JP7596695B2 - Behavior estimation device, behavior estimation method, and program - Google Patents

Description

The present invention relates to a behavior estimation device, a behavior estimation method, and a program.

There are times when the subject's living situation in the home is checked by others. Hereinafter, such checking is also called monitoring.
Specific examples of monitoring situations include various medical and nursing situations, such as checking the living conditions of a person with dementia to determine the degree of progression of dementia, or checking the living conditions of a person receiving care to consider optimal nursing care services. Note that there may be cases where the subject is both a person with dementia and a person receiving care.

The method for monitoring electrical appliances described in Patent Document 1 involves determining the power consumption of individual electrical appliances, such as home appliances (variable load appliances) whose power consumption (current) varies depending on the mode, settings, etc. (see Patent Document 1).

JP 2013-213825 A

However, the technology described in Patent Document 1 alone was insufficient in terms of estimating human behavior. In particular, when checking the living conditions of people with dementia or those receiving care, it was necessary to develop not only this technology but also other technologies for estimating behavior.

The present invention has been made in consideration of these circumstances, and aims to provide a behavior estimation device, a behavior estimation method, and a program that can effectively estimate the behavior of a subject based on power information.

As one configuration example, the behavior estimation device includes an information input unit that inputs input information including equipment separation power and individual circuit power obtained from power measured by a power sensor installed in the home of a subject to be monitored, a behavior estimation unit that uses machine learning results to estimate the behavior of the subject based on the input information input by the information input unit, and an information output unit that outputs information related to the estimation result by the behavior estimation unit, wherein the equipment separation power is obtained by performing a predetermined equipment separation process based on information on the total power amount of the home obtained by the power sensor and includes information representing the amount of power used by one or more devices used in the home, and a room corresponding to the individual circuit power is a room in which there is no equipment that obtains power from the equipment separation power .

As one configuration example, a behavior estimation method performed in a behavior estimation device includes an information input step of inputting input information including equipment separation power and individual circuit power obtained from power measured by a power sensor installed in the home of a subject to be monitored, a behavior estimation step of using machine learning results to estimate the behavior of the subject based on the input information input by the information input step, and an information output step of outputting information related to the estimation result by the behavior estimation step, wherein the equipment separation power is obtained by performing a predetermined equipment separation process based on information on the total power amount of the home obtained by the power sensor, and includes information representing the amount of power used by one or more devices used in the home, and a room corresponding to the individual circuit power is a room in which there is no device that obtains power from the equipment separation power .

As one configuration example, the program is for causing a computer to realize an information input function that inputs input information including equipment separation power and individual circuit power obtained from power measured by a power sensor installed in the home of a subject to be monitored, a behavior estimation function that uses machine learning results to estimate the behavior of the subject based on the input information input by the information input function, and an information output function that outputs information related to the estimation result by the behavior estimation function, wherein the equipment separation power is obtained by performing a predetermined equipment separation process based on information on the total amount of power in the home obtained by the power sensor, and includes information representing the amount of power used by one or more devices used in the home, and a room corresponding to the individual circuit power is a room in which there is no equipment that obtains power from the equipment separation power .

The behavior estimation device, behavior estimation method, and program of the present invention can effectively estimate the behavior of a subject based on power information.

1 is a diagram showing a schematic configuration of a behavior inference system according to an embodiment of the present invention; FIG. 2 is a diagram showing an outline of the flow of power-related information using a power sensor according to one embodiment (first embodiment) of the present invention. FIG. 1 is a diagram showing a schematic configuration of a behavior inference device according to an embodiment (first embodiment) of the present invention. FIG. 1 is a diagram showing a schematic configuration of a behavior inference device according to one embodiment (second embodiment) of the present invention.

The following describes an embodiment of the present invention with reference to the drawings.

[Power supply system]
FIG. 1 is a diagram showing a schematic configuration of a behavior inference system 1 according to an embodiment of the present invention.
In this embodiment, the behavior inference system 1 performs processing for checking the living conditions of a subject 511 living in a house 11, which is a residence.

The subject 511 may be any person, and in this embodiment, is a person receiving care at home. As another example, the subject 511 may be, for example, a person with dementia, or an elderly person living alone who is not receiving home care. Note that the subject 511 may be both a person with dementia and a person receiving care.

The house 11 has an entrance 111, a corridor 112, a living room 113, a toilet room 114, a bathroom 115, and a bedroom 116.
The house 11 includes a distribution board 231 at the entrance 111. The distribution board 231 may be provided at any location.
The entrance 111 has a door 131 .
The living room 113 has a door 132 .
The toilet room 114 has a door 133 .
The bathroom 115 has a door 134 .
The bedroom 116 has a door 135 .
In this embodiment, the behavior inference system 1 does not include the house 11 , but the behavior inference system 1 may be considered to include the house 11 .

Here, the configuration of the house 11 shown in FIG. 1 is an example for the purpose of explanation.
The layout and other configuration of the house 11 may be any configuration.
For example, there is a floor plan in which the toilet room 114 and the bathroom 115 are shared.

The behavior inference system 1 includes various sensors installed in a house 11 and a router device 212. The router device 212 is an example of a wireless device.
The behavior estimation system 1 also includes a behavior estimation server device 12, a home appliance separation server device 13, and a network 14.
The behavior estimation server device 12 includes a behavior estimation device 31 .
The home appliance separation server device 13 includes a home appliance separation device 51 .

Here, the network 14 may be a wired network or a wireless network.
The network 14 may be a dedicated network, or a public network such as the Internet.
In addition, if the network 14 is a public network, the behavior estimation system 1 may be considered not to include the network 14.

Various sensors installed in the house 11 will be described.
The behavior inference system 1 includes a power sensor 211 in a distribution board 231. The power sensor 211 may be provided in a location remote from the distribution board 231.
Furthermore, the behavior inference system 1 includes a temperature sensor 311 and a humidity sensor 312 in the living room 113. The temperature sensor 311 and the humidity sensor 312 may be provided in any desired location.
The behavior estimation system 1 also includes an entrance sensor 331 , a living room sensor 332 , a toilet sensor 333 , a bathroom sensor 334 , and a bedroom sensor 335 .
The behavior estimation system 1 also includes a toilet sensor 351 for a toilet (not shown) in the toilet room 114 and a pillow sensor 371 for a pillow (not shown) in the bedroom 116 .

For example, in a configuration in which the toilet room 114 and the bathroom 115 are shared, the toilet room sensor 333 and the bathroom sensor 334 are a common sensor.

Here, the combination of various sensors other than the power sensor 211 according to this embodiment is an example for the purpose of explanation.
The various sensors other than the power sensor 211 do not necessarily have to be the same as those in the present embodiment. For example, some or all of the various sensors other than the power sensor 211 according to the present embodiment do not have to be used, or one or more sensors not included in the various sensors other than the power sensor 211 according to the present embodiment may be used.

As the various sensors, an environmental sensor may be used, a human presence sensor may be used, or both of these sensors may be used.
In addition, the human presence sensor can be configured to consume special power when it detects human movement, so that the power consumption can be reflected in the home appliance separation result.
Furthermore, for example, Internet of Things (IoT) technology may be used in various sensors.

The power sensor 211 detects predetermined information related to power. In this embodiment, the predetermined information includes information that allows the total amount of power used in the house 11 to be grasped, and information that allows the amount of power used in a predetermined individual circuit in the house 11 to be grasped.
In this embodiment, the specified individual circuits include an individual circuit for the toilet room 114 and an individual circuit for the bathroom 115 .
For example, in a configuration in which the toilet room 114 and the bathroom 115 are shared, the individual circuit of the toilet room 114 and the individual circuit of the bathroom 115 become a common circuit.

Here, as the power sensor 211, for example, a smart meter may be used.
Additionally, the power sensor 211 may have multiple output channels.
As an example, the power sensor 211 may have four output channels, and two output channels may be used to output information for grasping the amount of transmitted power in two separate pieces, and the remaining two output channels may output information for grasping the amount of power of each of the two individual circuits. In this case, the total amount of power may be grasped by combining the information for grasping the amount of transmitted power outputted using the two output channels.

Moreover, in this embodiment, the power sensor 211 detects information regarding the strength of a signal wirelessly transmitted from the router device 212. In this embodiment, the signal is a beacon signal.
In addition, in this embodiment, received signal strength indication (RSSI) is used as information regarding signal strength, but other index information may be used.

Here, the strength of the signal detected by the power sensor 211 varies depending on the open/close state of a door (door 132 in the example of FIG. 1 ) between the power sensor 211 and the router device 212. Typically, the strength of the signal detected by the power sensor 211 is greater when there are fewer factors blocking the signal wirelessly transmitted from the router device 212, such as when the door is open. Such signal strength can be useful when estimating the behavior of the subject 511.
As another configuration example, a configuration in which the power sensor 211 does not detect information regarding the strength of a signal wirelessly transmitted from the router device 212 may be used.

The temperature sensor 311 detects the temperature.
The humidity sensor 312 detects humidity.
The temperature sensor 311 and the humidity sensor 312 may be a common sensor.

The entrance sensor 331 is an acceleration sensor that detects the opening and closing of the door 131 of the entrance 111 .
The living room sensor 332 is an acceleration sensor that detects the opening and closing of the door 132 to the living room 113 .
The toilet sensor 333 is an acceleration sensor that detects the opening and closing of the door 133 of the toilet 114 .
The bathroom sensor 334 is an acceleration sensor that detects the opening and closing of the door 134 to the bathroom 115 .
The bedroom sensor 335 is an acceleration sensor that detects the opening and closing of the door 135 to the bedroom 116 .
The toilet sensor 351 is an acceleration sensor that detects the opening and closing of the toilet seat. The toilet may be provided with a function to stop the heater of the toilet seat when the toilet seat is opened.
The pillow sensor 371 is an acceleration sensor that detects the movement of the pillow.

In this embodiment, the subject 511 is not fitted with a sensor, but as another configuration example, any sensor may be fitted to the subject 511. The sensor may be, for example, a sensor that detects the movement of the subject 511, a sensor that detects the body temperature, heart rate, pulse, etc. of the subject 511, or a sensor that detects an image in the line of sight of the subject 511.

The power sensor 211 and various other sensors (temperature sensor 311, humidity sensor 312, entrance sensor 331, living room sensor 332, toilet sensor 333, bathroom sensor 334, bedroom sensor 335, toilet sensor 351, and pillow sensor 371) each communicate wirelessly with the router device 212.
The power sensor 211 and various other sensors each transmit detected information to the router device 212 wirelessly.

In this embodiment, the router device 212 wirelessly transmits a predetermined beacon signal. The power sensor 211 and various other sensors each recognize the router device 212 by receiving the beacon signal transmitted from the router device 212, and start wireless communication with the router device 212.

The router device 212 receives information wirelessly transmitted from the power sensor 211 and each of the other various sensors, and transmits the received information to the home appliance separation server device 13 via the network 14 .
The communication between the router device 212 and the network 14 may be wired or wireless.

In this embodiment, the information detected by the power sensor 211 and various other sensors is transmitted to the outside (in this embodiment, the network 14) by the router device 212, but a configuration in which one or more of the power sensor 211 and various other sensors transmit the detected information to the outside (in this embodiment, the network 14) by other methods may also be used.

The home appliance separation server device 13 receives the information transmitted from the router device 212 .
In the home appliance separation server device 13, the home appliance separation device 51 performs a process of separating home appliances based on the received information on the total power amount, and obtains information on the result of the process of separating home appliances.
Then, the home appliance separation server device 13 transmits the information transmitted from the router device 212 and information on the result of the home appliance separation process to the behavior estimation server device 12.

Here, the result of the home appliance separation process includes information indicating the amount of power used by each of one or more electrical appliances used in the house 11.
Examples of such electrical appliances include washing machines, vacuum cleaners, air conditioners, microwave ovens, rice cookers, refrigerators, and high-heat appliances, including hair dryers and electric kettles.

The communication between home appliance separation server device 13 and network 14 may be wired communication or wireless communication.
Furthermore, the communication between the home appliance separation server device 13 and the behavior estimation server device 12 may be wired communication or wireless communication.
Furthermore, communication between the home appliance separation server device 13 and the behavior estimation server device 12 may be performed via a dedicated network, or may be performed via a public network such as the Internet.

The behavior estimation server device 12 receives the information transmitted from the home appliance separation server device 13 .
In the behavior estimation server device 12, the behavior estimation device 31 estimates information regarding the behavior of the subject person 511 based on the received information.
Then, the behavior estimation server device 12 performs, for example, storing the estimated information in a storage unit such as a predetermined database, or transmitting the estimated information to a predetermined device to provide it.

Here, for example, a public device may be used as the home appliance separation server device 13 having the home appliance separation device 51. In this case, the behavior estimation system 1 may be considered not to include the home appliance separation server device 13 having the home appliance separation device 51.
For example, a device using a technique (device separation estimation technique) as described in Patent Literature 1 may be used as the home appliance separation device 51. Home appliance separation is an example of device separation, and the electrical device is a home appliance device.

In the present embodiment, the home appliance separation server device 13 and the behavior estimation server device 12 are configured as separate entities. However, as another configuration example, the home appliance separation server device 13 and the behavior estimation server device 12 may be configured as an integrated entity.
In addition, in the present embodiment, the home appliance separation device 51 and the behavior inference device 31 are configured to be separate entities. However, as another configuration example, the home appliance separation device 51 and the behavior inference device 31 may be configured as an integrated device.

First Embodiment
The first embodiment will be described with reference to FIG. 2 and FIG.

FIG. 2 is a diagram showing an outline of the flow of power-related information using a power sensor 211 according to one embodiment (first embodiment) of the present invention.
In the example of FIG. 2, information detected by various sensors other than the power sensor 211 is omitted.

The power sensor 211 detects total power amount information 1011 , individual circuit power information 1012 , and wireless strength information 1013 .
The total power information 1011 is information that indicates the total amount of power (total power amount) used in the house 11 .
The individual circuit power information 1012 is information that indicates the amount of power used in each individual circuit. In this embodiment, the individual circuit power information 1012 includes information that indicates the amount of power used in the individual circuit of the toilet room 114 and information that indicates the amount of power used in the individual circuit of the bathroom 115.
The wireless strength information 1013 is information that indicates the strength of a wireless signal from the router device 212 .

The home appliance separation device 51 acquires separated home appliance power information 1051 based on the total power amount information 1011 by the home appliance separation function 1031 .
The separate home appliance power information 1051 is information that indicates the amount of power used by each of the specific electrical appliances used in the house 11 .

The behavior estimation device 31 acquires total power information 1011, individual circuit power information 1012, wireless strength information 1013, and home appliance separate power information 1051.

[Behavior estimation device]
FIG. 3 is a diagram showing a schematic configuration of a behavior inference device 31 according to one embodiment (first embodiment) of the present invention.
The behavior inference device 31 includes an information input unit 711 , a behavior inference unit 712 , and an information output unit 713 .
The information input unit 711 includes a power information input unit 731, a wireless information input unit 732, and an additional information input unit 733. Note that, for convenience of explanation, in this embodiment, input information other than the power information and the wireless information will be referred to as additional information, but this name may be arbitrary.
The behavior inferrer 712 includes a learning unit 751 and a learning result storage unit 752 .

The power information input unit 731 inputs power information and outputs the power information to the behavior estimation unit 712 .
In this embodiment, the power information includes total power amount information, individual circuit power information, and separate home appliance power information.

The wireless information input unit 732 inputs wireless information and outputs the wireless information to the behavior estimation unit 712 .
In this embodiment, the wireless information includes wireless strength information.

The additional information input unit 733 inputs the additional information, and outputs the additional information to the behavior estimation unit 712 .
In this embodiment, the additional information includes temperature information, humidity information, entrance acceleration information, toilet acceleration information, pillow acceleration information, living room acceleration information, toilet acceleration information, bathroom acceleration information, and bedroom acceleration information.

The temperature information is information that indicates the temperature detected by the temperature sensor 311 .
The humidity information is information that indicates the humidity detected by the humidity sensor 312 .
The entrance acceleration information is information that indicates the acceleration detected by the entrance sensor 331 .
The toilet acceleration information is information that represents the acceleration detected by the toilet sensor 351.
The pillow acceleration information is information that represents the acceleration detected by the pillow sensor 371.

The living acceleration information is information that represents the acceleration detected by the living sensor 332 .
The toilet acceleration information is information that indicates the acceleration detected by the toilet sensor 333 .
The bathroom acceleration information is information that represents the acceleration detected by the bathroom sensor 334.
The bedroom acceleration information is information that represents the acceleration detected by the bedroom sensor 335.

In this embodiment, for convenience of explanation, a distinction is made between the power information input unit 731, the wireless information input unit 732, and the additional information input unit 733, but this is not necessarily limited to such a configuration, and the information input unit 711 may be configured to input power information, wireless information, and additional information, and output this information to the behavior estimation unit 712.

In this embodiment, the information input to the information input unit 711 is output as is to the behavior estimation unit 712. However, as another configuration example, the information input to the information input unit 711 may be processed in the information input unit 711 and then output to the behavior estimation unit 712.
For example, the information input unit 711 may calculate a standard deviation, an average value, a minimum value, a maximum value, or the like for a predetermined period of time for the inputted predetermined information, and output the calculation result to the behavior estimation unit 712. Such a configuration may be useful in improving the accuracy of behavior estimation performed by the behavior estimation unit 712.

The behavior estimation section 712 estimates the behavior of the subject person 511 based on the information input from the information input section 711 , and outputs estimated information that is the result of the estimation to the information output section 713 .
The information output unit 713 receives the estimated information output from the behavior estimation unit 712 and outputs the estimated information.
In this embodiment, the estimated information includes sleep information, excretion information, bathing information, going out information, meal information, cleaning information, and laundry information.

The sleep information is information related to sleep performed by the subject 511. The sleep information may include information on the number of sleeps or the sleep duration.
The excretion information is information regarding excretion performed by the subject 511. The excretion information may include information regarding the number of excretion events or the excretion time.
The bathing information is information regarding bathing performed by the subject 511. The bathing information may include information regarding the number of times bathing has been performed or the bathing time.
The outing information is information relating to outings made by the subject person 511. The outing information may include information on the number of times the subject person goes out or the duration of the outing.
The meal information is information about meals taken by the subject 511. The meal information may include information about the number of meals or meal times.
The cleaning information is information about cleaning performed by the subject person 511. The cleaning information may include information about the number of cleanings or the cleaning time.
The laundry information is information about laundry done by the subject person 511. The laundry information may include information about the number of times laundry has been done or the laundry time.

The information output unit 713 receives the estimated information output from the behavior estimation unit 712 and outputs the estimated information in a predetermined format.
The specified manner may be, for example, a manner in which the estimated information is transmitted to a device other than the behavior estimation device 31, a manner in which the estimated information is displayed on a screen of a display unit, or a manner in which the estimated information is printed on paper or the like.

In addition, in the behavior inference device 31, the information output unit 713 may output information indicating a predetermined alarm (alarm information) when a predetermined condition is satisfied.
As a specific example, the behavior inference device 31 may output predetermined alarm information when it is inferred that the subject person 511 has gone outside.
For example, the behavior inference device 31 may output predetermined alarm information when the value of a predetermined index exceeds a predetermined threshold (or falls below a predetermined threshold).
As a specific example, the behavior inference device 31 may output predetermined alarm information when the temperature detected by the temperature sensor 311 exceeds a predetermined threshold.
The manner in which the alert information is output may be, for example, a manner in which the alert information is transmitted to a device other than the behavior estimation device 31, a manner in which the alert information is displayed on the screen of a display unit, or a manner in which the alert information is output as a sound.

Here, in this embodiment, the behavior inferrer 712 calculates output information from input information using the results of machine learning.
The learning unit 751 has a function of performing machine learning using input information for training, information on true values, etc. Note that a known technique may be used as the machine learning technique.
The learning unit 751 stores information on the results of machine learning in the learning result storage unit 752 .

The behavior estimation unit 712 uses the information stored in the learning result storage unit 752 to calculate output information based on the input information input from the information input unit 711, and outputs the calculated output information to the information output unit 713.

In this embodiment, the behavior estimation unit 712 is provided with a learning unit 751 having a function of performing machine learning. However, as another configuration example, the function of performing machine learning (in this embodiment, the learning unit 751) does not necessarily have to be provided in the behavior estimation unit 712, and information on the results of machine learning performed in a device other than the behavior estimation device 31 may be stored in the learning result storage unit 752.

In addition, in this embodiment, the behavior estimation unit 712 is configured to perform behavior estimation using the results of machine learning, but as another configuration example, a configuration in which behavior estimation is performed using a technique other than machine learning may be used.

Here, the input information to the behavior inference device 31 in the present embodiment is an example, and for example, less input information may be used, or more input information may be used. Also, each piece of input information may be, for example, the same information as in the present embodiment, or may be information different from that in the present embodiment.
As an example, the combination of input information may include total power information, individual circuit power information, home appliance separate power information, wireless strength information, temperature information, humidity information, entrance acceleration information, toilet acceleration information, and pillow acceleration information.

In addition, the output information to the behavior estimation device 31 in this embodiment is an example, and for example, less output information may be used, or more output information may be used. In addition, each piece of output information may be, for example, the same information as in this embodiment, or may be information different from that in this embodiment.

[Regarding the first embodiment]
As described above, the behavior inference device 31 in the behavior inference system 1 according to the present embodiment can effectively infer the behavior of the subject person 511 based on the power information.

Furthermore, in the behavior inference device 31 according to the present embodiment, the accuracy of inferring the behavior of the subject person 511 can be improved by using information from other sensors in addition to the power information.
For example, in general, home appliance separation technology can estimate home appliances that have high power consumption and characteristic power consumption patterns, but may not be able to estimate other appliances. In particular, home appliance separation technology often does not include appliances in toilets and bathrooms as targets for estimation.
Therefore, in the behavior inference device 31 according to this embodiment, by using information on the results of the separation of home appliances as well as information on the power consumption of the individual circuits in the toilet room and bathroom, it is possible to improve the accuracy of inferring the behavior of the subject person 511.
In particular, when the subject 511 is a person receiving care, information regarding the toilet room and bathroom is useful.

The behavior inference device 31 according to this embodiment can infer the living situation of a person by inferring the behavior of the person using power data.
The behavior inference device 31 according to this embodiment can grasp the overall behavior of people living their lives by deriving objective facts while taking privacy into consideration.
For example, in this embodiment, it is not necessary to attach a wearable sensor to the subject 511, so the subject 511 can be made more comfortable than if a wearable sensor were attached to the subject 511.
For example, in this embodiment, even if the subject 511 is an elderly person or the like who does not use a smartphone, the behavior can be estimated.
For example, in this embodiment, the subject 511 can continue living in a familiar home or the like, and can also be given motivation to maintain independence.

In the behavior estimation device 31 according to this embodiment, when estimating behavior, the power data may be divided by time period and the power data may be analyzed for each time period, thereby making it possible to improve the accuracy of estimating human behavior.
In addition, the behavior inference device 31 according to this embodiment may be configured to divide the different habits of each consumer into multiple categories and to create a separate machine learning model for each category, thereby enabling the device to accommodate a variety of consumers.

For example, in the past, a camera and a microphone for collecting sound were installed in a house to detect what activities a target person was doing in the house, and the detected video and audio were checked by a family member or the like who was far away to confirm the safety of the target person. However, in such a case, it is difficult to protect the privacy of the target person, and the installation of a camera and a microphone for collecting sound may be refused. In addition, although such a method provides high accuracy of information, it often takes time to check the video and audio.
In contrast, the behavior inference device 31 according to the present embodiment can suppress such problems by using power information.

For example, in the past, infrared sensors and environmental sensors were installed in a house, and the detected data was confirmed to estimate the living conditions of a subject in the house. For example, a threshold value was set for the time when the sensor reacted or the sensor value, and when these exceeded the threshold, a notification was sent to family members by e-mail or the like to be used to confirm the safety of the subject. However, while such a method can objectively confirm a single event such as sleep, if all of the subject's living behavior is to be grasped, the number of sensors to be installed would be enormous, and the cost would be very high. In addition, the habits of subjects are diverse, and some people may not fall under the threshold, and their behavior may not be estimated.
In contrast, the behavior inference device 31 according to the present embodiment can suppress such problems by using power information.

Second Embodiment
A second embodiment will be described with reference to FIG.

[Behavior estimation device]
FIG. 4 is a diagram showing a schematic configuration of a behavior inference device 31 according to one embodiment (second embodiment) of the present invention.
The configuration and operation of the behavior inference device 31 according to this embodiment differ from the configuration and operation shown in FIG. 3 according to the first embodiment in that the input information is different, but are the same in other respects.
In this embodiment, for convenience of explanation, the same reference numerals as those shown in FIG. 3 are used.

In the example of FIG. 4, the wireless information input section 732 and the additional information input section 733 of the information input section 711 are omitted from the illustration.
When wireless information (wireless strength information in this embodiment) is not used as the input information, the information input unit 711 does not need to be provided with the wireless information input unit 732 .

In this embodiment, individual circuit power information, home appliance separate power information, and information other than power are input to the information input unit 711 of the behavior estimation device 31.

[Regarding the second embodiment]
As described above, the behavior inference device 31 in the behavior inference system 1 according to the present embodiment can effectively infer the behavior of a subject based on power information.

<About the above embodiment>
[Configuration example]
As one configuration example, the behavior inference device (behavior inference device 31 in the example of FIG. 1) includes an information input unit (information input unit 711 in the example of FIG. 3) that inputs input information including separated appliance power (separate home appliance power information 1051 in the example of FIG. 2) and individual circuit power (individual circuit power information 1012 in the example of FIG. 2) obtained from power measured by a power sensor (power sensor 211 in the example of FIG. 1) provided in a residence (house 11 in the example of FIG. 1) of a subject to be monitored (subject 511 in the example of FIG. 1), a behavior inference unit (behavior estimation unit 712 in the example of FIG. 3) that uses a machine learning result to infer the behavior of the subject based on the input information input by the information input unit, and an information output unit (information output unit 713 in the example of FIG. 3) that outputs information regarding the estimation result by the behavior inference unit.

As one configuration example, in the behavior inference device, the input information includes the reception strength (in the example of FIG. 2, wireless strength information 1013) when a signal transmitted from a wireless device (in the example of FIG. 1, router device 212) installed in the home is received by a power sensor.
As one configuration example, in the behavior inference device, input information includes a total amount of power in a home measured by a power sensor (total amount of power information 1011 in the example of FIG. 2 ).

As one configuration example, in the behavior inference device, a room corresponding to the individual circuit power is a room in which there is no device that can obtain power through device separation power.
As one configuration example, in the behavior inference device, the room corresponding to the individual circuit power is at least one of a toilet room (in the example of FIG. 1, the toilet room 114) and a bathroom (in the example of FIG. 1, the bathroom 115) in a home. Note that the toilet room and the bathroom may be a common room.
In one configuration example, in the behavior inference device, the subject is a care recipient.

A behavior estimation method performed by the behavior estimation device as described above may be provided.
As one configuration example, the behavior estimation method is a behavior estimation method performed in a behavior estimation device, and includes an information input step of inputting input information including device separation power and individual circuit power obtained from power measured by a power sensor installed in the home of the subject to be monitored, a behavior estimation step of using machine learning results to estimate the subject's behavior based on the input information input by the information input step, and an information output step of outputting information related to the estimation result by the behavior estimation step.

A program for executing the processes performed by the behavior inference device as described above may be provided.
As one configuration example, it is a program for causing a computer to realize an information input function that inputs input information including device separated power and individual circuit power obtained from power measured by a power sensor installed in the home of the subject to be monitored, a behavior estimation function that uses machine learning results to estimate the subject's behavior based on the input information input by the information input function, and an information output function that outputs information regarding the estimation results by the behavior estimation function.

Here, a program for realizing any function in the behavior inference device 31 according to the embodiment described above may be recorded (stored) in a computer-readable recording medium, and the program recorded in this recording medium may be read into a computer system and executed to perform processing.
It should be noted that the "computer system" referred to here may include an operating system (OS) or hardware such as peripheral devices.
In addition, the term "computer-readable recording medium" refers to a storage device such as a flexible disk, a magneto-optical disk, a ROM (Read Only Memory), a writable non-volatile memory such as a flash memory, a portable medium such as a DVD (Digital Versatile Disc), or a hard disk built into a computer system.
Furthermore, the computer-readable recording medium is, for example, a non-transitory recording medium.

Furthermore, the term "computer-readable recording medium" includes a storage medium that holds a program for a certain period of time, such as a volatile memory (e.g., a DRAM (Dynamic Random Access Memory)) within a computer system that serves as a server or client when a program is transmitted via a network such as the Internet or a communication line such as a telephone line.
The above program may be transmitted from a computer system in which the program is stored in a storage device or the like to another computer system via a transmission medium, or by transmission waves in the transmission medium. Here, the "transmission medium" that transmits the program refers to a medium that has a function of transmitting information, such as a network (communication network) such as the Internet or a communication line (communication line) such as a telephone line.
The above program may be for implementing some of the above functions. Furthermore, the above program may be a so-called differential file (differential program) that can implement the above functions in combination with a program already recorded in the computer system.

The functions of any of the components in the behavior estimation device 31 described above may be realized by a processor. For example, each process in this embodiment may be realized by a processor that operates based on information such as a program and a computer-readable recording medium that stores information such as a program. Here, the functions of each part of the processor may be realized by individual hardware, or the functions of each part may be realized by integrated hardware. For example, the processor may include hardware, and the hardware may include at least one of a circuit that processes digital signals and a circuit that processes analog signals. For example, the processor may be configured using one or more circuit devices mounted on a circuit board, or one or both of one or more circuit elements. An IC (Integrated Circuit) or the like may be used as the circuit device, and a resistor or a capacitor may be used as the circuit element.

Here, the processor may be, for example, a CPU. However, the processor is not limited to a CPU, and various processors such as a GPU (Graphics Processing Unit) or a DSP (Digital Signal Processor) may be used. The processor may be, for example, a hardware circuit using an ASIC (Application Specific Integrated Circuit). The processor may be, for example, a plurality of CPUs, or a hardware circuit using a plurality of ASICs. The processor may be, for example, a combination of a plurality of CPUs and a hardware circuit using a plurality of ASICs. The processor may include, for example, one or more of an amplifier circuit or a filter circuit that processes an analog signal.

The above describes an embodiment of the present invention in detail with reference to the drawings, but the specific configuration is not limited to this embodiment, and design changes and the like that do not deviate from the gist of the present invention are also included.

1...Behavior estimation system, 11...House, 12...Behavior estimation server device, 13...Home appliance separation server device, 14...Network, 31...Behavior estimation device, 51...Home appliance separation device, 111...Entrance, 112...Corridor, 113...Living room, 114...Toilet room, 115...Bathroom, 116...Bedroom, 131-135...Door, 211...Power sensor, 212...Router device, 231...Distribution board, 311...Temperature sensor, 312...Humidity sensor, 331...Entrance sensor, 332...Living room sensor, 333 ...Toilet room sensor, 334...Bathroom sensor, 335...Bedroom sensor, 351...Toilet sensor, 371...Pillow sensor, 511...Subject, 711...Information input unit, 712...Behavior estimation unit, 713...Information output unit, 731...Power information input unit, 732...Wireless information input unit, 733...Additional information input unit, 751...Learning unit, 752...Learning result storage unit, 1011...Total power information, 1012...Individual circuit power information, 1013...Wireless strength information, 1031...Home appliance separation function, 1051...Home appliance separation power information

Claims (8)

an information input unit that inputs input information including device separation power and individual circuit power obtained from power measured by a power sensor installed in the home of a person to be monitored;
a behavior estimation unit that estimates the behavior of the subject based on the input information input by the information input unit using a machine learning result;
an information output unit that outputs information regarding a result of the estimation by the behavior estimation unit;
Equipped with
The device separation power is acquired by performing a predetermined device separation process based on information on the total power amount of the home acquired by the power sensor, and includes information representing the amount of power used by each of one or more devices used in the home,
The room corresponding to the individual circuit power is a room in which no device that can obtain power from the device separation power exists.
Behavioral estimation device. The room corresponding to the individual circuit power is at least one of a toilet room or a bathroom of the residence.
The behavior inference device according to claim 1 . The subject is a person receiving care.
The behavior inference device according to claim 1 or 2 . The input information includes a reception strength of a signal transmitted from a wireless device installed in the home when the signal is received by the power sensor.
The behavior inference device according to claim 1 . The input information includes a total amount of electricity in the home measured by the power sensor.
The behavior inference device according to claim 1 . the device separation process is performed by a server device separate from the behavior inference device.
The behavior inference device according to claim 1 . A behavior estimation method performed in a behavior estimation device, comprising:
An information input step of inputting input information including device separation power and individual circuit power obtained from power measured by a power sensor installed in the home of a person to be monitored;
a behavior estimation step of estimating the behavior of the subject based on the input information inputted by the information input step, using a result of the machine learning;
an information output step of outputting information related to the estimation result by the behavior estimation step;
having
The device separation power is acquired by performing a predetermined device separation process based on information on the total power amount of the home acquired by the power sensor, and includes information representing the amount of power used by each of one or more devices used in the home,
The room corresponding to the individual circuit power is a room in which no device that can obtain power from the device separation power exists.
Behavioral estimation methods. An information input function for inputting input information including device separation power and individual circuit power obtained from power measured by a power sensor installed in the home of the person to be monitored;
A behavior estimation function that estimates the behavior of the subject based on the input information input by the information input function using a machine learning result;
an information output function for outputting information regarding an estimation result by the behavior estimation function;
A program for causing a computer to realize the above,
The device separation power is acquired by performing a predetermined device separation process based on information on the total power amount of the home acquired by the power sensor, and includes information representing the amount of power used by each of one or more devices used in the home,
The room corresponding to the individual circuit power is a room in which no device that can obtain power from the device separation power exists.
program.

Images