Life data tabulating device

Abstract

To provide a technique for grasping a slowly progressing abnormality in a target person.SOLUTION: In a life situation grasping system 100 of one embodiment, a life data tabulating device 30 outputs a use degree R of water calculated by tabulating the number of times of use of a water meter device 10 installed in a residence H of a target person for each specified time zone for each comparison period, as life data for grasping a life situation. As a result, an average behavior pattern for one day related to water use can be specified for each comparison period such as for every month or every 10 days. Then, a slowly progressing abnormality of the target person can be grasped based on whether or not there is a change in behavior pattern by comparing the life data from a plurality of comparison periods.SELECTED DRAWING: Figure 3

Description

[Technical field]

[0001] The present disclosure relates to a life data compilation device for understanding the living conditions of a subject. [Background technology]

[0002] 2. Description of the Related Art Known conventional life data compilation devices are those that output the amount of use of specific facilities and equipment as life data (see, for example, Patent Document 1). [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2022-147117 A (Claim 1, paragraphs

[0013] and

[0035] ) Summary of the Invention [Problem to be solved by the invention]

[0004] However, the above-mentioned conventional life data compilation device has a problem in that, although it can grasp sudden abnormalities of the subject, it cannot grasp slowly progressing abnormalities in which the subject's life becomes gradually difficult due to aging or worsening of a chronic illness. Therefore, this application discloses a technology for grasping slowly progressing abnormalities of the subject. [Means for solving the problem]

[0005] A first aspect of the invention of the present disclosure is a lifestyle data collection device that collects and outputs lifestyle data based on the use of specific equipment or devices in a subject's residence in order to grasp the lifestyle situation of the subject, the lifestyle data collection device comprising: a usage determination means that determines whether the specific equipment or devices was used every day in each of a plurality of specified time periods that are equal parts of a day; and a data calculation unit that collects the number of days on which the specific equipment or devices was used for each specified time period for each comparison period consisting of a predetermined plurality of days, and calculates the lifestyle data as the degree of usage of the specific equipment or devices for each specified time period per comparison period. Effect of the Invention

[0006] The life data collection device according to the first aspect of the present disclosure collects the number of times a specific facility or device is used in a specified time period in the subject's residence for each comparison period, and outputs the calculated usage of the facility or device as life data for grasping the living situation. This makes it possible to identify an average daily behavior pattern regarding the use of the specific facility or device for each comparison period, such as for example, for each month or every 10 days. By comparing the life data from multiple comparison periods, it becomes possible to grasp the gradually progressing abnormality of the subject based on the presence or absence of a change in the behavior pattern. [Brief description of the drawings]

[0007] [Figure 1] FIG. 1 is a schematic diagram showing the overall configuration of a living situation monitoring system according to an embodiment of the present invention. [Diagram 2] A block diagram showing the electrical configuration of a water meter device and a life data collection device. [Diagram 3] A block diagram showing the control configuration of the life data compilation device. [Figure 4] A graph showing an example of water usage during a specified time period from 1:00 to 2:00 in January [Diagram 5] A graph showing an example of lifestyle data [Figure 6] A conceptual diagram of a 24-hour period divided into 24 sections, each shifted by one hour within a limited time frame. [Figure 7] A graph showing an example of lifestyle data for the latest collection month and the comparison month [Figure 8] FIG. 8A is an example of the first index data, and FIG. 8B is an example of the second index data. [Figure 9] Flowchart of the lifestyle data collection program [Figure 10] Flowchart of a program executed by the usage determination process [Figure 11] Flowchart of the program executed by data processing [Figure 12] Flowchart of a program executed by the living situation determination process [Figure 13] 1 is a flowchart of a program executed by the first calculation process. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0008] A living situation grasping system 100 according to an embodiment of the present disclosure will be described with reference to FIGS. 1 to 13. The living situation grasping system 100 of this embodiment is a system for grasping living situations, and includes a plurality of water meter devices 10 installed at each residence H of a subject, a life data collecting device 30 connected to the plurality of water meter devices 10 via a communication network 101, and an external terminal 60 that can be arbitrarily connected to the life data collecting device 30 via the communication network 101. Here, the subject whose living situation is grasped by the living situation grasping system 100 is a resident of the residence H in which the living situation grasping system 100 is installed, and may be one or more people, regardless of age or gender. In addition, the residence H may be any type as long as the water usage can be known from the outside by the water meter device 10, and examples thereof include a detached house, each dwelling unit of a collective housing such as an apartment, and each private room of a nursing home.

[0009] The water meter device 10 includes a water meter 11 and a communication device 12. The communication device 12 generates usage data T1, which includes identification data for identifying the water meter device 10 and date and time data, from the measurement result of the water usage measured by the water meter 11 and is transmitted to the life data compilation device 30 at predetermined intervals (e.g., 10 minutes) that are shorter than a specified time period described below.

[0010] 2, the life data compilation device 30 is, for example, a server computer and has at least a communication circuit 31, a control unit 32 including a CPU 32A and a memory 32B, and a storage medium 40. The life data compilation device 30 may be a cloud server.

[0011] The storage medium 40 is composed of a RAM, a hard disk, a flash memory, etc., and has an identification data storage unit 41 and a program storage unit 42, etc., as shown in Fig. 3. The identification data storage unit 41 stores information about a user of the living situation awareness system 100 and information such as the identification data of the water meter device 10 of the subject's residence H, whose living situation is to be understood, in association with each other. The program storage unit 42 stores a life data compilation program PG1, which will be described later, etc.

[0012] The life data compilation program PG1 may be stored on a CD-ROM, a USB memory, or the like, and may be read and executed by the CPU 32A, or the CPU 32A may execute the life data compilation program PG1 by using a service such as an application via the communications network 101.

[0013] By executing the life data compilation program PG1, the CPU 32A functions as a control block for the data acquisition unit 50, usage determination means 51, data calculation unit 52, life situation determination unit 53, information transmission unit 58, life data storage unit 44, index data storage unit 45, etc. shown in Figure 3.

[0014] The data acquisition unit 50 acquires the usage data T1 from the multiple water meter devices 10 via the communication circuit 31 and stores the data in the usage data storage unit 43.

[0015] The usage determination means 51 operates, for example, once a month at a predetermined date and time, and obtains one month's usage data T1 for the month immediately preceding the operating month (hereinafter referred to as the "latest tally month") for each identification data, and generates one month's usage data for the latest tally month.

[0016] Specifically, the usage determination means 51 divides a day into 24 specified time slots, each of which is one hour, and reads out the water usage U for each specified time slot for each day of the latest tally month from the usage data storage unit 43. Then, it is determined whether the usage U for each specified time slot exceeds a predetermined standard usage Uc, and if it exceeds the standard usage Uc, it is determined that water was used during that specified time slot. For example, FIG. 4 shows an example of water usage U for a specified time slot from 1:00 to 2:00 on each day of the latest tally month. Here, if the standard usage Uc is, for example, 0.1 [L], it is determined that water was used on the day marked with a circle on the horizontal axis. Then, the usage determination means 51 sets a count value F of "1" on the day on which it is determined that water was used during the specified time slot from 1:00 to 2:00, and sets a count value F of "0" on the day on which it is determined that water was not used, thereby generating usage data for one month for the specified time slot from 1:00 to 2:00. Similarly, use / non-use data for one month is generated for each of the specified time periods other than the specified time period from 1:00 to 2:00.

[0017] In this embodiment, the length of the specified time period is one hour, but it is not limited to this and may be any number of hours or may be in minutes. In addition, the period during which the usage data T1 is collected is referred to as the "comparison period" in this disclosure. In this embodiment, the comparison period is one month, but it is not limited to this and may be any number of days. Furthermore, the usage determination means 51 may be activated depending on the length of the comparison period.

[0018] The data calculation unit 52 operates, for example, every month after the use determination means 51 generates use data for the latest tally month. The data calculation unit 52 then obtains the use data for one month of the latest tally month, and calculates the usage level R for each specified time slot by dividing the integrated value ΣF, which is the sum of the count values ​​F for all days of the month, by the number of days in the month. Specifically, for example, the integrated value ΣF for one month of the specified time slot in FIG. 4 described above is 15, and since one month has 30 days, the usage level R is calculated to be 0.5. The data calculation unit 52 then generates life data summarizing the usage levels R for each specified time slot from 0:00 to 24:00 for one month of the latest tally month, as shown in FIG. 5, and stores the data in the life data storage unit 44.

[0019] An example of life data is shown in both a bar graph and a line graph in Figure 5. Comparisons between life data are made by comparing the shapes of the line graphs of the life data, as described below. Therefore, in the following explanation, unless otherwise specified, life data refers to line graph data, and life data will be referred to as "waveform data" as appropriate.

[0020] The living situation assessment unit 53 compares the living data of the latest tally month with living data of a previous month (hereinafter referred to as a “comparison month”) to be compared with the latest tally month, as will be described in detail later, and assesses the similarity between the living data.

[0021] Living situation assessment unit 53 has index calculation unit 54 and similarity determination unit 57. Index calculation unit 54 calculates index data indicating the similarity between two life data to be compared. As shown in FIG. 3, index calculation unit 54 has first calculation unit 55 and second calculation unit 56.

[0022] The first calculation section 55 has a limiting calculation section 55A, a sum calculation section 55B, and a first index data calculation section 55C, and calculates the first index data M1 using dynamic time warping (DTW).

[0023] In a general dynamic time warping method, the distance between two points of two waveform data is simply calculated by exhaustively searching for the path with the shortest distance, and the accumulated distance is calculated as the DTW distance. In contrast, in this embodiment, the dynamic time warping method used by the limited calculation unit 55A and the like divides one piece of life data into multiple sections, applies the dynamic time warping method to each section, and calculates the DTW distance only for a specified period included in the section.

[0024] Specifically, the limited calculation unit 55A sets a limited time frame consisting of five specified time periods, and divides the life data into 24 overlapping sections shifted by one hour in the limited time frame. Then, the limited calculation unit 55A calculates a DTW distance (hereinafter referred to as a "partial DTW distance Dp") that targets only the five specified time periods of each section. That is, as shown in FIG. 6, the horizontal axis of the life data from 0:00 to 24:00 is divided into 24 overlapping sections, 11:00 to 4:00, 0:00 to 5:00, 1:00 to 6:00, 21:00 to 1:00, and 22:00 to 2:00, and calculates a partial DTW distance Dp for each section. Then, the summation calculation unit 55B integrates and sums up the partial DTW distances Dp of the 24 sections to calculate a summation DTW distance Ds.

[0025] The limited time frame may be set to include two or more and four or less specified time periods, or may be set to include six or more specified time periods.

[0026] The first index data calculation unit 55C calculates the first index data from the sum total DTW distance Ds calculated by the sum total calculation unit 55B. Here, since the usage level of each specified time period takes a value between 0 and 1, the maximum possible distance between two points of life data is 1. Since each section includes five specified time periods, the maximum value of the partial DTW distance Dp is 5, and the maximum value of the sum of the 24 sections (hereinafter referred to as the "maximum sum total DTW distance Dm") is 120. Therefore, the first index data calculation unit 55C calculates the first index data M1 by subtracting the sum total DTW distance Ds from the maximum sum total DTW distance Dm and dividing the result by the maximum sum total DTW distance Dm. As a result, the first index data M1 takes a value between 0 and 1, and approaches 1 as the waveforms of the two life data are more similar. Note that the first index data M1 may be calculated not only between the life data of the latest collection month and the comparison month, but also between the life data of the comparison months (see FIG. 8A).

[0027] The second calculation unit 56 calculates the second index data M2 using the correlation coefficient. Specifically, the second index data M2 is calculated by using the formula of Pearson's product moment correlation coefficient, and dividing the covariance of the degree of use of the life data of the latest compilation month and the degree of use of the life data of the comparison month by the product of their respective standard deviations. As a result, the second index data M2 takes a value between -1 and 1, and approaches 1 as the waveforms of the two life data are more similar. Note that the second index data M2 may be calculated not only between the life data of the latest compilation month and the comparison month, but also between the life data of the comparison months (see FIG. 8B). Note that the type of correlation coefficient is not limited to the Pearson's product moment correlation coefficient.

[0028] Here, the comparison month refers to the life data of a specified past month, and can be freely specified by the user, for example. The comparison month may be, for example, the month immediately preceding the most recent tally month, or the month several months ago, or the month one year ago or several years ago. The comparison month may also be the average of the past several months or the past year, or the monthly average of the same season as the most recent tally month. Multiple comparison months may also be specified. The comparison month may be freely set by the user.

[0029] In this embodiment, the past six months immediately preceding the latest tally month are designated as the months to be compared, and the first calculation unit 55 and the second calculation unit 56 calculate the first and second index data M1, M2 between the life data of each of those months and the latest tally month. The calculated first and second index data M1, M2 are stored in the index data storage unit 45 for each identification data.

[0030] The similarity determination unit 57 determines whether or not the behavior pattern has changed between the life data of the latest counting month and the life data of the comparison month based on the first and second index data M1, M2. As described above, the closer the value of the first and second index data M1, M2 is to 1, the higher the similarity between the life data. In this embodiment, the first and second thresholds Mt1, Mt2 are set in advance, and when both the first condition that the first index data M1 exceeds the first threshold Mt1 and the second condition that the second index data M2 exceeds the second threshold Mt2 are satisfied, it is determined that the similarity between the compared life data is high and that there is no change in the behavior pattern. When neither the first condition nor the second condition is satisfied, it is determined that there is a change in the behavior pattern. The first and second thresholds Mt1, Mt2 are stored in the index data storage unit 45 in advance. In this embodiment, for example, the first threshold value Mt1 is 0.87, and the second threshold value Mt2 is 0.70.

[0031] The similarity determination unit 57 also determines that there is a change in the behavior pattern when either the first condition or the second condition is satisfied. Here, the dynamic time warping method tends to easily detect a change in the amplitude of the waveforms of the two life data, but does not easily detect a change in the phase, and when only the phase changes, the change is difficult to detect, and there is a risk that the similarity is high. On the other hand, the correlation coefficient tends to easily detect a change in the phase of the waveforms of the two life data, but does not easily detect a change in the amplitude, and when only the amplitude changes, the change is difficult to detect, and there is a risk that the similarity is high. Therefore, when the first condition is satisfied and the second condition is not satisfied, the similarity determination unit 57 determines that the change in phase that could not be detected by the dynamic time warping method has been detected by the correlation coefficient, and determines that there is a change in the behavior pattern. Then, when the second condition is satisfied and the first condition is not satisfied, the similarity determination unit 57 determines that the change in amplitude that could not be detected by the correlation coefficient has been detected by the dynamic time warping method, and determines that there is a change in the behavior pattern.

[0032] The information transmitting unit 58 operates, for example, after the similarity determining unit 57 has determined whether or not there has been a change in the behavior pattern. The information transmitting unit 58 transmits the lifestyle data of the latest tally month and the comparison month (see FIG. 7 ) and the determination result of whether or not there has been a change in the behavior pattern of the latest tally month compared to the comparison month, via the communication circuit 51 to the external terminal 60 of the user linked to each identification data.

[0033] The result of the determination of the presence or absence of a change in the behavior pattern may be, for example, only a comment indicating whether or not the behavior pattern has changed in the latest collection month compared to the comparison month, or the first and second index data M1, M2 may be transmitted together. Also, the first and second index data M1, M2 between the lifestyle data of the past six months may be included (see FIG. 8A and FIG. 8B).

[0034] The external terminal 60 is owned by the user or the like, and may be a general communication means capable of communicating with the life data compilation device 30, such as a personal computer, a tablet, a mobile information terminal such as a smartphone, etc. As described above, the external terminal 60 receives the life data and the determination result of the presence or absence of changes in the behavior pattern between the latest compilation month and the comparison month from the life data compilation device 30 once a month, and grasps the life situation of the subject. In addition, the external terminal 60 may be configured to access the life data compilation device 30 and freely view the past life data stored in the life data storage unit 44.

[0035] An example of the life data compilation program PG1 executed by the CPU 32A of the life data compilation device 30 is shown in Figs. 9 to 13. As shown in Fig. 9, the life data compilation program PG1 first acquires water usage measurement results from a plurality of water meter devices 10 in steps S11 and S12. Specifically, when the water meter devices 10 transmit water usage measurement results for each predetermined period (YES in S11), the results are stored in the usage data storage unit 43 by classification data (S12). Here, the CPU 32A executing steps S11 and S12 corresponds to the data acquisition unit 50.

[0036] Then, when a predetermined date and time arrives once a month (YES in S13), the CPU 32A executes a usage determination process (S14). Until this date and time arrives once a month (NO in S13), steps S11 and S12 are repeated to collect the measurement results of the water usage. After executing the usage determination process, the CPU 32A successively executes a data calculation process (S15) and a life situation determination process (S16), and finally executes a transmission process (S17) to exit the life data compilation program PG1. Here, the CPU 32A when executing the usage determination process (S14) corresponds to the usage determination means 51, the CPU 32A when executing the data calculation process (S15) corresponds to the data calculation unit 52, the CPU 32A when executing the life situation determination process (S16) corresponds to the life situation determination unit 53, and the CPU 32A when executing the transmission process (S17) corresponds to the information transmission unit 58.

[0037] Fig. 10 shows an example of a program for the usage determination process (S14 in Fig. 9). In the usage determination process, first, in step S21, the amount of water usage U for each specified time period for each day in the most recent tally month is obtained. Then, it is determined whether the amount of water usage U for each specified time period exceeds a standard amount of water usage Uc (S22), and on days when the amount of water usage exceeds the standard amount of water usage Uc (YES in S22), the count value F is set to 1 (S23). On the other hand, on days when the amount of water usage does not exceed the standard amount of water usage Uc (NO in S22), the count value F is set to 0 (S24). In this way, usage data for all specified time periods for one month is generated.

[0038] Fig. 11 shows an example of a program for the data calculation process (S15 in Fig. 9). In the data calculation process, first, in step S31, an integrated value ΣF is calculated by summing up the count values ​​F for all days in a month for each specified time period from the use presence / absence data for that month in the latest tally month. Next, in step S32, the integrated value ΣF is divided by the number of days in that month to calculate the usage level R for each specified time period. Then, in step S33, life data is generated that summarizes the usage levels R for each specified time period from 0:00 to 24:00 for that month in the latest tally month.

[0039] An example of a program for the living situation assessment process (S16 in FIG. 9) is shown in Fig. 12. In the living situation assessment process, first, first and second index data M1, M2 serving as indices for determining the similarity between the living situation data of the latest compilation month and the living situation data of a comparison month are calculated by executing a first calculation process and a second calculation process (S41, S42).

[0040] In the first calculation process (S41), the first index data M1 is calculated using the dynamic time warping method. Specifically, as shown in Fig. 13, first, in step S51, a limited time frame consisting of five specified time periods is set for the life data, and the life data is divided into 24 overlapping sections shifted by one hour in the limited time frame, and a partial DTW distance Dp is calculated for only the five specified time periods of each section. Then, in step S52, each partial DTW distance Dp of the 24 sections is integrated and summed up to calculate a total DTW distance Ds. Next, in step S53, the total DTW distance Ds is subtracted from the maximum total DTW distance Dm, which is the maximum value of the sum of the 24 sections, and the result is divided by the maximum total DTW distance Dm to calculate the first index data M1. It should be noted that the CPU 32A when executing step S51 corresponds to a limited calculation unit 55A, the CPU 32A when executing step S52 corresponds to a sum calculation unit 55B, and the CPU 32A when executing step S53 corresponds to a first index data calculation unit 55C.

[0041] On the other hand, in the second calculation process (S42 in FIG. 12), the Pearson product-moment correlation coefficient is calculated as the second index data M2.

[0042] Then, after the first and second index data M1, M2 are calculated, in step S43 of Fig. 12, it is determined whether or not a first condition, that the first index data M1 exceeds a first threshold value Mt1, is satisfied. If the first condition is not satisfied (NO in S43), it is determined that there is a change in the behavior pattern (S46). On the other hand, if the first condition is satisfied (YES in S43), it is further determined whether or not a second condition, that the second index data M2 exceeds a second threshold value Mt2, is satisfied (S44). If the condition is satisfied (YES in S44), it is determined that there is no change in the behavior pattern (S45), and if the condition is not satisfied (NO in S44), it is determined that there is a change in the behavior pattern (S46).

[0043] Then, when it is determined in step S44 whether or not there has been a change in the behavior pattern, the transmission process (S17 in FIG. 9) is executed as described above, and the life data for the latest collection month and the past six months, as well as the determination result of whether or not there has been a change in the behavior pattern, are transmitted to the external terminal 60.

[0044] The configuration of the living situation grasping system 100 of this embodiment has been described above. In the living situation grasping system 100 of this embodiment, the life data compilation device 30 converts the amount of water usage collected from the water meter device 10 installed in the residence H of the subject into a water usage level R for each specified time period of one hour during a comparison period of one month, and outputs the converted amount as life data for grasping the living situation. As a result, the presence or absence of an action of whether or not water was used for each specified time period is specified, so that an average daily behavior pattern regarding water usage can be specified for each comparison period. Then, by comparing the life data for multiple comparison periods, it becomes possible to grasp a gradually progressing abnormality in the subject based on the presence or absence of a change in the behavior pattern.

[0045] In this embodiment, the amount of water usage measured by the water meter device 10 is converted into the number of times water is used to grasp behavioral patterns. For example, even if a person takes a bath at the same time of day, the amount of water usage varies greatly between a day when the person soaks in the bathtub and a day when the person only takes a shower, making it difficult to determine whether the change in the amount of water usage is due to an abnormality in the subject. However, in this embodiment, the amount of water usage is converted into the water usage degree R, so that both a day when the person soaks in the bathtub and a day when the person only takes a shower are counted as one use of water, making it easier to detect behavioral patterns related to water usage.

[0046] In this embodiment, the presence or absence of water usage during a specified time period is determined based on whether the water usage amount U for each specified time period on each day during the most recent month for which water usage amounts were collected exceeds a reference usage amount Uc. Setting the reference usage amount Uc can prevent, for example, a case in which a small amount of stagnant water in the pipes is distributed regardless of the subject's water usage from being determined to have been used by the subject.

[0047] Furthermore, in this embodiment, a count value F is set to 1 on a day when it is determined that water was used, and 0 on a day when it is determined that water was not used. Furthermore, the count values ​​F for all days in the most recent tally month are summed up for each specified time period to calculate an integrated value ΣF, which is then divided by the number of days in the month to calculate the usage level R. Then, life data is generated that summarizes the usage levels R for each specified time period from 0:00 to 24:00 for the most recent tally month. This makes it possible to detect daily behavioral patterns related to water use, such as a tendency to use water at 4:00 a.m. In addition, by comparing with past life data, it is possible to detect changes in behavioral patterns over a long span of time. For example, if water use in the morning has increased since last month, it is possible to estimate behavioral patterns such as a change in the time spent at home, or if the number of times water is used at night has increased, it is possible to estimate behavioral patterns such as a change in the number of times the toilet is used. Furthermore, by comparing this with lifestyle data from the same month one or several years ago, or lifestyle data generated by dividing a year into three or four periods and calculating the average usage R for the same season, it becomes possible to detect any abnormalities in the subject by excluding the effects of changes in water usage due to seasonal changes that are not an abnormality for the subject, such as different bathing times in winter and summer.

[0048] In this embodiment, the life data compilation device 30 generates waveform data as the life data. By making the change in the usage level R for each specified time period of the day into waveform data, the user can easily visually detect the change in the daily behavior pattern when viewing the life data. For example, if the subject's pace of life shifts from morning to night, it may seem as if water usage has stopped during the time period when it was used up until then, when comparing only the usage level R for the same specified time period. However, in this embodiment, by comparing two waveform data superimposed on each other, it is clear at a glance that the time period during which water usage has simply shifted.

[0049] In this embodiment, the life data compilation device 30 measures the similarity between two pieces of life data as waveform data using dynamic time compaction and a correlation coefficient to determine whether a change has occurred in a behavior pattern. The dynamic time compaction method tends to easily detect a change in amplitude between two pieces of waveform data but not a change in phase, and the correlation coefficient tends to easily detect a change in phase between two pieces of waveform data but not an change in amplitude, so that a determination based on only one of them may miss a change in amplitude or phase. However, in this embodiment, the dynamic time compaction method and the correlation coefficient are combined to measure the similarity between two pieces of life data, so that both a change in amplitude and a change in phase can be detected, and the presence or absence of a change in a behavior pattern can be objectively determined without relying on human senses.

[0050] Here, in this embodiment, a limited time frame consisting of five specified time periods is set, and the life data is divided into 24 overlapping sections shifted by one hour in the limited time period, and the dynamic time warping method is applied to each section. Here, if the dynamic time compaction method is applied to the entire life data, even if there is a change in amplitude in the same specified time period, the change may not be detected due to the influence of the difference in amplitude between completely different specified time periods. In contrast, according to the configuration of this embodiment, the life data is divided into multiple sections and the dynamic time compaction method is applied only to the close time periods, so that the DTW distance can be calculated without being influenced by completely different time periods, and the change in amplitude in the same specified time period can be reliably detected. Moreover, since the dynamic time compaction method is applied to each section divided into overlapping sections, the change in amplitude occurring at the boundary between the sections can be reliably detected without missing anything.

[0051] Furthermore, in this embodiment, the partial DTW distances Dp calculated in each section are summed up to calculate a total DTW distance Ds, and the total DTW distance Ds is subtracted from the maximum total DTW distance Dm, which is the sum of the maximum DTW distances that each section can have, and the result is divided by the maximum total DTW distance Dm to obtain the first index data M1. As a result, the first index data M1 takes a value between 0 and 1, and the closer the waveforms of the two life data are to 1, the easier it is to determine the degree of similarity.

[0052] [Other embodiments] In the above embodiment, the data calculation unit 52 calculates the usage level R by accumulating the count value F for each specified time period for all days in the comparison period, but it does not have to be for all days in the comparison period. For example, it may be accumulated for each day of the week, or may be accumulated separately for weekdays and holidays. This makes it possible to grasp differences in behavioral patterns for each day of the week, such as the tendency for water usage to be higher in a certain time period on a specific day of the week than on other days of the week. It also makes it possible to grasp the tendency to go out on weekdays and holidays, as well as changes therein.

[0053] In addition, the count value F for a particular day in the most recent tally month may be compared with the lifestyle data for that month or with past lifestyle data. This makes it possible to ascertain behavioral patterns, such as not using water on a particular day during a period when water is used heavily on other days.

[0054] In the above embodiment, the behavior pattern is grasped only from the subject's own life data, but it may be compared with the average life data of a group (age, sex, occupation, etc.) whose behavior pattern is estimated to be similar to that of the subject. By comparing with this life data, deviation from the general behavior pattern trend can be detected.

[0055] In addition, the change in the life data may be grasped by taking into consideration information on the area where the subject lives, meteorological information, etc. According to this, for example, when water usage in winter increases suddenly, it is possible to estimate whether this is simply due to a sudden increase in the rate of people staying at home or due to an abnormality in the subject, taking into consideration information on whether the subject lives in a snowless area such as Okinawa or in a heavy snow area such as Hokkaido. Also, when water usage increases during the day, it is possible to estimate whether the subject is unable to go out due to a decline in physical strength, taking into consideration information on long-term bad weather, if there is such information.

[0056] In addition to the lifestyle data, information on water usage may also be considered. For example, if water usage increases during the night and the water usage is approximately equal to the water usage per toilet visit, it can be determined that the number of toilet visits has increased.

[0057] In the above embodiment, the similarity determination unit 57 was configured to transmit the information to the external terminal 60 whether or not it determined that there was a change in the behavior pattern. However, the information may be transmitted only when it is determined that there was a change in the behavior pattern, or the user may be free to set whether or not to transmit the information when it is determined that there was no change.

[0058] In the above embodiment, the use determination means 51 is configured to operate according to the length of the comparison period, but it does not have to operate according to the length of the comparison period. For example, even when generating life data in units of a one-week comparison period, the use determination means 51 may be configured to operate once a month. In this case, it is sufficient to generate life data for the number of weeks from the previous execution to the current execution. Furthermore, the use determination means 51 may be configured to be executed irregularly at a timing requested by the user, rather than periodically.

[0059] In the above embodiment, the water usage amount U is converted into the water usage level R to grasp the living conditions, but the gas or electricity usage amount may be converted into the gas or electricity usage level. In these cases, the usage amount is obtained from a gas meter or a power meter instead of the water meter 11. Also, the living conditions may be grasped from the usage levels of these multiple facilities by collecting the usage amounts of the multiple facilities.

[0060] In the above embodiment, the water meter device 10 has the water meter 11 and the communication device 12 and is installed at each subject's residence H, but it may be configured such that only the water meter 11 is installed at each residence H, and one communication device 12 is installed for each of multiple residences H to acquire usage data from multiple water meters 11. Also, the water meter 11 and the communication device 12 may be integrated.

[0061] In addition, although this specification and drawings disclose specific examples of the technology included in the scope of the claims, the technology described in the claims is not limited to these specific examples, but includes various modifications and variations of the specific examples, as well as parts of the specific examples taken alone.

[0062] <Additional Notes> The following describes the feature groups extracted from the above-described embodiments, while indicating their effects, etc., as necessary. Note that, in the following, for ease of understanding, the corresponding configurations in the above-described embodiments are appropriately indicated in parentheses, but these feature groups are not limited to the specific configurations indicated in parentheses.

[0063] [Feature 1] The life data collecting device (30) collects and outputs life data based on the use of specific equipment or devices in a residence (H) of a subject in order to grasp the living conditions of the subject, and includes: a usage determination means (51) for determining whether the specific equipment or devices was used every day in each of a plurality of specified time periods obtained by equally dividing a day; and a data calculation unit (52) for collecting the number of days on which the specific equipment or devices was used for each specified time period for each comparison period consisting of a predetermined plurality of days, and calculating the life data as a usage degree (R) of the specific equipment or devices for each specified time period for each comparison period.

[0064] [Feature 2] The data calculation unit (52) is the life data compilation device (30) according to feature 1, which calculates the usage degree (R) by dividing the number of days during the comparison period when the specific facility / equipment was used by the number of days in the comparison period.

[0065] [Feature 3] The specific facility / equipment is at least one or more specific facilities selected from the group consisting of water, gas, and electricity, and the usage determination means (51) is configured to acquire time information and a measurement result of facility usage (U) output from the specific facility, and determine that the specific facility has been used on condition that the facility usage (U) for each specified time period exceeds a predetermined standard usage (Uc). This is the life data aggregation device (30) described in feature 1 or 2.

[0066] [Feature 4] The data calculation unit (52) is the life data compilation device (30) according to any one of features 1 to 3, which outputs the life data as waveform data specifying a change in the degree of use (R) over the course of the plurality of specified time periods.

[0067] [Feature 5] The life data collection device (30) according to feature 4 includes an index calculation device (54) that calculates and outputs index data (M1, M2) of changes in the life data over a plurality of comparison periods, the index calculation device (54) including a first calculation unit (55) that calculates the index data (M1) based on a DTW distance between the life data calculated using a dynamic time warping method.

[0068] [Feature 6] The first calculation unit (55) is a life data collection device (30) according to feature 5, comprising: a limited calculation unit (55A) that uses a limited time frame consisting of a series of M specified time periods that are smaller than N, which is the total number of specified time periods included in a day, divides the life data into the N intervals by shifting the limited time frame by one of the specified time periods, and calculates the DTW distance by applying the dynamic time warping method to each of the intervals; and a sum calculation unit (55B) that calculates the sum of the DTW distances of the N intervals.

[0069] [Feature 7] The first calculation unit (55) is the life data collection device (30) described in feature 6, which calculates the index data (M1) by subtracting the sum of the DTW distances of the N sections from a maximum sum DTW distance (Dm), which is the sum of the maximum DTW distances that each of the sections can have, and dividing the result by the maximum sum DTW distance (Dm).

[0070] [Feature 8] The life data collection device (30) according to any one of features 5 to 7, wherein the index calculation device (54) includes a second calculation unit (56) that calculates, as the index data (M2), a product-moment correlation coefficient between the two life data to be compared to calculate the DTW distance.

[0071] The lifestyle data aggregation device (30) of feature 1 aggregates the number of times that a specific facility or device is used in a specified time period in the subject's residence (H) for each comparison period, and outputs the calculated facility or device usage level (R) as lifestyle data for grasping the lifestyle situation. This makes it possible to identify an average daily behavior pattern regarding the use of a specific facility or device for each comparison period, such as for example, for each month or every 10 days. By comparing lifestyle data from multiple comparison periods, it becomes possible to grasp the gradually progressing abnormality of the subject based on the presence or absence of a change in the behavior pattern.

[0072] Here, the usage rate (R) can be calculated by dividing the number of days on which the specific facility or equipment was used in a comparison period consisting of a predetermined number of days by the number of days in the comparison period, as in Feature 2. The specific facility or equipment can be at least one of water, gas, or electricity, as in Feature 3.

[0073] Furthermore, the use of the facility / device for each specified time period may be determined to have been made on the condition that the facility usage (U) output from the water meter (11), gas meter, or power meter exceeds a predetermined reference usage (Uc) for each specified time period, as in feature 3. This prevents, for example, the subject from determining that the subject has used the water supply when a small amount of stagnant water in the pipes is distributed regardless of the subject's water usage.

[0074] If the lifestyle data is output as waveform data that specifies the change in the usage level (R) over a number of specified time periods, as in Feature 4, the change in the usage level (R) can be easily detected visually. For example, if the subject's lifestyle shifts from a morning person to an evening person, it may appear as if water usage has stopped during the time periods when water usage had occurred up until then when comparing only the numerical values ​​for the same time periods, but in this embodiment, by overlaying and comparing two waveform data, it is clear at a glance that the time periods when water usage had occurred up until then have simply shifted.

[0075] Furthermore, if the life data is waveform data, it is possible to determine changes in life situations by measuring the similarity between two pieces of life data using dynamic time compaction or a correlation coefficient, as in Features 5 and 8. Since dynamic time compaction makes it easy to detect a shift in the amplitude of two pieces of waveform data, and correlation coefficients make it easy to detect a shift in the phase of two pieces of waveform data, by measuring the similarity of the life data using both dynamic time compaction and the correlation coefficient, it is possible to objectively determine changes in life situations without relying on human senses.

[0076] In addition, when measuring the similarity of two waveform data by the dynamic time compaction method, the DTW distance is usually calculated by calculating the distance between two points of two waveform data in a brute force manner. However, in feature 6, a limited time frame consisting of a plurality of specified time periods is set, and the life data is divided into N sections, which is the total number of overlapping specified time periods, by shifting the data by the specified time periods in the limited time frame, and the dynamic time compaction method is applied to each section. Then, the sum of the DTW distances (Dp) of the N sections is calculated. Here, if the dynamic time compaction method is applied to the entire life data as usual, even if there is a change in amplitude in the same time period, the change may not be detected due to the influence of the difference in amplitude between completely different time periods. However, if the life data is divided into multiple sections and the dynamic time compaction method is applied only to the close time periods as in the present disclosure, the DTW distance can be calculated without being influenced by the completely different time periods, and the change in amplitude in the same time period can be reliably detected. Moreover, since the overlapping sections are divided and the dynamic time compaction method is applied to each section, the change in amplitude occurring at the boundary between the sections can be reliably detected without missing any changes.

[0077] In this case, the index data (M1) may be calculated by subtracting the sum of the DTW distances (Dp) of N sections from the maximum sum DTW distance (Dm), which is the sum of the maximum DTW distances that each section can have, and dividing the result by the maximum sum DTW distance (Dm), as in Feature 7. As a result, the index data (M1) takes a value between 0 and 1, and approaches 1 as the waveforms of the two life data are more similar, making it easier to determine the degree of similarity. [Explanation of symbols]

[0078] 10. Consumption Measuring Device 11 Water meter 30 Monitoring device (life data collection device) 51 Use determination means 52 Data Calculation Section 54 Index calculation device 55 1st calculation section 55A Limited operation section 55B Summation calculation section 56 2nd calculation section Dm Maximum total DTW distance Dp Partial DTW distance Ds Total DTW distance H residence M1 First indicator date M2 Second indicator R Usage U usage (device usage) Uc Baseline Usage

Claims

1. A life data collection device that collects and outputs life data based on the use of specific facilities and devices in a subject's residence in order to grasp the living situation of the subject, a usage determination means for determining whether the specific facility or equipment has been used every day during each of a plurality of specified time periods that are obtained by equally dividing a day; a data calculation unit that calculates the life data as a usage degree of the specific facility / device for each of the specified time periods by tallying up the number of days on which the specific facility / device was used for each of the specified time periods for each comparison period, the life data being calculated as a usage degree of the specific facility / device for each of the specified time periods for each comparison period; A lifestyle data collection device comprising:

2. The life data compilation device according to claim 1 , wherein the data calculation unit calculates the degree of use by dividing the number of days during the comparison period when the specific facility or equipment was used by the number of days in the comparison period.

3. The specific facility or equipment is at least one of water, gas, or electricity. The use determination means 2. The life data compilation device according to claim 1, configured to acquire time information and a measurement result of equipment usage output from the specific equipment, and determine that the specific equipment has been used on condition that the equipment usage for each of the specified time periods exceeds a predetermined standard usage.

4. 4. The life data compilation device according to claim 1, wherein the data calculation unit outputs the life data as waveform data specifying a change in the degree of use over the course of the plurality of specified time periods.

5. An index calculation device that calculates and outputs index data of changes in the life data over a plurality of comparison periods; The life data compilation device according to claim 4 , wherein the index calculation device includes a first calculation unit that calculates the index data based on a DTW distance between the life data calculated using a dynamic time warping method.

6. The first calculation unit is a limited calculation unit that uses a limited time frame that is a series of M specified time periods that are smaller than N, which is a total number of specified time periods included in one day, shifts the limited time frame by one specified time period, divides the life data into the N sections, and applies the dynamic time warping method to each of the sections to calculate the DTW distance; The life data compilation device according to claim 5 , further comprising: a sum calculation unit that calculates a sum of the DTW distances of the N sections.

7. The life data collection device of claim 6, wherein the first calculation unit calculates the index data by subtracting the sum of the DTW distances of the N sections from a maximum sum DTW distance, which is the sum of the maximum DTW distances that each of the sections can have, and dividing the result by the maximum sum DTW distance.

8. The life data compilation device according to any one of claims 5 to 7, wherein the index calculation device includes a second calculation unit that calculates, as the index data, a product-moment correlation coefficient between the two life data to be compared in order to calculate the DTW distance.

Images