Information Processing System
By integrating sensors and machine learning models in the electric vacuum cleaner, user attribute information is automatically estimated, which solves the problem of cumbersome collection of user attribute information, improves the accuracy and detailed information, and reduces the burden of user input.
Patent Information
- Application Number
- CN202110263542.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Priority Date
- 2020-05-29
- Filing Date
- 2021-03-11
- Publication Date
- 2024-11-26
- Estimated Expiration
- 2041-03-11
AI Technical Summary
In the prior art, the collection process of user attribute information is complicated, and the user input information is inaccurate or unwilling to input, resulting in a large burden on information collection and it is difficult to obtain detailed usage status.
By integrating sensors and machine learning models in the electric vacuum cleaner, the user's attribute information is estimated based on the use data of the electric vacuum cleaner, including family structure, gender, age, employment methods, etc., reducing the user's input needs.
It realizes the automatic collection of user attribute information, reduces the burden of user input, improves the accuracy and detail of information, and reduces the complexity of information collection.
Smart Images

Figure CN113729546B_ABST
Abstract
Description
Technical Field
[0001] Embodiments of the present invention relate to an information processing system. Background Art
[0002] There is known an electric vacuum cleaner that receives data related to a cleaning operation unit from an external device via a communication interface and executes control based on the received data.
[0003] In addition, it may be desirable to collect user attribute information for product development, service provision, etc. In this case, it may be necessary to ask the user to complete a questionnaire or register on the Internet, and the burden of collecting information may be heavy.
[0004] Patent Document 1: Japanese Patent Application Publication No. 2019-000442 Summary of the invention
[0005] An object of the present invention is to provide an information processing system capable of reducing the burden of collecting attribute information of a user.
[0006] The information processing system of the embodiment includes an acquisition unit and an estimation unit. The acquisition unit acquires data sent from an electric vacuum cleaner. The estimation unit estimates the attributes of the user of the electric vacuum cleaner based on the input information obtained from the data acquired by the acquisition unit by using a model that learns to output an estimation result of the user's attributes when information related to the electric vacuum cleaner is input.
[0007] Effect of the invention:
[0008] According to the present invention, it is possible to estimate the attributes of a user of a vacuum cleaner. BRIEF DESCRIPTION OF THE DRAWINGS
[0009] Figure 1 It is a diagram showing the overall configuration of the information processing system according to the first embodiment.
[0010] Figure 2 It is a perspective view showing an example of the electric vacuum cleaner according to the first embodiment.
[0011] Figure 3 This is a block diagram showing the configuration of a sensor and a control unit of the electric vacuum cleaner according to the first embodiment.
[0012] Figure 4 This is a diagram conceptually showing a part of data transmitted from the electric vacuum cleaner according to the first embodiment.
[0013] Figure 5 This is a block diagram showing the configuration of a server according to the first embodiment.
[0014] Figure 6 It is a diagram showing an example of the content of user registration information according to the first embodiment.
[0015] Figure 7 This is a diagram showing an example of an estimation model related to unsupervised learning according to the first embodiment.
[0016] Figure 8 This is a diagram showing an example of an estimation model related to supervised learning according to the first embodiment.
[0017] Fig. 9 This is a diagram showing an example of the content of user attribute information according to the first embodiment.
[0018] Fig.10 1 is a flowchart showing the flow of processing performed by the electric vacuum cleaner according to the first embodiment.
[0019] Fig.11 This is a flowchart showing the processing flow of the server according to the first embodiment.
[0020] Fig.12 This is a block diagram showing the configuration of a server according to the second embodiment.
[0021] Fig.13 This is a block diagram showing the configuration of a server according to the third embodiment.
[0022] Marking Description:
[0023] 100, 100A, 100B...Electric vacuum cleaner, 200...Server, 201...Information acquisition unit (acquisition unit), 202...Information conversion unit, 203...Learning unit, 204...User attribute estimation unit (estimation unit), M...Estimation model. DETAILED DESCRIPTION
[0024] Hereinafter, the information processing system of the implementation mode will be described with reference to the accompanying drawings. In the following description, the same mark is attached to the structure having the same or similar function. In addition, the repeated description of these structures is sometimes omitted. The so-called "based on XX" means "at least based on XX", and can also include the case where it is based on other elements in addition to XX. The so-called "based on XX" is not limited to the case where XX is used directly, but can also include the case where it is based on the result of calculating and processing XX. The so-called "XX or YY" is not limited to the case where one of XX and YY is used, but can also include the case where both XX and YY are used. This is also the same when the selected elements are more than 3. "XX" and "YY" are arbitrary elements (for example, arbitrary information).
[0025] (First Embodiment)
[0026] <1. Overall structure of the information processing system>
[0027] Figure 1It is a diagram showing the overall structure of an information processing system (user information collection system) 1 of an implementation method. The information processing system 1 includes, for example, an electric vacuum cleaner 100, a server 200, and a terminal device 300 installed in each household. However, the so-called "information processing system" in this specification may not include the electric vacuum cleaner 100 and the terminal device 300 but only refer to the server 200. The network NW described later may utilize, for example, the Internet, a cellular network, a Wi-Fi network, LPWA (Low Power Wide Area), a WAN (Wide Area Network), a LAN (Local Area Network), other public lines, dedicated lines, etc. depending on the situation.
[0028] The electric vacuum cleaner 100 is placed in the home of the user U. The electric vacuum cleaner 100 can communicate with the server 200 via, for example, a wireless router R and a network NW placed in the home of the user U. The details of the electric vacuum cleaner 100 will be described later.
[0029] The server 200 is composed of one or more server devices SD (e.g., cloud servers). The server 200 may also be referred to as a "server system". The server 200 may also include an information processing unit for edge computing or fog computing, such as an information processing unit included in a router in the network NW. Details of the server 200 will be described later.
[0030] The terminal device 300 is a device such as a personal computer and can communicate with the server 200 via the network NW. The terminal device 300 includes a display device 300a such as a liquid crystal display or an organic EL display (Electro Luminescence) display. The terminal device 300 and the server device SD may be provided integrally in one device.
[0031] <2.Electric vacuum cleaner>
[0032] <2.1 Overall structure of electric vacuum cleaner>
[0033] Figure 2 1 is a perspective view showing an example of an electric vacuum cleaner 100. The electric vacuum cleaner 100 of this embodiment is a so-called stick-type electric vacuum cleaner, and is a cordless electric vacuum cleaner having a built-in battery 15. The electric vacuum cleaner 100 includes, for example, a cleaner body 10, an extension pipe 20, and a head (suction port body) 30.
[0034] The vacuum cleaner body 10 includes, for example, a body casing 11, a handle (grip portion) 12, a dust collecting device 13, an electric blower 14, a battery 15, a circuit board 16, and a communication module 17 (see Figure 3 ).
[0035] The main body case 11 forms the outer contour of the cleaner body 10. The main body case 11 accommodates the electric blower 14, the battery 15, the circuit board 16, the communication module 17, etc. The main body case 11 has an extension tube connection portion 11a to which one end of an extension tube 20 described later is connected.
[0036] The handle 12 is provided at the upper rear end of the main body housing 11. The handle 12 is a portion held by the user U when the electric vacuum cleaner 100 is used to clean a floor (surface to be cleaned). The handle 12 has an operation button 12a which is operated by the user U.
[0037] The dust collecting device 13 is detachably mounted on the main body shell 11. The dust collecting device 13 is a device that separates dust contained in the air sucked into the vacuum cleaner body 10 due to the action of the electric blower 14 described later. For example, the dust collecting device 13 is a centrifugal separation type dust collecting device including a separation unit 13a and a filter 13b. The separation unit 13a separates dust contained in the sucked air from the air by centrifugation (cyclone separation). The filter 13b separates a small amount of dust remaining in the air that has passed through the separation unit 13a (filtration separation). However, the dust collecting device 13 is not limited to a centrifugal separation type dust collecting device, and may also be a filtering type dust collecting device including a dust collecting paper bag, etc.
[0038] The electric blower 14 includes a motor called a fan motor or a main motor, and a turbine rotated by the motor, and is driven to generate negative pressure. The electric blower 14 uses the generated negative pressure to suck dust-containing air from the suction port 31a of the head 30 described later into the dust collector 13, and exhausts the air separated from the dust in the dust collector 13 to the outside of the electric vacuum cleaner 100.
[0039] The battery 15 is a power supply unit that supplies power to the electric blower 14 and the brush motor 34 described later. The circuit board 16 includes a printed wiring board having a wiring pattern and a plurality of electronic components mounted on the printed wiring board. The circuit board 16 implements the control unit 80 (see FIG. 1 ) that controls the operation of the electric vacuum cleaner 100. Figure 3 The communication module 17 includes, for example, a high-frequency circuit and an antenna, and is capable of wireless communication with the router R.
[0040] In addition, a blower accessory 18 is detachably mounted on the cleaner body 10. The blower accessory 18 is a nozzle connected to the exhaust air passage of the cleaner body 10. When the blower accessory 18 is mounted, the user U can use the exhaust air blown from the blower accessory 18 for cleaning. The blower accessory 18 is an example of an "accessory".
[0041] Next, the extension tube 20 is described. The extension tube 20 is formed into an elongated shape, for example, and has a first end 20a and a second end 20b. The first end 20a of the extension tube 20 is airtightly connected to the extension tube connection portion 11a of the vacuum cleaner body 10. The second end 20b of the extension tube 20 is airtightly connected to the head 30. Inside the extension tube 20, a connection wiring is provided to electrically connect the vacuum cleaner body 10 and the head 30.
[0042] Next, the head 30 will be described. The head 30 is a part that moves along the floor. The head 30 includes, for example, a head case 31 , a connection pipe 32 , a rotating brush 33 , a brush motor 34 , and a lamp 35 .
[0043] The head housing 31 is formed in a horizontally long shape, that is, in a strip shape in the left-right direction. The head housing 31 accommodates a brush motor 34. The head housing 31 has a suction port 31a at the lower part facing the ground. The suction port 31a is an opening for sucking dust on the ground by driving the electric blower 14.
[0044] The connecting pipe 32 is a portion that airtightly connects the head housing 31 to the second end 20b of the extension pipe 20, and is rotatably connected to the head housing 31. By connecting the head housing 31 to the extension pipe 20 with the connecting pipe 32, an air path is formed from the suction port 31a of the head housing 31 to the vacuum cleaner body 10 via the extension pipe 20.
[0045] The rotating brush 33 is provided at the suction port 31a and arranged along the ground. The rotating brush 33 is rotatably provided relative to the head housing 31. The rotating brush 33 plays a role in making dust float from the ground or making the front end of the fluff of a carpet stand up. The brushed motor 34 is connected to the rotating brush 33 via a rotating drive mechanism not shown in the figure, and drives the rotating brush 33 (to rotate it). The lamp 35 is provided at the head housing 31, and is lit according to the detection result of the illumination sensor 65 described later when the head 30 is inserted into a dark place (under a bed or sofa or in the gap of furniture, etc.).
[0046] Here, various accessories 36 can be detachably mounted on the second end portion 20b of the extension tube 20 instead of the head 30. The accessories 36 are, for example, a gap nozzle 36A, a round brush 36A, or a head 36C for cleaning a bed. These accessories 36 can also be detachably mounted on the extension tube connection portion 11a of the main body housing 11 instead of the extension tube 20.
[0047] <2.2 Configuration of sensor group and control unit of electric vacuum cleaner>
[0048] Figure 31 is a block diagram showing the structure of the sensor group and the control unit 80 of the electric vacuum cleaner 100. The electric vacuum cleaner 100 includes, for example, an electric blower current detection unit 41, a power detection unit 42, a battery remaining detection unit 43, a battery temperature sensor 44, a main body acceleration sensor 45, a main body gyro sensor 46, a gripping force detection unit 47, a garbage sensor 48, a camera 49, a dust collector detection switch 50, an air supply accessory detection switch 51, a brush motor current detection unit 61, an encoder 62, a head acceleration sensor 63, a head gyro sensor 64, an illumination sensor 65, an extension tube angle sensor 66, and an accessory detection switch 67. Hereinafter, these are sometimes collectively referred to as the "sensor group SU". However, the electric vacuum cleaner 100 does not need to include all of these sensors and switches.
[0049] The electric blower current detection unit 41 includes a shunt resistor, etc., and detects the current value of the current flowing from the battery 15 to the motor of the electric blower 14. The current value of the current flowing to the motor of the electric blower 14 is proportional to the load of the electric blower 14. For example, when the filter 13b is clogged and the load of the electric blower 14 increases, the current value detected by the electric blower current detection unit 41 increases. Therefore, the electric blower current detection unit 41 can also be said to be a detection unit that detects the load of the electric blower 14 (for example, the clogging of the filter 13b). In addition, from another point of view, when the head 30 floats from the ground and the load of the electric blower 14 decreases, the current value detected by the electric blower current detection unit 41 decreases. Therefore, the electric blower current detection unit 41 can also be said to be a detection unit that detects the state of the head 30 relative to the ground (for example, the up and down of the head 30).
[0050] When the vacuum cleaner 100 is connected to an external commercial power source (eg, a household outlet) for charging, the power source detection unit 42 detects that the vacuum cleaner 100 is connected to the commercial power source and whether the AC frequency of the commercial power source is 50 Hz or 60 Hz.
[0051] The battery remaining detection unit 43 detects the remaining capacity of the battery 15 based on the voltage of the battery 15. The battery temperature sensor 44 is provided adjacent to the battery 15 and detects the temperature of the battery 15. The temperature of the battery 15 is substantially the same as the temperature (room temperature) of the surrounding environment of the electric vacuum cleaner 100 immediately after the operation of the electric vacuum cleaner 100 is started. Therefore, the battery temperature sensor 44 can also be said to be a detection unit for detecting the room temperature of the residence of the user U. In addition, the electric vacuum cleaner 100 may also have a sensor that is different from the battery temperature sensor 44 and is specifically used to detect the room temperature.
[0052] The main body acceleration sensor 45 is provided in the cleaner main body 10 , and detects the acceleration of the cleaner main body 10 . The main body gyro sensor 46 is provided in the cleaner main body 10 , and detects the angular velocity of the cleaner main body 10 .
[0053] The gripping force detection unit 47 includes a pressure sensor provided on the handle 12, and detects the force (gripping force) of the user holding the handle 12 and / or the position (gripping position) of the user U in the handle 12. Here, in most cases, the gripping force of men is greater than that of women. In addition, the gripping position is an element that shows different habits of each user U.
[0054] The garbage sensor 48 is provided, for example, in the air intake passage of the vacuum cleaner body 10. The garbage sensor 48 has, for example, a light emitting portion and a light receiving portion separately arranged on the left and right inner side surfaces of the air intake passage, and the light from the light emitting portion to the light receiving portion is blocked by the garbage passing through the air intake passage, thereby detecting the garbage sucked in. For example, the garbage sensor 48 determines that garbage is being sucked in during the period when the light from the light emitting portion to the light receiving portion is blocked. On the other hand, the garbage sensor 48 determines that garbage is not being sucked in during the period when the light from the light emitting portion to the light receiving portion is not blocked. The garbage sensor 48 (or the control unit 80 of the electric vacuum cleaner 100, or the information conversion unit 202 of the server 200 described later) measures the amount of garbage sucked in based on the ratio of the time when garbage is sucked in to the time when garbage is not sucked in (in other words, the time when garbage is sucked in during the operation time).
[0055] The camera 49 is, for example, disposed in the dust collecting device 13 of the vacuum cleaner body 10 to photograph the garbage accumulated inside the dust collecting device 13. Alternatively, the camera 49 may be disposed in the head 30 to photograph the garbage in front of the head 30 (for example, the garbage before being sucked in by the head 30). In addition, the camera 49 may also be disposed in the air intake duct of the vacuum cleaner body 10 to photograph the garbage sucked in by the electric vacuum cleaner 100. The camera 49 is a detection unit that detects the type of garbage. In addition, the so-called "the camera 49 detects the type of garbage" means, for example, that the shooting data captured by the camera 49 is subjected to image recognition (for example, graphic matching or image recognition using a neural network) based on the control unit 80 of the electric vacuum cleaner 100 or the information conversion unit 202 of the server 200 described later to determine the type of garbage.
[0056] The dust collecting device detection switch 50 is provided at a position adjacent to the dust collecting device 13, and detects the installation and removal state of the dust collecting device 13 relative to the vacuum cleaner body 10. The air supply accessory detection switch 51 is provided at a position adjacent to the air supply accessory 18 when the air supply accessory 18 is installed on the vacuum cleaner body 10, and detects the installation state of the air supply accessory 18 relative to the vacuum cleaner body 10.
[0057] The brushed motor current detection unit 61 includes a shunt resistor and the like, and detects the current value of the current flowing from the battery 15 to the brushed motor 34. The current value of the current flowing to the brushed motor 34 is proportional to the load of the brushed motor 34. For example, when the floor is a carpet and the load of the electric blower 14 increases, the current value detected by the brushed motor 34 increases. Therefore, the brushed motor current detection unit 61 can also be said to be a detection unit for detecting the type of floor (floor / mat / carpet, etc.).
[0058] The encoder 62 is provided on the rotating brush 33 (or the brushed motor 34) to detect the rotation speed of the rotating brush 33 (or the brushed motor 34). For example, when the floor is a carpet and the rotation speed of the rotating brush 33 decreases, the rotation speed detected by the encoder 62 decreases. Therefore, the encoder 62 can also be said to be a detection unit for detecting the type of floor (floor / mat / carpet, etc.). In addition, the detection unit for detecting the type of floor can also be a detection unit that replaces the brushed motor current detection unit 61 or the encoder 62, or coexists with the brushed motor current detection unit 61 or the encoder 62, and irradiates light to the floor and detects the type of floor based on the characteristics of the reflected light.
[0059] The head acceleration sensor 63 is provided in the head 30 , and detects the acceleration of the head 30 . The head gyro sensor 64 is provided in the head 30 , and detects the angular velocity of the head 30 .
[0060] The illuminance sensor 65 is provided on the head 30, and when the head 30 is inserted into a dark place (under a bed or sofa, or in a gap between furniture, etc.), it detects that the illuminance is below a predetermined value. When the illuminance sensor 65 detects that the illuminance is below a predetermined value, the control unit 80 described later turns on the lamp 35 to illuminate the dark place.
[0061] The extension tube angle sensor 66 is provided on the connection tube 32 of the head 30, and detects the angle ( Figure 2 For example, when the height of the user using the electric vacuum cleaner 100 is high, the angle of the extension tube 20 relative to the head 30 becomes larger. On the other hand, when the height of the user using the electric vacuum cleaner 100 is low, the angle of the extension tube 20 relative to the head 30 becomes smaller. Therefore, the extension tube angle sensor 66 can also be said to be a detection unit for detecting the height of the user. Here, in most cases, the height of men is higher than that of women.
[0062] The accessory detection switch 67 is provided at the second end portion 20b of the extension tube 20 and / or the extension tube connection portion 11a of the cleaner body 10, and detects the attachment state of the accessory 36 with respect to the second end portion 20b of the extension tube 20 or the extension tube connection portion 11a of the cleaner body 10. Furthermore, the control unit 80 described later detects the power-on state of the object when the attachment of the object is detected by the accessory detection switch 67. Furthermore, the control unit 80 determines that the object is the head 30 when the attached object is powered on. On the other hand, when the attached object is not powered on, the control unit 80 determines that the object is the accessory 36. In the following description, the so-called "detection result of the accessory detection switch 67" is used to include the detection result of the power-on state by the control unit 80 described above. Furthermore, the accessory detection switch 67 can also detect the type of the accessory 36 (gap nozzle 36A, round brush 36B or bed cleaning head 36C).
[0063] Next, the control unit 80 is described. The control unit 80 includes, for example, an operation receiving unit 81, a motor control unit 82, an information recording unit 83, and an information output unit 84. These functional units are implemented by executing a program (software) by a hardware processor such as a CPU (Central Processing Unit) installed in the electric vacuum cleaner 100. However, all or part of these functional units may be implemented by hardware (circuit unit; including circuits) such as ASIC (Application Specific Integrated Circuit), PLD (Programmable Logic Device) or FPGA (Field Programmable Gate Array), or may be implemented by the collaboration of software and hardware.
[0064] Furthermore, the control unit 80 includes a storage unit 85. The storage unit 85 is implemented by, for example, a RAM (Random Access Memory), a ROM (Read Only Memory), a HDD (Hard Disk Drive), a flash memory, or a combination of a plurality of these. The storage unit 85 stores identification information I1 and status information I2.
[0065] The operation accepting unit 81 accepts the operation of the electric vacuum cleaner 100 by the user U according to the operation of the operation button 12a of the electric vacuum cleaner 100 by the user U. The operation of the user U accepted by the operation accepting unit 81 is, for example, turning on / off the power of the electric vacuum cleaner 100, the operation mode (weak / strong / automatic, etc.) of the electric blower 14, and turning on / off the rotation of the rotating brush 33.
[0066] The motor control unit 82 supplies power from the battery 15 to the electric blower 14 based on the operation of the user U accepted by the operation accepting unit 81, and drives the electric blower 14. Similarly, the motor control unit 82 supplies power from the battery 15 to the brush motor 34 based on the operation of the user U accepted by the operation accepting unit 81, and drives the brush motor 34.
[0067] The information recording unit 83, for example, refers to the control command output from the motor control unit 82 and the timer not shown in the figure, and establishes a correspondence between the operating state of the electric vacuum cleaner 100 and the date and time information, and adds it to the state information I2 of the storage unit 85. The "operating state of the electric vacuum cleaner 100" includes, for example, the on / off state of the power supply of the electric vacuum cleaner 100 and the operating mode of the electric blower 14. The date and time information includes information indicating the day of the week and the time. Furthermore, the information recording unit 83 establishes a correspondence between the detection results of the sensor group SU and the date and time information, and adds it to the state information I2 of the storage unit 85. The detection results of the sensor group SU are the detection results of the various sensors and switches mentioned above. Regarding this information, the original data can be recorded, and it can also be recorded in a state where necessary calculations (processing) have been performed.
[0068] The information output unit 84 sends the state information I2 stored in the storage unit 85 to the server 200 via the communication module 17. The information output unit 84 sends the state information I2 to the server 200, for example, at a predetermined period. At this time, the information output unit 84 links the state information I2 with the identification information I1 stored in the storage unit 85 and sends it to the server 200. The identification information I1 is a device ID assigned to each electric vacuum cleaner 100 in order to identify the electric vacuum cleaner 100. The state information I2 and the identification information I1 are examples of data sent from the electric vacuum cleaner 100 to the server 200. Hereinafter, the state information I2 and the identification information I1 are collectively referred to as "data D".
[0069] Alternatively, instead of the above configuration, the information recording unit 83 may be omitted, and the information output unit 84 may send the operating state of the electric vacuum cleaner 100 and the detection results of the sensor group SU to the server 200 in real time. In this case, the server 200 may also establish a correspondence between the operating state of the electric vacuum cleaner 100 and the detection results of the sensor group SU and the date and time information.
[0070] Figure 4 1 is a diagram conceptually showing a part of data D transmitted from the electric vacuum cleaner 100 . Figure 4 1 shows an example of data D obtained in a household of two persons, a husband who is a company employee and a wife who is a housewife. In addition, the "operation time of the electric vacuum cleaner" in the figure is actually an indicator calculated by the server 200 in this embodiment, but is included here for the sake of easy communication. Figure 4 Indicated by Zhonglai.
[0071] Figure 4 It can be seen from the data D shown that the cleaning was performed on Monday afternoon, Wednesday afternoon, Friday afternoon, which are working days, and Sunday morning, which is a rest day. In addition, it can be seen that the height of the user who cleaned on Sunday morning is higher than that of the user who cleaned on Monday afternoon, Wednesday afternoon, and Friday afternoon, and the gripping force of the user who cleaned on Sunday morning is greater than that of the user who cleaned on Monday afternoon, Wednesday afternoon, and Friday afternoon. By analyzing this information, it can be inferred that there are at least two family members, and these family members include males and females. It is highly likely that there are males who rest on Saturdays and Sundays, and females who are unemployed or work part-time.
[0072] <3. Server>
[0073] Figure 5 2 is a block diagram showing the structure of the server 200. The server 200 includes, for example, an information acquisition unit 201, an information conversion unit 202, a learning unit 203, a user attribute estimation unit 204, an information recording unit 205, and an information output unit 206. These functional units are implemented by executing a program (software) by a hardware processor such as a CPU included in the server 200. However, all or part of these functional units may be implemented by hardware (circuitry; including circuits) such as ASIC, PLD, or FPGA, or may be implemented by the cooperation of software and hardware.
[0074] The server 200 further includes a storage unit 207. The storage unit 207 is implemented by, for example, a RAM, a ROM, a HDD, a flash memory, or a combination of these. The storage unit 207 stores the accumulated information I11, the user registration information I12, the estimation model (learning model) M, and the user attribute information I13.
[0075] The information acquisition unit 201 acquires (obtains) data D transmitted from the electric vacuum cleaner 100. In addition, when learning the estimation model M, the information acquisition unit 201 collects data D from a plurality of electric vacuum cleaners 100 used in a plurality of households. The information acquisition unit 201 accumulates the acquired data D as accumulated information I11 in the storage unit 207. The information acquisition unit 201 accumulates the data D for each electric vacuum cleaner 100 for a predetermined period (e.g., 3 months), for example. The information acquisition unit 201 is an example of an "acquisition unit".
[0076] The information conversion unit 202 generates input information to be input to the estimation model M described later, based on the accumulated information I11 in which the data D of each electric vacuum cleaner 100 is accumulated. The estimation model M is a model for estimating the attributes of the user U (type of cohabiting household (e.g., size of cohabiting household), presence or absence of children, presence or absence of pets, gender, age, employment type, marital status, housing type (e.g., size of housing), floor type of room, residential area, etc.).
[0077] (a. Input information for estimating attributes of a cohabiting household)
[0078] The information conversion unit 202 generates, for example, information related to the operating time of the electric vacuum cleaner 100, information related to the time until a specified amount of garbage is accumulated inside the electric vacuum cleaner 100, information related to the time from the end of the operation to the resumption of the operation of the electric vacuum cleaner 100, information related to the number of times the power of the electric vacuum cleaner 100 is turned on / off, information related to the number of times the head 30 of the electric vacuum cleaner 100 is lifted, information related to the day of the week and the time period when the electric vacuum cleaner 100 is used, information related to the angle of the extension tube 20 of the electric vacuum cleaner 100 relative to the head 30, information related to the gripping force or gripping position of the handle 12 by the user U of the electric vacuum cleaner 100, and information related to the number of electric vacuum cleaners 100 in the same residence, as input information to be input into the estimation model M for estimating attributes about a cohabiting household (hereinafter referred to as "input information for estimating a cohabiting household"). In addition, the number of people living in the same household can also be regarded as the number of users U (number of users) who use the electric vacuum cleaner 100.
[0079] The "information related to the operation time" is the average value or maximum value of the operation time of the electric vacuum cleaner 100 in each cleaning, or the cumulative value of the operation time of the electric vacuum cleaner 100 in the above-mentioned prescribed period (for example, 3 months). The so-called "each cleaning" means that even if the power of the electric vacuum cleaner 100 is set to be disconnected, if the time until the next power is turned on is within the prescribed time (for example, within 10 minutes), these operations are also calculated as the same cleaning. The so-called "operation time" is the time when the power of the electric vacuum cleaner 100 is set to be turned on (that is, the time when the electric blower 14 rotates). The above-mentioned average value, maximum value or cumulative value is calculated, for example, based on the "operation status of the electric vacuum cleaner" and date and time information contained in the data D. Here, in most cases, the more people living together in a family, the larger the house or the house with more rooms. In addition, the above-mentioned average value, maximum value or cumulative value tends to become larger as the house is larger and / or the number of rooms is more.
[0080] The information on the time until the specified amount of garbage is accumulated is the time interval from when the dust collector detection switch 50 detects the removal of the dust collector 13 at a certain time to when the same removal is detected next time (see Figure 4 The data D may be an average value of T1 in the data D, or an average value of the time from a certain moment (for example, the moment when the removal of the dust collector 13 is detected by the dust collector detection switch 50) to the time when the garbage detected by the garbage sensor 48 reaches a specified amount, etc. These average values are calculated based on, for example, the detection result of the dust collector detection switch 50, the detection result of the garbage sensor 48, and the date and time information contained in the data D. The act of removing the dust collector 13 from the vacuum cleaner body 10 can be regarded as the moment when garbage accumulates in the dust collector 13 and the accumulated garbage is discarded by the user U. Here, generally, the more people living in the same household, the more garbage is generated. Moreover, the more garbage is generated, the shorter the above average value is.
[0081] "Information about the time from the end of operation to the restart of operation" is the time interval from the end of a certain cleaning to the start of the next cleaning (refer to Figure 4The average value is calculated, for example, based on the "operating state of the electric vacuum cleaner" and date and time information contained in the data D. Here, generally, the more people there are in a cohabiting household, the more garbage is generated and the frequency of cleaning increases. In addition, the more people there are in a cohabiting household, the more cases of cleaning at the timing desired by each cohabitant, and the frequency of cleaning increases. And, the more the frequency of cleaning increases, the shorter the above-mentioned average value. "Information related to the time from the end of operation to the resumption of operation" can be information for estimating the existence of a user U with a tendency to clean for a shorter time per time and the existence of a user U with a tendency to clean for a longer time per time. In other words, "information related to the time from the end of operation to the resumption of operation" is information useful for estimating the number of people using the electric vacuum cleaner 100.
[0082] The so-called "information related to the number of times the power is turned on / off" is the average or maximum value of the number of times the power of the electric vacuum cleaner 100 is turned on / off in each cleaning, or the cumulative value of the number of times the power of the electric vacuum cleaner 100 is turned on / off in the above-mentioned prescribed period (for example, 3 months). These average values, maximum values or cumulative values are calculated, for example, based on the "operating state of the electric vacuum cleaner" and date and time information contained in the data D. Here, generally, the more people there are in a cohabiting household, the more furniture and the more rooms there are, so the number of times the power of the electric vacuum cleaner 100 is turned on / off increases. The "information related to the number of times the power is turned on / off" can be used as information for estimating the existence of a user U with a tendency to have a shorter cleaning time per time, the existence of a user U with a tendency to have a longer cleaning time per time (or the existence of a user U who frequently stops the power, the existence of a user U who rarely stops the power). In other words, the "information related to the number of times the power is turned on / off" is information useful for estimating the number of users of the electric vacuum cleaner 100.
[0083] The so-called "number of times the head is lifted" is the average or maximum number of times the head 30 is lifted in each cleaning, or the cumulative value of the number of times the head 30 is lifted in the above-mentioned specified period (for example, 3 months). These average values, maximum values or cumulative values are calculated based on, for example, the detection results of the electric blower current detection unit 41, the detection results of the head acceleration sensor 63, the detection results of the head gyro sensor 64, and date and time information contained in the data D. Here, generally, the more people living together in a household, the more furniture, the more rooms, and the more cases of housing types including stairs, so the number of times the head 30 is lifted increases.
[0084] The so-called "information related to the day of the week and time period of use" is, for example, information indicating whether the day of the week in which the electric vacuum cleaner 100 is used is Monday to Friday or Saturday, Sunday, or holiday, information indicating whether the time period in which the electric vacuum cleaner 100 is used is morning, daytime, or nighttime, etc. In addition, the so-called "information related to the time period" in this specification may also be information indicating the time itself. Such information is derived, for example, based on the "operating state of the electric vacuum cleaner" and date and time information contained in the data D. Here, generally, the more people there are in a cohabiting household, the more cases there are of full-time housewives, non-employed elderly people, and students. In the case of full-time housewives, non-employed elderly people, and students, there are more opportunities to clean in the morning and daytime on weekdays. In other words, the day of the week and time period in which the electric vacuum cleaner 100 is used has different tendencies depending on whether the user U is dedicated to doing housework, working together, or living alone. That is, the "information related to the day of the week and time period of use" is information useful for estimating the number of users of the electric vacuum cleaner 100.
[0085] The so-called "information related to the angle of the extension tube relative to the head" is, for example, information indicating whether the difference between the average value of the angle of the extension tube 20 relative to the head 30 in a certain cleaning and the average value of the angle of the extension tube 20 relative to the head 30 in other cleanings exceeds a specified amount. This information is derived, for example, based on the detection results of the extension tube angle sensor 66 contained in the data D and date and time information. Here, usually, if the number of people living together in a household is more than 2, the possibility of including both men and women is high. And, in the case where a household living together includes both men and women, sometimes the difference in the angle of the extension tube 20 relative to the head 30 exceeds a specified amount. That is, the "information related to the angle of the extension tube relative to the head" is information useful for estimating the number of people using the electric vacuum cleaner 100.
[0086] The so-called "information related to the gripping force of the user on the handle" is, for example, information indicating whether the difference between the average value (or gripping position) of the gripping force of the user in a certain cleaning and the average value (or gripping position) of the gripping force of the user in other cleanings exceeds a specified amount. This information is derived, for example, based on the detection result of the gripping force detection unit 47 contained in the data D and the date and time information. Here, usually, if the number of people living together in a household is more than 2, the possibility that the cohabiting household includes both men and women is high. And, in the case where the cohabiting household includes both men and women, sometimes the difference in the gripping force of the users exceeds a specified amount. In addition, if the number of people living together in a household is more than 2, the possibility of the existence of users holding the handle 12 in different gripping positions becomes higher. That is, the "information related to the gripping force of the user on the handle" is information useful for estimating the number of people using the electric vacuum cleaner 100.
[0087] The so-called "information related to the number of electric vacuum cleaners in the same residence" is, for example, information indicating whether there are more than two electric vacuum cleaners 100 in the residence of user U. This information is derived, for example, based on the identification information I1 (device ID) of the electric vacuum cleaner 100 contained in the data D and the user registration information I12 stored in the storage unit 207. Here, generally, if there are more than two electric vacuum cleaners 100, it is highly likely that the preferences for the categories of electric vacuum cleaners 100 (categories such as pole type / can type / sweeping robot, or categories such as suction-oriented type / portability-oriented type, etc.) are different for multiple cohabitants. In other words, when there are more than two electric vacuum cleaners 100, the possibility that the number of people living in the cohabiting household is more than two becomes higher. That is, the "information related to the number of electric vacuum cleaners in the same residence" is information useful for estimating the number of people using the electric vacuum cleaner 100.
[0088] Here, the number of people living together in the household (or the number of people using the electric vacuum cleaner 100) is an attribute representing the family structure of the user U. Furthermore, by adding the number of people living together in the household (or the number of people using the electric vacuum cleaner 100) and the presence or absence of children and / or the attributes related to their ages described later, an attribute representing a more detailed family structure can be inferred.
[0089] For example, if the user U of the electric vacuum cleaner 100 is estimated to be one person, the family structure of the user U can be considered to be a single-person household. For example, if the number of people estimated to be a cohabiting household is two and the age group is the same generation, the family structure of the user U can be considered to be a household consisting of only a married couple.
[0090] For example, when the number of people estimated as a cohabiting household is multiple and includes children and adults, it can be considered that the family structure of the user U is a family of parents and children (a two-generation household, child-rearing age group). In this case, in order to estimate the presence or absence of children, the "input information for estimating a cohabiting household" may also include information such as "input information for estimating children, etc.", "input information for estimating age," and "input information for estimating employment method" described later.
[0091] For example, when the estimated adults include adults of child-rearing age and elderly adults, it can be considered that the family structure of the user U is a three-generation family. In this case, in order to estimate the age group of the adults, the "input information for estimating cohabiting households" may also include information such as "input information for estimating children, etc.", "input information for estimating age," and "input information for estimating employment methods," which will be described later.
[0092] Figure 61 is a diagram showing an example of the content of user registration information I12. In the user registration information I12, the user ID, the device ID of the electric vacuum cleaner 100, the model information of the electric vacuum cleaner 100, and the model category of the electric vacuum cleaner 100 are registered in correspondence. The user registration information I12 is obtained, for example, by the user registering when the product is purchased. The information conversion unit 202 obtains information indicating the number of electric vacuum cleaners 100 registered in correspondence with the same user ID by referring to the user ID based on the identification information I1 (device ID) included in the data D received from the electric vacuum cleaner 100.
[0093] (b. Input information for estimating the presence or absence of children or pets)
[0094] Back to Figure 5 The explanation continues. For example, the information conversion unit 202 generates information related to the category of garbage sucked by the electric vacuum cleaner 100 based on the accumulated information I11 as input information to the estimation model M in order to estimate the presence or absence of children and / or the presence or absence of pets of the user U (hereinafter referred to as "input information for estimating children, etc."). For example, the information conversion unit 202 performs graphic matching on the captured data captured by the camera 49 or uses image recognition such as a neural network to generate information related to the category of garbage. "Information related to the category of garbage" is information indicating whether a specific category of garbage exists in excess of a specified amount. Here, generally, when there are children in a cohabiting household, there is more garbage containing sebum. In addition, when there are pets, there is more garbage containing pet hair. "Garbage containing sebum" and / or "pet hair" are examples of the above-mentioned specific categories of garbage.
[0095] (c. Input information for estimating gender)
[0096] For example, the information conversion unit 202 generates one or more of information related to the angle of the extension tube 20 of the electric vacuum cleaner 100 relative to the head 30, information related to the gripping force of the user U of the electric vacuum cleaner 100 on the handle 12, information related to the acceleration acting on the electric vacuum cleaner 100, information related to the number of times the filter of the electric vacuum cleaner 100 is clogged, and information related to the type of garbage sucked by the electric vacuum cleaner 100, as input information (hereinafter referred to as "input information for gender estimation") input to the estimation model M for estimating the gender of the user U. In addition, "information related to the angle of the extension tube relative to the head", "information related to the gripping force of the user on the handle", and "information related to the type of garbage sucked" are as described above. Here, generally, when the user U is a female, the garbage often contains long hair.
[0097] The "information related to acceleration" is an average value or a maximum value of the acceleration acting on the vacuum cleaner body 10 during the above-mentioned specified period (for example, 3 months). These average values and maximum values are calculated, for example, based on the detection results of the main body acceleration sensor 45 included in the data D. Here, generally, when the user U is a male, the acceleration acting on the vacuum cleaner body 10 becomes higher than when the user U is a female.
[0098] The "information on the number of times the filter is clogged" is the number of times the load of the electric blower 14 exceeds a certain point in time (for example, the moment when the dust collector 13 is removed by the dust collector detection switch 50) during the above-mentioned prescribed period (for example, three months), and the user U is notified of the filter being clogged (for example, a notification lamp not shown is turned on by the control of the control unit 80). This number is calculated, for example, based on the detection result of the electric blower current detection unit 41 and date and time information included in the data D. Here, generally, when the user U is a female, the garbage contains more long hairs than when the user U is a male, and the number of times the filter is clogged increases.
[0099] (d. Input information for estimating age)
[0100] For example, the information conversion unit 202 derives information indicating the model of the electric vacuum cleaner 100 based on the accumulated information I11 as input information input to the estimation model M in order to estimate the age of the user U (hereinafter referred to as "input information for age estimation"). This information is derived, for example, based on the identification information I1 (device ID) of the electric vacuum cleaner 100 contained in the data D, and the user registration information I12 stored in the storage unit 207. Here, generally, the elderly have a tendency to prefer models that emphasize lightness over models that emphasize suction power. In addition, the elderly have a tendency to prefer canister types over pole types. In addition, the elderly have a tendency to prefer filter types (dust collection paper bag types) as dust collection methods over centrifugal types. In addition, popular models are sometimes different according to age groups such as 20-year-olds / 30-year-olds / 40-year-olds. Information indicating the model of the electric vacuum cleaner 100 is an example of "information that can determine the model."
[0101] (e. Input information for inferring employment method)
[0102] For example, the information conversion unit 202 generates information related to the day of the week and time period when the electric vacuum cleaner 100 is used based on the accumulated information I11 as input information (hereinafter referred to as "input information for estimating the employment mode") to be input into the estimation model M in order to estimate the employment mode (full-time / part-time / non-employed) of the user U. The "information related to the day of the week and time period when the electric vacuum cleaner 100 is used" is as described above.
[0103] (f. Input information for estimating housing type)
[0104] For example, the information conversion unit 202 generates one or more of information related to the operating time of the electric vacuum cleaner 100, information related to the number of times the electric vacuum cleaner 100 as a wired type is reconnected to an external power supply (commercial power supply), information related to accessories installed and removed from the electric vacuum cleaner 100, information related to the lighting state of the lamp provided on the head of the electric vacuum cleaner 100, and information related to the number of electric vacuum cleaners 100 in the same residence, as input information (hereinafter referred to as "input information for residence type estimation") input to the estimation model M for estimating the residence type of the user U. The "information related to the operating time" and the "information related to the number of electric vacuum cleaners in the same residence" are as described above. Here, generally, when there are two or more electric vacuum cleaners 100, it is likely that the residence is large and has many rooms.
[0105] The “information on the number of times of reconnection to the external power source” is the average or maximum number of times of reconnection to the external power source in each cleaning, or the cumulative value of the number of times of reconnection in the above-mentioned prescribed period (for example, 3 months). The “number of times of reconnection to the external power source” is, for example, the number of times the plug of the electric vacuum cleaner 100 is plugged in and out of a household socket. This number is calculated, for example, based on the detection result of the power source detection unit 42 and date and time information contained in the data D. Here, generally, the larger the room and the more rooms there are, the greater the number of times of reconnection.
[0106] "Information on attached and detached accessories" is information indicating the number of times the accessory 36 or the air supply accessory 18 is attached during the above-mentioned predetermined period (e.g., 3 months). The attachment of the accessory 36 or the air supply accessory 18 is derived, for example, based on the detection result of the accessory detection switch 67, the detection result of the air supply accessory detection switch 51, and date and time information contained in the data D. Here, generally, in most cases, the larger the house, the more areas to be cleaned using the accessory 36 or the air supply accessory 18.
[0107] The "information on the lighting state of the lamp" is the average or maximum number of times the lamp 35 is lit in each cleaning, or the cumulative value of the number of times the lamp 35 is lit in the above-mentioned predetermined period (for example, 3 months). These average values, maximum values, or cumulative values are calculated, for example, based on the detection results of the illumination sensor 65 and date and time information included in the data D. Here, generally, in most cases, the larger the house, the more furniture, etc., and the more dark places to be cleaned that need to light the lamp 35 increase.
[0108] (g. Input information for estimating ground type)
[0109] For example, the information conversion unit 202 generates one or more of information related to the load state of the brushed motor 34 of the electric vacuum cleaner 100, information related to the acceleration acting on the head 30 of the electric vacuum cleaner 100, and information related to the detection result of the garbage sensor 48 based on the accumulated information I11, as input information input into the estimation model M for estimating the floor type of the room (hereinafter referred to as "input information for floor type estimation").
[0110] The "information on the load state of the brushed motor" is information indicating the time change of the load state of the brushed motor 34. This information is derived, for example, based on the detection result of the brushed motor current detection unit 61 and date and time information contained in the data D. Here, generally, among carpets, mats, and floors, the load acting on the brushed motor 34 is the highest when the carpet is used, the second highest when the mat is used, and the lowest when the floor is used.
[0111] "Information related to the acceleration acting on the head" is information indicating the time change of the acceleration acting on the head 30, etc. This information is derived, for example, based on the detection result of the head acceleration sensor 63 and date and time information contained in the data D. Here, generally, among carpets, mats, and floors, the resistance to the head 30 is the highest on carpets, the second highest on mats, and the lowest on floors. Therefore, on the floor, the head 30 moves smoothly, and the acceleration of the head 30 tends to increase. On the other hand, on the carpet, it is difficult for the head 30 to move smoothly, and the acceleration of the head 30 tends to decrease.
[0112] The "information related to the detection result of the garbage sensor" is information indicating the time when the garbage sensor 48 detects garbage in a time series, for example. This information is derived, for example, from the detection result of the garbage sensor 48 and date and time information contained in the data D. Here, generally, garbage is found all over the carpet, while on the floor, garbage tends to be concentrated in the corners of the room. Therefore, when cleaning the floor, the detection result of the garbage sensor 48 shows the time when garbage is not found and the time when garbage is found intensively.
[0113] (h. Input information for estimating the area of residence)
[0114] For example, the information conversion unit 202 derives one or more of information related to the temperature detected in the electric vacuum cleaner 100 and information related to the frequency of an external power supply connected to the electric vacuum cleaner 100 based on the accumulated information I11 as input information to be input into the estimation model M in order to estimate the residential area of the user U (hereinafter referred to as "input information for residential area estimation").
[0115] "Temperature-related information" is information indicating the temperature of the battery 15 immediately after the operation of the electric vacuum cleaner 100 is started in each cleaning operation, etc. This information is derived, for example, based on the detection results of the battery temperature sensor 44 and date and time information contained in the data D. As described above, the temperature of the battery 15 immediately after the operation of the electric vacuum cleaner 100 is started can be considered to be substantially the same as the room temperature. Here, the server 200 may obtain temperature information of various places from the server of the Meteorological Agency, and use the comparison result between the obtained temperature information of various places and the detection result of the battery temperature sensor 44 as input information for estimating the residential area. According to such a structure, the area where the residence of the user U is located can be estimated based on the comparison information of the temperature of various places and the room temperature.
[0116] The “information about the frequency of the external power source” is information indicating whether the AC frequency of the external power source connected to the electric vacuum cleaner 100 is 50 Hz or 60 Hz. This information is derived, for example, based on the detection result of the power source detection unit 42 included in the data D. Here, the frequency of the commercial power source in eastern Japan is 50 Hz, and the frequency of the commercial power source in western Japan is 60 Hz.
[0117] Next, the learning unit 203 will be described. The learning unit 203 generates an estimation model M by machine learning based on the various input information described above. The estimation model M is learned to output an estimation result of the attribute of the user U of the vacuum cleaner 100 when information on the vacuum cleaner 100 is input.
[0118] In the present embodiment, as an estimation model M for estimating the attributes of the user U, the learning unit 203 generates an estimation model MA for estimating attributes related to a cohabiting household (e.g., the size of the cohabiting household or the family structure), an estimation model MB for estimating attributes related to the presence or absence of children or pets, an estimation model MC for estimating attributes related to gender, an estimation model MD for estimating attributes related to age, an estimation model ME for estimating attributes related to employment method, an estimation model MF for estimating attributes related to residence (e.g., residence type), an estimation model MG for estimating attributes related to the floor type of the room, and an estimation model MH for estimating attributes related to the residential area. In the present specification, "learning" may also refer to either unsupervised learning or supervised learning. Below, an example of generating an estimation model M by unsupervised learning and an example of generating an estimation model M by supervised learning are described respectively. The estimation model M is not limited to a neural network, and may also be a model based on a mathematical formula and a threshold.
[0119] First, unsupervised learning will be described. For example, the learning unit 203 generates an estimation model M by clustering the various input information described above. The clustering method is, for example, the K-means method, but other methods may also be used.
[0120] Figure 7 is a diagram showing an example of an estimation model M obtained by unsupervised learning. Figure 7 This is an example of an estimation model MA for estimating attributes related to the household of the user U. Figure 7 The example shown is the result of generating a model in which the "information related to the operating time" and the "information related to the time until garbage accumulation" obtained by the information conversion unit 202 are given as two input information (parameters X1, X2), and the number of clusters is set to 3 (single / nuclear family / extended family). That is, the first group G1 with a relatively short operating time and a relatively long time until garbage accumulation is classified as a single, the third group G3 with a relatively long operating time and a relatively short time until garbage accumulation is classified as an extended family, and the second group G2 located between them is classified as a nuclear family. In addition, in this specification, "extended family" is used to expand the meaning of family relative to the nuclear family. A single person can also be regarded as a "single family". A nuclear family can also be regarded as a "two-generation family". An extended family can also be regarded as a "three-generation family".
[0121] As described above, based on the information related to the operating time of the electric vacuum cleaner 100, the information related to the time until a specified amount of garbage is accumulated inside the electric vacuum cleaner 100, the information related to the time from the end of the operation of the electric vacuum cleaner 100 to the resumption of operation, the information related to the number of times the power of the electric vacuum cleaner 100 is turned on / off, the information related to the number of times the head 30 of the electric vacuum cleaner 100 is lifted, the information related to the day of the week and the time period when the electric vacuum cleaner 100 is used, the information related to the angle of the extension tube 20 relative to the head 30 of the electric vacuum cleaner 100, the information related to the user's gripping force (or gripping position) on the handle 12 of the electric vacuum cleaner 100, and the information related to the number of electric vacuum cleaners 100 in the same residence, one or more input information (for example, two or more input information) are clustered to derive an estimation model MA for estimating attributes related to the cohabiting household of the user U. However, instead of this, clustering may be performed based on one piece of input information included in the above-mentioned various information to derive the estimated model M, or clustering may be performed based on three or more pieces of input information included in the above-mentioned various information to derive the estimated model M.
[0122] Here, as the user U, there are users who clean carefully and users who give priority to time and clean roughly. Therefore, if, for example, the attributes such as the number of people living together and the size of the house are estimated based only on the information related to the operation time of the electric vacuum cleaner 100, the accuracy of the estimation is sometimes difficult to improve. However, for example, by supplementing the use of two or more input information such as "the operation time of the electric vacuum cleaner 100" and "the information related to the time until the garbage is accumulated", it can be estimated that the user U is a type of user who cleans carefully and the number of people living together is small (or the house is small) when the operation time of the electric vacuum cleaner 100 is long and there is not much garbage, and the user U is a type of user who gives priority to time and cleans roughly when there is a large number of people living together (or the house is large) when a large amount of garbage is obtained even if the operation time of the electric vacuum cleaner 100 is short. In this way, by using two or more (more specifically, three or more) input information, the estimation accuracy can be further improved.
[0123] Although several examples of input information have been described above, the input information is not limited to the above examples. The input information is information included in the data D received from the electric vacuum cleaner 100, and various information that cannot be described as specific examples can be appropriately adopted.
[0124] Next, other estimation models MB to MH are described. By clustering the above-mentioned information on the types of garbage sucked by the electric vacuum cleaner 100 as input information, an estimation model MB that estimates attributes related to the presence or absence of children or the presence or absence of pets is derived. The number of clusters is set to 2 (the presence or absence of children, or the presence or absence of pets), for example.
[0125] An estimation model MC for estimating gender-related attributes is derived by clustering one or more (e.g., two or more) of the information related to the angle of the extension tube 20 relative to the head 30 of the electric vacuum cleaner 100, the information related to the gripping force of the handle 12 of the electric vacuum cleaner 100 by the user, the information related to the acceleration acting on the electric vacuum cleaner 100, and the information related to the number of times the filter of the electric vacuum cleaner 100 is clogged as input information. The number of clusters is set to 2 (male / female).
[0126] The information indicating the model of the vacuum cleaner 100 described above is clustered as input information to derive an estimation model MD for estimating attributes related to age. The number of clusters is set to, for example, three (young people / child-rearing age group / elderly people).
[0127] By clustering the information on the days and hours of the week when the vacuum cleaner 100 is used as input information, an estimation model ME for estimating attributes related to the employment style is derived. The number of clusters is set to, for example, three (full-time / part-time / non-employed).
[0128] An estimation model MF for estimating the attributes related to the residence is derived by clustering one or more (e.g., two or more) of the information related to the operation time of the electric vacuum cleaner 100, the number of times the electric vacuum cleaner 100 as a wired type is reconnected to the external power supply, the information related to the accessories installed and removed from the electric vacuum cleaner 100, the information related to the lighting state of the lamp provided on the head of the electric vacuum cleaner 100, and the information related to the number of electric vacuum cleaners 100 in the same residence as input information. The number of clusters is set to, for example, three (large residence / medium residence / small residence).
[0129] By clustering one or more (for example, two or more) of the information about the load state of the brush motor 34 of the electric vacuum cleaner 100, the information about the acceleration acting on the head 30 of the electric vacuum cleaner 100, and the information about the detection result of the garbage sensor 48 as input information, an estimation model MG for estimating the attribute related to the floor type of the room is derived. The number of clusters is set to three (floor / mat / carpet), for example.
[0130] By clustering one or more (e.g., two or more) of the above-mentioned information related to the temperature detected in the electric vacuum cleaner 100 and the information related to the frequency of the external power supply connected to the electric vacuum cleaner 100 as input information, an estimation model MH for estimating the attribute related to the residential area is derived. The number of clusters is set to, for example, 9 (Hokkaido / Tohoku / Kanto Koshinetsu / Tokai / Hokuriku / Kansai / Chugoku / Shikoku / Kyushu Okinawa) or 2 (Eastern Japan / Western Japan).
[0131] on the other hand, Figure 8 : is a diagram showing an example of an inference model M used in supervised learning. The inference model M used in supervised learning is composed of a neural network including an input layer, a hidden layer (intermediate layer) and an output layer. The inference model M is learned by adjusting the weighting coefficients between the nodes included in the input layer, the hidden layer and the output layer through deep learning using training data. The training data is, for example, data in which the above-mentioned various input information collected from the electric vacuum cleaner 100 of users who assist in the questionnaire and users who register on the network (hereinafter referred to as "specific users") is matched with the correct data of the user attributes entered by the above-mentioned specific users through the questionnaire or registered on the network.
[0132] In the present embodiment, the output information about the estimated model M in the supervised learning corresponds to the classification of the various clusters described in the above-mentioned unsupervised learning. For example, the estimation model MA for estimating attributes related to a cohabiting household is learned to use one or more (for example, two or more) of the information related to the operating time of the electric vacuum cleaner 100, the information related to the time until a specified amount of garbage is accumulated inside the electric vacuum cleaner 100, the information related to the time from the end of the operation of the electric vacuum cleaner 100 to the restart of the operation, the information related to the number of times the power of the electric vacuum cleaner 100 is turned on / off, the information related to the number of times the head 30 of the electric vacuum cleaner 100 is lifted, the information related to the days of the week and the time period when the electric vacuum cleaner 100 is used, the information related to the angle of the extension tube 20 relative to the head 30 of the electric vacuum cleaner 100, the information related to the user's gripping force (or gripping position) on the handle 12 of the electric vacuum cleaner 100, and the information related to the number of electric vacuum cleaners 100 in the same residence as input information, and output as output information an estimated result indicating one of single / nuclear family / extended family. Furthermore, part or all of the input information is not limited to the information calculated or derived by the information conversion unit 202, but may be the data D itself (eg, time series data indicating the detection results of the sensor group SU) received from the electric vacuum cleaner 100. This is also the same for the other estimation models MB to MH.
[0133] The user attribute estimation unit 204 estimates the user attributes of the user U of the determination target cleaning machine 100 (hereinafter referred to as the determination target cleaning machine 100) based on the data D received from the electric vacuum cleaner 100 of the determination target of the user attribute (hereinafter referred to as the "determination target cleaning machine 100") using the estimation model M obtained by the learning unit 203. Here, the so-called "estimation" is not limited to determining the most likely attribute candidate, but also includes the case where the probability of each of a plurality of attribute candidates (for example, the probability of being single: 13%, the probability of a nuclear family: 65%, the probability of a large family: 22%) is output.
[0134] The user attribute estimation unit 204 inputs input information based on the data D obtained from the judgment target cleaning machine 100 to the estimation model M, and obtains the estimation result of the user attribute of the user U of the judgment target cleaning machine 100 as the output information of the estimation model M. In the present embodiment, the user attribute estimation unit 204 inputs the above-mentioned cohabitation household estimation input information, children, etc. estimation input information, gender estimation input information, age estimation input information, employment type estimation input information, housing type estimation input information, ground type estimation input information, and residential area estimation input information as input information based on the data D obtained from the judgment target cleaning machine 100 to the corresponding estimation models MA to MG, respectively, and outputs the estimation result of the type of cohabitation household (for example, the size of the cohabitation household or the family structure), the presence or absence of children, the presence or absence of pets, the gender, age, employment type, housing type, ground type, and residential area of the user U of the judgment target cleaning machine 100 as the output information of the estimation models MA to MG.
[0135] The information recording unit 205 stores the user attributes estimated by the user attribute estimation unit 204 in the storage unit 207 as user attribute information I13.
[0136] Fig. 9 1 is a diagram showing an example of the content of user attribute information I13. In the user attribute information I13, for example, user ID, type of cohabiting household, presence or absence of children, presence or absence of pets (not shown), gender, age, employment type, housing type (e.g., size of housing), ground type (not shown), and residential area (not shown) are registered in correspondence.
[0137] The information output unit 206 transmits the user attribute information I13 obtained through the above-mentioned processing to the terminal device 300. Thus, the user attribute information I13 can be used for product development and service provision.
[0138] <4. Processing Flow>
[0139] Next, the flow of processing will be described.
[0140] Fig.10 1 is a diagram showing the process flow of the electric vacuum cleaner 100. First, the control unit 80 determines whether the power of the electric vacuum cleaner 100 is turned on (S101). When the power of the electric vacuum cleaner 100 is turned off (OFF), the control unit 80 waits until the power of the electric vacuum cleaner 100 is turned on.
[0141] On the other hand, when the electric vacuum cleaner 100 is powered on ( S101 : YES), the control unit 80 transmits the detection result detected by the sensor group SU to the server 200 in a predetermined cycle or in real time ( S102 ).
[0142] Next, the control unit 80 determines whether the power of the electric vacuum cleaner 100 is turned off (S103). When the power of the electric vacuum cleaner 100 is turned on (S103: No), the control unit 80 repeats the process of S102. On the other hand, when the power of the electric vacuum cleaner 100 is turned off (S103: Yes), the control unit 80 ends a series of processes. And, the electric vacuum cleaner 100 repeats the above-mentioned processes (S101 to S103) for a predetermined period of time, for example.
[0143] Fig.11 2 is a diagram showing the flow of processing by the server 200. As a premise, the data D transmitted from the electric vacuum cleaner 100 to the server 200 is acquired by the information acquisition unit 201 and accumulated as the accumulated information I11.
[0144] First, the information conversion unit 202 derives the above-mentioned various input information based on the accumulated information I11 (S201). Next, the user attribute estimation unit 204 inputs the derived various input information into the estimation models MA to MH respectively, thereby obtaining the estimation result of the user attribute as output information (S202). Next, the information output unit 206 outputs the user attribute estimated by the user attribute estimation unit 204 to the terminal device 300.
[0145] <5. Function>
[0146] As a comparative example, it is possible to collect user attribute information by asking users to input information such as gender, age, family composition, or cleaning time on a questionnaire or the Internet. However, in these cases, inputting information may be troublesome and users may not input information, or the information input by users may not match the actual situation. In addition, it is difficult to confirm the detailed usage status of users in the form of questionnaires.
[0147] On the other hand, in the present embodiment, the information processing system 1 includes an information acquisition unit 201 that acquires data D transmitted from the electric vacuum cleaner 100, and a user attribute estimation unit 204 that estimates the attributes of the user U of the electric vacuum cleaner 100 based on input information obtained from the data D acquired by the information acquisition unit 201 using a model M developed by machine learning. According to such a structure, even without input from the user U, the attributes of the user U can be estimated based on the use result of the electric vacuum cleaner 100 by the user U. Thus, the burden of collecting attribute information of the user U can be reduced.
[0148] (Second Embodiment)
[0149] Next, the second embodiment will be described. The second embodiment is different from the first embodiment in that the electric vacuum cleaner 100A is a canister type. The configuration other than the following description is the same as that of the first embodiment.
[0150] Fig.12 : is a diagram showing the structure of the server 200 of the second embodiment. In the present embodiment, the electric vacuum cleaner 100A is a canister-type electric vacuum cleaner. In the electric vacuum cleaner 100A, as a detection unit included in the sensor group SU, there are a main body ground detection unit 71 and a cable lead-out amount detection unit 72. The main body ground detection unit 71 includes, for example, a switch that contacts the ground when the electric vacuum cleaner 100A is located on the ground, and detects whether the electric vacuum cleaner 100A is in contact with or floating relative to the ground. The cable lead-out amount detection unit 72 includes, for example, an encoder that detects the amount of rotation of a reel on which a cable is wound, and detects the length of the cable (the amount of cable lead-out) led out from the electric vacuum cleaner 100A. The detection results of the main body ground detection unit 71 and the cable lead-out amount detection unit 72 are sent to the server 200 as part of the data D.
[0151] In the present embodiment, for example, the information conversion unit 202 includes information related to the usage of the electric vacuum cleaner 100A as one of the input information for gender estimation. The "information related to the usage" is information indicating whether the vacuum cleaner body 10 of the electric vacuum cleaner 100A is cleaning in a state where the vacuum cleaner body 10 is touching the ground, or in a state where the vacuum cleaner body 10 of the electric vacuum cleaner 100A is lifted from the ground by the user U, etc. This information is derived, for example, based on the "operating state of the electric vacuum cleaner 100A" contained in the data D, the detection result of the main body contact detection unit 71 (or the detection result of the head acceleration sensor 63), and date and time information. Here, generally, when the user U is a male, the vacuum cleaner body 10 of the electric vacuum cleaner 100A is more likely to be cleaned in a state where the vacuum cleaner body 10 is lifted from the ground than when the user U is a female.
[0152] In the present embodiment, for example, the information conversion unit 202 includes information related to the movement of the vacuum cleaner body 10 of the electric vacuum cleaner 100A as one of the input information for estimating the housing type. The vacuum cleaner body 10 is a part having wheels and placed on the ground. The vacuum cleaner body 10 is an example of a "main body". "Information related to the movement of the vacuum cleaner body 10" is an average value or maximum value of the distance moved by the vacuum cleaner body 10 of the electric vacuum cleaner 100A during each cleaning, or a cumulative value of the distance moved by the vacuum cleaner body 10 of the electric vacuum cleaner 100A during the above-mentioned specified period (for example, 3 months). The above-mentioned average value, maximum value, or cumulative value is calculated, for example, based on the detection results of the main body acceleration sensor 45, the detection results of the main body gyro sensor 46, and date and time information contained in the data D. Here, usually, in most cases, the movement of the head 30 of the canister type electric vacuum cleaner 100A is roughly the same when the room is small and when it is large, but in most cases, the moving distance of the vacuum cleaner body 10 is shorter when the room is small, and the moving distance of the vacuum cleaner body 10 is longer when the room is large.
[0153] In the present embodiment, for example, the information conversion unit 202 includes information related to the amount of cable drawn out of the electric vacuum cleaner 100A as one of the input information for estimating the housing type. The "information related to the amount of cable drawn out" is the average value or maximum value of the amount of cable drawn out of the electric vacuum cleaner 100A in each cleaning, or the cumulative value of the amount of cable drawn out of the electric vacuum cleaner 100A in the above-mentioned prescribed period (for example, 3 months). The above-mentioned average value, maximum value or cumulative value is calculated, for example, based on the detection result of the cable draw-out amount detection unit 72 included in the data D and the date and time information. Here, usually, in most cases, the cable of the electric vacuum cleaner 100A is not drawn out much when the room is small, but is drawn out longer when the room is large.
[0154] In this embodiment, the user attribute estimation unit 204 estimates the gender of the user U based on the information about the usage of the electric vacuum cleaner 100A. In addition, the user attribute estimation unit 204 may estimate the age of the user U based on the information about the usage of the electric vacuum cleaner 100A. That is, compared with young people, the elderly tend to lift the vacuum cleaner body 10 of the electric vacuum cleaner 100A less often to clean.
[0155] In this embodiment, the user attribute estimation unit 204 estimates the housing type (e.g., the size of the room) of the user U based on the information related to the movement of the vacuum cleaner body 10 of the electric vacuum cleaner 100A and / or the information related to the amount of cable extension of the electric vacuum cleaner 100A. In addition, the user attribute estimation unit 204 may also estimate the attributes (number of people) of the cohabiting household based on the information related to the movement of the vacuum cleaner body 10 of the electric vacuum cleaner 100A and / or the information related to the amount of cable extension of the electric vacuum cleaner 100A.
[0156] According to such a configuration, similarly to the first embodiment, it is possible to reduce the burden of collecting the attribute information of the user U.
[0157] (Third Embodiment)
[0158] Next, the third embodiment will be described. The third embodiment is different from the first embodiment in that the electric vacuum cleaner 100B is a cleaning robot. The configuration other than that described below is the same as that of the first embodiment.
[0159] Fig.13 : is a diagram showing the structure of the server 200 of the third embodiment. In this embodiment, the electric vacuum cleaner 100B is a sweeping robot that automatically drives and collects garbage. In the electric vacuum cleaner 100B, as a detection unit included in the sensor group SU, there is a map generation unit 75, a fall prevention sensor 76 and an illumination sensor 65. The map generation unit 75 generates map information of the room based on the result of the automatic driving of the electric vacuum cleaner 100B. The fall prevention sensor 76 is a sensor for detecting whether there is no height difference (such as a step) in front of the direction of the electric vacuum cleaner 100B. The illumination sensor 65 is similar to the first embodiment, and detects when the illumination is below a specified amount, such as when the electric vacuum cleaner 100B enters under the bed. The map information generated by the map generation unit 75, the detection result of the fall prevention sensor 76, and the detection result of the illumination sensor 65 are sent to the server 200 as part of the data D.
[0160] In this embodiment, for example, information conversion unit 202 includes information on a map generated by map generation unit 75 of vacuum cleaner 100B as one of the input information for estimating a cohabiting household. "Information on a map" is information indicating the number of rooms in a house and the size of each room.
[0161] In this embodiment, for example, the information conversion unit 202 includes information related to the days and time periods when the electric vacuum cleaner 100B operates as one of the input information for estimating the employment mode. This information is derived, for example, based on the "operating state of the electric vacuum cleaner 100B" and date and time information contained in the data D. Here, usually, in most cases, the user U uses the electric vacuum cleaner 100B as a cleaning robot to clean while he is out. Therefore, the days and time periods when the electric vacuum cleaner 100B operates can be regarded as the time when the user U is not at home (time at work).
[0162] In this embodiment, for example, the information conversion unit 202 includes information related to the detection result of the fall prevention sensor 76 of the electric vacuum cleaner 100B as one of the input information for estimating the type of residence. The so-called "information related to the detection result of the fall prevention sensor 76" is information indicating whether there is a height difference that can be regarded as a step, etc. When there is a height difference that can be regarded as a step, the possibility that the residence is large increases.
[0163] In the present embodiment, for example, the information conversion unit 202 includes information related to the detection result of the illumination sensor 65 provided in the electric vacuum cleaner 100B as input information (hereinafter referred to as "input information for marriage estimation") input to the estimation model M in order to estimate the attribute related to the presence or absence of marriage (married / unmarried). The so-called "information related to the detection result of the illumination sensor 65" is information indicating the size of the area where the illumination is continuously below a specified amount, etc. The size of the area where the illumination is continuously below a specified amount can be regarded as the size of the bed placed in the residence of the user U. Therefore, if the size of the area where the illumination is continuously below a specified amount is the size of a single bed, the possibility of being unmarried increases. In addition, when there are more than two areas where the illumination is continuously below a specified amount, the possibility of being married (the possibility of having more than two beds) increases.
[0164] In this embodiment, the user attribute estimation unit 204 estimates the type of the cohabiting household (single / nuclear / extended family) based on information related to the map generated by the map generation unit 75. In addition, the user attribute estimation unit 204 estimates attributes related to the employment method based on information related to the day of the week and the time period when the electric vacuum cleaner 100B is operated.
[0165] In this embodiment, the user attribute estimation unit 204 estimates the attribute related to the housing type based on the information related to the detection result of the fall prevention sensor 76. In addition, the user attribute estimation unit 204 estimates the attribute related to the presence or absence of marriage based on the information related to the detection result of the illumination sensor 65.
[0166] According to such a configuration, similarly to the first embodiment, it is possible to reduce the burden of collecting the attribute information of the user U.
[0167] According to at least one embodiment described above, there is an acquisition unit for acquiring data sent from the electric vacuum cleaner, and an estimation unit for estimating the attributes of the user of the electric vacuum cleaner based on the data acquired by the acquisition unit using a model developed by machine learning, thereby reducing the burden of collecting the user's attribute information.
[0168] Several embodiments of the present invention have been described, but these embodiments are presented as examples and are not intended to limit the scope of the invention. These embodiments can be implemented in various other forms, and various omissions, substitutions, and changes can be made without departing from the scope of the invention. These embodiments and their variations are included in the scope and spirit of the invention, and similarly, are included in the invention described in the claims and their equivalents.
Claims
1. An information processing system, wherein: have: an acquisition unit that acquires data transmitted from the electric vacuum cleaner; and an estimating unit that estimates the attributes of the user of the electric vacuum cleaner based on the input information obtained from the data acquired by the acquiring unit, using a model that is learned to output an estimated result of the user's attributes when information related to the electric vacuum cleaner is input; The input information includes: information about the gripping force or gripping position of the handle of the electric vacuum cleaner detected by a sensor of the electric vacuum cleaner, information about the angle of the extension tube relative to the head of the electric vacuum cleaner, or information about the acceleration acting on the electric vacuum cleaner. The estimating unit estimates the number of people living in a cohabiting household as an attribute of the user.
2. The information processing system according to claim 1, wherein: The input information includes first information and second information, the first information being information related to the gripping force or gripping position of the handle of the electric vacuum cleaner detected by the sensor, information related to the angle of the extension tube relative to the head of the electric vacuum cleaner, or information related to the acceleration acting on the electric vacuum cleaner, and the second information being information obtained in association with the first information and related to the day of the week and time period when the electric vacuum cleaner is used, The estimating unit estimates the number of persons in the cohabiting household, and when estimating that there are two or more persons in the cohabiting household, estimates an attribute related to the employment method of each member of the two or more persons in the cohabiting household as an attribute of the user.
3. The information processing system according to claim 1, wherein: The input information further includes: information related to the operation time of the electric vacuum cleaner, information related to the time until a specified amount of garbage is accumulated inside the electric vacuum cleaner, information related to the time from the end of the operation of the electric vacuum cleaner to the restart of the operation, information related to the number of times the power of the electric vacuum cleaner is turned on / off, information related to the number of times the head of the electric vacuum cleaner is lifted, information related to the day of the week and time period when the electric vacuum cleaner is used, information related to the number of electric vacuum cleaners registered in the server corresponding to the identification information of the electric vacuum cleaner and the same user, or information related to a map generated by the electric vacuum cleaner as a cleaning robot, The estimating unit further estimates an attribute related to the cohabiting household as an attribute of the user.
4. The information processing system according to claim 3, wherein: The estimating unit estimates an attribute related to the cohabiting household based on information related to an operation time of the electric vacuum cleaner and information related to a time required for a predetermined amount of garbage to accumulate inside the electric vacuum cleaner.
5. The information processing system according to claim 1, wherein: The input information further includes information on the type of garbage sucked by the electric vacuum cleaner, the information being generated based on image data captured by a camera provided in the electric vacuum cleaner and indicating whether a specific type of garbage exists in an amount exceeding a specified amount. The estimating unit further estimates an attribute related to the presence or absence of children or the presence or absence of pets as the attribute of the user.
6. The information processing system according to claim 1, wherein: The input information includes: information about the angle of the extension tube relative to the head of the electric vacuum cleaner, information about the user's grip on the handle of the electric vacuum cleaner, information about the acceleration acting on the electric vacuum cleaner, information about the number of times the filter of the electric vacuum cleaner is clogged, information about the type of garbage sucked by the electric vacuum cleaner, or information about the usage method of the canister type electric vacuum cleaner, wherein the information about the type of garbage sucked by the electric vacuum cleaner is generated based on the shooting data of the camera of the electric vacuum cleaner, indicating whether a specific type of garbage exists in excess of a specified amount, The estimating unit further estimates an attribute related to gender as an attribute of the user.
7. The information processing system according to claim 1, wherein: The input information also includes information capable of identifying the model of the electric vacuum cleaner. The estimating unit further estimates an attribute related to age as an attribute of the user.
8. The information processing system according to claim 1, wherein: The input information further includes: information about the day of the week and time period when the electric vacuum cleaner is used, or information about the day of the week and time period when the electric vacuum cleaner as a sweeping robot works. The estimating unit further estimates an attribute related to employment style as an attribute of the user.
9. The information processing system according to claim 1, wherein: The input information further includes: information related to the operating time of the electric vacuum cleaner, information related to the number of times the wired electric vacuum cleaner is reconnected to the external power supply, information related to accessories installed and removed from the electric vacuum cleaner, information related to the lighting state of the light provided on the head of the electric vacuum cleaner, information related to the number of electric vacuum cleaners registered in the server corresponding to the identification information of the electric vacuum cleaner and the same user, information related to the movement of the main body of the canister type electric vacuum cleaner, information related to the lead-out amount of the cable of the canister type electric vacuum cleaner, information related to a map generated by the electric vacuum cleaner as a sweeping robot, or information related to the detection result of the fall prevention sensor of the electric vacuum cleaner as a sweeping robot, The estimating unit further estimates an attribute related to a residence as an attribute of the user.
10. The information processing system according to claim 1, wherein: The input information may further include: information related to a load state of a brush motor of the electric vacuum cleaner, information related to an acceleration acting on a head of the electric vacuum cleaner, or information related to a detection result of a garbage sensor provided in the electric vacuum cleaner. The estimating unit further estimates an attribute related to a floor type of a room as an attribute of the user.
11. The information processing system according to any one of claims 1 to 10, wherein: The input information may further include information related to the temperature detected in the electric vacuum cleaner or information related to the frequency of an external power source connected to the electric vacuum cleaner. The estimating unit further estimates an attribute related to a residential area as an attribute of the user.
