Equipment status prediction method and equipment status prediction system

Abstract

To provide a device state prediction method and a device state prediction system configured to predict an occurrence of a predetermined event and how a device is used at the occurrence of the event on the basis of past records of status of use of the device.SOLUTION: In a program for predicting a device state, a prediction model generation unit 203 generates operation data indicating operation state of a device, a first model which predicts a state of the device at a second time which is later than a first time on the basis of operation data obtained at the first time, and a second model which predicts how the device is used at the second time on the basis of the operation data obtained at the first time, to be stored in a storage unit 201. A prediction execution unit 204 applies the acquired operation data obtained at the first time to the first model to predict a state of the device at the second time, and applies the acquired operation data obtained at the first time to the second model to predict how the device is used at the second time, and outputs a result of predicting the state of the device at the second time and a result of predicting how the device is used at the second time.SELECTED DRAWING: Figure 2

Description

[Technical field]

[0001] The present invention relates to a technique for predicting the state of an appliance such as a refrigerator. [Background technology]

[0002] One example of a technique for predicting when an equipment will fail is described in Japanese Patent Laid-Open No. 10-267509 (Patent Document 1). Patent Document 1 states that "an operating state management device that manages equipment based on data related to the operating state of the equipment includes a database construction means that constructs a database related to past operating states by classifying and storing information related to the operating state of the equipment according to the operating conditions of the equipment under which the information was obtained, and a failure time prediction means that predicts when the equipment will fail based on information related to the current operating state of the equipment and past information in the database under the same operating conditions." [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Publication No. 10-267509 Summary of the Invention [Problem to be solved by the invention]

[0004] Generally, when a problem occurs such as an appliance not performing as expected, the cause is not necessarily a malfunction of the appliance, but may be due to the way the appliance is used by the user. Taking a refrigerator as an example of an appliance, for example, a problem may occur in which the temperature inside the refrigerator does not drop sufficiently in summer, but generally, this is often due to the user opening and closing the door too many times and for how long the door is open. Therefore, by predicting in advance the appliances that are used in a way that is likely to cause such problems in the future, it is believed that the occurrence of problems can be avoided and the number of times service personnel are called out to deal with the problems can be reduced.

[0005] Patent Document 1 describes a technology for predicting when a device such as a low-temperature showcase, refrigerator, or freezer will fail based on information about the past operating state of the device itself. However, it does not disclose a technology for predicting the occurrence of a specific event (such as a temperature rise in summer) in the device itself and how the device will be used when the event will occur, based on the past usage conditions of many devices, not limited to the device itself. [Means for solving the problem]

[0006] In order to achieve at least one of the above-mentioned objects, the present invention provides an appliance state prediction method executed by an appliance state prediction system having a processor and a storage device, the storage device holding operation data indicating an operating status of an appliance, a first model for predicting a state of the appliance at a second time period after a first time period based on the operation data at a first time period, and a second model for predicting a usage manner of the appliance by a user at the second time period based on the operation data at the first time period, the appliance state prediction method including a first step in which the processor predicts a state of the appliance at the second time period by applying the acquired operation data at the first time period to the first model, a second step in which the processor predicts a usage manner of the appliance by the user at the second time period by applying the acquired operation data at the first time period to the second model, and a third step in which the processor outputs a prediction result of the state of the appliance at the second time period and a prediction result of the usage manner of the appliance by the user at the second time period. the appliance is a refrigerator having at least one of a refrigerator compartment and a freezer compartment, the second period is a summer period, the operation data includes data on climate, data on how the appliance is used by a user, and data on a control state of the appliance, the first model is a model that predicts a rise in temperature inside the appliance in the second period based on at least one of the data on climate, data on how the appliance is used by a user, and data on a control state of the appliance in the first period, and the second model is a model that predicts a rise in temperature inside the appliance in the second period based on at least the data on how the appliance is used by the user in the first period. In the first step, the processor predicts a rise in temperature inside the device in the second future time period by applying at least one of the acquired data on the climate in the first time period, the data on the usage of the device by a user, and the data on a control state of the device to the first model, and if it is predicted that the temperature inside the device in the second future time period will rise, in the second step, the processor predicts the usage of the device by the user in the second future time period by applying the acquired data on the usage of the device by the user in the first time period to the second model. It is characterized by: Effect of the Invention

[0007] According to one aspect of the present invention, it is possible to predict the occurrence of a predetermined event, such as a temperature rise in summer, in an appliance based on the past usage records of a large number of appliances, and to predict how the appliance will be used when the event occurs. Problems, configurations, and effects other than those described above will become apparent from the description of the following embodiments. [Brief description of the drawings]

[0008] [Figure 1] 1 is a block diagram showing an example of a configuration of an equipment state prediction system according to an embodiment of the present invention. [Diagram 2] FIG. 2 is an explanatory diagram illustrating an example of a software configuration of a prediction program stored in a data collection and analysis server according to an embodiment of the present invention. [Diagram 3] FIG. 2 is an explanatory diagram showing an example of the concept of a prediction model generated by a data collection and analysis server according to an embodiment of the present invention. [Figure 4] FIG. 2 is an explanatory diagram showing an example of a cause of a temperature rise in a refrigerator in an embodiment of the present invention. [Diagram 5] FIG. 2 is an explanatory diagram showing an example of driving data held by a data collection and analysis server according to the embodiment of the present invention. [Figure 6] FIG. 10 is an explanatory diagram showing an example of a model of the relationship between operation data for the same period and an internal temperature rise in an embodiment of the present invention. [Figure 7] FIG. 2 is an explanatory diagram showing an example of a model of the relationship between operation data at different times and an internal temperature rise in an embodiment of the present invention. [Figure 8] FIG. 11 is an explanatory diagram showing an example of processing in which the data collection and analysis server according to the embodiment of the present invention generates a prediction model for predicting a temperature rise in summer from data observed in spring. [Figure 9] FIG. 11 is an explanatory diagram showing an example of a process in which the data collection and analysis server according to the embodiment of the present invention generates a model showing the relationship between usage in spring and usage in summer. [Figure 10] FIG. 11 is an explanatory diagram showing an example of a process in which the data collection and analysis server according to the embodiment of the present invention generates a model showing the relationship between the weather and usage and various sensor data. [Figure 11] FIG. 11 is an explanatory diagram showing an example of a process in which the data collection and analysis server according to the embodiment of the present invention predicts a temperature rise in summer. [Figure 12] FIG. 2 is an explanatory diagram showing a first example of utilization of prediction results by the data collection and analysis server according to the embodiment of the present invention. [Figure 13]FIG. 13 is an explanatory diagram showing an example of a screen displayed in a first example of utilization of prediction results by the data collection and analysis server according to the embodiment of the present invention. [Figure 14] FIG. 11 is an explanatory diagram showing a second example of utilization of prediction results by the data collection and analysis server according to the embodiment of the present invention. [Figure 15] FIG. 13 is an explanatory diagram showing an example of a screen displayed in a second example of utilization of prediction results by the data collection and analysis server according to the embodiment of the present invention. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0009] Hereinafter, an embodiment of the present invention will be described with reference to the drawings.

[0010] FIG. 1 is a block diagram showing an example of the configuration of an equipment state prediction system according to an embodiment of the present invention.

[0011] The appliance state prediction system 100 of this embodiment includes a data collection and analysis server 110 connected to a network 150 , and a plurality of refrigerators 120 that communicate with the data collection and analysis server 110 via the network 150 .

[0012] The data collection and analysis server 110 is a computer having a processor 112, a memory 114, a network interface (I / F) 113, and an information storage unit 111, all of which are connected to each other.

[0013] The processor 112 executes programs stored in the memory 114 .

[0014] The memory 114 is, for example, a semiconductor memory, and stores a program executed by the processor 112, data referenced by the processor 112, data output from the processor 112, etc. The memory 114 of this embodiment stores at least a prediction program 115, which is a program executed by the processor 112. The processing realized by the processor 112 executing the prediction program 115 will be described later. The processing based on the prediction program 115 described below is executed by the processor 112.

[0015] The network I / F 113 is connected to a network 150 and communicates with the refrigerator 120 and the like.

[0016] The information storage unit 111 may be a storage device such as a hard disk drive (HDD) or a solid state drive (SSD). For example, operation data acquired from each refrigerator 120 is stored in the information storage unit 111. In addition, the information storage unit 111 may store a prediction model generated by learning the operation data based on the prediction program 115, and a result predicted using the prediction model.

[0017] Each refrigerator 120 is an appliance whose state is predicted by the appliance state prediction system 100 of this embodiment, and is used to store food and the like by keeping the interior of the refrigerator at a low temperature. A typical configuration of one of the plurality of refrigerators 120 will be described below. The configurations of the other refrigerators 120 may be similar to that shown in Fig. 1, so illustrations and descriptions thereof will be omitted.

[0018] The refrigerator 120 includes a processor 124, a memory 127, a network I / F 121, a sensor group 126, an external I / F 125, a control device 123, and a controlled device 122, which are connected to each other.

[0019] The processor 124 executes programs stored in the memory 127 .

[0020] The memory 127 is, for example, a semiconductor memory, and stores the programs executed by the processor 124, data referenced by the processor 124, etc. The memory 127 in this embodiment stores a control program 128, which is a program executed by the processor 124, and temporary operating data storage 129 generated during processing by the processor 124.

[0021] The sensor group 126 includes one or more sensors. The sensor group 126 may include, for example, a temperature sensor that measures the temperature inside the refrigerator 120, and a sensor that detects the opening and closing of the door of the refrigerator 120. If the refrigerator 120 has a plurality of compartments, such as a so-called freezer and a refrigerator, the sensor group 126 may include a sensor that measures the temperature of each compartment and a sensor that detects the opening and closing of the door of each compartment. The sensor group 126 may also include a sensor that measures the room temperature outside the refrigerator.

[0022] Processor 124 obtains data indicating the operating status of refrigerator 120 measured by sensor group 126 as operation data in accordance with control program 128, and stores the data in operation data temporary storage 129 of memory 127. Then, processor 124 transmits the contents of operation data temporary storage 129 to data collection and analysis server 110 at an appropriate timing (for example, periodically). The transmitted data is stored in information storage unit 111 of data collection and analysis server 110.

[0023] Furthermore, when the processor 124 receives a control instruction from the data collection and analysis server 110, the processor 124 executes processing according to the control instruction according to the control program 128, and outputs a control signal for controlling the control device 123 if necessary. The control device 123 controls the controlled device 122 according to the control signal. The controlled device 122 is, for example, an operation panel of the refrigerator 120 and a compressor for cooling the refrigerator 120. The control device 123 may refer to the temperature inside the refrigerator obtained from the sensor group 126 and control the compressor so that the temperature inside the refrigerator approaches a set temperature, or may control the compressor so that a defrosting operation is performed at a predetermined timing. Furthermore, if the controlled device 122 includes a door-open alarm (buzzer, etc.), the control device 123 may control the alarm to output an alarm according to the door-open time obtained from the sensor group 126.

[0024] Network I / F 121 is connected to network 150 and executes communication with data collection and analysis server 110 and the like via network 150. Specifically, for example, transmission of operation data from refrigerator 120 and reception of control instructions from data collection and analysis server 110 and the like are realized by such communication.

[0025] The external I / F 125 is connected to devices external to the refrigerator 120. For example, the external I / F 125 may be connected to a sensor external to the refrigerator 120, such as an outside air temperature sensor (not shown), and data measured by the sensor may be stored in the memory 127 as part of the operation data. Note that such external devices are not essential, and therefore the external I / F 125 may not be used.

[0026] The appliance state prediction system 100 of this embodiment may further include a terminal device used by a user of the refrigerator 120 and a terminal device used by a person in charge of supporting the user of the refrigerator 120. In Fig. 1, a smartphone 140 is illustrated as an example of a terminal device used by a user, and an information reference / control terminal 130 is illustrated as an example of a terminal device used by a support person.

[0027] The smartphone 140 can communicate with the data collection and analysis server 110 and the like via the network 150. For example, a user can transmit a control instruction to the data collection and analysis server 110 and the like via the smartphone 140. Furthermore, the smartphone 140 can display a status (for example, information on the status of the refrigerator 120 predicted by the data collection and analysis server 110) to the user according to information on status display received from the data collection and analysis server 110 and the like.

[0028] The information reference / control terminal 130 can communicate with the data collection / analysis server 110, the refrigerator 120, the smartphone 140, etc. via the network 150. For example, a support staff member can refer to information such as the prediction results from the data collection / analysis server 110 via the information reference / control terminal 130, and can send control instructions to the refrigerator 120 as necessary, and can send warnings to the refrigerator 120 or the smartphone 140.

[0029] The data collection and analysis server 110 may acquire weather-related information from an external organization 160 via the network 150. The external organization 160 may be, for example, a company or a government agency that provides weather-related information. The acquired weather-related information may be the temperature, humidity, and weather (e.g., the presence or absence of sunshine, the presence or absence of rain, etc.) of each region, or may include statistical information such as the minimum temperature, maximum temperature, average temperature, hours of sunshine, and amount of precipitation over a certain period of time.

[0030] FIG. 2 is an explanatory diagram showing an example of the software configuration of the prediction program 115 held by the data collection and analysis server 110 according to the embodiment of the present invention.

[0031] The prediction program 115 of this embodiment includes a storage unit 201 , a feature amount acquisition unit 202 , a prediction model generation unit 203 , and a prediction execution unit 204 .

[0032] The memory unit 201 refers to the operation data of the refrigerator 120 stored in the information storage unit 111. The operation data includes past data 211 and latest data 212. The past data 211 includes past operation data and information indicating whether the temperature inside the refrigerator 120 has increased as a result of the past operation. The latest data 212 includes the operation data, but does not include information indicating whether the temperature inside the refrigerator 120 has increased as a result of the operation because the information has not yet been obtained.

[0033] As preprocessing 213, feature acquisition unit 202 extracts feature values ​​of past data 211 and latest data 212. Prediction model generation unit 203 executes model learning 214 using the feature values ​​of past data 211 as explanatory variables and whether or not the corresponding temperature inside refrigerator 120 has increased as a target variable, and generates prediction model 215 that predicts whether or not the temperature has increased from the feature values ​​of past operation data.

[0034] The prediction execution unit 204 executes prediction 216 by inputting the feature amount of the latest data 212 into a prediction model 215 , and obtains a prediction result 217 .

[0035] FIG. 3 is an explanatory diagram showing an example of the concept of the prediction model 215 generated by the data collection and analysis server 110 according to the embodiment of the present invention.

[0036] The subject of prediction in this embodiment is a temperature rise inside the refrigerator 120 in summer. In summer, the load increases due to an increase in the frequency of use of the refrigerator 120 and a rise in the temperature of the installation environment, and this causes events such as a rise in temperature inside the refrigerator (i.e., the inside of the refrigerator does not cool to a desired temperature), and there is a tendency for repair requests for the refrigerator 120 to increase. In particular, problems related to the freezing function, such as the contents of the freezer compartment of the refrigerator 120 melting, occur, and inquiries to the support department increase. For this reason, there is a need to predict the temperature rise inside the refrigerator 120 in summer at an earlier point in time (for example, in spring).

[0037] The data collection and analysis server 110 of this embodiment models the relationship between the operating data from the spring of the previous year or any year before and the temperature rise in summer using machine learning or the like, and uses the model to predict the temperature rise in that year (i.e., the upcoming summer) from the operating data from that spring as of the spring of the current year.

[0038] In the example of Fig. 3, a temperature rise prediction model 215 is generated by modeling the relationship between spring operation data and summer temperature rise based on the operation data in the spring and the inside temperature in the summer of two refrigerators 120, individual 1 and individual 2, purchased before the spring of the previous year. Then, the temperature rise prediction model 215 is applied to the operation data in the spring of this year of three refrigerators 120 including individual 2 and individual 3 and individual 4, purchased before the spring of this year. As a result, a temperature rise in this summer is predicted as of the spring of this year.

[0039] For the sake of explanation, four refrigerators 120 are shown in FIG. 3, but in practice modeling and prediction are performed for a larger number of refrigerators 120.

[0040] FIG. 4 is an explanatory diagram showing an example of factors that cause a temperature rise inside the refrigerator 120 in the embodiment of the present invention.

[0041] Risk factors for a rise in temperature inside refrigerator 120 include factors mainly caused by the equipment (i.e., refrigerator 120 itself), factors caused by the environment, and factors caused by usage. Here, among risk factors, factors mainly caused by the equipment include, for example, a failure risk specific to refrigerator 120 (for example, caused by the design of refrigerator 120 or the quality of parts used, etc.) and aging of refrigerator 120. Factors caused by the environment include, for example, the installation environment (for example, the distance between the refrigerator and surrounding objects such as walls), the room temperature of the room in which refrigerator 120 is placed, and the outside air temperature. Also, factors caused by usage include, for example, the number of times the door is opened, the time for which it is opened, and the amount of items packed inside.

[0042] The manner in which the magnitude of the risk of temperature rise changes over time (e.g., whether there is any change, whether there is any change related to the season, etc.) may differ depending on the factor. For example, among the risks caused by equipment, the risk of equipment-specific failure is thought to basically not change. On the other hand, the risk caused by aging deterioration is thought to gradually increase over time, regardless of the season.

[0043] Among the risks arising from the environment, the installation environment, such as the distance between refrigerator 120 and surrounding objects, basically does not change unless the arrangement of refrigerator 120 itself and its surrounding objects is changed. On the other hand, the room temperature and the outside air temperature change with the seasons. Also, the risks arising from usage are considered to include risks arising from basic usage that does not change with the seasons, and risks arising from usage that changes with the seasons.

[0044] An actual temperature rise is believed to occur when the risk of a temperature rise, which is the combination of the risks of each of the above factors, exceeds a certain limit. As mentioned above, there are seasonal risk factors that make the risk higher in summer than in spring, and for the same individual, there is the risk factor of deterioration due to aging that increases over time, so the risk of a temperature rise in spring does not exceed the limit and an actual temperature rise does not occur, but the risk of a temperature rise in the following summer may exceed the limit and an actual temperature rise may occur.

[0045] FIG. 5 is an explanatory diagram showing an example of driving data held by the data collection and analysis server 110 according to the embodiment of the present invention.

[0046] Operation data 500 shown in FIG. 5 includes device ID 501, time 502, freezer compartment temperature 503, refrigerator compartment temperature 504, vegetable compartment temperature 505, indoor temperature 506, indoor humidity 507, freezer compartment door opening time 508, number of times the freezer compartment door is opened 509, refrigerator compartment left door opening time 510, number of times the left refrigerator door is opened 511, refrigerator compartment right door opening time 512, number of times the right refrigerator door is opened 513, refrigerator compartment temperature adjustment 514, freezer compartment temperature adjustment 515, rapid ice making mode 516, power saving mode 517, motor rotation speed 518, compressor temperature 519, and defrosting operation 520.

[0047] Device ID 501 is identification information of each refrigerator 120. Time 502 indicates the time when data was acquired by sensor group 126. Freezer compartment temperature 503, refrigerator compartment temperature 504, and vegetable compartment temperature 505 indicate temperatures measured by a sensor in sensor group 126 that measures the temperature of the freezer compartment, a sensor that measures the temperature of the refrigerator compartment, and a sensor that measures the temperature of the vegetable compartment, respectively, when refrigerator 120 has a freezer compartment, a refrigerator compartment, and a vegetable compartment as compartments.

[0048] Indoor temperature 506 and indoor humidity 507 indicate the temperature and humidity of the room in which refrigerator 120 is installed, measured by a temperature sensor and a humidity sensor, respectively, of sensor group 126 that are installed outside refrigerator 120 .

[0049] Freezer door open time 508 and freezer door open count 509 are the door open time and the open count, respectively, measured by a sensor of sensor group 126 that detects opening and closing of the freezer door.

[0050] Refrigerator left door open time 510 and refrigerator left door open count 511 are the open time and the open count of the left door, respectively, measured by a sensor that detects the opening and closing of the left door of the refrigerator, among sensor group 126, when the refrigerator has a so-called double door. Refrigerator right door open time 512 and refrigerator right door open count 513 are the open time and the open count of the right door, respectively, measured by a sensor that detects the opening and closing of the right door of the refrigerator, among sensor group 126.

[0051] Refrigerator temperature adjustment 514 and freezer temperature adjustment 515 indicate the temperatures of the refrigerator and freezer, respectively, adjusted by the user. The values ​​of refrigerator temperature adjustment 514 and freezer temperature adjustment 515 may be specific temperatures, or may be set values ​​of the cooling strength of the refrigerator and freezer. An example of the latter is shown in FIG. 5.

[0052] Quick ice-making mode 516 indicates the mode set by the user for the ice-making function of refrigerator 120. In the example of Fig. 5, quick ice-making mode 516 indicates whether or not the quick ice-making mode (i.e., a mode that increases the cooling strength of the ice-making function to shorten the time required for making ice) is set.

[0053] The power saving mode 517 indicates a power consumption mode set by a user for the refrigerator 120. In the example of Fig. 5, the power saving mode 517 indicates whether or not a power saving mode (i.e., a mode for suppressing power consumption) is set.

[0054] Motor rotation speed 518 indicates the rotation speed of the motor that drives the compressor for compressing the refrigerant. Compressor temperature 519 indicates the temperature of the compressor for compressing the refrigerant. These are measured by a sensor of sensor group 126 that detects the rotation of the motor and a temperature sensor attached to the compressor.

[0055] The defrosting operation 520 indicates whether or not the refrigerator is performing a defrosting operation.

[0056] The items of sensor data stored as driving data shown in FIG. 5 are just an example, and actual driving data may not include at least any of the above items, or may further include items other than those above.

[0057] In the example of Fig. 5, column 521 stores operation data measured by sensor group 126 of refrigerator 120 identified by device ID "00001" at 12:35 on January 1, 2020. In this example, the freezer temperature at that time is -18.5°C, the refrigerator temperature is 3.3°C, the vegetable temperature is 3.5°C, the room temperature is 18.5°C, the room humidity is 55%, the freezer door opening time and opening count are 15 seconds and 2 times, respectively, the left refrigerator door opening time and opening count are 0 seconds and 0 times, respectively, the right refrigerator door opening time and opening count are 3 seconds and 1 time, respectively, the refrigerator temperature control is medium, the freezer temperature control is strong, the quick ice making mode is set, the power saving mode is not set, the motor speed is 3000 rpm, the compressor temperature is 50.1°C, and the defrosting operation is stopped.

[0058] In this example, the operation data is recorded every minute, so the door open time and the number of times the door was opened are the total open time and the number of times the door was opened for one minute, for example, from 12:34 to 12:35. The temperature, humidity, motor rotation speed, etc. may be measured values ​​at one point in time during that one minute, or may be the average value of values ​​measured multiple times during that one minute.

[0059] Similarly, column 522 stores operation data measured by sensor group 126 of refrigerator 120 identified by device ID "00001" at 12:36 on January 1, 2020. Columns 523 and 524 store operation data measured by sensor group 126 of refrigerator 120 identified by device ID "00002" at 12:35 and 12:36 on January 1, 2020, respectively. Although omitted in FIG. 5, operation data at other times is also stored for each refrigerator 120, and operation data at each time for other refrigerators 120 is also stored. In this manner, operation data at each time for each refrigerator 120 is stored in information storage unit 111.

[0060] FIG. 6 is an explanatory diagram showing an example of a model of the relationship between operation data for the same period and an internal temperature rise in an embodiment of the present invention.

[0061] The operation data in this embodiment can be classified into data related to climate, data related to how the user uses the refrigerator 120, and various sensor data related to the control of the refrigerator 120. In the example of Fig. 5, indoor temperature 506 and indoor humidity 507 are classified as data related to climate. When information related to climate is acquired from external agency 160, the information also corresponds to data related to climate. In principle, the user cannot control the data related to climate. However, there are cases where the user can control indoor temperature 506 and indoor humidity 507.

[0062] 5, data from freezer door open time 508 to power saving mode 517 corresponds to data on how the user uses refrigerator 120 (hereinafter, also simply referred to as data on usage). The usage can be controlled by the user.

[0063] 5, data from freezer temperature 503 to vegetable compartment temperature 505 and from motor rotation speed 518 to defrosting operation 520 correspond to various sensor data related to the control of refrigerator 120 (hereinafter also simply referred to as various sensor data). These are data indicating the status of control of each part in refrigerator 120 that refrigerator 120 performs according to control program 128 in response to the weather and how the user uses the refrigerator 120 (for example, control amount, control setting value, and status of refrigerator 120 resulting from the control, etc.). For this reason, the user cannot control the various sensor data, at least directly.

[0064] Next, the relationships between the classified data will be explained. Below, weather-related data will be classified as a, usage-related data as b, and various sensor data as c. If the objective variable corresponding to the temperature rise inside the storage unit during a certain period is X, the above data from a to c for the same period will be used as explanatory variables, and a model of the following formula (1) can be generated by performing machine learning or the like.

[0065] X = f1(a,b,c) (1)

[0066] On the other hand, data b relating to a user's usage during a certain period is expressed by the model shown in the following equation (2), which takes into account the individual differences in usage of each user, family composition, etc., as well as seasonal fluctuations, using weather data a for the same period as input.

[0067] b = f2(a) (2)

[0068] Furthermore, various sensor data c observed as a result of control of the refrigerator 120 during a certain period is expressed by a model of the following equation (3) based on the internal factors and environmental factors of each refrigerator 120, using as input data a related to the climate and data b related to usage during the same period.

[0069] c = f3(a,b) (3)

[0070] FIG. 7 is an explanatory diagram showing an example of a model of the relationship between operation data at different times and an internal temperature rise in the embodiment of the present invention.

[0071] Here, spring and the following summer will be given as an example of different periods, but the present invention can also be applied to other periods that follow each other, regardless of which specific season each period corresponds to and how far apart they are.

[0072] In the following explanation, the temperature rise in summer will be denoted as X, the data related to the summer climate as a, the data related to how you use it in summer as b, and the various sensor data in summer as c. Meanwhile, the temperature rise in the previous spring will be denoted as X', the data related to the spring climate as a', the data related to how you use it in spring as b', and the various sensor data in spring as c'.

[0073] A model for data from a different time period can be generated by replacing the explanatory variables of the model for data from the same time period shown in FIG. 6 with explanatory variables from a different time period.

[0074] As a model for obtaining the rise X in the temperature inside the storage unit in summer, a model of the following formula (4) can be generated, in which data a', b', and c' observed in the previous spring are used as explanatory variables. Although there is not necessarily a causal relationship between data a', b', and c' observed in the spring and the rise X in the temperature inside the storage unit in summer, the prediction model generation unit 203 can generate a model showing the relationship by learning data observed in the past.

[0075] X = F1(a', b', c') (4)

[0076] The above formula (4) is an example of a model that predicts the occurrence of a temperature rise in a later period based on driving data in a certain period, and the model generated in this embodiment is not limited to this. For example, in the above example, a temperature rise in summer is predicted based on driving data in spring, but a model that predicts a temperature rise in summer based on driving data in a period before spring may be generated. This is similar to other prediction models described later. In addition, in the example of formula (4) above, the explanatory variables include all of the data a' related to the spring climate, the data b' related to how to use the spring, and the various sensor data c' in spring, but a prediction model based on explanatory variables that do not include some of these may be generated.

[0077] The model of formula (4) above can predict a temperature rise in summer from data observed in spring, but even if this model predicts that the temperature will rise in summer, it is not possible to know how users are expected to use the equipment at that time. If it were possible to know in advance how users are expected to use the equipment in summer, it might be possible to suppress the temperature rise by using the equipment in a way that differs from the expected way when summer actually arrives. Therefore, the prediction model generation unit 203 also generates a model shown in the following formula (5) that shows the relationship between the usage b' in spring and the usage b in summer.

[0078] b = F2(a,b') (5)

[0079] As a result, for refrigerator 120 predicted to experience a temperature rise in summer using the model of formula (4) above, prediction execution unit 204 predicts how the refrigerator 120 will be used in summer from how it will be used in spring using model (5) above. At this time, as data a relating to summer climate to be input to the model, for example, a summer forecast value obtained from external organization 160 may be used. Then, in order to reduce the risk of temperature rise, measures can be taken, such as alerting the user to avoid using the refrigerator in the predicted summer.

[0080] In addition, by informing the support staff that a temperature rise of the refrigerator 120 is predicted, the support staff can provide prompt support by utilizing the prediction result. In addition, the manufacturer of the refrigerator 120 can use the result in product development so that appropriate control corresponding to the predicted usage is performed.

[0081] The above formula (5) is an example of a model that predicts how to use a device in a later period based on how the device is used in a certain period, and the model generated in this embodiment is not limited to this. For example, in the example of formula (5), how the device will be used in summer b is predicted based on data b' related to how the device will be used in spring and data a related to the climate in summer, but a model that predicts how the device will be used in summer based only on data b' related to how the device will be used in spring may be generated. Alternatively, a prediction model may be generated based on explanatory variables that include at least one of data a' related to the climate in spring and various sensor data c' in spring in addition to data b' related to how the device will be used in spring.

[0082] Furthermore, the prediction model generation unit 203 generates a model shown in the following formula (6) which indicates the relationship between the usage behavior of the user under certain weather conditions and various sensor information. This may be the same as the model shown in formula (3) described with reference to FIG. 5.

[0083] c = F3(a,b) (6)

[0084] For example, in the case of a refrigerator 120 in which a temperature rise actually occurred in summer, if various sensor data c obtained by applying data a related to the climate at that time and data b related to how the user uses the refrigerator to the model of the above formula (6) is significantly different from various sensor data actually obtained from the refrigerator 120, and the result of comparing the two satisfies a predetermined condition, it is suspected that the temperature rise is not due to how the user uses the refrigerator 120 but due to a malfunction or deterioration over time of the refrigerator 120. For this reason, the data collection and analysis server 110 can take measures such as providing support promptly by outputting information related to the malfunction of the refrigerator 120 to the user of the refrigerator 120 or a support staff member of the refrigerator 120.

[0085] FIG. 8 is an explanatory diagram showing an example of a process in which the data collection and analysis server 110 according to the embodiment of the present invention generates a prediction model for predicting a temperature rise in summer from data observed in spring.

[0086] The process of generating a prediction model (i.e., the model of the above formula (4)) for predicting a temperature rise in summer from data observed in spring is executed, for example, periodically. When the process starts (step 801), the prediction model generation unit 203 extracts data for a predetermined period from the past data 211 stored in the storage unit 201 (step 802). For example, the prediction model generation unit 203 may extract operation data of each refrigerator 120 for a predetermined period in spring of a certain year and operation data for a predetermined period in summer of the same year.

[0087] Next, the prediction model generating unit 203 creates a response variable and an explanatory variable for each individual refrigerator 120 as preprocessing for the extracted operation data (step 803). Specifically, the prediction model generating unit 203 executes the following steps 804 to 807. First, the prediction model generating unit 203 refers to the internal temperature in summer contained in the operation data of each individual refrigerator, and judges whether an increase in the internal temperature in summer has occurred (step 804). For example, the total time and the duration during which the temperature of the freezer, ice maker, refrigerator, etc. of each individual refrigerator exceeds a predetermined threshold may be tallied, and it may be judged that an internal temperature increase has occurred if these meet a predetermined condition. A flag is attached to the individual refrigerator for which it is judged that a temperature increase has occurred. This judgment result corresponds to the response variable of the prediction model to be generated.

[0088] Next, the prediction model generation unit 203 collects operation data for the spring period, specifically data related to the spring climate, data related to usage in the spring, and various sensor data for the spring (step 805). Next, the prediction model generation unit 203 extracts feature amounts from the collected operation data for the spring (step 806). These feature amounts correspond to explanatory variables of the prediction model to be generated. Next, the prediction model generation unit 203 generates a pair of a learning objective variable and an explanatory variable for each individual refrigerator 120 (step 807).

[0089] Next, the prediction model generation unit 203 performs machine learning using the generated pairs of objective variables and explanatory variables (step 808), and updates the prediction model that has been held up to that point to a prediction model newly obtained by machine learning (step 809). This completes the process of generating a prediction model that predicts a temperature rise in summer from data observed in spring (step 810).

[0090] In addition, the method of extracting the feature amount of the driving data and the method of machine learning in the above processing are not limited, and any method including a known method can be used. The same applies to the processing in Figs. 9 and 10 described later.

[0091] FIG. 9 is an explanatory diagram showing an example of a process in which the data collection and analysis server 110 according to the embodiment of the present invention generates a model showing the relationship between usage in spring and usage in summer.

[0092] The process of generating a model showing the relationship between usage in spring and usage in summer (i.e., the model of the above formula (5)) is executed, for example, periodically. When the process starts (step 901), the prediction model generation unit 203 extracts operation data for a predetermined period from the past data 211 stored in the storage unit 201 (step 902). For example, the prediction model generation unit 203 may extract operation data for a predetermined period in spring of a certain year and operation data for a predetermined period in summer of the same year for each refrigerator 120.

[0093] Next, the prediction model generation unit 203 creates a response variable and an explanatory variable for each individual refrigerator 120 as preprocessing for the extracted operation data (step 903). Specifically, the prediction model generation unit 203 executes the following steps 904 to 907. First, the prediction model generation unit 203 aggregates data on usage during the summer period included in the operation data of each individual refrigerator (step 904). The aggregated data on usage corresponds to the response variable of the model to be generated.

[0094] Next, the prediction model generation unit 203 aggregates data on usage in the spring and data on the climate in the summer (step 905). Next, the prediction model generation unit 203 extracts feature quantities from the aggregated data on usage in the spring and data on the climate in the summer (step 906). These feature quantities correspond to explanatory variables of the model to be generated. Next, the prediction model generation unit 203 generates a pair of a learning objective variable and an explanatory variable for each individual refrigerator 120 (step 907).

[0095] Next, the prediction model generation unit 203 performs machine learning using the generated pairs of objective variables and explanatory variables (step 908), and updates the model that has been held up until that point to a model newly obtained by machine learning (step 909). This completes the process of generating a model showing the relationship between usage in spring and usage in summer (step 910).

[0096] FIG. 10 is an explanatory diagram showing an example of a process in which the data collection and analysis server 110 according to the embodiment of the present invention generates a model showing the relationship between the weather and usage, and various sensor data.

[0097] The process of generating a model (i.e., the model of the above formula (6)) indicating the relationship between the climate and usage and various sensor data is executed, for example, periodically. When the process is started (step 1001), the prediction model generation unit 203 extracts operation data for a predetermined period from the past data 211 stored in the storage unit 201 (step 1002). For example, the prediction model generation unit 203 may extract operation data for each refrigerator 120 for a predetermined period in the summer of a certain year.

[0098] Next, the prediction model generation unit 203 creates a response variable and an explanatory variable for each individual refrigerator 120 as preprocessing for the extracted operation data (step 1003). Specifically, the prediction model generation unit 203 executes the following steps 1004 to 1007. First, the prediction model generation unit 203 aggregates various sensor data for the summer period included in the operation data of each individual refrigerator (step 1004). The aggregated various sensor data correspond to the response variable of the model to be generated.

[0099] Next, the prediction model generation unit 203 collects data on the climate during the summer and data on usage (step 1005). Next, the prediction model generation unit 203 extracts feature quantities from the collected data on the climate during the summer and data on usage (step 1006). These feature quantities correspond to explanatory variables of the model to be generated. Next, the prediction model generation unit 203 generates a pair of a learning objective variable and an explanatory variable for each individual refrigerator 120 (step 1007).

[0100] Next, the prediction model generation unit 203 performs machine learning using the generated pairs of objective variables and explanatory variables (step 1008), and updates the model held up until that point to a model newly obtained by machine learning (step 1009). This completes the process of generating a model showing the relationship between the climate and usage and various sensor data (step 1010).

[0101] FIG. 11 is an explanatory diagram showing an example of a process in which the data collection and analysis server 110 according to the embodiment of the present invention predicts a temperature rise in summer.

[0102] When the prediction process starts (step 1101), the prediction execution unit 204 extracts data for a predetermined period from the operating data stored in the storage unit 201. For example, when predicting a temperature rise in the summer of the year that has not yet arrived in the spring, the prediction execution unit 204 extracts the operating data for that spring included in the latest data 212.

[0103] Next, the prediction execution unit 204 performs pre-processing on the extracted operation data for each individual refrigerator 120 (step 1103). Specifically, the prediction execution unit 204 executes the following steps 1104 to 1106. First, the prediction execution unit 204 collects the acquired spring operation data, specifically, data on the spring climate, data on how to use the refrigerator in the spring, and various sensor data in the spring (step 1104). Next, the prediction execution unit 204 extracts features of the collected spring operation data (step 1105). This extraction of features can be performed in the same manner as in step 806 of FIG. 8. Then, the prediction execution unit 204 applies the extracted features to the prediction model of the above formula (4) as explanatory variables (step 1106).

[0104] Next, the prediction execution unit 204 acquires a prediction result obtained by applying the above explanatory variables to the model (step 1107). As a result, it is determined whether or not a temperature rise will occur in summer for each individual refrigerator 120.

[0105] Next, the prediction execution unit 204 executes prediction of usage and various sensor data for the individual determined as "present" in the prediction result of summer temperature rise (step 1108). Specifically, the prediction execution unit 204 executes the following steps 1109 to 1111. First, the prediction execution unit 204 extracts features to be explanatory variables using data on spring usage included in the driving data extracted in step 1102 that has been collected (step 1109). At this time, the prediction execution unit 204 may obtain a summer weather forecast from an external organization 160 or the like, and extract features of the data on spring usage and the data on summer weather generated from the summer weather forecast. This extraction of features can be executed in the same manner as step 906 in FIG. 9.

[0106] Next, the prediction execution unit 204 applies the feature amounts extracted in step 1109 as explanatory variables to the prediction model of the above formula (5) (step 1110). Furthermore, the prediction execution unit 204 applies the predicted value of the data related to summer usage obtained in step 1110 to the prediction model of the above formula (6) (step 1111). For example, the prediction execution unit 204 extracts feature amounts of the data related to summer weather based on the summer weather forecast and the predicted value of the data related to summer usage obtained in step 1110, and applies them to the prediction model of formula (6). This extraction of feature amounts can be performed in the same manner as in step 1006 of FIG. 10.

[0107] Then, the prediction execution unit 204 obtains, as a result of step 1108, predicted values ​​of the summer usage data and predicted values ​​of various summer sensor data (step 1112), and outputs the obtained prediction results (i.e., prediction results based on the prediction models of equations (4), (5), and (6)) (step 1113).

[0108] FIG. 12 is an explanatory diagram showing a first example of utilization of the prediction result by the data collection and analysis server 110 according to the embodiment of the present invention.

[0109] Specifically, Fig. 12 shows an example in which the result of predicting a temperature rise in refrigerator 120 in summer is utilized in the work of a support staff member for refrigerator 120. In this example, information reference / control terminal 130 sets prediction conditions according to input from an operator, and calls prediction program 115 of data collection / analysis server 110 (step 1201). The conditions set here may be, for example, the time period to be predicted (a predetermined period in summer in the example of Fig. 12), the time period to obtain data to be input to the prediction model (a predetermined period in spring in the example of Fig. 12), etc. Furthermore, when refrigerators 120 to be predicted are limited, the conditions for the limitation (for example, the model or the area where the refrigerator is installed) may be set.

[0110] The data collection and analysis server 110 extracts spring operating data 1203 from the stored operating data 1202 in accordance with the set conditions, applies it to the prediction model 1204, thereby predicting a temperature rise in summer, and outputs the result to the information reference and control terminal 130 (step 1205). Here, the stored operating data 1202 is the operating data stored in the memory unit 201 shown in Fig. 2, the extracted spring operating data 1203 corresponds to the latest data 212, the prediction model 1204 corresponds to the prediction model 215 (specifically, the prediction model of equations (4) to (6)), and step 1205 corresponds to the processing in Figs. 8 to 10.

[0111] The information reference and control terminal 130 tabulates the prediction results obtained from the data collection and analysis server 110 and presents the analysis results to the support staff (step 1206). The support staff can check the presented tabulation and analysis results and take appropriate measures (step 1207). The results of the tabulation and analysis executed in step 1206 and an example of a screen for presenting them will be described with reference to FIG. 13.

[0112] In the above example, the information reference / control terminal 130 executes step 1206, but the data collection / analysis server 110 may collect and analyze the prediction results and transmit the results to the information reference / control terminal 130, which may then present the results to the support staff.

[0113] FIG. 13 is an explanatory diagram showing an example of a screen displayed in a first example of utilization of a prediction result by the data collection and analysis server 110 according to the embodiment of the present invention.

[0114] A temperature rise prediction report screen 1301 for summer 2021 shown in FIG. 13 is an example of a screen displayed by a display device (not shown) of the information reference / control terminal 130. In this example, the number of registered refrigerators of a specific model installed nationwide by region, and the number and percentage of those determined to be at high risk of temperature rise in summer 2021 based on operation data in spring 2021 are displayed. The risk of temperature rise may be the likelihood of a temperature rise in summer predicted based on a prediction model. Here, when the operator selects one of the regions and instructs to view a list of devices at high risk of temperature rise, for example, screen 1302 is displayed.

[0115] A screen 1302 is a temperature rise prediction report screen when the user selects Nara prefecture in the Kansai region. In this example, a list of information on each refrigerator 120 of the registered model installed in Nara prefecture is displayed. This list includes, for example, the ID of the refrigerator, purchase date, usage history, temperature rise risk, risk factors for summer predicted from spring operation data, and the number of complaints. Here, the usage history is the period that has elapsed from the purchase date to the present time, the temperature rise risk is a risk value predicted by the prediction model, and the number of complaints is the number of complaints that users have already made to a support staff member about each refrigerator.

[0116] Here, the summer risk factors predicted from the spring driving data are estimated from the predicted summer usage based on the model of formula (5), and for example, the higher the predicted number of door openings of an individual is compared to the national average (or the average for the region to which the individual belongs), the higher the risk that it will cause a temperature rise may be determined. A specific example will be described later with reference to Fig. 15. The summer risk factors predicted from the spring driving data may be displayed in order of calculated risk.

[0117] FIG. 14 is an explanatory diagram showing a second example of utilization of the prediction result by the data collection and analysis server 110 according to the embodiment of the present invention.

[0118] Specifically, Fig. 14 shows an example in which the result of predicting a temperature rise in the refrigerator 120 in the summer is utilized to give advice to the user on how to use the refrigerator 120. In this example, the process (step 1201) in which the information reference and control terminal 130 sets prediction conditions and calls the prediction program 115 is similar to that shown in Fig. 12. Also, the process in which the data collection and analysis server 110 predicts a temperature rise in the summer according to the set conditions and outputs the result to the information reference and control terminal 130 (step 1205) is also similar to that shown in Fig. 12.

[0119] The information reference / control terminal 130 refers to the prediction results obtained from the data collection / analysis server 110, extracts individual refrigerators 120 that are at high risk of temperature rise in summer, and creates a list of risk factors for each (step 1401). Then, the information reference / control terminal 130 transmits an alert to the refrigerators 120 at high risk or a notification to the terminal device (e.g., smartphone 140) of the user of the refrigerator 120.

[0120] The refrigerator 120 may display the received notification, for example, on an operation panel. Alternatively, the smartphone 140 may display the received notification. The user can check the displayed alert, etc., and take measures such as changing the way the refrigerator is used (step 1403). An example of a screen displayed to the user via the smartphone 140 will be described with reference to FIG. 15.

[0121] Furthermore, in step 1401, information reference / control terminal 130 may compare the predicted values ​​of the various sensor data in summer based on the prediction model of above formula (6) obtained from data collection / analysis server 110 with the values ​​of the various sensor data actually obtained in summer, and if a predetermined condition is satisfied, such as the deviation between the two is greater than a predetermined standard, it may determine that there is a high risk of a breakdown in refrigerator 120, and output information about the breakdown of refrigerator 120 to a support staff member. Alternatively, information reference / control terminal 130 may transmit information about the breakdown of refrigerator 120 to refrigerator 120 or smartphone 140, and refrigerator 120 or smartphone 140 may output the information to a user.

[0122] In the above example, the information reference and control terminal 130 executes steps 1402 and 1403, but the data collection and analysis server 110 may execute steps 1402 and 1403 and transmit the results to the refrigerator 120 or the smartphone 140.

[0123] FIG. 15 is an explanatory diagram showing an example of a screen displayed in a second example of utilization of a prediction result by the data collection and analysis server 110 according to the embodiment of the present invention.

[0124] A refrigerator usage navigation screen 1501 shown in FIG. 15 is an example of a screen that is displayed on a display device (not shown) of a smartphone 140 of a user of a refrigerator 120 that has been determined to have a high risk of temperature rise in summer based on a prediction model, based on a notification from the information reference / control terminal 130.

[0125] The refrigerator usage navigation screen 1501 may include, for example, a message display section 1502 that notifies and calls attention to the high risk of the destination user's refrigerator 120 rising in temperature in summer, a risk value display section 1503 that displays a specific predicted risk value, and an advice display section 1504 that displays advice for reducing the risk.

[0126] Risk value display section 1503 displays, for example, a risk value predicted for each risk factor and its evaluation. The risk factors may include, for example, the number of times and the opening time of the refrigerator door, freezer door, and vegetable compartment door of the refrigerator 120, the degree of packing of items in the refrigerator, freezer, and vegetable compartment, the usage status of the ice compartment, the number of times quick ice is made, etc. The risk value here may be a result of comparing a predicted value (for example, the number of times the door is opened) of data on how to use the refrigerator 120 calculated for each risk factor based on the prediction model of equation (5) with the national average.

[0127] 15, the risk value is calculated so that if the predicted value is the same as the national average, the value is 50, if the predicted value is higher than the national average, the value is higher than 50, and if the predicted value is lower than the national average, the value is lower than 50, with the maximum value being 100 and the minimum value being 0. The risk value evaluation for each risk factor may be, for example, a rank of the magnitude of the risk value, such as "high," "slightly high," "slightly low," or "low."

[0128] In this example, the risk value is the result of comparing the predicted value of data regarding the usage of each individual refrigerator with the national average, but depending on the selection of the user, the result of comparing with the average of the area in which the refrigerator 120 is installed may be used as the risk value instead of the national average. Also, depending on an instruction from the user, the most recent usage history of the refrigerator 120 may be displayed. Although the usage history is not shown in the figure, information such as the number of times the door was opened and the opening time for the most recent few days may be displayed.

[0129] Advice display section 1504 displays advice for reducing risk factors with high risk value rankings among the risk factors displayed in risk value display section 1503. For example, if the risk value of the door open time is evaluated to be high, advice for shortening the door open time may be displayed.

[0130] In this embodiment, an example is shown in which the temperature inside the refrigerator 120 is predicted to rise in summer. However, for devices other than refrigerators, a model can be created similarly based on sensor data indicating past usage conditions to predict devices that will experience a predetermined event, such as a failure to achieve the expected performance, and a model to predict how the device will be used at that time. This makes it possible to predict devices that will experience problems in the future for any type of device and to propose improvements in usage.

[0131] Furthermore, the system according to the embodiment of the present invention may be configured as follows.

[0132] (1) An appliance state prediction method executed by an appliance state prediction system (e.g., appliance state prediction system 100) having a processor (e.g., processor 112) and a storage device (e.g., information storage unit 111), in which the storage device holds operation data (e.g., operation data 500) indicating an operating status of the appliance, a first model (e.g., a model of formula (4)) for predicting a state of the appliance at a second time period after the first time period based on the operation data at the first time period, and a second model (e.g., a model of formula (5)) for predicting a usage manner by a user of the appliance at the second time period based on the operation data at the first time period, The measurement method includes a first step (e.g., steps 1102 to 1107) in which a processor predicts a state of the appliance at a second time period by applying the acquired operating data at a first time period to a first model, a second step (e.g., steps 1109 to 1110, 1112) in which the processor predicts how the appliance will be used by a user at the second time period by applying the acquired operating data at the first time period to a second model, and a third step (e.g., step 1113) in which the processor outputs a prediction result of the state of the appliance at the second time period and a prediction result of how the appliance will be used by the user at the second time period.

[0133] This makes it possible to predict not only the occurrence of a specific event, but also the usage of devices related to the occurrence of the event, and to utilize this information in measures to reduce the risk of the event occurring.

[0134] (2) In the above (1), the appliance is a refrigerator (e.g., refrigerator 120) having at least one of a refrigerator compartment and a freezer compartment, the second period is a summer period, the operation data includes data on climate, data on how the appliance is used by a user, and data on a control state of the appliance, the first model is a model that predicts a rise in temperature inside the appliance in the second period based on at least one of the data on climate, data on how the appliance is used by a user, and data on a control state of the appliance in the first period, and the second model is a model that predicts a rise in temperature inside the appliance in the second period based on at least the data on how the appliance is used by a user in the first period. The method includes the steps of: (1) predicting how a user of the device will use the device at a time period; and (2) predicting an increase in temperature inside the device at a second time period in the future by applying at least one of the acquired data on climate for the first time period, data on how the user of the device will use the device, and data on the state of the control of the device to the first model; and (3) predicting an increase in temperature inside the device at a second time period in the future if an increase in temperature inside the device at a second time period in the future is predicted; and (4) predicting how a user of the device will use the device at a second time period in the future by applying the acquired data on how the user of the device will use the device at the first time period to the second model.

[0135] This makes it possible to predict the occurrence of a temperature rise inside the refrigerator, as well as how the refrigerator will be used when a temperature rise is predicted to occur, and to use this information to take measures to reduce the risk of temperature rise.

[0136] (3) In (2) above, the storage device further holds a predicted value of data related to the climate for a second time period in the future, and the second model is a model that predicts how the user of the device will use the device in the second time period (e.g., c in formula (5)) based on data related to how the user of the device will use the device in the first time period (e.g., b' in formula (5)) and data related to the climate for the second time period (e.g., a in formula (5)), and in the second step, the processor predicts how the user of the device will use the device in the second time period in the future by applying the acquired data related to how the user of the device will use the device in the first time period and the predicted value of the data related to the climate for the second time period in the future to the second model.

[0137] This allows for accurate prediction of future usage.

[0138] (4) In the above (3), the storage device further retains a third model (e.g., a model of equation (6)) that calculates the control state of the equipment for the second period based on data related to the climate for the second period and data related to how the equipment is used by a user, and the equipment state prediction method further includes a fourth step (e.g., steps 1109 to 1113) in which the processor calculates the control state of the equipment for the second period by applying the acquired data related to the climate for the second period and the acquired data related to how the equipment is used by a user for the second period to the third model, compares the control state of the equipment for the second period calculated based on the third model with the control state of the equipment actually acquired for the second period, and outputs information regarding an equipment failure if the magnitude of deviation between the two satisfies a predetermined condition.

[0139] This makes it possible to estimate the occurrence of a malfunction in the device and use the information to take measures against the malfunction.

[0140] (5) In the above (4), the method further includes a fifth step (e.g., FIG. 8) in which the processor generates a first model by learning the temperature inside the device at a second time period in the past and at least one of data related to the climate at a first time period prior to the second time period in the past, data related to how the device was used by a user, and data related to a control state of the device; a sixth step (e.g., FIG. 9) in which the processor generates a second model by learning the data related to how the device was used by a user at the second time period in the past, data related to the climate at the second time period in the past, and data related to how the device was used by a user at a first time period prior to the second time period in the past; and a seventh step (e.g., FIG. 10) in which the processor generates a third model by learning the data related to the control state of the device at the second time period in the past, data related to the climate at the second time period in the past, and data related to how the device was used by a user at the second time period in the past.

[0141] This will generate a suitable model.

[0142] (6) In the above (2), the method further includes an eighth step (e.g., steps 1206, 1402, 1403) in which the processor aggregates, for a plurality of devices, results of prediction of how the device will be used by the user at a second future time period based on the second model, and compares, for devices predicted to have a temperature rise based on the first model, the predicted how the device will be used by the user based on the second model with the aggregated how the plurality of devices will be used by the users, thereby estimating risk factors that will cause an increase in temperature, and outputting information indicating the estimated risk factors.

[0143] This allows for a valid risk assessment to be carried out and used to implement appropriate countermeasures.

[0144] (7) In the eighth step of (6) above, the processor aggregates the results of predictions of how the multiple devices will be used by users for each region in which the devices are installed, and presumes that the greater the difference between the usage by users of the multiple devices predicted to have a rise in temperature based on the first model and the usage by users of the multiple devices predicted based on the second model, the higher the risk of that usage.

[0145] This allows for a valid risk assessment to be carried out and used to implement appropriate countermeasures.

[0146] (8) In (2) above, the climate data includes at least one of the temperature and humidity around the appliance, the data on how the appliance is used by the user includes at least one of the number of times the appliance door is opened and the duration for which it is open, and the data on the control status of the appliance includes at least one of the setting value of the cooling strength of the appliance, the setting value of the operating mode of the appliance, the rotation speed of the motor that drives the compressor for cooling the appliance, and the temperature inside the appliance.

[0147] This allows for the generation of appropriate predictive models and the performance of risk assessments based on these models.

[0148] The present invention is not limited to the above-mentioned embodiment, but includes various modifications. For example, the above-mentioned embodiment has been described in detail for a better understanding of the present invention, and is not necessarily limited to those including all of the configurations described. Also, it is possible to replace a part of the configuration of one embodiment with the configuration of another embodiment, and it is also possible to add the configuration of another embodiment to the configuration of one embodiment. Also, it is possible to add, delete, or replace a part of the configuration of each embodiment with another configuration.

[0149] In addition, the above-mentioned configurations, functions, processing units, processing means, etc. may be realized in part or in whole by hardware, for example, by designing them as integrated circuits. In addition, the above-mentioned configurations, functions, etc. may be realized in software by a processor interpreting and executing a program that realizes each function. Information such as a program, table, file, etc. that realizes each function can be stored in a storage device such as a non-volatile semiconductor memory, a hard disk drive, or an SSD (Solid State Drive), or in a computer-readable non-transitory data storage medium such as an IC card, an SD card, or a DVD.

[0150] In addition, the control lines and information lines are those considered necessary for the explanation, and not all control lines and information lines in the product are necessarily shown. In reality, it can be considered that almost all components are connected to each other. [Explanation of symbols]

[0151] 100 Equipment Condition Prediction System 110 Data collection and analysis server 111 Information Storage Unit 112, 124 processors 113, 121 Network I / F 114, 127 Memory 115 Prediction Program 120 Refrigerator 122 Controlled Equipment 123 Control device 125 External I / F 126 Sensor Group 128 Control Program 129 Temporary storage of operation data 130 Information reference and control terminal 140 Smartphones 150 Network 160 External Organizations

Claims

1. An equipment state prediction method executed by an equipment state prediction system having a processor and a storage device, comprising: the storage device holds operation data indicating an operating status of an appliance, a first model for predicting a state of the appliance at a second time period after the first time period based on the operation data at the first time period, and a second model for predicting a usage manner of the appliance by a user at the second time period based on the operation data at the first time period; The apparatus state prediction method includes: a first step of the processor predicting a state of the equipment at the second time by applying the acquired operating data at the first time to the first model; a second step of the processor predicting a usage manner of the device by a user during the second time period by applying the acquired operating data during the first time period to the second model; a third step of the processor outputting a prediction result of a state of the device at the second time period and a prediction result of a usage of the device by a user at the second time period, the device is a refrigerator having at least one of a refrigeration compartment and a freezer compartment, the second period is a summer period, The operational data includes data on weather, data on how the user of the device uses the device, and data on the state of control of the device; the first model is a model that predicts a rise in temperature inside the device during the second time period based on at least one of data related to the climate during the first time period, data related to a usage of the device by a user, and data related to a control state of the device; The second model is a model for predicting how the user of the device will be used in the second time period based on at least data regarding how the user of the device will be used in the first time period; In the first step, the processor predicts a temperature rise in the device during the second time period in the future by applying at least one of the acquired data on the climate during the first time period, data on how the device is used by a user, and data on a state of control of the device to the first model; An equipment status prediction method characterized in that, in the second step, when it is predicted that the temperature inside the equipment will rise at the second time period in the future, the processor predicts how the equipment will be used by the user at the second time period in the future by applying the acquired data regarding how the equipment will be used by the user at the first time period to the second model.

2. The equipment state prediction method according to claim 1, The storage device further stores a predicted value of the weather-related data for the second time period in the future; the second model is a model that predicts how the user of the device will use the device in the second time period based on data on how the user will use the device in the first time period and data on the weather in the second time period; An equipment status prediction method characterized in that in the second step, the processor predicts how the user of the equipment will use the equipment in the second period in the future by applying the acquired data regarding how the user of the equipment will use the equipment in the first period and a predicted value of the weather data for the second period in the future to the second model.

3. The equipment state prediction method according to claim 2, The storage device further stores a third model that calculates a state of control of the device during the second time period based on data related to the weather during the second time period and data related to how the device is used by a user; The equipment status prediction method further includes a fourth step of: the processor calculates a control state of the equipment for the second period by applying the acquired data regarding the weather for the second period and the acquired data regarding how the user of the equipment for the second period to the third model; comparing the control state of the equipment for the second period calculated based on the third model with the control state of the equipment actually acquired for the second period; and outputting information regarding a failure of the equipment if the magnitude of deviation between the two satisfies a predetermined condition.

4. The equipment state prediction method according to claim 3, a fifth step of generating the first model by the processor learning at least one of the temperature inside the appliance during the second time period in the past, and data related to the climate during the first time period prior to the second time period in the past, data related to how the appliance was used by a user, and data related to a control state of the appliance; a sixth step of generating the second model by the processor learning data on how the user of the device used the device in the second time period in the past, data on the weather in the second time period in the past, and data on how the user of the device used the device in the first time period prior to the second time period in the past; The method for predicting an equipment status further comprises a seventh step of generating the third model by the processor learning data regarding the control state of the equipment during the second period in the past, data regarding the weather during the second period in the past, and data regarding how the equipment was used by a user during the second period in the past.

5. The equipment state prediction method according to claim 1, The processor, aggregating results of predictions of how users of the devices will use the devices at the second time period in the future based on the second model for a plurality of the devices; estimating risk factors that cause a rise in temperature by comparing a usage manner by a user of the device predicted based on the second model for the device whose temperature is predicted to rise based on the first model with a usage manner by the users of the aggregated plurality of devices; An equipment status prediction method, further comprising an eighth step of outputting information indicating the estimated risk factors.

6. The equipment state prediction method according to claim 5, In the eighth step, the processor aggregating results of predictions of how the plurality of devices will be used by users for each region in which the devices are installed; An equipment state prediction method, characterized in that the greater the difference between the usage by a user of an equipment predicted to have a rise in temperature based on the first model, predicted based on the second model, and the aggregated usage by users of the plurality of equipment, the higher the risk of that usage is estimated to be.

7. The equipment state prediction method according to claim 1, The weather data includes at least one of the temperature and humidity of the surroundings of the device; The data on how the user of the device uses the device includes at least one of the number of times the door of the device is opened and the duration of the door being opened; An equipment status prediction method, characterized in that the data regarding the control status of the equipment includes at least one of a set value of the cooling strength of the equipment, a set value of the operation mode of the equipment, a rotation speed of a motor that drives a cooling compressor of the equipment, and a temperature inside the equipment.

8. An equipment state prediction system having a processor and a storage device, the storage device holds operation data indicating an operating status of an appliance, a first model for predicting a state of the appliance at a second time period after the first time period based on the operation data at the first time period, and a second model for predicting a usage manner of the appliance by a user at the second time period based on the operation data at the first time period; The processor, a first step of predicting a state of the equipment at the second time by applying the acquired operating data at the first time to the first model; a second step of predicting how a user of the device will be used during the second period by applying the acquired operational data during the first period to the second model; a third step of outputting a prediction result of a state of the device at the second time period and a prediction result of a usage of the device by a user at the second time period; the device is a refrigerator having at least one of a refrigeration compartment and a freezer compartment, the second period is a summer period, The operational data includes data on weather, data on how the user of the device uses the device, and data on the state of control of the device; the first model is a model that predicts a rise in temperature inside the device during the second time period based on at least one of data related to the climate during the first time period, data related to a usage of the device by a user, and data related to a control state of the device; The second model is a model for predicting how the user of the device will be used in the second time period based on at least data regarding how the user of the device will be used in the first time period; In the first step, the processor predicts a temperature rise in the device during the second time period in the future by applying at least one of the acquired data on the climate during the first time period, data on how the device is used by a user, and data on a state of control of the device to the first model; An equipment status prediction system characterized in that, in the second step, when it is predicted that the temperature inside the equipment will rise during the second period in the future, the processor predicts how the equipment will be used by the user during the second period in the future by applying the acquired data regarding how the equipment will be used by the user during the first period to the second model.

Images