acquisition means and the second data acquisition means based on date and time information, and a health information estimation means for estimating a user's health information based on the data array, wherein the health information estimation means selects a model corresponding to the data structure from a plurality of health information estimation models to estimate the health condition of the user.

[0012] According to one aspect of the embodiment, a health information estimation system estimates a user's health condition based on the acquired data by selecting a model from among multiple health information estimation models according to a data structure determined based on the data array of data detected by sensors installed in a toilet space, which is a wet space, or data acquired externally. This allows the health information estimation system to appropriately estimate information about a person's health. For example, the health information estimation system can provide health information personalized to the user's condition by selecting an optimal health information estimation model from among multiple health information estimation models based on the structure of data acquired from sensors installed in the wet space.

[0013] In one aspect of the embodiment, in the health information estimation system, the data structure determination means determines the data structure based on the data array constructed by associating the data acquired by the data acquisition means based on personal information or date information.

[0014] According to one aspect of the embodiment, a health information estimation system can appropriately determine a data structure by determining the data arrangement of data associated with personal information or date information. This allows the health information estimation system to appropriately estimate information about a person's health. For example, by associating acquired data with date information or personal information, the health information estimation system can appropriately provide health information even when there are multiple users.

[0015] In one aspect of the embodiment, in the health information estimation system, the data structure determination means determines the data structure based on a combination of at least one data item in the data array.

[0016] According to one aspect of the embodiment, a health information estimation system can appropriately determine a data structure by determining the data structure based on the combination of data in a data array. This allows the health information estimation system to appropriately estimate information related to a person's health. For example, the health information estimation system can provide health information corresponding to any combination of data, even if the data array contains missing data.

[0017] In one aspect of the embodiment, in the health information estimation system, the data structure determination means uses a past data array within a specified date and time range to fill in any missing data in the latest data array, and determines the data structure based on the filled-in data array.

[0018] According to one aspect of the embodiment, a health information estimation system can appropriately determine a data structure by filling in missing data in a data sequence and determining a data structure based on the filled data sequence. This allows the health information estimation system to appropriately estimate information related to a person's health. For example, even if a data sequence contains missing data, the health information estimation system can provide more accurate health information by filling in the missing data with past data from a specified date and time range.

[0019] In one aspect of the embodiment, in the health information estimation system, the data structure determination means determines the data structure based on a time series analysis result for each piece of data acquired by the data acquisition means within a predetermined date and time range.

[0020] According to one aspect of the embodiment, a health information estimation system can appropriately determine a data structure by determining the data structure based on the results of time series analysis of each piece of data acquired within a predetermined date and time range. This allows the health information estimation system to appropriately estimate information related to a person's health. For example, the health information estimation system can estimate health information based on time series trends, rather than just features obtained from a single time.

[0021] In one aspect of the embodiment, in the health information estimation system, the data structure determination means determines the data structure based on the results of time series analysis of each piece of data acquired by the first data acquisition means and the second data acquisition means within a specified date and time range.

[0022] According to one aspect of the embodiment, a health information estimation system determines a data structure based on the results of time series analysis of each piece of data acquired within a predetermined date and time range, thereby enabling the system to appropriately estimate information related to a person's health. For example, by basing the data on the time series trends, the system can estimate health information based on time series trends, rather than just features obtained from a single time point, with greater accuracy.

[0023] In a health information estimation system according to one aspect of the embodiment, the health information estimation means selects a model corresponding to the latest data structure from the plurality of health information estimation models and estimates the health state of the user.

[0024] According to one aspect of the embodiment, a health information estimation system can estimate a user's health status based on the most recent data by selecting a model corresponding to the most recent data structure from among multiple health information estimation models. This allows the health information estimation system to appropriately estimate information about a person's health. For example, by selecting a model corresponding to the most recent data structure, the health information estimation system can accurately estimate the user's current health information.

[0025] In one aspect of the embodiment, the health information estimation system selects a model from the plurality of health information estimation models in accordance with the data structure based on the time series analysis results for each piece of data acquired by the data acquisition means within a specified date and time range, and estimates the health status of the user.

[0026] According to one aspect of the embodiment, a health information estimation system can appropriately estimate a user's health condition by selecting a model according to a data structure based on the results of time-series analysis of each data. This allows the health information estimation system to appropriately estimate information about a person's health. For example, the health information estimation system can estimate health information based on time-series trends, rather than just features obtained from a single time.

[0027] In one aspect of the embodiment, the health information estimation means selects a model from the plurality of health information estimation models in accordance with a plurality of data structures based on the time series analysis results for each piece of data acquired by the data acquisition means within a specified date and time range, and estimates the health status of the user.

[0028] According to one aspect of the embodiment, a health information estimation system can appropriately estimate a user's health condition by selecting a model according to multiple data structures based on the results of time series analysis of each data. This allows the health information estimation system to appropriately estimate information related to a person's health. For example, by using multiple data structures, the health information estimation system can estimate health information more accurately, for example, based on the time series trends of each data.

[0029] In one aspect of the embodiment, in the health information estimation system, the health information estimation means corrects the health information estimation model based on a correction value calculated in accordance with the data acquired by the data acquisition means.

[0030] According to one aspect of the embodiment, a health information estimation system can appropriately estimate a user's health status by correcting a health information estimation model based on a correction value calculated according to acquired data. This allows the health information estimation system to appropriately estimate information about a person's health. For example, by adjusting coefficients in the data structure according to the amount of data, the health information estimation system can allow a user to refer to health information at the time or timing that is most effective for the user, thereby estimating health information with greater accuracy.

[0031] In one aspect of the embodiment, the health information estimation system selects a model from the plurality of health information estimation models in accordance with a plurality of data structures based on the time series analysis results for each piece of data acquired by the first data acquisition means and the second data acquisition means within a predetermined date and time range, and estimates the health status of the user.

[0032] According to one aspect of the embodiment, a health information estimation system can appropriately estimate a user's health condition by selecting a model according to multiple data structures based on the results of time series analysis of each data. This allows the health information estimation system to appropriately estimate information related to a person's health. For example, by using multiple data structures, the health information estimation system can estimate health information more accurately, for example, based on the time series trends of each data.

[0033] In one aspect of the embodiment, in the health information estimation system, the health information estimation means estimates the health information of the user at a predetermined cycle.

[0034] According to one aspect of the embodiment, a health information estimation system estimates a user's health information at a predetermined cycle, thereby enabling the user's health information to be estimated at an appropriate timing. This allows the health information estimation system to appropriately estimate information about a person's health. For example, the health information estimation system can estimate health information at an effective timing when the user is most likely to take action.

[0035] In one aspect of the embodiment, in the health information estimating system, the data structure determining means prompts the user to fill in the external data to fill in any missing data in the data array within a predetermined date and time range.

[0036] According to one aspect of the embodiment, a health information estimation system prompts a user to input external data to compensate for missing data sequences, thereby increasing the likelihood that the missing data will be filled in and accurately estimating the user's health information. This allows the health information estimation system to accurately estimate information about a person's health. For example, by having the user input external data, the health information estimation system can estimate health information with greater accuracy.

[0037] An information estimation system according to one aspect of the embodiment includes a data acquisition means for acquiring data detected by sensors placed in a water-related space, a data structure determination means for determining a data structure based on a data array or data group formed by associating the data acquired by the data acquisition means based on date and time information, and an information estimation means for estimating a user's health information or service information to be provided to the user based on the data acquired by the data acquisition means, wherein the information estimation means uses a model corresponding to the data structure from among a plurality of information estimation models.

[0038] According to one aspect of the embodiment, an information estimation system determines a data structure based on a data array or a data group formed by associating data detected by sensors installed in a water-related space based on date and time information, and when estimating a user's health information or service information to be provided to the user based on the data, the system uses a model corresponding to the data structure from among multiple information estimation models to estimate the user's health information or service information to be provided to the user based on the acquired data. This allows the information estimation system to appropriately estimate information related to a person's health. Furthermore, by determining a data structure from the acquired data array or data group and using an appropriate model corresponding to the data structure, the information estimation system can provide health information or service information without imposing a burden on the user, such as (additional) data entry.

[0039] In an information estimation system according to one aspect of the embodiment, the sensor is a sensor that detects urine components among the excrement excreted into a toilet device, the data acquisition means acquires urine component data detected by the sensor, and the data structure determination means determines a urine data structure based on a urine component array that is constructed by associating the urine component data based on date and time information.

[0040] According to one aspect of the embodiment, an information estimation system acquires urine component data obtained by detecting urine components among excrement discharged into a toilet device using a sensor, and determines a urine data structure based on a urine component sequence that associates the urine component data based on date and time information, thereby estimating a user's health information or service information to be provided to the user based on the acquired data. This allows the information estimation system to appropriately estimate information related to a person's health.

[0041] In one aspect of the embodiment, in the information estimation system, the sensor is a sensor installed in the toilet space that acquires biometric data of the user, the data acquisition means acquires the biometric data detected by the sensor, and the data structure determination means determines the biometric data structure based on a biometric data array that is constructed by associating the biometric data based on date and time information.

[0042] According to an information estimation system according to one aspect of the embodiment, biometric data detected by a sensor installed in the toilet space is acquired, and a biometric data structure is determined based on a biometric data sequence constructed by associating the biometric data based on date and time information, thereby making it possible to estimate health information of a user or service information to be provided to the user according to the acquired data. This allows the information estimation system to appropriately estimate information related to a person's health. [Effects of the Invention]

[0043] According to one aspect of the embodiment, information about a person's health can be appropriately estimated. [Brief explanation of the drawings]

[0044] [Figure 1] FIG. 1 is a diagram illustrating an example of an estimation process according to the embodiment. [Figure 2] FIG. 2 is a diagram illustrating an example of a processing flow in the health information estimating system. [Figure 3] FIG. 3 is a diagram illustrating an example of data and processing. [Figure 4] FIG. 4 is a diagram illustrating an example of the configuration of a health information estimating system according to the embodiment. [Figure 5] FIG. 5 is a block diagram illustrating an example of the configuration of a health information estimation device according to the embodiment. [Figure 6] FIG. 6 is a diagram illustrating an example of a model information storage unit according to the embodiment. [Figure 7] FIG. 7 is a diagram illustrating an example of a processing flow in the health information estimating system. [Figure 8]FIG. 8 is a diagram illustrating an example of data and processing. [Figure 9] FIG. 9 is a diagram illustrating an example of a processing flow in the health information estimating system. [Figure 10] FIG. 10 is a diagram illustrating an example of data and processing. [Figure 11] FIG. 11 is a diagram illustrating an example of data and processing. [Figure 12] FIG. 12 is a diagram illustrating an example of data and processing. [Figure 13] FIG. 13 is a diagram illustrating an example of data and processing. [Figure 14] FIG. 14 is a diagram illustrating an example of data and processing. [Figure 15] FIG. 15 is a diagram illustrating an example of data and processing. [Figure 16] FIG. 16 is a diagram illustrating an example of data and processing. [Figure 17] FIG. 17 is a diagram illustrating an example of data and processing. [Figure 18] FIG. 18 is a diagram illustrating an example of data and processing. [Figure 19] FIG. 19 is a diagram illustrating an example of data and processing. [Figure 20] FIG. 20 is a diagram illustrating an example of data and processing. [Figure 21] FIG. 21 is a diagram illustrating an example of data and processing. [Figure 22] FIG. 22 is a diagram illustrating an example of a processing flow in the health information estimating system. [Figure 23] FIG. 23 is a diagram illustrating an example of data and processing. [Figure 24] FIG. 24 is a diagram illustrating an example of data and processing. [Figure 25] FIG. 25 is a diagram illustrating an example of data and processing. DETAILED DESCRIPTION OF THE INVENTION

[0045] Hereinafter, embodiments of a health information estimation system and an information estimation system disclosed in the present application will be described in detail with reference to the accompanying drawings. Note that the present invention is not limited to the following embodiments.

[0046] <1. Information processing example> First, an overview of the information processing executed in the health information estimation system 1 according to the embodiment will be described with reference to Fig. 1. Fig. 1 is a diagram showing an example of the estimation processing according to the embodiment. In Fig. 1, a toilet space PS1 is described as an example of a wet space, but the wet space is not limited to the toilet space PS1 and the health information estimation system 1 can be applied to any wet space as long as it is applicable.

[0047] For example, the health information estimation system 1 may be applied not only to toilet spaces but also to any wet space such as a bathroom space PS2, a space with a vanity (washing space), a kitchen space, etc. Hereinafter, when wet spaces such as the toilet space PS1, bathroom space PS2, washing space, and kitchen space are described without distinction, they may be referred to as "wet space PS."

[0048] The health information estimation system 1 estimates health information about a user of the bathroom space PS (also referred to simply as "user") using data based on detection by sensor devices 5 arranged in the bathroom space PS (also referred to as "bathroom sensor data"). For example, health information is information indicating a person's health-related condition (also referred to as "health condition"). The health condition here indicates a person's health-related condition, and examples include bowel condition, nocturia, overactive bladder, menstrual cycle, stress, frailty, etc. Furthermore, bowel condition includes various things that can indicate a person's bowel condition, such as constipation, diarrhea, and intestinal environment. Note that the above is merely an example, and health information is not limited to the above, as long as it is related to a person's health.

[0049] First, an overview of the processing in the health information estimation system 1, including the transmission and reception of information between each device in the health information estimation system 1, will be explained using Figure 1, and then the details of each information processing will be explained using Figures 2 and 3. In Figure 1, an example will be explained in which a user U1, who is a user of the bathroom space PS, is the subject of health information estimation.

[0050] First, the health information estimation device 100 acquires data detected by a sensor device 5 arranged in the bathroom space PS (bathroom sensor data) from the sensor device 5 (step S1). For example, the health information estimation device 100 acquires data detected by a sensor device 5 arranged in a toilet space PS1, which is provided with a toilet device 50 that functions as a toilet bowl (also referred to as "toilet sensor data"), from the sensor device 5. In FIG. 1, the health information estimation device 100 acquires data about a user U1 detected by the sensor device 5 arranged in the toilet space PS1 (toilet sensor data) from the sensor device 5.

[0051] Note that bathroom sensor data such as toilet sensor data may include biometric data, gas data, and feces data of the user based on detection by the sensor device 5. Bathroom sensor data is not limited to toilet sensor data and may include data acquired about various bathroom spaces PS. For example, if the bathroom spaces PS targeted by the health information estimation system 1 include a bathroom space PS2, the bathroom sensor data will include data acquired about the bathroom space PS2 (also referred to as "bathroom sensor data"). For example, if the bathroom spaces PS targeted by the health information estimation system 1 include a space with a vanity (washing space), the bathroom sensor data will include data acquired about the washing space (also referred to as "washing space sensor data"). For example, if the bathroom spaces PS targeted by the health information estimation system 1 include a kitchen space, the bathroom sensor data will include data acquired about the kitchen space (also referred to as "kitchen sensor data"). But details will be given later.

[0052] 1 shows a case where one toilet device 50 is provided in the toilet space PS1 for the sake of explanation, multiple toilet devices 50 may be provided in the toilet space PS1, and a washbasin device, a men's urinal device, etc. may also be provided in addition to the toilet device 50. Furthermore, the sensor device 5 may be placed not only in the toilet space PS1, but also in various wet spaces PS such as the bathroom space PS2, kitchen space, and washroom space.

[0053] The health information estimation device 100 acquires data held by the server device 2 (also referred to as "external data") from the server device 2 (step S2). For example, the external data includes the user's dietary data, sleep data, exercise data, etc. In FIG. 1, the health information estimation device 100 acquires external data about user U1 from the server device 2. For example, the health information estimation device 100 acquires the dietary data, sleep data, exercise data, etc. of user U1 as the external data of user U1.

[0054] If external data is not used in the health information estimation process, health information estimation device 100 may not perform step S2. Steps S1 and S2 may be performed at any time before step S3. For example, step S2 may be performed before step S1, and each of steps S1 and S2 may be performed multiple times.

[0055] The health information estimation device 100 then estimates the user's health information using the acquired information (step S3). For example, the health information estimation device 100 estimates the user's health information using plumbing sensor data and external data. In FIG. 1, the health information estimation device 100 estimates the health information of user U1 using plumbing sensor data, etc. of user U1. For example, the health information estimation device 100 determines a data structure based on a data array constructed by associating data acquired about user U1 based on date and time information. The health information estimation device 100 then selects a model from multiple health information estimation models (also simply referred to as "models") that corresponds to the data structure for user U1, and performs an estimation process to estimate the health state of user U1 using the selected model.

[0056] As described above, the health information estimation system 1 including the health information estimation device 100 executes processing, for example, according to a processing flow as shown in Figure 2. Figure 2 is a diagram showing an example of a processing flow in the health information estimation system. As shown in the first processing flow PF1 in Figure 2, in the health information estimation system 1, the health information estimation device 100 acquires plumbing sensor data (referred to as "plumbing sensor data" in the figure) using data acquisition means (e.g., corresponding to acquisition unit 131 in Figure 5).

[0057] Then, based on the data array of the acquired plumbing sensor data, the health information estimation device 100 determines the data structure of the plumbing sensor data using a data structure determination means (e.g., corresponding to determination unit 132 in Figure 5 ).The health information estimation device 100 then selects a health information estimation model (model) according to the determined data structure of the plumbing sensor data using a health information estimation means (e.g., corresponding to estimation unit 133 in Figure 5 ), and estimates the user's health information using the selected model.

[0058] For example, as shown in dataset DS1 in FIG. 3, the health information estimation device 100 determines the data structure of the acquired plumbing sensor data based on the data array of the data, and selects a health information estimation model (model) according to the determined data structure. FIG. 3 is a diagram showing an example of data and processing. Data set DS1 in FIG. 3 includes a data group in which data acquired at a date and time corresponding to a time is associated with the time. FIG. 3 shows a case in which plumbing sensor data of types corresponding to data #1 to #6 has been acquired, as shown in the column labeled "sensor-acquired data." Note that when each dataset used for processing by the health information estimation device 100, such as dataset DS1 or DS2, is referred to without distinction, it will be referred to as dataset DS. For example, dataset DS is stored in storage unit 120 (data storage unit 121, etc.).

[0059] The types of plumbing sensor data corresponding to data #1 to #6 in FIG. 3 may be data of any type. For example, data #1 may be gas data of an odorless gas, and data #2 may be gas data of a foul-smelling gas. Furthermore, for example, data #3 may be stool data on the amount of stool (also simply referred to as "stool"), data #4 may be stool data on the shape of stool (stool), and data #5 may be stool data on the color of stool (stool). Furthermore, for example, data #4 may be biological data on blood flow. Note that the above is merely an example, and data #1 to #6 may be plumbing sensor data of various types. For example, data #1 to #6 may each be data on a different excretion type, and may include stool shape (stool shape), stool color, stool amount, odorless gas (amount), foul-smelling gas (amount), urine color, urine amount, urine flow rate, urine components, etc.

[0060] The "xx" placed in the elements (cells) from the third row onwards in the columns corresponding to each of data #1 to #6 indicates that the data was acquired at the corresponding time. In Fig. 3, this indicates that data of type Data #1, data of type Data #2, data of type Data #4, and data of type Data #5 were acquired at the time corresponding to "2024 / 3 / 1 8:00". In other words, this indicates that data of type Data #3 and data of type Data #6 were not acquired at the time corresponding to "2024 / 3 / 1 8:00".

[0061] 3 also shows that data types #2 to #5 were acquired at the time corresponding to "2024 / 3 / 1 12:00". FIG. 3 also shows that data types #3 to #6 were acquired at the time corresponding to "2024 / 3 / 2 9:30". FIG. 3 also shows that data types #1, #3 to #6 were acquired at the time corresponding to "2024 / 3 / 3 18:00".

[0062] The date and time indicated by the time may be a reference date and time, in which case, the data may be acquired within a predetermined date and time range (time period) corresponding to that date and time. For example, the predetermined date and time range may be a date and time range including several hours before and after the reference time (date and time), and can be set arbitrarily. For example, data assigned with "xx" in a row corresponding to "2024 / 3 / 1 8:00" may be data of that type acquired within a predetermined range from "2024 / 3 / 1 8:00". For example, when "xx" is assigned to a row corresponding to "2024 / 3 / 1 8:00", the data of that type may be data acquired during a predetermined period including 8:00 on March 1, 2024 (e.g., 7:00 to 9:00 on March 1, 2024).

[0063] 3, health information estimation device 100 determines the data structure shown in the eighth column based on the data in the first to seventh columns from the left in dataset DS1, i.e., the arrangement of data (data array) in the seven columns of time and sensor-acquired data (#1 to #6). For example, because the data corresponding to "2024 / 3 / 1 8:00" has a data array including data types #1, #2, #4, and #5, health information estimation device 100 determines that the data structure of the data corresponding to "2024 / 3 / 1 8:00" is data structure a.

[0064] Furthermore, for example, for the data corresponding to "2024 / 3 / 1 12:00", the health information estimation device 100 determines the data structure of the data corresponding to "2024 / 3 / 1 12:00" to be data structure b, because the data array includes data types #2 to #5. Similarly, the health information estimation device 100 determines the data structures of the data corresponding to "2024 / 3 / 2 9:30" and "2024 / 3 / 3 18:00" to be data structures c and d, respectively.

[0065] For example, the health information estimation device 100 may determine a data structure corresponding to a data sequence using a list of information (also referred to as a "data structure list") that associates data structures with data sequence patterns. For example, the health information estimation device 100 may search the data structure list to identify data structure a associated with a data sequence containing data types #1, #2, #4, and #5, and determine the identified data structure a as the data structure for data corresponding to "2024 / 3 / 1 8:00." Note that the above is merely an example, and the health information estimation device 100 may determine a data structure using any information as appropriate.

[0066] The health information estimation device 100 then selects a health information estimation model shown in column 9 based on the data structure shown in column 8. For example, for the data corresponding to "2024 / 3 / 1 8:00," the health information estimation device 100 selects model A as the health information estimation model corresponding to "2024 / 3 / 1 8:00" because the data structure is data structure a. Similarly, for the data corresponding to "2024 / 3 / 1 12:00," "2024 / 3 / 2 9:30," and "2024 / 3 / 3 18:00," the health information estimation device 100 selects models B, C, and D as the health information estimation models corresponding to "2024 / 3 / 1 12:00," "2024 / 3 / 2 9:30," and "2024 / 3 / 3 18:00," respectively, because the data structures are data structures b, c, and d. The health information estimation device 100 then estimates health information using the selected model.

[0067] In this way, the health information estimation device 100 performs estimation processing using a model selected based on the data structure of the data acquired about the user. This allows the health information estimation system to appropriately estimate information about a person's health. The health information estimation device 100 may acquire multiple health information estimation models to be used as selection candidates in any manner. For example, the health information estimation device 100 may acquire (receive) multiple health information estimation models from another device, such as a model providing device, and perform estimation processing using a model selected from the acquired multiple health information estimation models. Alternatively, the health information estimation device 100 may generate multiple health information estimation models and perform estimation processing using a model selected from the generated multiple health information estimation models. Examples of models used by the health information estimation device 100 for estimation processing and estimation processing will be described in detail below.

[0068] The health information estimation device 100 may transmit information based on the estimation process to the user (step S4). For example, the health information estimation device 100 may transmit the user's health information estimated by the estimation process to the terminal device 10 used by the user. In FIG. 1, the health information estimation device 100 may transmit information indicating the health condition of user U1 estimated by the estimation process to the terminal device 10 used by user U1. In this case, the terminal device 10 receiving the information from the health information estimation device 100 notifies user U1 of the health information by displaying the received information or outputting it as audio. Note that the recipient of the information related to the estimation result is not limited to user U1, but may also be user U1's family, other entities related to user U1's health management (such as medical institutions such as medical interview centers), etc.

[0069] <2. Example of health information estimation system configuration> Next, an example of the configuration of health information estimation system 1 will be described with reference to Fig. 4. Fig. 4 is a diagram showing an example of the configuration of a health information estimation system according to an embodiment. Fig. 4 shows an example of the configuration of health information estimation system 1, and other example configurations will be described later.

[0070] Health information estimation system 1 includes a health information estimation device 100, a terminal device 10, a sensor device 5, and a server device 2. Note that health information estimation system 1 may include multiple health information estimation devices 100, multiple terminal devices 10, multiple sensor devices 5, and multiple server devices 2.

[0071] The health information estimation device 100 is an information processing device (computer) that estimates health information of a user. The health information estimation device 100 is communicatively connected to the sensor device 5, terminal device 10, server device 2, etc. via a predetermined network such as the Internet, either wired or wirelessly. The health information estimation device 100 receives various plumbing sensor data, such as toilet sensor data, from the sensor device 5. The health information estimation device 100 receives external data from the server device 2. The health information estimation device 100 receives data for compensating for missing data (also referred to as "compensation data") from the terminal device 10. The health information estimation device 100 performs estimation processing using information acquired from at least one of the sensor device 5, terminal device 10, and server device 2.

[0072] The health information estimation device 100 determines a data structure based on a data array formed by associating data acquired from at least one of the sensor device 5, the terminal device 10, and the server device 2 based on date and time information. The health information estimation device 100 estimates the user's health information based on the data array. For example, the health information estimation device 100 selects a model from multiple health information estimation models according to the data structure and estimates the user's health condition.

[0073] The terminal device 10 is an information processing device (device) used by a user. The terminal device 10 is realized by, for example, a smartphone, a mobile phone, a PDA (Personal Digital Assistant), a tablet terminal, a notebook PC (Personal Computer), etc. In the example shown in Fig. 1, the terminal device 10 is a smartphone used by a user.

[0074] Terminal device 10 transmits data (information) to health information estimation device 100 in response to a user's operation. Terminal device 10 also outputs information received from health information estimation device 100. Terminal device 10 is communicably connected to server device 2, health information estimation device 100, etc. via a predetermined network such as the Internet, and transmits and receives information between server device 2, health information estimation device 100, etc. In response to a request from health information estimation device 100, terminal device 10 transmits compensation data to health information estimation device 100 to compensate for missing data. Terminal device 10 displays information received from health information estimation device 100. Terminal device 10 may also output information received from health information estimation device 100 as audio.

[0075] The sensor device 5 detects various types of wet-related sensor data, such as toilet sensor data. The sensor device 5 is disposed in the wet-related space PS and performs detection within the wet-related space PS. For example, the sensor device 5 is disposed in a toilet space PS1 in which a toilet device 50 is provided. The toilet space PS1 in which the toilet device 50 is provided is not limited to a user's residence, but may also be a toilet (public toilet) provided in a facility such as a building. The sensor device 5 is not limited to being disposed in the toilet space PS1, but may also be disposed in any wet-related space PS, such as a bathroom space PS2, a kitchen space, or a washroom space.

[0076] For example, the sensor device 5 placed in the toilet space PS1 has a gas sensor. In this case, the sensor device 5 detects gas data using the gas sensor, and the water-related sensor data may include the gas data. The gas sensor of the sensor device 5 detects gases emitted from inside the user's body to outside the toilet space PS1. For example, the gas sensor of the sensor device 5 detects the user's bowel movements (farts, etc.), gases emitted from the user's excrement, etc. For example, the sensor device 5 may have a gas sensor (odorless gas sensor) that detects odorless gases composed of hydrogen, methane, carbon dioxide, etc. Furthermore, for example, the sensor device 5 may have a gas sensor (malodorous gas sensor) that detects odorous gases composed of hydrogen sulfide, methyl mercaptan, etc.

[0077] The sensor device 5 is not limited to the above, and may have a gas sensor that detects any gas. For example, the sensor device 5 may have a gas sensor (health-related gas sensor) that detects gases derived from intestinal fermentation that indicate a high level of health (also called "health-related gases"). For example, the sensor device 5 may have a gas sensor (health-related gas sensor) that detects at least one of health-related gases such as hydrogen, carbon dioxide, acetic acid, methane, ethanol, and water.

[0078] For example, the sensor device 5 may have a gas sensor (odor gas sensor) that detects gases (also called "odor gases") that are derived from intestinal putrefaction and indicate poor health. For example, the sensor device 5 may have a gas sensor (odor gas sensor) that detects at least one of odor gases such as ammonia, trimethylamine, hydrogen sulfide, methyl mercaptan, indole, and skatole.

[0079] The sensor device 5 arranged in the toilet space PS1 may also have a feces sensor. In this case, the sensor device 5 detects feces data using a gas sensor, and the feces data may be included in the bathroom sensor data. For example, the feces sensor of the sensor device 5 is a sensor that captures images of feces using a line sensor, a camera (two-dimensional image sensor), or the like. In this case, the feces sensor of the sensor device 5 captures images of feces (sometimes simply referred to as "feces") excreted by the user and generates an image including the user's feces (feces). Note that the feces sensor of the sensor device 5 can be any sensor that can detect at least one of the feces properties, such as the color, shape, and amount of feces. Note that the determination of the user's feces properties may be performed by the sensor device 5 or by the health information estimation device 100. The feces sensor may also detect the user's urine. For example, the feces sensor may detect the color, amount, etc. of the user's urine. In this case, the sensor device 5 detects urine data using a gas sensor, and the bathroom sensor data may include urine data.

[0080] The above is merely an example, and the sensor device 5 may include various other sensors. For example, the sensor device 5 arranged in the toilet space PS1 may include a sensor (sitting detection sensor) that detects sitting on the toilet seat. Furthermore, for example, the sensor device 5 may include a sensor that detects biometric data of a user in the bathroom space PS. For example, the sensor device 5 may include sensors that detect biometric data such as the user's blood flow, heart rate (pulse), breathing, body temperature, brain waves, and muscle (electromyography). For example, when detecting the user's blood flow in the toilet space PS1, the sensor device 5 may include a blood flow sensor provided on the toilet seat or the like of the toilet device 50. In this case, the sensor device 5 detects blood flow data using the blood flow sensor, and the bathroom sensor data may include the blood flow data. In this way, the bathroom sensor data may include various biometric data such as blood flow data, heart rate data, breathing data, body temperature data, brain wave data, and muscle data. Furthermore, the sensor device 5 may include a human body detection sensor that detects the presence or absence of a person in the bathroom space PS.

[0081] The sensor device 5 has a communication function and can transmit and receive information to and from the health information estimation device 100 and the like via a predetermined network such as the Internet. The sensor device 5 may also have the function of an information analysis means. The sensor device 5 transmits data (information) collected by detection to the health information estimation device 100. The sensor device 5 transmits various water-related sensor data, such as toilet sensor data, to the health information estimation device 100.

[0082] Server device 2 is a data storage device (service providing device) that stores data (external data) used by health information estimation device 100 for processing and provides the external data to health information estimation device 100. Server device 2 may be a cloud including multiple server devices. For example, server device 2 can transmit and receive information to and from health information estimation device 100, terminal device 10, sensor device 5, etc. via a predetermined network such as the Internet. Server device 2 transmits external data to health information estimation device 100. In response to a request from health information estimation device 100, server device 2 transmits external data corresponding to the request from health information estimation device 100 to health information estimation device 100.

[0083] The above is merely an example, and health information estimation system 1 can have any device configuration as long as it can achieve the desired processing. For example, health information estimation system 1 may have various sensors for collecting water-related sensor data, such as toilet sensor data. For example, health information estimation system 1 may have a door sensor that detects the opening and closing of the toilet door as an entry detection sensor for collecting information on users entering and exiting the toilet space. In this case, the door sensor communicates with sensor device 5 and transmits information on users entering and exiting the toilet to sensor device 5. As such, the above system configuration is merely an example, and health information estimation system 1 may have any system configuration as long as it can achieve the desired processing.

[0084] Furthermore, the health information estimation apparatus 100 may receive data detected by the sensor device 5 via the terminal device 10. For example, the health information estimation apparatus 100 may receive data about a user detected by the sensor device 5 via the user's terminal device 10. In this case, the terminal device 10 is communicably connected to the sensor device 5 via a predetermined network such as the Internet, and transmits and receives information to and from the sensor device 5. Note that any communication method may be employed, and the terminal device 10 may communicate with the sensor device 5 via a predetermined wireless communication function such as Bluetooth (registered trademark) or Wi-Fi (Wireless Fidelity) (registered trademark).

[0085] Terminal device 10 receives bathroom sensor data, such as toilet sensor data, detected for a user using terminal device 10 from sensor device 5. Terminal device 10 then transmits the bathroom sensor data, such as toilet sensor data, received from sensor device 5 to health information estimation device 100. In this case, terminal device 10 transmits the bathroom sensor data, such as toilet sensor data, received from sensor device 5 to health information estimation device 100 along with information identifying the user using terminal device 10 (e.g., user ID). Having received the bathroom sensor data and information identifying the user (e.g., user ID) from terminal device 10, health information estimation device 100 associates the bathroom sensor data with the information identifying the user (e.g., user ID) and registers them in storage unit 120 (e.g., data storage unit 121).

[0086] <3. Functional configuration of the health information estimation device> The functional configuration of the health information estimation device 100 will be described below with reference to Fig. 5. Fig. 5 is a block diagram showing an example of the configuration of the health information estimation device according to an embodiment.

[0087] 5, health information estimation apparatus 100 includes communication unit 110, storage unit 120, and control unit 130. Health information estimation apparatus 100 may also include an input unit (e.g., a keyboard or mouse) for accepting various operations from an administrator of health information estimation apparatus 100, and a display unit (e.g., a liquid crystal display) for displaying various information.

[0088] The communication unit 110 is realized by, for example, a communication circuit. The communication unit 110 is connected to a predetermined network such as the Internet via a wired or wireless connection, and transmits and receives information to and from an external information processing device. For example, the communication unit 110 transmits and receives information to and from other devices having communication functions, such as the server device 2, the terminal device 10, and the sensor device 5, via the predetermined network such as the Internet.

[0089] The storage unit 120 is realized by, for example, a semiconductor memory element such as a random access memory (RAM) or a flash memory, or a storage device such as a hard disk or an optical disk. For example, the storage unit 120 is a computer-readable recording medium that non-temporarily records data used by the health information estimation program.

[0090] As shown in FIG. 5, the storage unit 120 according to the embodiment includes a data storage unit 121, a model information storage unit 122, and a user information storage unit 123.

[0091] Data storage unit 121 according to the embodiment stores various data used for processing by health information estimation device 100. Data storage unit 121 stores various data used to estimate health information of users who use the wet space. Data storage unit 121 stores various data acquired about users who use the wet space.

[0092] The data storage unit 121 stores bathroom sensor data such as toilet sensor data. The data storage unit 121 associates each piece of data based on detection by the sensor device 5 with information on the date and time the data was acquired and stores it as bathroom sensor data. For example, the data storage unit 121 stores toilet sensor data acquired for the toilet space PS1. For example, the data storage unit 121 stores toilet sensor data including gas data, feces data, urine data, biological data, etc. detected in the toilet space PS1. For example, the data storage unit 121 stores gas data including data on one of various gases such as odorless gas and foul-smelling gas, and feces data including data on at least one of the amount, shape, and color of feces.

[0093] In this way, the data storage unit 121 may store information such as stool shape, stool color, stool amount, odorless gas (amount), malodorous gas (amount), urine color, urine amount, urine flow rate, and urine components. For example, various types of data such as stool shape, stool color, stool amount, odorless gas (amount), malodorous gas (amount), urine color, urine amount, urine flow rate, and urine components may be generated by the sensor device 5 or by the health information estimation device 100 through analytical processing. For example, data such as stool shape, stool color, stool amount, urine color, urine amount, urine flow rate, and urine components may be generated based on images of stool or urine. This point will be described using an example of stool shape. The health information estimation device 100 may determine stool shape using AI (artificial intelligence) technology. For example, the health information estimation device 100 may determine stool shape using a shape determination model, which is an AI model (machine learning model) generated by machine learning.

[0094] In this case, the shape determination model is trained in advance using training data that indicates classification judgments. This training data includes multiple combinations of images containing stool (also called "stool images") and labels (correct answer information) that indicate the shape (one of Types 1 to 7) of the mass (stool) contained in the stool images. For example, Types 1 to 7 may be seven classifications of stool shape based on the Bristol Scale. For example, Types 1 to 7 correspond to round, hard, cracked, banana-shaped, soft, muddy, and watery, respectively.

[0095] For example, the shape determination model is a model that receives a stool image as input and outputs information indicating the shape of the mass (stool) contained in the input stool image. For example, the shape determination model is trained to output label (stool shape) information corresponding to the input stool image when a stool image is input. The shape determination model is trained using various methods related to so-called supervised learning, as appropriate. In this case, the shape determination model may be stored in the memory unit 120, and the health information estimation device 100 may determine the shape of the stool using the shape determination model stored in the memory unit 120. For example, the health information estimation device 100 may perform a learning process to generate the shape determination model.

[0096] Note that the above is merely an example, and the health information estimation device 100 may determine the shape of stool using various information as appropriate. Furthermore, the seven levels of round, hard, cracked, banana-shaped, soft, muddy, and watery are merely examples of shapes, and the health information estimation device 100 may determine other shapes or may determine the shape using six levels or fewer. Furthermore, the sensor device 5 may determine (analyze) the shape of stool. In this case, the sensor device 5 may determine (analyze) the shape of stool using a shape determination model or the like, and the health information estimation device 100 may acquire data indicating the shape of stool from the sensor device 5 as bathroom sensor data. Furthermore, data on types of stool other than shape, such as stool color, stool volume, odorless gas (amount), foul-smelling gas (amount), urine color, urine volume, urine flow rate, and urine components, may also be generated using an AI model in the same way as the stool shape data.

[0097] The data storage unit 121 stores various kinds of bathroom sensor data in addition to toilet sensor data. For example, the data storage unit 121 stores bathroom sensor data acquired for the bathroom space PS2. For example, the data storage unit 121 stores bathroom sensor data including biological data detected in the bathroom space PS2. For example, the data storage unit 121 stores washroom sensor data acquired for the washroom space. For example, the data storage unit 121 stores washroom sensor data including biological data detected in the washroom space. For example, the data storage unit 121 stores kitchen sensor data acquired for the kitchen space. For example, the data storage unit 121 stores kitchen sensor data including biological data detected in the kitchen space.

[0098] For example, the data storage unit 121 may store, as bathroom sensor data, the analysis results of the data detected by the sensor device 5. For example, the data storage unit 121 may store data such as bathroom sensor data, such as toilet sensor data, obtained from the analysis of sensor information detected by the sensor device 5.

[0099] Furthermore, the data storage unit 121 stores external data. For example, the data storage unit 121 stores external data (external information) acquired from the server device 2. The data storage unit 121 stores various data related to the user's behavior. The data storage unit 121 stores exercise data (exercise information) related to exercise. For example, the data storage unit 121 stores various exercise data related to the user's exercise. For example, the data storage unit 121 stores exercise data indicating the content of the user's exercise. For example, the data storage unit 121 stores exercise data indicating the amount of exercise the user performed. For example, the data storage unit 121 stores exercise data indicating the duration of the user's exercise.

[0100] The data storage unit 121 stores meal data (meal information) related to meals. For example, the data storage unit 121 stores meal data indicating the content of the user's meals. For example, the data storage unit 121 stores meal data indicating the amount of food the user eats. For example, the data storage unit 121 stores meal data indicating the time of the user's meals. The data storage unit 121 stores sleep data (sleep information) related to sleep. For example, the data storage unit 121 stores sleep data indicating the content of the user's sleep. For example, the data storage unit 121 stores sleep data indicating the amount of sleep the user gets. For example, the data storage unit 121 stores sleep data indicating the duration of sleep.

[0101] The data storage unit 121 is not limited to the above, and may store various types of information depending on the purpose. For example, when there are multiple users, the data storage unit 121 stores information about the users in association with information identifying the users (e.g., user IDs, etc.). For example, the data storage unit 121 stores information about the cleaning of the sensor device 5. For example, the data storage unit 121 stores data indicating the number of cleanings. Furthermore, the data storage unit 121 may be divided into a plumbing sensor data storage unit that stores plumbing sensor data based on detection by the sensor device 5, and an external data storage unit that stores external data.

[0102] The model information storage unit 122 according to the embodiment stores information related to models. For example, the model information storage unit 122 stores information on a plurality of models (health information estimation models) used to estimate a user's health information. FIG. 6 is a diagram illustrating an example of the model information storage unit according to the embodiment. The model information storage unit 122 shown in FIG. 6 includes items such as "model ID," "health information estimation model," "data structure," and "model data."

[0103] "Model ID" indicates identification information for identifying the model. "Health Information Estimation Model" indicates the content of the model identified by the model ID. The content of the model indicated by the information in the "Health Information Estimation Model" item may include various information related to the content of the model, such as the name, purpose, and type of the model.

[0104] "Data structure" indicates a data structure associated with a health information estimation model. "Data structure" indicates a data structure for which the corresponding health information estimation model is used. The information in the "Data structure" item may include various information related to the contents of the data structure, such as the data array (pattern) corresponding to the data structure.

[0105] "Model data" refers to the data of a model. Figure 6 shows an example in which conceptual information such as "MDT1" is stored in "Model data," but in reality, it contains various information that makes up the model, such as information about the network included in the model and functions.

[0106] In the example shown in Figure 6, model A, which is a model identified by model ID "M1", is a health information estimation model used when the data structure is data structure a. Also, it is shown that the model data of model A is model data MDT1.

[0107] For example, model A may be a model that, in response to input of data corresponding to data structure a, outputs health information of a user estimated from the data. In this case, model A may be a machine learning model (also called an "AI model") trained by machine learning.

[0108] For example, when the data structure of data acquired about a user is data structure a, model A may be an AI model that outputs health information estimated about the user in response to input of the data. For example, model A may be an AI model that outputs a score (value) indicating a certain health condition (e.g., bowel condition, nocturia, overactive bladder, menstrual cycle, stress, frailty, etc.) in response to input of data corresponding to data structure a. Note that, as described above, bowel condition may be estimated into more specific conditions such as constipation, diarrhea, and bowel environment. In this case, for example, when model A estimates bowel condition, model A may be an AI model that outputs a score (value) indicating the good or bad state of at least one of constipation, diarrhea, bowel environment, etc. in response to input of data corresponding to data structure a.

[0109] Model A may also be an AI model that, in response to input of data corresponding to data structure a, outputs information indicating a classification of the user's health condition corresponding to the input. For example, Model A may be a classifier (identifier) ​​that, in response to input of data corresponding to data structure a, outputs information indicating whether the user's health condition corresponding to the input corresponds to (is classified as) good (health condition), bowel condition, nocturia, overactive bladder, menstrual cycle, stress, frailty, etc. In this case, bowel condition may mean poor bowel condition, nocturia may mean nocturia, overactive bladder may mean overactive bladder, menstrual cycle may mean an irregular menstrual cycle or other undesirable condition, stress may mean a state of excessive stress, and frailty may mean a state of frailty. In this way, Model A may be a function that generates new information based on data, such as an AI model. The term "function" here is not limited to an AI model, but may also include various functions such as programs and algorithms, as long as they generate new information based on data.

[0110] Model A is not limited to the above-described function, and may be any model capable of estimating a user's health information. For example, Model A may be information indicating the health condition of a user from whom data corresponding to data structure a has been acquired. In this case, Model A may be information (also referred to as "type information") indicating the type of health condition to which a user from whom data corresponding to data structure a has been acquired is classified. For example, Model A may be type information indicating whether a user from whom data corresponding to data structure a has been acquired is classified as being in good health, intestinal condition, nocturia, overactive bladder, menstrual cycle, stressed, frail, etc.

[0111] Furthermore, model B, which is a model identified by model ID "M2", is a health information estimation model used when the data structure is data structure b. Furthermore, it is indicated that the model data of model B is model data MDT2. Note that model B, like model A, may be a function such as an AI model, or may be information indicating the user's health condition, such as type information.

[0112] Furthermore, model C, which is a model identified by model ID "M3", is a health information estimation model used when the data structure is data structure c. Furthermore, it is indicated that the model data of model C is model data MDT3. Note that model C, like model A, may be a function such as an AI model, or may be information indicating the user's health condition, such as type information.

[0113] Furthermore, model D, which is a model identified by model ID "M4", is a health information estimation model used when the data structure is data structure d. Furthermore, it is indicated that the model data of model D is model data MDT4. Note that model D, like model A, may be a function such as an AI model, or may be information indicating the user's health condition, such as type information.

[0114] The model information storage unit 122 is not limited to the above, and may store various types of information depending on the purpose. For example, the above models A to D are merely examples, and the model information storage unit 122 may store other models, or may store five or more models.

[0115] The user information storage unit 123 according to the embodiment stores user information. For example, the user information storage unit 123 stores various pieces of user information relating to users who use the bathroom space. For example, the user information storage unit 123 stores information identifying a user (user ID) in association with the user's information.

[0116] The user information storage unit 123 stores various information related to the attributes of a user identified by a user ID. For example, the user information storage unit 123 stores various types of attribute information of a user, such as demographic attribute information such as age and gender, and psychographic attribute information such as lifestyle, interests, etc. For example, the attribute information may include weight, height, mental state, medical history, etc.

[0117] The user information storage unit 123 may store various types of information according to the purpose, without being limited to the above. For example, the user information storage unit 123 stores user information collected by user input. In this case, the user information storage unit 123 may store information indicating data (compensated data) compensated by the user.

[0118] Note that the above is merely an example, and storage unit 120 stores various types of information used in processing. In addition to the above, storage unit 120 stores various types of information used in the health information estimation process. For example, storage unit 120 stores list information (data structure list) that associates data structures with data array patterns.

[0119] Control unit 130 is realized, for example, by a central processing unit (CPU) or a graphics processing unit (GPU) executing a program (e.g., a health information estimation program according to the present disclosure) stored in health information estimation device 100 using RAM or the like as a work area. Control unit 130 is also realized, for example, by an integrated circuit such as an application specific integrated circuit (ASIC) or a field programmable gate array (FPGA).

[0120] The control unit 130 has an acquisition unit 131, a determination unit 132, an estimation unit 133, and a transmission unit 134, and realizes or executes the functions and actions of information processing described below. Note that the internal configuration of the control unit 130 is not limited to the configuration shown in Fig. 4, and may be any other configuration that performs the information processing described below.

[0121] The acquisition unit 131 acquires various types of information. The acquisition unit 131 acquires various types of information from the storage unit 120. The acquisition unit 131 receives various types of information from various computers such as the sensor device 5 and the terminal device 10. The acquisition unit 131 acquires various types of data (information) used in the estimation process, such as plumbing sensor data, external data, and supplementary data. For example, the acquisition unit 131 acquires plumbing sensor data from the data storage unit 121. For example, the acquisition unit 131 acquires external data from the data storage unit 121. For example, the acquisition unit 131 acquires model information from the model information storage unit 122. For example, the acquisition unit 131 acquires user information from the user information storage unit 123.

[0122] The acquisition unit 131 functions as a data acquisition means for acquiring data (bathroom sensor data, etc.) detected by the sensor device 5 arranged in the bathroom space PS. The acquisition unit 131 functions as a first data acquisition means for acquiring data (toilet sensor data, etc.) detected by the sensor device 5 arranged in the toilet space. The acquisition unit 131 functions as a second data acquisition means for acquiring external data.

[0123] The acquisition unit 131 receives bathroom sensor data from the sensor device 5 of each bathroom space PS, thereby acquiring the bathroom sensor data from the sensor device 5 of each bathroom space PS. The acquisition unit 131 stores (registers) the bathroom sensor data acquired from the sensor device 5 of each bathroom space PS in the data storage unit 121. The acquisition unit 131 acquires toilet sensor data from the sensor device 5 of the toilet space PS1. The acquisition unit 131 acquires toilet sensor data including gas data, feces data, urine data, biological data, etc. from the sensor device 5 of the toilet space PS1. For example, the acquisition unit 131 acquires gas data including data on one of various gases such as odorless gas and foul-smelling gas, and feces data including data on at least one of the amount, shape, and color of feces from the sensor device 5 of the toilet space PS1.

[0124] The acquisition unit 131 acquires bathroom sensor data from the sensor device 5 in the bathroom space PS2. The acquisition unit 131 acquires bathroom sensor data including biological data and the like from the sensor device 5 in the bathroom space PS2. The acquisition unit 131 acquires washroom sensor data from the sensor device 5 in the washroom space. The acquisition unit 131 acquires washroom sensor data including biological data and the like from the sensor device 5 in the washroom space. The acquisition unit 131 acquires kitchen sensor data from the sensor device 5 in the kitchen space. The acquisition unit 131 acquires kitchen sensor data including biological data and the like from the sensor device 5 in the kitchen space.

[0125] The determination unit 132 executes a determination process to determine various pieces of information. For example, the determination unit 132 executes the determination process using various pieces of information stored in the storage unit 120. For example, the determination unit 132 executes the determination process using various pieces of information acquired by the acquisition unit 131.

[0126] The determination unit 132 performs the determination process using information received from various computers such as the server device 2, the sensor device 5, and the terminal device 10. For example, the determination unit 132 performs the determination process based on a plumbing log including sensor information detected by the sensor device 5. For example, the determination unit 132 performs the determination process based on external data acquired from the server device 2.

[0127] The determination unit 132 functions as a data structure determination means that determines a data structure based on a data array configured by associating, based on date and time information, the data acquired by the acquisition unit 131. The determination unit 132 determines a data structure based on a data array configured by associating, based on date and time information, data including at least one of plumbing sensor data and external data.

[0128] The determination unit 132 determines a data structure based on a data array formed by associating the data acquired by the acquisition unit 131 based on personal information or date information. The determination unit 132 determines a data structure based on a combination of at least one piece of data in the data array.

[0129] The determination unit 132 compensates for any missing data in the latest data sequence using a past data sequence within a predetermined date and time range, and determines a data structure based on the compensated data sequence. The determination unit 132 determines the data structure based on the time series analysis results for each piece of data acquired by the acquisition unit 131 within the predetermined date and time range. The determination unit 132 prompts the user to compensate for any missing data in the data sequence within the predetermined date and time range. For example, the determination unit 132 prompts the user to compensate for the external data by causing the transmission unit 134 to transmit information requesting the user to compensate for the external data.

[0130] The estimation unit 133 performs an estimation process to estimate various pieces of information. For example, the estimation unit 133 performs the estimation process using various pieces of information stored in the storage unit 120. For example, the estimation unit 133 performs the estimation process using various pieces of information acquired by the acquisition unit 131. For example, the estimation unit 133 performs the estimation process using various pieces of information acquired by the determination unit 132.

[0131] The estimation unit 133 performs the estimation process using information received from various computers such as the server device 2, the sensor device 5, and the terminal device 10. For example, the estimation unit 133 performs the estimation process based on a plumbing log including sensor information detected by the sensor device 5. For example, the estimation unit 133 performs the estimation process based on external data acquired from the server device 2.

[0132] Estimation unit 133 functions as a health information estimation means for estimating the health information of the user based on the data array. Estimation unit 133 selects a model from multiple health information estimation models according to the data structure and estimates the health condition of the user.

[0133] For example, if the health information estimation model is a function such as an AI model, estimation unit 133 inputs data into the selected model and estimates the user's health condition using the information output by the model. For example, if the selected model is an AI model that outputs a score (value) indicating the health condition in response to input data, estimation unit 133 inputs data into the selected model and estimates whether the user of the data is in that health condition based on the score output by the model.

[0134] In this case, for example, when the selected model is an AI model that outputs a score (value) indicating nocturia in response to input data, if the score output by the model is equal to or greater than a predetermined threshold when data is input into the selected model, the estimation unit 133 estimates that the user of the data has nocturia. Also, for example, when the selected model is an AI model that outputs a score (value) indicating nocturia in response to input data, if the score output by the model is less than a predetermined threshold when data is input into the selected model, the estimation unit 133 estimates that the user of the data does not have nocturia.

[0135] Furthermore, for example, when the selected model is an AI model that outputs information indicating a classification of a user's health condition in response to input of data, the estimation unit 133 inputs data into the selected model and estimates the health condition of the user of the data based on the information output by the model. For example, by inputting data into the selected model, the estimation unit 133 estimates whether the health condition of the user of the data is good (health condition), intestinal condition, nocturia, overactive bladder, menstrual cycle, stress, frailty, etc. based on the information output by the model.

[0136] The estimation unit 133 may execute a learning process to learn an AI model used in the estimation process. For example, the estimation unit 133 uses learning data (teacher data, etc.) to learn the AI ​​model. This learning data includes multiple pieces of learning data that combine data with labels (correct answer information) that indicate the health status of the user corresponding to the data. When the estimation unit 133 executes the learning process, the storage unit 120 may store the learning data used in the learning process.

[0137] For example, when an AI model is used to estimate whether a user has a certain health condition, it is a model that takes data as input and outputs information indicating the user's health condition corresponding to the input data (the value (score) of that health condition).For example, when an AI model is used to estimate the classification of a user's health condition, it is a model that takes data as input and outputs information indicating the user's health condition corresponding to the input data (the classification result of the health condition).

[0138] For example, when data is input, the AI ​​model is trained to output label information corresponding to the input data. The AI ​​model is trained using various methods related to so-called supervised learning, as appropriate. Note that the AI ​​model training is not limited to the above, and any learning method can be adopted as long as it can generate a desired AI model. Furthermore, if the health information estimation device 100 does not perform the learning process but acquires an AI model from another device (such as a model providing device), the estimation unit 133 does not need to perform the learning process. In this way, the health information estimation device 100 may generate a model to be used in the estimation process, or may use the model to be used in the estimation process to acquire it from another device such as a model providing device.

[0139] Furthermore, when the health information estimation model is information indicating the user's health condition, such as type information, estimation unit 133 estimates the user's health condition based on the health condition indicated by the selected model. For example, when the health information estimation model is type information indicating the type (classification) of the user's health condition, estimation unit 133 estimates that the health condition indicated by the selected model is the user's health condition.

[0140] In this case, for example, when the type information, which is a health information estimation model, indicates that the user's health condition is good, the estimation unit 133 estimates that the user's health condition is good. For example, when the type information, which is a health information estimation model, indicates that the user's health condition is poor intestinal health, the estimation unit 133 estimates that the user's health condition is poor intestinal health. For example, when the type information, which is a health information estimation model, indicates that the user's health condition is stressed and overactive bladder, the estimation unit 133 estimates that the user's health condition is stressed and overactive bladder.

[0141] The estimation unit 133 selects a model corresponding to the latest data structure from among the plurality of health information estimation models and estimates the health condition of the user. The estimation unit 133 selects a model from among the plurality of health information estimation models according to the data structure based on the time series analysis results for each piece of data acquired by the acquisition unit 131 within a predetermined date and time range and estimates the health condition of the user.

[0142] The estimation unit 133 selects a model from among multiple health information estimation models according to multiple data structures based on the time series analysis results for each piece of data acquired by the acquisition unit 131 within a predetermined date and time range, and estimates the user's health condition. The estimation unit 133 corrects the health information estimation model based on a correction value calculated according to the data acquired by the acquisition unit 131. The estimation unit 133 estimates the user's health information at a predetermined cycle. For example, the estimation unit 133 estimates the user's health information at a predetermined cycle, such as at 12:00 every day, every Monday, or the first day of every month. For example, the estimation unit 133 may estimate the user's health information at a predetermined cycle, such as when new data is acquired.

[0143] Furthermore, the estimation unit 133 may have a function as information generation means for generating various pieces of information to be notified by the transmission unit 134. In this case, the estimation unit 133 generates various pieces of information using information acquired from other devices, information stored in the storage unit 120, etc. The estimation unit 133 generates information using information acquired by the acquisition unit 131. The estimation unit 133 generates information using an estimation result.

[0144] The estimation unit 133 generates various information such as a screen (image information) to be provided to an external information processing device by appropriately using various techniques. The estimation unit 133 generates a screen (image information) to be provided to the terminal device 10. For example, the estimation unit 133 generates a screen (image information) to be provided to the terminal device 10 based on information stored in the storage unit 120.

[0145] The estimation unit 133 may generate a screen (image information) or the like by any process as long as the screen (image information) or the like to be provided to an external information processing device can be generated. For example, the estimation unit 133 generates a screen (image information) to be provided to the terminal device 10 by appropriately using various technologies related to image generation, image processing, etc. For example, the estimation unit 133 generates a screen (image information) to be provided to the terminal device 10 by appropriately using various technologies such as Java (registered trademark).

[0146] The transmitting unit 134 transmits information to an external information processing device. For example, the transmitting unit 134 transmits various information to computers such as the server device 2, the sensor device 5, and the terminal device 10. The transmitting unit 134 notifies the user of information related to the user's health information. The transmitting unit 134 notifies the user by transmitting information estimated by the estimation unit 133 to the terminal device 10. The transmitting unit 134 transmits information generated by the estimation unit 133 to the terminal device 10. The transmitting unit 134 prompts the user to fill in external data to fill in missing data arrays in response to an instruction from the determining unit 132. The transmitting unit 134 transmits information requesting the user to fill in external data to the terminal device 10 used by the user.

[0147] <4. Other processing examples> 2 and 3 are merely examples, and health information estimation system 1 may execute various processes other than those described above. This point will be explained using several examples. Note that explanations of points similar to those described above will be omitted where appropriate.

[0148] For example, health information estimation system 1 may execute processing according to a processing flow as shown in FIG. 7, rather than being limited to that shown in FIG. 2. FIG. 7 is a diagram showing an example of a processing flow in a health information estimation system. As shown in second processing flow PF2 in FIG. 7, in health information estimation system 1, health information estimation device 100 acquires bathroom sensor data using first data acquisition means (e.g., corresponding to acquisition unit 131 in FIG. 5). Health information estimation device 100 also acquires external data using first data acquisition means (e.g., corresponding to acquisition unit 131 in FIG. 5).

[0149] Then, based on the data arrangement of the acquired plumbing sensor data and external data, the health information estimation device 100 determines the data structures of the plumbing sensor data and external data using a data structure determination means (e.g., corresponding to determination unit 132 in Figure 5 ).The health information estimation device 100 then selects a health information estimation model (model) according to the determined data structures of the plumbing sensor data and external data using a health information estimation means (e.g., corresponding to estimation unit 133 in Figure 5 ), and estimates the user's health information using the selected model.

[0150] For example, as shown in dataset DS2 in FIG. 8, health information estimation device 100 determines the data structure of the acquired plumbing sensor data and external data based on the data arrangement of the data, and selects a health information estimation model (model) according to the determined data structure. FIG. 8 is a diagram showing an example of data and processing. Data set DS2 in FIG. 8 includes a data group in which a time, which is date and time information, is associated with data acquired at the corresponding date and time. FIG. 8 illustrates a case in which, as shown in the column labeled "Sensor Acquired Data," plumbing sensor data of types corresponding to data #1 to #3 is acquired, and as shown in the column labeled "External Data," external data of types corresponding to data #1 to #3 is acquired. Hereinafter, data #1 to #3 of the plumbing sensor data may be referred to as plumbing data #1 to #3, and data #1 to #3 of the external data may be referred to as external data #1 to #3.

[0151] The plumbing sensor data of the types corresponding to plumbing data #1 to #3 in FIG. 8 may be any type of data. For example, plumbing data #1 may be gas data of odorless gas, and plumbing data #2 may be gas data of foul-smelling gas. Furthermore, for example, plumbing data #3 may be urine data. Note that the above is merely an example, and plumbing data #1 to #3 may be plumbing sensor data of various types.

[0152] The types of external data corresponding to external data #1 to #3 in Fig. 8 may be any type of data. For example, external data #1 may be dietary data, external data #2 may be sleep data, and external data #3 may be exercise data. Note that the above is merely an example, and external data #1 to #3 may be various types of external data.

[0153] 8 shows that at the time corresponding to "2024 / 3 / 1 8:00", data of the type of plumbing data #1, data of the type of plumbing data #2, data of the type of external data #1, and data of the type of external data #2 were acquired. In other words, at the time corresponding to "2024 / 3 / 1 8:00", data of the type of plumbing data #3 and data of the type of external data #3 were not acquired.

[0154] FIG. 8 also shows that data of the following types was acquired at the time corresponding to "2024 / 3 / 1 12:00": plumbing data #2, #3, and external data #1, #2. FIG. 8 also shows that data of the following types was acquired at the time corresponding to "2024 / 3 / 2 9:30": plumbing data #3, and external data #1 to #3. FIG. 8 also shows that data of the following types was acquired at the time corresponding to "2024 / 3 / 3 18:00": plumbing data #1, #3, and external data #1 to #3.

[0155] 8, the health information estimation device 100 determines the data structure shown in the eighth column based on the data in the first to seventh columns from the left in dataset DS2, i.e., the arrangement of data in the seven columns (time, first sensor acquired data (#1 to #3), and external data (#1 to #3)). For example, the health information estimation device 100 determines that the data structure of the data corresponding to "2024 / 3 / 1 8:00" is data structure a, because the data array for the data corresponding to "2024 / 3 / 1 8:00" includes data types of plumbing data #1 and #2 and external data #1 and #2.

[0156] Furthermore, for example, for the data corresponding to "2024 / 3 / 1 12:00", the health information estimation device 100 determines the data structure of the data corresponding to "2024 / 3 / 1 12:00" to be data structure b, because the data array includes data of types Plumbing Data #2 and #3 and External Data #1 and #2. Similarly, the health information estimation device 100 determines the data structures of the data corresponding to "2024 / 3 / 2 9:30" and "2024 / 3 / 3 18:00" to be data structures c and d, respectively.

[0157] For example, the health information estimation device 100 may determine a data structure corresponding to a data sequence using a list (data structure list) that associates data structures with data sequence patterns. For example, the health information estimation device 100 searches the data structure list to identify data structure a associated with a data sequence including data of the following types: plumbing data #1, #2, and external data #1, #2, and determines the identified data structure a as the data structure for the data corresponding to "2024 / 3 / 1 8:00."

[0158] The health information estimation device 100 then selects a health information estimation model shown in column 9 based on the data structure shown in column 8. For example, for the data corresponding to "2024 / 3 / 1 8:00," the health information estimation device 100 selects model A as the health information estimation model corresponding to "2024 / 3 / 1 8:00" because the data structure is data structure a. Similarly, for the data corresponding to "2024 / 3 / 1 12:00," "2024 / 3 / 2 9:30," and "2024 / 3 / 3 18:00," the health information estimation device 100 selects models B, C, and D as the health information estimation models corresponding to "2024 / 3 / 1 12:00," "2024 / 3 / 2 9:30," and "2024 / 3 / 3 18:00," respectively, because the data structures are data structures b, c, and d. The health information estimation device 100 then estimates health information using the selected model.

[0159] Furthermore, for example, health information estimation system 1 may execute processing according to a processing flow as shown in FIG. 9. FIG. 9 is a diagram showing an example of a processing flow in the health information estimation system. As shown in third processing flow PF3 in FIG. 9, in health information estimation system 1, health information estimation device 100 acquires toilet sensor data (referred to as "toilet sensor data" in the figure) as bathroom sensor data using first data acquisition means (e.g., corresponding to acquisition unit 131 in FIG. 5). Health information estimation device 100 also acquires external data using first data acquisition means (e.g., corresponding to acquisition unit 131 in FIG. 5).

[0160] Then, based on the data arrangement of the acquired toilet sensor data and external data, the health information estimation device 100 determines the data structures of the toilet sensor data and external data using a data structure determination means (e.g., corresponding to the determination unit 132 in FIG. 5 ).The health information estimation device 100 then selects a health information estimation model (model) according to the determined data structures of the toilet sensor data and external data using a health information estimation means (e.g., corresponding to the estimation unit 133 in FIG. 5 ), and estimates the user's health information using the selected model.

[0161] In addition, in the above example, as shown in datasets DS1, DS2, etc., a case was described in which the data structure was determined and a model was selected based solely on whether or not each type of data was acquired. However, the health information estimation device 100 may also determine the data structure and select a model based on the amount and content of each type of data.

[0162] For example, as shown in dataset DS3 in Figure 10, the health information estimation device 100 determines the data structure of the acquired toilet sensor data based on the data array of the toilet sensor data, and selects a health information estimation model (model) according to the determined data structure. Figure 10 is a diagram showing an example of data and processing. Data set DS3 in Figure 10 includes a data group that associates five types of data: the amount of odorless gas detected by the first detection sensor (odorless gas sensor), the amount of malodorous gas detected by the second detection sensor (malodorous gas sensor), and the amount, shape, and color of stool detected by the third detection sensor (stool sensor).

[0163] If the elements (cells) in the third row and onward in the columns corresponding to the amount of odorless gas, the amount of odorous gas, the amount of stool, the shape of stool, and the color of stool are not blank, it indicates that the corresponding data has been acquired. The acquired data (record) in the third row of Fig. 10 indicates that the amount of odorless gas and the amount of odorous gas have been acquired. In other words, the acquired data (record) in the third row of Fig. 10 indicates that data for the three types of stool amount, shape of stool, and color of stool has not been acquired. Furthermore, the acquired data (record) in the third row of Fig. 10 indicates that data indicating that the amount of odorless gas is 1 and the amount of odorous gas is 1 have been acquired.

[0164] The acquired data (record) on the fourth line of FIG. 10 indicates that the amount of odorless gas and the amount of malodorous gas were acquired. The acquired data (record) on the fourth line of FIG. 10 indicates that data was acquired indicating that the amount of odorless gas is 0.5 and the amount of malodorous gas is 2. The acquired data (record) on the fifth line of FIG. 10 indicates that the amount of odorless gas, the amount of malodorous gas, the amount of stool, the shape of stool, and the color of stool were acquired. The acquired data (record) on the third line of FIG. 10 indicates that data was acquired indicating that the amount of odorless gas is 1, the amount of malodorous gas is 1, the amount of stool is 4, the shape of stool is type 2, and the color of stool is brown.

[0165] 10, health information estimation device 100 determines the data structure shown in the sixth column based on the data in the first to fifth columns from the left in dataset DS3. For example, because the acquired data (record) in the third row of Fig. 10 has a data array in which the amount of odorless gas is 1 and the amount of malodorous gas is 1, health information estimation device 100 determines the data structure of the acquired data (record) in the third row of Fig. 10 to be gas type (data structure #11).

[0166] Furthermore, for example, the health information estimation apparatus 100 determines the data structures for the acquired data (records) on the fourth and fifth lines of FIG. 10 as a malodorous gas type (data structure #12) and a gas and feces type (data structure #13).

[0167] The health information estimation device 100 then selects the health information estimation model shown in the seventh column based on the data structure shown in the sixth column. For example, for the acquired data (record) on the third row of FIG. 10, the data structure is gas type (data structure #11), so the health information estimation device 100 selects gas type (model #11) as the health information estimation model corresponding to the acquired data (record) on the third row of FIG. 10. Similarly, for the acquired data (record) on the fourth and fifth rows of FIG. 10, the data structure is malodorous gas type (data structure #12) and gas + feces type (data structure #13), respectively, so the health information estimation device 100 selects malodorous gas type (model #12) and gas + feces type (model #13) as the health information estimation models corresponding to the acquired data (record) on the fourth and fifth rows of FIG. 10. The health information estimation device 100 then estimates health information using the selected model.

[0168] Furthermore, the data such as the plumbing sensor data and external data may be associated with various information other than date and time information. For example, the data such as the plumbing sensor data and external data may be associated with personal information of the user corresponding to the data. This point will be explained using FIG. 11 as an example.

[0169] For example, as shown in dataset DS4 in FIG. 11, the health information estimation device 100 determines a data structure using data that associates acquired bathroom sensor data and external data with personal information (identification information such as a user ID) of the user corresponding to the data, and selects a health information estimation model (model) based on the determined data structure. FIG. 11 is a diagram showing an example of data and processing. Dataset DS4 in FIG. 11 includes a data group that associates personal information and a time (date and time information) with data acquired at the corresponding date and time for the user indicated by the personal information. This allows the health information estimation device 100 to appropriately identify which user the estimated health information corresponds to. Note that other aspects are similar to those described in FIG. 8 and other figures, and therefore detailed description will be omitted.

[0170] As described above, health information estimation system 1 determines a data structure according to the data arrangement (combination) of data acquired at a date and time corresponding to the date and time information. For example, as shown in dataset DS11 in Fig. 12, health information estimation system 1 determines the data structure of the plumbing sensor data and external data based on the data arrangement of the plumbing sensor data and external data acquired at a date and time corresponding to the date and time information, and selects a health information estimation model (model) according to the determined data structure.

[0171] FIG. 12 is a diagram showing an example of data and processing. In this way, in FIG. 12, the data structure is determined according to the data arrangement (combination) of data corresponding to the horizontal direction of dataset DS11, i.e., the combination of acquired data types. For example, the health information estimation device 100 determines that the data structure of the data corresponding to "2024 / 3 / 1 8:00" is data structure a because the data arrangement includes data of the following types: plumbing data #1, #2, and external data #1, #2.

[0172] Furthermore, for example, the health information estimation device 100 determines that the data structure of the data corresponding to "2024 / 3 / 1 12:00" is data structure b because the data array includes data of the following types: plumbing data #2, #3, and external data #1, #2.

[0173] Note that the above is merely an example, and health information estimation system 1 may determine the data structure according to the date and time corresponding to the date and time information and the data arrangement (combination) of the data acquired at that date and time. For example, as shown in dataset DS12 in Figure 13, health information estimation system 1 determines the data structure of the date and time, bathroom sensor data, and external data based on the date and time corresponding to the date and time information and the data arrangement of the bathroom sensor data and external data acquired at that date and time, and selects a health information estimation model (model) according to the determined data structure.

[0174] FIG. 13 is a diagram showing an example of data and processing. As shown in FIG. 13, the data structure is determined according to the data arrangement (combination) of the date and time and data corresponding to the horizontal direction of the data set DS12, i.e., the combination of the date and time and the type of acquired data. For example, for the data corresponding to "2024 / 3 / 1 8:00", the health information estimation device 100 determines the data structure of the data corresponding to "2024 / 3 / 1 8:00" as data structure a. 朝 is decided.

[0175] Furthermore, for example, the data corresponding to "2024 / 3 / 1 12:00" is a data array containing data of types #2 and #3 of plumbing data and #1 and #2 of external data at 12:00 AM, i.e., in the daytime. Therefore, the health information estimation apparatus 100 defines the data structure of the data corresponding to "2024 / 3 / 1 12:00" as data structure b. 昼 is decided.

[0176] The health information estimation device 100 then selects the health information estimation model shown in column 9 based on the data structure shown in column 8. For example, the health information estimation device 100 selects the health information estimation model shown in column 9 based on the data structure shown in column 8. For example, for data corresponding to "2024 / 3 / 1 8:00", the health information estimation device 100 selects the health information estimation model shown in column 9 based on the data structure shown in column 8. 朝 Therefore, model A is used as the health information estimation model corresponding to "2024 / 3 / 1 8:00". 朝 Similarly, for the data corresponding to "2024 / 3 / 1 12:00", health information estimation apparatus 100 selects the data structure b 昼 Therefore, model B is used as the health information estimation model corresponding to "2024 / 3 / 1 12:00". 昼 Then, health information estimation device 100 estimates health information using the selected model.

[0177] Health information estimation system 1 may also supplement data. For example, as shown in dataset DS13 in Fig. 14, health information estimation system 1 supplements data using past data, determines the data structures of the plumbing sensor data and external data based on the data arrangement of the supplemented data, and selects a health information estimation model (model) according to the determined data structure.

[0178] FIG. 14 is a diagram illustrating an example of data and processing. In FIG. 14, the health information estimation device 100 supplements two pieces of data, one corresponding to "2024 / 3 / 1 8:00" and the other corresponding to "2024 / 3 / 1 12:00," with the most recent data corresponding to "2024 / 3 / 1 12:00" and the other corresponding to "2024 / 3 / 1 8:00." Specifically, the health information estimation device 100 supplements the data by using the data of the plumbing-related data #1 type acquired at "2024 / 3 / 1 8:00" as the data of the plumbing-related data #1 type that has not been acquired at "2024 / 3 / 1 12:00," which is the most recent data.

[0179] In this case, the health information estimation device 100 determines that the data structure of the data corresponding to "2024 / 3 / 1 12:00" after the compensation is data structure e, because the data array includes data of types Plumbing Data #1-#3 and External Data #1 and #2. In this way, the health information estimation device 100 changes the data structure from data structure b before the compensation to data structure e after the compensation.

[0180] Then, for the data corresponding to "2024 / 3 / 1 12:00", the data structure is data structure e, so the health information estimation device 100 selects model E as the health information estimation model corresponding to "2024 / 3 / 1 12:00". In this way, the health information estimation device 100 changes the model from pre-compensation model B to post-compensation model E. The health information estimation device 100 then estimates health information using the selected model.

[0181] In the above example, the health information estimation system 1 determines the data structure and selects a model based on the combination of data corresponding to the horizontal direction of the dataset DS, i.e., the combination of acquired data types. However, this is not limiting. For example, the health information estimation device 100 may determine the data structure and select a model based on the data arrangement (combination) of data corresponding to the vertical direction of the dataset DS, i.e., the time series combination of acquired data. An example of this point is described below.

[0182] For example, as shown in dataset DS14 in Figure 15, health information estimation device 100 determines the data structure of the acquired data by type based on the data arrangement of the acquired data by type, and selects a health information estimation model (model) according to the determined data structure. Figure 15 is a diagram showing an example of data and processing. Data set DS14 in Figure 15 includes a data group in which a time, which is date and time information, is associated with data acquired on the date and time corresponding to that time. Figure 15 shows a case in which plumbing sensor data of types corresponding to plumbing data #1 to #3 has been acquired.

[0183] 15 shows that data of the plumbing data #1 type was acquired twice, at a time corresponding to "2024 / 3 / 1 8:00" (hereinafter also referred to as "first acquisition timing") and at a time corresponding to "2024 / 3 / 3 18:00" (hereinafter also referred to as "fourth acquisition timing"). In other words, data of the plumbing data #1 type was not acquired at a time corresponding to "2024 / 3 / 1 12:00" (hereinafter also referred to as "second acquisition timing") and at a time corresponding to "2024 / 3 / 2 9:30" (hereinafter also referred to as "third acquisition timing").

[0184] The health information estimation device 100 performs time series analysis based on information indicating the acquisition timing of data of the plumbing data #1 type, and determines the data structure of the data of the plumbing data #1 type. In Figure 15, the health information estimation device 100 determines the data structure of the data of the plumbing data #1 type to be data structure a' based on the time series analysis result that data was acquired at the first acquisition timing and the fourth acquisition timing for the data of the plumbing data #1 type.

[0185] Furthermore, for data of the plumbing data #2 type, the health information estimation device 100 determines the data structure of the data of the plumbing data #2 type to be data structure b' based on the time series analysis result that data was acquired at the first acquisition timing and the second acquisition timing. Furthermore, for data of the plumbing data #3 type, the health information estimation device 100 determines the data structure of the data of the plumbing data #3 type to be data structure c' based on the time series analysis result that data was acquired at the first to third acquisition timings.

[0186] For example, the health information estimation device 100 may determine a data structure corresponding to a data sequence using a list (data structure list) that associates data structures with data sequence patterns for each data type. For example, the health information estimation device 100 searches the data structure list for plumbing data #1 to identify data structure a' associated with a data sequence including data from the first acquisition timing and the fourth acquisition timing, and determines the identified data structure a' as the data structure of the data of plumbing data #1.

[0187] The health information estimation device 100 then selects a health information estimation model for each data type based on the data structure of each data type. For example, for plumbing data #1, the data structure is data structure a', so the health information estimation device 100 selects model A' as the health information estimation model corresponding to plumbing data #1. Similarly, for plumbing data #2 and #3, the data structures are data structures b' and c', so the health information estimation device 100 selects models B' and C' as the health information estimation models corresponding to plumbing data #2 and #3, respectively. The health information estimation device 100 then estimates health information using the selected model. For example, if the data for the plumbing data #2 type is gas data on foul-smelling gases, the health information estimation device 100 may estimate the quality of the user's intestinal environment using the data for the plumbing data #2 type acquired for the user and model B'.

[0188] Furthermore, when acquiring external data, the health information estimation system 1 may determine the data structure and select a model according to the data arrangement (combination) of data corresponding to the vertical direction of the dataset DS, i.e., the time series combination of acquired data, as in Fig. 15. This point will be briefly explained below.

[0189] For example, as shown in dataset DS15 in Figure 16, health information estimation device 100 determines the data structure of the acquired data type based on the data arrangement of the acquired data type, and selects a health information estimation model (model) according to the determined data structure. Figure 16 is a diagram showing an example of data and processing. Data set DS15 in Figure 16 includes a data group in which a time, which is date and time information, is associated with data acquired on the date and time corresponding to that time. Figure 16 shows a case in which plumbing sensor data of types corresponding to plumbing data #1 to #3 and external data of types #1 to #3 have been acquired.

[0190] 16 is the same as the plumbing data #1 to #3 in FIG. 15, and therefore description thereof will be omitted. FIG. 16 shows that data of the external data #3 type was acquired twice, at the time corresponding to "2024 / 3 / 2 9:30" (third acquisition timing) and at the time corresponding to "2024 / 3 / 3 18:00" (fourth acquisition timing). In other words, data of the external data #3 type was not acquired at the time corresponding to "2024 / 3 / 1 8:00" (first acquisition timing) and at the time corresponding to "2024 / 3 / 1 12:00" (hereinafter also referred to as "second acquisition timing").

[0191] The health information estimation device 100 performs time series analysis based on information indicating the acquisition timing of data of the external data #3 type, and determines the data structure of the data of the external data #3 type. In Figure 16, the health information estimation device 100 determines the data structure of the data of the external data #3 type to be data structure d' based on the time series analysis result that data was acquired at the third acquisition timing and the fourth acquisition timing for the data of the external data #3 type.

[0192] The health information estimation device 100 then selects a health information estimation model for each data type based on the data structure of that data type. For example, because external data #3 has data structure d', the health information estimation device 100 selects model D' as the health information estimation model corresponding to external data #3. The health information estimation device 100 then estimates health information using the selected model. Note that, as with external data #3, the health information estimation device 100 may also determine the data structure for external data #1 and #2 based on the data arrangement of that type of data, and select a model based on the determined data structure, as with external data #3.

[0193] When determining the data structure and selecting a model based on the combination of data corresponding to the horizontal direction of the dataset DS, i.e., the combination of acquired data types, the health information estimation device 100 may perform processing using the model selected for the most recent data. This point will be explained using FIG. 17, which is a diagram showing an example of data and processing. The dataset DS16 in FIG. 17 is similar to the dataset DS2 in FIG. 8, and therefore will not be described in detail.

[0194] 17, the most recent data for the user is the data acquired at the time corresponding to "2024 / 3 / 3 18:00." Therefore, health information estimation device 100 estimates the user's health information using model D, which is a health information estimation model selected for the data acquired at the time corresponding to "2024 / 3 / 3 18:00."

[0195] Furthermore, when determining the data structure and selecting a model based on the data arrangement (combination) of data corresponding to the vertical direction of dataset DS, i.e., the time-series combination of acquired data, health information estimation device 100 may perform processing using the model selected for each data type. This point will be explained using FIG. 18, which is a diagram showing an example of data and processing. Dataset DS17 in FIG. 18 is similar to dataset DS14 in FIG. 15, and therefore will not be described in detail.

[0196] 18, model A' is selected for the data of plumbing data #1, model B' is selected for the data of plumbing data #2, and model C' is selected for the data of plumbing data #3. Therefore, the health information estimation device 100 estimates health information using model A' for the data of plumbing data #1, model B' for the data of plumbing data #2, and model C' for the data of plumbing data #3.

[0197] Furthermore, when determining the data structure and selecting a model based on the data arrangement (combination) of data corresponding to the vertical direction of dataset DS, i.e., the time-series combination of acquired data, health information estimation device 100 may select a model for the entire data type. This point will be explained using FIG. 19, which shows an example of data and processing. Dataset DS21 in FIG. 19 is similar to dataset DS14 in FIG. 15 except for the health information estimation model, and therefore detailed explanation other than the selection of the health information estimation model will be omitted.

[0198] 19, the data structure of plumbing data #1 is determined to be data structure a', the data structure of plumbing data #2 is determined to be data structure b', and the data structure of plumbing data #3 is determined to be data structure c'. The health information estimation device 100 then selects model ABC' as the health information estimation model to be used for processing based on the combination of data structure a' of plumbing data #1, data structure b' of plumbing data #2, and data structure c' of plumbing data #3.

[0199] For example, the health information estimation device 100 may determine a health information estimation model corresponding to a data structure combination pattern using a list (health information estimation model list) that associates health information estimation models with data structure combination patterns. For example, the health information estimation device 100 searches the health information estimation model list, identifies model ABC' that corresponds to the combination of data structure a' of bathroom data #1, data structure b' of bathroom data #2, and data structure c' of bathroom data #3, and selects the identified model ABC' as the health information estimation model to be used for processing.

[0200] The health information estimation device 100 then estimates health information using the selected model. For example, the health information estimation device 100 estimates health information using plumbing data #1, plumbing data #2, plumbing data #3, and model ABC'. For example, the health information estimation device 100 inputs plumbing data #1, plumbing data #2, and plumbing data #3 into model ABC', and estimates health information based on the information output by model ABC'.

[0201] The health information estimation device 100 may weight the data for each type (calculate weight values) and perform estimation using a model that reflects the corrections made by the weight values. This point will be explained using FIG. 20, which is a diagram showing an example of data and processing. Data set DS22 in FIG. 20 is the same as data set DS21 in FIG. 19, except that weights are added to the data for each type. Therefore, detailed explanation other than the weighting will be omitted.

[0202] In data set DS22 in Figure 20, the data structure of plumbing data #1 is determined to be data structure a', the data structure of plumbing data #2 is determined to be data structure b', and the data structure of plumbing data #3 is determined to be data structure c'. The health information estimation device 100 also calculates weight values, which are correction values, from statistical values ​​of each type of data. For example, the health information estimation device 100 calculates larger weight values ​​the more data is acquired for each type of data. In Figure 20, the health information estimation device 100 calculates the weight values ​​for plumbing data #1 to be 0.3, for plumbing data #2 to be 0.3, and for plumbing data #3 to be 0.4.

[0203] Then, the health information estimation device 100 determines model A as a health information estimation model to be used for processing based on the combination of the data structure a' of the bathroom data #1 and a weight value of 0.3, the data structure b' of the bathroom data #2 and a weight value of 0.3, and the data structure c' of the bathroom data #3 and a weight value of 0.4. 0.3 B 0.3 C 0.4The health information estimation device 100 then selects model ABC'. The health information estimation device 100 then estimates health information using the selected model. The health information estimation device 100 may select model ABC' as the health information estimation model to use for processing based on a combination of data structure a' of bathroom data #1, data structure b' of bathroom data #2, and data structure c' value of 0.4 of bathroom data #3. In this case, the health information estimation device 100 may use weight values ​​of 0.3 for bathroom data #1, 0.3 for bathroom data #2, and 0.4 for bathroom data #3 as weight values ​​for model ABC'. For example, the health information estimation device 100 may input data from bathroom data #1, data from bathroom data #2, data from bathroom data #3, weight values ​​of 0.3 for bathroom data #1, weight values ​​of 0.3 for bathroom data #2, and weight values ​​of 0.4 for bathroom data #3 into model ABC' and estimate health information based on the information output by model ABC'.

[0204] Furthermore, when determining the data structure and selecting a model based on the data arrangement (combination) of data corresponding to the vertical direction of dataset DS, i.e., the time series combination of acquired data, health information estimation device 100 may select a model based on a combination of the type of plumbing sensor data and the type of external data. This point will be explained using FIG. 21, which is a diagram showing an example of data and processing. Dataset DS23 in FIG. 21 is similar to dataset DS15 in FIG. 16 except for the health information estimation model, and therefore detailed explanation other than the selection of the health information estimation model will be omitted.

[0205] In dataset DS23 in Figure 21, the data structure of plumbing data #1 is determined to be data structure a', and the data structure of external data #3 is determined to be data structure d'. Then, based on the combination of data structure a' of plumbing data #1 and data structure d' of external data #3, health information estimation device 100 selects model AD as the health information estimation model to use for processing. Then, health information estimation device 100 estimates health information using the selected model.

[0206] Furthermore, for example, health information estimation system 1 may execute processing according to a processing flow as shown in Fig. 22. Fig. 22 is a diagram showing an example of a processing flow in the health information estimation system. Note that the fourth processing flow PF11 in Fig. 22 is similar to the third processing flow PF3 in Fig. 9 except that supplementary data is acquired from the user, and therefore detailed description of the process other than the acquisition of supplementary data from the user will be omitted.

[0207] When determining the data structure, if there is data that needs to be supplemented, the health information estimation device 100 requests the user to supplement the data. The health information estimation device 100 then adds the supplemented data acquired from the user to external data and performs processing.

[0208] In this way, health information estimation system 1 can estimate a user's health information based on not only fecal gas but also any kind of plumbing sensor data. For example, health information estimation system 1 can estimate a user's health information based on various types of data, such as feces and urine data obtained in the toilet, and lifestyle behavior data obtained in the bathroom (bath) or kitchen. In the above-described process, health information estimation system 1 structures each piece of data and processes it using statistics and algorithms to estimate health information.

[0209] For example, in conventional systems, when it is determined that additional data other than existing excrement data is needed to estimate health information, the system can request the additional data from the user via a device such as a smartphone, but health information is not determined unless the user responds to the request for additional data, so there is room for improvement. Therefore, in health information estimation system 1, when it is determined that additional data other than the data already obtained is needed to estimate health information, the system requests the additional data from the user, allowing the system to appropriately supplement the data and perform processing.

[0210] <4-1. Examples of other information estimation> The information estimation described above is merely an example, and health information estimation system 1 may estimate various types of information, not limited to the information described above. Some examples of this point are described below. The following describes an example in which health information estimation system 1 estimates at least one of a user's health information and service information to be provided to the user based on data detected by sensors installed in a bathroom space.

[0211] In the following example, health information estimation system 1 determines a data structure based on a data array or data group constructed by associating data detected by sensors placed in a bathroom space based on date and time information. When estimating a user's health information or service information to be provided to the user based on the data, health information estimation system 1 selects and uses a model from multiple information estimation models according to the data structure.

[0212] First, an example of information processing executed by the health information estimation system 1, which is an information estimation system, will be described based on the example shown in Fig. 23. Fig. 23 is a diagram showing an example of data and processing. Note that explanations of points similar to those described above will be omitted as appropriate.

[0213] For example, the health information estimation device 100 determines the data structure of the acquired bathroom sensor data based on the data array of the data, as shown in dataset DS31 in Fig. 23, and selects an estimation model according to the determined data structure. Note that the estimation model here may be the same as or different from the health information estimation model described above.

[0214] For example, the estimation model may be a model that, in response to input of data having a data structure associated with the estimation model, outputs at least one of the user's health information estimated from the data or service information to be provided to the user. In this case, the estimation model may be a machine learning model (AI model) trained by machine learning.

[0215] For example, the estimation model may be an AI model that outputs at least one of health information or service information estimated for a user in response to input of data acquired about the user that has a data structure associated with the estimation model. For example, the estimation model may be an AI model that outputs a score (value) indicating a certain health condition (e.g., bowel condition, nocturia, overactive bladder, menstrual cycle, stress, frailty, excess salt intake, etc.) in response to input of data having a data structure associated with the estimation model.

[0216] Furthermore, for example, the estimation model may be an AI model that, in response to input of data having a data structure associated with the estimation model, outputs service information to be provided to a user who is the source of the data. Note that the estimation model may be any AI model as long as it can output at least one of health information or service information estimated for a user. For example, the estimation model may be an LLM (Large Language Model). In this way, the estimation model may be various AI models, such as a generative AI such as an LLM, as long as it can output desired information.

[0217] Furthermore, the estimation model may be an AI model that, in response to input of data having a data structure associated with the estimation model, outputs information indicating a classification of the user's health condition corresponding to the input. For example, the estimation model may be a classifier (identifier) ​​that, in response to input of data having a data structure associated with the estimation model, outputs information indicating whether the user's health condition corresponding to the input corresponds to (is classified as) good (health condition), bowel condition, nocturia, overactive bladder, menstrual cycle, stress, frailty, excessive salt intake, etc. In this case, for example, excessive salt intake may simply mean excessive salt intake. In this way, the estimation model may be a function that generates new information based on data, such as an AI model. The function here is not limited to an AI model, but includes various functions such as programs and algorithms, as long as it generates new information based on data.

[0218] The estimation model is not limited to the above-described function, and may be any model capable of estimating at least one of the user's health information or the service information to be provided to the user. For example, the estimation model may be information indicating the health condition of a user from whom data having a data structure associated with the estimation model has been acquired. In this case, the estimation model may be information (type information) indicating the type of health condition a user from whom data having a data structure associated with the estimation model has been acquired is classified into. For example, the estimation model may be type information indicating whether a user from whom data having a data structure associated with the estimation model has been acquired is classified into good (health condition), bowel condition, nocturia, overactive bladder, menstrual cycle, stress, frailty, excessive salt intake, etc.

[0219] In FIG. 23, the health information estimation system 1 performs processing using data (urine data) indicating urine components detected by a sensor (also referred to as a "urine sensor") that detects urine components among the excrement excreted into the toilet apparatus 50. In this case, the sensor apparatus 5 has a urine sensor that detects urine components. For example, the urine sensor may be provided on the toilet seat of the toilet apparatus 50. Note that the urine sensor is a known sensor and will not be described in detail, but any sensor may be used as the urine sensor as long as it can detect the desired urine components.

[0220] 23 shows an example in which urine data is used as biological data, but health information estimation system 1 may perform the following processing using various biological data other than urine data. For example, health information estimation system 1 may use various data (biological data) other than urine, such as stool (defecation), fecal gas, skin gas, and blood flow. For example, health information estimation system 1 may perform processing using multiple types of biological data such as stool (defecation), urine, fecal gas, skin gas, and blood flow. Sensor device 5 has various sensors for acquiring biological data used by health information estimation system 1 for processing.

[0221] The various sensors included in the sensor device 5 may be arranged in any position as long as the desired detection is possible. For example, the various sensors included in the sensor device 5 may be arranged not only in the toilet device 50 but also in a remote control that operates the sensor device 5, etc. For example, a sensor that detects biometric data may be arranged in the remote control and detect the biometric data of the user from the user's finger that touches the remote control. The various sensors included in the sensor device 5 may be either contact sensors (contact sensors) or non-contact sensors (non-contact sensors).

[0222] 23 shows a case where component data (also referred to as "urine component data") related to urine, such as urea, sodium, potassium, etc., is acquired based on biological data detected by a urine sensor, as shown in the column labeled "first detection sensor." Data set DS31 in Fig. 23 includes a data group in which, for a time that is date and time information, urine component data acquired on a date and time corresponding to that time is associated with the time.

[0223] 23, health information estimation device 100 of health information estimation system 1 determines the data structure shown in the fifth column based on the data (records) in the first to fourth columns from the left in dataset DS31, i.e., the arrangement of time and urine component data (data array). For example, health information estimation device 100 determines the data structure based on the time of each data array in dataset DS31.

[0224] For example, the health information estimation apparatus 100 determines that the data structure of the data corresponding to "2024 / 11 / 19 8:00" is data structure #31 because the data corresponding to "2024 / 11 / 19 8:00" is a data array of data acquired at 8:00 (8:00 AM), i.e., in the morning. For example, data structure #31 is a urine-morning type data structure.

[0225] Furthermore, for example, the data corresponding to "2024 / 11 / 19 12:00" is a data array of data acquired at 12:00 (12:00 AM), i.e., during the day, so the health information estimation apparatus 100 determines that the data structure of the data corresponding to "2024 / 11 / 19 12:00" is data structure #32. For example, data structure #32 is a urine-daytime type data structure.

[0226] For example, the health information estimation device 100 determines that the data structure of the data corresponding to "2024 / 11 / 20 8:00" is data structure #31 because the data is a data array of data acquired at 8:00 (8:00 AM), i.e., in the morning. For example, data structure #31 is a urine-morning type data structure. Note that the above is merely an example, and the health information estimation device 100 may determine the data structure based on urine component data rather than time.

[0227] The health information estimation device 100 then selects the estimation model shown in the sixth column based on the data structure shown in the fifth column. For example, for the data corresponding to "2024 / 11 / 19 8:00", the data structure is data structure #31, so the health information estimation device 100 selects model #31 as the estimation model corresponding to "2024 / 11 / 19 8:00". For example, model #31 is the estimation model used for urine - morning type (data structure #31).

[0228] Furthermore, for the data corresponding to "2024 / 11 / 19 12:00", the data structure is data structure #32, so the health information estimation device 100 selects model #31 as the estimation model corresponding to "2024 / 11 / 19 12:00". For example, model #32 is an estimation model used for urine - daytime type (data structure #32). For the data corresponding to "2024 / 11 / 20 8:00", the data structure is data structure #31, so the health information estimation device 100 selects model #31 as the estimation model corresponding to "2024 / 11 / 20 8:00".

[0229] For example, for data corresponding to "2024 / 11 / 19 8:00," the health information estimation device 100 uses model #31 to estimate the user's health information or service information to be provided to the user. For example, if model #31 is an estimation model that outputs information indicating whether a user is classified as having high salt intake, the health information estimation device 100 inputs the data corresponding to "2024 / 11 / 19 8:00" into model #31. Then, model #31, to which the data corresponding to "2024 / 11 / 19 8:00" is input, outputs information indicating that the user of the data corresponding to "2024 / 11 / 19 8:00" is classified as having high salt intake. In this case, the health information estimation device 100 estimates that the user of the data corresponding to "2024 / 11 / 19 8:00" is consuming high salt intake.

[0230] For example, if model #31 is an estimation model that outputs service information to be provided to a user, health information estimation device 100 inputs data corresponding to "2024 / 11 / 19 8:00" into model #31. Then, model #31, to which data corresponding to "2024 / 11 / 19 8:00" has been input, outputs service information that suggests health-related advice to the user, such as "You're consuming too much salt, so try using reduced-salt soy sauce."

[0231] Model #31 may output various information, not limited to the above. For example, model #31 may output the uniform resource locator (URL) of an electronic commerce (EC) site selling products such as soy sauce. This allows health information estimation device 100 to estimate service information to be provided to the user of the data corresponding to "2024 / 11 / 19 8:00." In this way, health information estimation device 100 performs estimation processing using a model selected based on the data structure of the data acquired for the user. This allows the health information estimation system to appropriately estimate information related to a person's health.

[0232] The health information estimation device 100 may transmit information based on the estimation process to a user. For example, the health information estimation device 100 may transmit at least one of the user's health information estimated by the estimation process or service information provided to the user to the terminal device 10 used by the user. For example, the health information estimation device 100 may transmit service information estimated by the estimation process for user U1, such as "Your salt intake is too high, so try using reduced-salt soy sauce," to the terminal device 10 used by user U1. Furthermore, for example, the health information estimation device 100 may transmit service information including the URL of an e-commerce site selling products such as soy sauce to the terminal device 10 used by user U1.

[0233] While FIG. 23 shows an example in which estimation processing is performed based on the data array of each data, health information estimation system 1 may also perform processing based on multiple data (records). This point will be explained using FIG. 24. FIG. 24 is a diagram showing an example of data and processing. Note that explanations of points similar to those described above in FIG. 23 etc. will be omitted as appropriate. Hereinafter, multiple data (records) may be referred to as a data group. In FIG. 24, the column labeled "Estimation Model 2" in dataset DS32 indicates an estimation model used for multiple data (records).

[0234] 24, health information estimation apparatus 100 selects model #30 as the estimation model for the data group corresponding to "2024 / 11 / 19 8:00" to "2024 / 11 / 23 8:00." For example, model #30 is a one-week model (estimation model) used for multiple data (data group) acquired within a period spanning multiple days (within one week).

[0235] For example, for a data group corresponding to "2024 / 11 / 19 8:00" to "2024 / 11 / 23 8:00," health information estimation device 100 uses model #30 to estimate a user's health information or service information to be provided to the user. For example, if model #30 is an estimation model that outputs health information, health information estimation device 100 inputs the data group corresponding to "2024 / 11 / 19 8:00" to "2024 / 11 / 23 8:00" into model #30. Then, model #30, to which the data group corresponding to "2024 / 11 / 19 8:00" to "2024 / 11 / 23 8:00" is input, outputs health information indicating the health status of the user from whom the data group was obtained.

[0236] As a result, the health information estimation device 100 estimates the health information of the user from which the data group corresponding to "2024 / 11 / 19 8:00" to "2024 / 11 / 23 8:00" was obtained. Furthermore, for the data group corresponding to "2024 / 11 / 19 8:00" to "2024 / 11 / 23 8:00", the health information estimation device 100 may use model #30 to estimate service information to be provided to the user. Note that the processes for estimating information and providing the estimated information are similar to those in FIG. 23, and therefore will not be described again.

[0237] Furthermore, the health information estimation system 1 is not limited to the above, and may perform processing such as that shown in FIG. 25. FIG. 25 is a diagram showing an example of data and processing. FIG. 25 shows an example in which the first urine data obtained in a day is used as data for processing. Note that explanations of points similar to those described above in FIGS. 23 and 24 will be omitted where appropriate.

[0238] In Fig. 25, as shown in the column labeled "first detection sensor," urine component data for sodium and potassium is acquired based on biological data detected by a urine sensor, but the urine component data may also include urea, as in Fig. 23 and Fig. 24. Data set DS33 in Fig. 25 includes a data group in which a time, which is date and time information, is associated with urine component data acquired on a date and time corresponding to that time.

[0239] 25, health information estimation device 100 determines a data structure based on the time of each data sequence in dataset DS33. For example, for data corresponding to "2024 / 11 / 19 8:00," health information estimation device 100 determines the data structure of the data corresponding to "2024 / 11 / 19 8:00" to be data structure #33 because the data sequence is data first acquired at 8:00 (8:00 AM), i.e., on November 19th. For example, data structure #33 is the data structure adopted as data to be used for processing (also referred to as the "adopted data structure").

[0240] Furthermore, for example, the health information estimation apparatus 100 determines that the data structure of the data corresponding to "2024 / 11 / 19 12:00" is data structure #34 because the data is the data array of data acquired for the second time at 12:00 (12:00 AM), i.e., on November 19th. For example, data structure #34 is a data structure that is not adopted as data to be used for processing (also referred to as a "rejected data structure").

[0241] For example, for data corresponding to "2024 / 11 / 20 8:00," health information estimation device 100 determines the data structure of the data corresponding to "2024 / 11 / 20 8:00" to be data structure #33 because the data sequence is data first acquired at 8:00 (8:00 AM), i.e., on November 20th. For example, for data corresponding to "2024 / 11 / 21 9:00," health information estimation device 100 determines the data structure of the data corresponding to "2024 / 11 / 21 9:00" to be data structure #33 because the data sequence is data first acquired at 9:00 (9:00 AM), i.e., on November 21st.

[0242] Furthermore, for example, for data corresponding to "2024 / 11 / 22 12:00," the data structure of the data corresponding to "2024 / 11 / 22 12:00" is determined to be data structure #33 because the data structure is for data first acquired at 12:00 (12:00 AM), i.e., on November 22nd. As such, in the example of FIG. 25, even data acquired at the same time may have different data structures depending on the data acquisition status on that day. For example, for data corresponding to "2024 / 11 / 23 8:00," the data structure is determined to be data structure #33 because the data structure is for data first acquired at 8:00 (8:00 AM), i.e., on November 23rd.

[0243] In FIG. 25, the column labeled "Estimation Model" in the data set DS33 indicates the estimation model used for the data associated with the data structure #33 (adopted data structure) adopted as the data to be used in the process.

[0244] In FIG. 25, health information estimation device 100 selects model #35 as the estimation model for the data group (also referred to as the "data group of the adopted data structure") associated with data structure #33 (adopted data structure) (lines 3, 5-8). For example, model #35 is an estimation model (sodium-potassium model) used for the data group of the adopted data structure. For example, model #35 is an estimation model that estimates information based on sodium and potassium in urine.

[0245] For example, model #35 may be a model that estimates health information such as a user's blood pressure based on the ratio of sodium to potassium (also referred to as the "Na / K ratio") in the data group of the adopted data structure. For example, model #35 may be a model that estimates that the higher the Na / K ratio in the data group of the adopted data structure, the higher the user's blood pressure. For example, model #35 may be a model that estimates that the user has high blood pressure if the Na / K ratio in the data group of the adopted data structure is equal to or greater than a predetermined threshold.

[0246] For example, for a data group with the adopted data structure in FIG. 25, health information estimation device 100 uses model #35 to estimate a user's health information or service information to be provided to the user. For example, if model #35 is an estimation model that outputs health information, health information estimation device 100 inputs the data group with the adopted data structure in FIG. 25 into model #35. Then, model #35, to which the data group with the adopted data structure in FIG. 25 has been input, outputs health information indicating the health status of the user from whom the data group was obtained. In this way, health information estimation device 100 estimates the health information of the user from whom the data group with the adopted data structure in FIG. 25 was obtained.

[0247] For example, the information obtained from urine varies depending on conditions such as urine components, data acquisition time, and data acquisition period (once, one day, one week), and it is desirable to vary the optimal estimation model accordingly. Therefore, as described above, the health information estimation device 100 can provide at least one of health information and service information based on various urine-related information and using an optimal estimation model depending on the data used for processing.

[0248] As described above, the data structure referred to in this application may include various things. For example, the data structure may be based on a pattern of the presence or absence of data for each item (type) in a data array. Also, for example, the data structure may be based on a pattern of date and time information (time) in a data array. Also, for example, the data structure may be based on a pattern of values ​​(numeric values) for each item (type) in a data array. Also, for example, the data structure may be based on a pattern of ratios of numerical values ​​(components, etc.) for each item (type) in a data array. Also, the data structure may be based on a pattern that combines the above-mentioned classifications. In this way, the data structure may indicate classifications according to patterns based on various information.

[0249] <5.Other> As described above, the health information estimation system 1 is not limited to cases where a single user (user U1 in FIG. 1) uses a bathroom space PS, but can also be applied to cases where multiple users use the same bathroom space PS. In other words, the health information estimation system 1 is not limited to bathroom spaces PS located in a residence where one person lives, but can also be applied to users of bathroom spaces PS located in facilities (buildings, etc.) shared by an unspecified number of people. An example of this case will be briefly described below.

[0250] As described above, when it is assumed that multiple users will use the bathroom space PS, the health information estimation system 1 may identify the users using the bathroom space PS in any manner. In this case, in FIG. 1, the users may be identified using an operation device installed in the bathroom space PS. For example, the health information estimation system 1 identifies the users when the users operate the operation device of the bathroom space PS and specify the users. In the example of FIG. 1, user U1 operates the operation device of the toilet device 50 and specifies user U1, and the health information estimation system 1 identifies the user as user U1.

[0251] For example, the health information estimation system 1 identifies a user by having the user select a user from a group of users displayed on the operation device of the wet space PS. Also, in FIG. 1, the user U1 may be identified as a user using the wet space PS by pairing the terminal device 10 of the user U1 with the operation device of the wet space PS, for example, via Bluetooth. The user may also be identified based on the detection results of various sensors installed in the wet space PS. Note that the above is just one example, and any process may be used to identify the user of the wet space PS as long as it is possible to identify the user.

[0252] As described above, if it is possible to identify the user of the wet space PS, each wet space PS, such as the toilet space PS1, may be located in any location. For example, the toilet space PS1 may be a so-called public toilet. For example, the toilet space PS1 is not limited to a home where a family or other person resides, but may also be a toilet space in a facility where multiple people reside (such as a nursing home or senior citizens' home), an office building, a store such as a department store, an amusement park, a stadium, a park, or a parking lot. In this way, the wet space PS may be located in any location as long as processing by the health information estimation system 1 is applicable.

[0253] Further advantages and modifications will readily occur to those skilled in the art. Therefore, the invention in its broader aspects is not limited to the specific details and representative embodiments shown and described above. Accordingly, various modifications may be made without departing from the spirit or scope of the general inventive concept as defined by the appended claims and their equivalents.

[0254] The above-described embodiments and modifications may have the following configurations, but are not limited to these. (1) a data acquisition means for acquiring data detected by a sensor disposed in the water-related space; a data structure determination means for determining a data structure based on a data array formed by associating the data acquired by the data acquisition means based on date and time information; health information estimation means for estimating health information of a user based on the data sequence; Equipped with The health information estimation means selects a model corresponding to the data structure from among a plurality of health information estimation models and estimates the health state of the user. A health information estimation system characterized by: (2) a first data acquisition means for acquiring data detected by a sensor disposed in the water-related space; a second data acquisition means for acquiring external data; a data structure determination means for determining a data structure based on a data array formed by associating the data acquired by at least one of the first data acquisition means and the second data acquisition means based on date and time information; health information estimation means for estimating health information of a user based on the data sequence; Equipped with The health information estimation means selects a model corresponding to the data structure from among a plurality of health information estimation models and estimates the health state of the user. A health information estimation system characterized by: (3) a first data acquisition means for acquiring data detected by a sensor disposed in the toilet space; a second data acquisition means for acquiring external data; a data structure determination means for determining a data structure based on a data array formed by associating the data acquired by at least one of the first data acquisition means and the second data acquisition means based on date and time information; health information estimation means for estimating health information of a user based on the data sequence; Equipped with The health information estimation means selects a model corresponding to the data structure from among a plurality of health information estimation models and estimates the health state of the user. A health information estimation system characterized by: (4) The data structure determination means The data structure is determined based on the data array configured by associating the data acquired by the data acquisition means based on personal information or date information. The health information estimation system according to (1) is characterized in that: (5) The data structure determination means determining the data structure based on a combination of at least one piece of data in the data array; The health information estimation system according to any one of (1) to (4), characterized in that: (6) The data structure determination means A past data array within a predetermined date range is used to fill in the gaps in the latest data array, and the data structure is determined based on the filled-in data array. The health information estimation system according to any one of (1) to (5), characterized in that: (7) The data structure determination means A data structure is determined based on the time series analysis results for each piece of data acquired by the data acquisition means within a predetermined date and time range. The health information estimation system according to (1) or (4) above. (8) The data structure determination means A data structure is determined based on a time series analysis result for each piece of data acquired by the first data acquisition means and the second data acquisition means within a predetermined date and time range. The health information estimation system according to (2) or (3) above. (9) The health information estimation means A model corresponding to the latest data structure is selected from the plurality of health information estimation models to estimate the health condition of the user. The health information estimation system according to any one of (1) to (8), characterized in that: (10) The health information estimation means A model is selected from the plurality of health information estimation models according to the data structure based on the time series analysis results for each piece of data acquired by the data acquisition means within a predetermined date and time range, and the health state of the user is estimated. The health information estimation system according to any one of (1), (4), and (7). (11) The health information estimation means A model is selected from the plurality of health information estimation models according to a plurality of data structures based on the time series analysis results for each piece of data acquired by the data acquisition means within a predetermined date and time range, and the health state of the user is estimated. A health information estimation system according to any one of (1), (4), (7), and (10). (12) The health information estimation means The health information estimation model is corrected based on the correction value calculated in accordance with the data acquired by the data acquisition means. A health information estimation system according to any one of (1), (4), (7), (10), and (11). (13) The health information estimation means Selecting a model from the plurality of health information estimation models in accordance with a plurality of data structures based on the time series analysis results for each piece of data acquired by the first data acquisition means and the second data acquisition means within a predetermined date and time range, and estimating the health state of the user. The health information estimation system according to any one of (2), (3), and (8). (14) The health information estimation means Estimating the health information of the user at a predetermined cycle The health information estimation system according to any one of (1) to (13), characterized in that: (15) The data structure determination means Prompting the user to fill in the external data to fill in the missing data array within a predetermined date and time range The health information estimation system according to any one of (2), (3), (8), and (13), characterized in that: (16) a data acquisition means for acquiring data detected by a sensor disposed in the water-related space; a data structure determination means for determining a data structure based on a data array or a data group formed by associating the data acquired by the data acquisition means based on date and time information; an information estimation means for estimating health information of a user or service information to be provided to the user based on the data acquired by the data acquisition means; Equipped with The information estimation means uses a model corresponding to the data structure from among a plurality of information estimation models. An information estimation system characterized by: (17) the sensor is a sensor for detecting urine components among excrement excreted into the toilet device, the data acquisition means acquires urine component data detected by the sensor, The data structure determination means determines a urine data structure based on a urine component sequence that is constructed by associating the urine component data based on date and time information. The information estimation system according to (16) above. (18) the sensor is provided in the toilet space and acquires biometric data of the user; the data acquisition means acquires biological data detected by the sensor, The data structure determination means determines a biometric data structure based on a biometric data sequence that is constructed by associating the biometric data based on date and time information. The information estimation system according to (16) above. [Explanation of symbols]

[0255] 1. Health Information Estimation System (Information Estimation System) 2. Server device 5. Sensor Device 10 Terminal equipment (device) 50 Toilet equipment 100 Health information estimation device 110 Communications Department 120 Storage section 121 Data storage unit 122 Model information storage unit 123 User information storage unit 130 Control Unit 131 Acquisition Department 132 Decision Section 133 Estimation Department 134 Transmitter

Claims

1. a data acquisition means for acquiring data detected by a sensor disposed in the water-related space; a data structure determination means for determining a data structure based on a data array formed by associating the data acquired by the data acquisition means based on date and time information; health information estimation means for estimating health information of a user based on the data sequence; Equipped with The health information estimation means selects a model corresponding to the data structure from among a plurality of health information estimation models and estimates the health state of the user. A health information estimation system characterized by:

2. a first data acquisition means for acquiring data detected by a sensor disposed in the water-related space; a second data acquisition means for acquiring external data; a data structure determination means for determining a data structure based on a data array formed by associating the data acquired by at least one of the first data acquisition means and the second data acquisition means based on date and time information; health information estimation means for estimating health information of a user based on the data sequence; Equipped with The health information estimation means selects a model corresponding to the data structure from among a plurality of health information estimation models and estimates the health state of the user. A health information estimation system characterized by:

3. a first data acquisition means for acquiring data detected by a sensor disposed in the toilet space; a second data acquisition means for acquiring external data; a data structure determination means for determining a data structure based on a data array formed by associating the data acquired by at least one of the first data acquisition means and the second data acquisition means based on date and time information; health information estimation means for estimating health information of a user based on the data sequence; Equipped with The health information estimation means selects a model corresponding to the data structure from among a plurality of health information estimation models and estimates the health state of the user. A health information estimation system characterized by:

4. The data structure determination means The data structure is determined based on the data array configured by associating the data acquired by the data acquisition means based on personal information or date information. The health information estimation system according to claim 1 .

5. The data structure determination means The data structure is determined based on a combination of at least one piece of data in the data array. The health information estimation system according to claim 1 .

6. The data structure determination means A past data array within a predetermined date range is used to fill in the gaps in the latest data array, and the data structure is determined based on the filled-in data array. The health information estimation system according to claim 1 .

7. The data structure determination means A data structure is determined based on the time series analysis results for each piece of data acquired by the data acquisition means within a predetermined date and time range. The health information estimation system according to claim 1 .

8. The data structure determination means A data structure is determined based on a time series analysis result for each piece of data acquired by the first data acquisition means and the second data acquisition means within a predetermined date and time range. The health information estimation system according to claim 2 .

9. The health information estimation means A model corresponding to the latest data structure is selected from the plurality of health information estimation models to estimate the health condition of the user. The health information estimation system according to claim 1 .

10. The health information estimation means A model is selected from the plurality of health information estimation models according to the data structure based on the time series analysis results for each piece of data acquired by the data acquisition means within a predetermined date and time range, and the health state of the user is estimated. The health information estimation system according to claim 1 .

11. The health information estimation means A model is selected from the plurality of health information estimation models according to a plurality of data structures based on the time series analysis results for each piece of data acquired by the data acquisition means within a predetermined date and time range, and the health state of the user is estimated. The health information estimation system according to claim 1 .

12. The health information estimation means The health information estimation model is corrected based on the correction value calculated in accordance with the data acquired by the data acquisition means. The health information estimation system according to claim 1 .

13. The health information estimation means Selecting a model from the plurality of health information estimation models in accordance with a plurality of data structures based on the time series analysis results for each piece of data acquired by the first data acquisition means and the second data acquisition means within a predetermined date and time range, and estimating the health state of the user. The health information estimation system according to claim 2 .

14. The health information estimation means Estimating the health information of the user at a predetermined cycle The health information estimation system according to claim 1 .

15. The data structure determination means Prompting the user to fill in the external data to fill in the missing data array within a predetermined date and time range The health information estimation system according to claim 2 .

16. a data acquisition means for acquiring data detected by a sensor disposed in the water-related space; a data structure determination means for determining a data structure based on a data array or a data group formed by associating the data acquired by the data acquisition means based on date and time information; an information estimation means for estimating health information of a user or service information to be provided to the user based on the data acquired by the data acquisition means; Equipped with The information estimation means uses a model corresponding to the data structure from among a plurality of information estimation models. An information estimation system characterized by:

17. the sensor is a sensor for detecting urine components among excrement excreted into the toilet device, the data acquisition means acquires urine component data detected by the sensor, The data structure determination means determines a urine data structure based on a urine component sequence that is constructed by associating the urine component data based on date and time information.

17. The information estimation system according to claim 16.

18. the sensor is provided in the toilet space and acquires biometric data of the user; the data acquisition means acquires biological data detected by the sensor, The data structure determination means determines a biometric data structure based on a biometric data sequence that is constructed by associating the biometric data based on date and time information.

17. The information estimation system according to claim 16.

Images