JP7078437B2 - Servers, programs, and equipment systems - Google Patents

Description

The present invention relates to servers, programs, and equipment systems.

In Patent Document 1, the communication terminal 1 reads the operation information 13a stored in the home electric appliance 10, and the communication terminal 1 uses the operation information 3a and the defect status 3c and 23a to estimate the defect content and perform a failure diagnosis. The methods that enable it are disclosed (0012, 0015, 0020, 0025, FIG. 1, etc.).

Japanese Unexamined Patent Publication No. 2016-95601

Since Patent Document 1 has the purpose of failure diagnosis, it is configured to perform this diagnosis when a failure actually occurs in the home electric appliance 10, and does not perform any diagnosis before the failure.

The server of the present invention made in view of the above circumstances is
A server that has a calculation unit
Acquire information via the network Acquire the operation history of the device via the communication unit and store it in the storage unit as teacher data .
When the arithmetic unit diagnoses the device using recent data including the operation history and detects a sign of failure and / or a pseudo-failure of the device, the detected failure sign / or pseudo-failure. Notify the output unit of the user of the device, including the output of the abnormality type and the output requesting the user to input whether or not the abnormality type actually occurred .
It is characterized in that the feedback from the user of the device returned in response to the output for which the input is requested is reflected in the teacher data in the storage unit .

Further, the device system of the present invention made in view of the above circumstances is
A server having a calculation unit, a device, a network, an output unit of a user of the device, and a storage unit that acquires the operation history of the device via a communication unit that acquires information via the network and stores it as teacher data. And is a device system that includes
When the arithmetic unit diagnoses the device using recent data including the operation history and detects a sign of failure and / or a pseudo-failure of the device, the detected failure sign / or pseudo-failure. Notify the output unit of the user of the device, including the output of the abnormality type and the output requesting the user to input whether or not the abnormality type actually occurred .
It is characterized in that the feedback from the user of the device returned in response to the output for which the input is requested is reflected in the teacher data in the storage unit .

The figure which shows the refrigerator system as an example of the equipment system of an embodiment. Side sectional view of the refrigerator of the embodiment A flowchart showing the flow of transmission of the operation history to the server of the refrigerator of the embodiment. (A) is a diagram showing the diagnostic data of the device stored in the server of the embodiment. (B) is a diagram showing abnormal teacher data including past operation history (teacher data) of the device, including signs of failure or pseudo-failure accumulated in the server of the embodiment. The figure which shows the teacher data (operation history) which concerns on the abnormality type "door open continuation" stored in the storage part of an embodiment. The figure which shows the teacher data (operation history) which concerns on the abnormality type "refrigerant leakage" stored in the storage part of an embodiment. A flowchart showing the flow of diagnostic processing of the server of the embodiment. The figure which shows the request output of the confirmation which is notified when the pseudo failure is detected in the refrigerator of an embodiment. The figure which shows the request output of the inspection request which is notified when the sign of failure is detected in the refrigerator of an embodiment.

Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings.
The various components of the present invention do not necessarily have to be independent of each other, and one component is composed of a plurality of members, a plurality of components are composed of one member, and one of the components. It is allowed that a part and a part of another component overlap with each other. For example, the server 3 may have the storage unit 31, or may exist as a separate device that is arranged so as to be able to communicate with the server 3.

In addition, not all of the components of each embodiment are essential to the present invention, and as long as there is no technical problem, some of the components of each embodiment can be selected and configured, and different embodiments can be configured. Configurations can be added, deleted, or converted or replaced with configurations of different embodiments.

By the way, when trying to make a predictive diagnosis that recognizes a sign that a device will fail in a relatively near future, the diagnosis includes the detection state of the sensor provided in the device and the operation of electrical components such as a refrigeration cycle and a fan. It is conceivable to acquire an operation history consisting of a part or all of a state or the like over a certain period of time and use this for diagnosis.

It is expected that the predictive diagnosis when the device is operating without receiving input or disturbance can be performed by judging whether or not there is a sign of failure of the device. However, in the actual usage environment of the device by the user, a defect caused by input or disturbance by the user or an operation error (for example, when the device is a refrigerator, the door keeps opening) (hereinafter referred to as a pseudo failure). Call.) Can occur. Although such a defect is not a malfunction of the equipment, it may adversely affect the function of the equipment at least with the passage of time, so it is preferable to diagnose the presence or absence of such a defect.

In the embodiment, by diagnosing the sign of failure, preferably the presence or absence of a pseudo-failure, the user can be notified in advance of the possibility of future trouble, so that the user can perform the operation before the failure actually occurs. It is possible to obtain an opportunity to take some measures at a time when the increase in the influence due to mistakes can be suppressed.

[Refrigerator system 4]
FIG. 1 is a diagram showing a refrigerator system 4 as an example of the equipment system in the present embodiment. The refrigerator system 4 is a refrigerator 1 (refrigerator 1a at Mr. A's house, refrigerator 1b at Mr. B's house, etc.) as an example of equipment arranged in the user's house 6 (Mr. A's house 6a, Mr. B's house 6b, ...). ()) A total of one or more (preferably a plurality), a terminal 5 (terminal 5a at Mr. A's house, terminal 5b at Mr. B's house, ...) That can be used by those users, and a wide area network 2 such as the Internet. , A server 3 and a call center 7 where a service person works.

<Terminal 5>
The terminal 5 has an output unit that outputs information perceptible to the user (for example, a display unit that displays characters and images, a pronunciation unit that outputs voice), and an input unit that accepts input from the user.

The terminal 5 can output the contents analyzed by the server 3 for the refrigerator 1 from the output unit in a manner perceptible to the user, and can receive the user's command or the like for this output via the input unit. If such a function can be achieved, an output unit or an input unit may be arranged in the refrigerator 1 instead of the terminal 5. The terminal 5 is preferable in terms of portability to the user, and this embodiment will be described mainly using the terminal 5.

<Server 3>
The server 3 has a storage unit 31, a calculation unit 32, and a communication unit 33.
The storage unit 31 can store abnormal teacher data including an operation history (teacher data) of the refrigerator 1 that has been determined to be a failure in the past or a pseudo-failure described later for a certain period of time. The details will be described later.
The calculation unit 32 uses the abnormal teacher data stored in the storage unit 31 to include an operation history (recent data) of the refrigerator 1 currently in operation for a certain amount of time from the present to the past. The diagnostic data is analyzed to determine whether the currently operating refrigerator 1 is showing signs of failure in the future or showing pseudo-failure. As the arithmetic unit 32, for example, using a support vector machine, a neural network, or other known artificial intelligence, is the refrigerator 1, which is a source of recent data, showing some kind of failure sign or pseudo-failure? It is possible to classify whether or not.
The communication unit 33 enables information to be exchanged between the server 3 and the refrigerator 1, the terminal 5, and the call center 7 via the wide area network 2.

<Call center 7>
The call center 7 can perform the work of inputting the abnormal teacher data into the storage unit 31. The abnormal teacher data can be created by using the operation history of the device collected by the communication unit 33, the service person belonging to the call center, or the like. The abnormal teacher data does not necessarily have to be stored in the storage unit 31, and may be stored in a mode in which the arithmetic unit 32 can learn or refer to it.

The call center 7 can dispatch a service person under the device 1 in response to a request from the server 3 or a user (input unit of the terminal 5 or the device 1).

<Equipment>
The device is installed in, for example, the user's house 6, and the type may include home appliances such as a lighting, a washing machine, an air conditioner, an AV device, or a refrigerator 1. As in the present embodiment, the equipment belonging to the equipment system 4 is composed of the same type (refrigerator in the present embodiment) because at least a part of the types of possible failures and pseudo-failures are common to each other. At least, it is configured so that it can be distinguished whether or not the types of devices are the same.

(Refrigerator 1)
FIG. 2 is a side sectional view of the refrigerator 1 as an example of the device of the present embodiment. The refrigerator 1 receives the sensor detection values described later, outputs control signals to electrical products such as refrigeration cycles and fans, detects user input signals through the input unit 101, and wide-area network 2. It has a control board 10 for performing communication via the above, and autonomous control is performed by this. The place where the input unit 101 is arranged is not particularly limited, but may be provided in the refrigerating room door 111 as in the present embodiment or may be arranged in any storage room. Further, the input unit 101 can include an output unit that outputs information in a manner perceptible to the user.

(Storage room)
As a storage chamber, the refrigerator 1 includes a refrigerating chamber 11 controlled to maintain the temperature in the refrigerating temperature zone, a freezing chamber 12 and an ice making chamber 13 controlled to maintain the temperature in the refrigerating temperature zone, and a refrigerating temperature zone. Among them, it has a vegetable chamber 14 controlled to maintain a relatively high temperature, and a decompression chamber 15 capable of maintaining a temperature in a chilled temperature zone and reducing the pressure to a pressure below atmospheric pressure. The decompression chamber 15 is installed in the refrigerating chamber 11.

(Storage room door)
The refrigerating room 11, the freezing room 12, the ice making room 13, the vegetable room 14, and the decompression room 15 have doors that can open and close the openings of the respective storage rooms, such as the refrigerating room door 111, the freezing room door 121, and the ice making room door 131. A vegetable room door 141 and a decompression room door 151 are provided.

A packing is attached to the edge of each door on the side in contact with the opening of each storage chamber in order to improve the adhesion to the opening when the door is closed.

(Supply of cold air)
The refrigerator 1 constitutes a refrigerating cycle that generates cold air, and includes a compressor 161 that compresses the refrigerant, a radiator that dissipates the compressed refrigerant toward the outside air (not shown), and a decompressor that depressurizes the refrigerant (not shown). (Illustrated), the first evaporator 162 and the second evaporator 163 as an evaporator for evaporating the refrigerant are connected in a ring shape by a pipe.

The cold air generated by the evaporators 162 and 163 is boosted by the first fan 164 and the second fan 165, respectively, and blown to either storage chamber. In the present embodiment, the cold air boosted by the first fan 164 is sent to the refrigerating room 11 and also sent to the vegetable room 14 through an air passage (not shown) communicating the refrigerating room 11 and the vegetable room 14. The cold air in the vegetable compartment 14 circulates through a first return air passage (not shown) communicating with the first evaporator 162.
Further, the cold air boosted by the second fan 165 is sent to the freezing chamber 12 and the ice making chamber 13 and circulates through the second return air passage (not shown) communicating with the second evaporator 163.

(Heat supply)
For example, storage chambers and evaporators 162 and 163 may be physically too cold or frosted for some reason or physically. In such a case, it is desirable to heat the overcooled region or the frosted region to eliminate the overcooled or frosted region. In the present embodiment, a decompression chamber heater 153, a first evaporator heater 166, and a second evaporator heater 167 as examples of electrical products are arranged below the decompression chamber 15 and in the vicinity of the evaporators 162 and 163 in the storage chamber, respectively. is doing. These heaters may be an energized type that energizes and generates heat, or may be one in which a refrigerant flows by switching the flow path of the refrigeration cycle and the refrigerant dissipates heat. The energized type is preferable. Further, a heater may be arranged in another storage room. Overcooling and frost formation can be distinguished or estimated using sensors and reference to storage room target temperature settings described below.

In the vicinity of the evaporators 162 and 163, an evaporator temperature sensor (not shown) for detecting the temperature of each of the evaporators 162 and 163 is arranged.

(Sensor)
The refrigerator 1 can detect information outside and inside the refrigerator, as well as the refrigerating cycle and the driving state of electric products by using various sensors.
The refrigerator 1 includes a refrigerating room temperature sensor 114 provided in the refrigerating room 11, a freezing room temperature sensor 114 provided in the freezing room 12, an ice making room temperature sensor 134 provided in the ice making room 13, and a vegetable room temperature sensor provided in the vegetable room 14. It has 144. In addition to these temperature sensors, a humidity sensor that detects humidity may be arranged.

The refrigerator 1 has an outside temperature sensor 171 and an outside humidity sensor 172 that detect the temperature and humidity outside the refrigerator, respectively.

Each door is provided with a refrigerating room door sensor, a freezing room door sensor, an ice making room door sensor, and a vegetable room door sensor as door sensors (not shown) capable of detecting the opening and closing of the door. A decompression chamber door sensor that detects the opening and closing of the decompression chamber door 151 may be provided. Various known ones can be adopted as each door sensor.
The door sensor can have, for example, a magnet provided on one of the opening edge side of each storage chamber and each door, and a sensor for detecting magnetic flux. For example, the door sensor configured in this way can determine whether they are close to each other (door closed) or not (door open). With such a door sensor, there is a risk that the state in which the packing is inadequately adhered to the door and cold air leaks more easily (half-door) than when the door is completely closed is recognized as the same state as when the door is closed.

The refrigerator 1 includes a compressor sensor (not shown) that detects the motor rotation speed of the compressor 161, a first evaporator temperature sensor (not shown) that detects the temperature of each of the evaporators 162 and 163, and a second evaporator temperature sensor. It has a first fan sensor (not shown) that detects the rotation speed of the first fan, and a second fan sensor (not shown) that detects the rotation speed of the second fan.

[Send operation history]
FIG. 3 is a flowchart showing the flow of transmission of the operation history to the server 3 of the refrigerator 1 of the embodiment. The refrigerator 1 transmits various sensor detection values, electrical products, and refrigeration cycle information to the server 3.
When the refrigerator 1 is supplied with power, the temperature, humidity, door opening / closing or pressure, energization amount of electrical components such as refrigeration cycle, and motor rotation speed detected by various sensor values described in the present specification. Various states such as are stored in a memory (not shown) as an operation history (step S101). The refrigerator 1 can count the elapsed time, and for example, when a predetermined time (not necessarily constant) has elapsed (step S110, Yes), the stored information is transmitted to the server 3 (step S120).

Further, the refrigerator 1 can also transmit the model number of the refrigerator 1, a unique identification number assigned to each individual, and the like.

The operation history transmitted by each refrigerator 1 to the server 3 is transmitted every predetermined time unit (not necessarily constant). For example, the operation history can be sequentially transmitted every 5, 10, 30, and 60 minutes and stored in the storage unit 31 of the server 3, or a plurality of operation histories stored in the memory can be collectively transmitted to the server 3. ..

The operation history of the refrigerator 1 is stored in the storage unit 31 for at least a certain period of time (recent data), and the calculation unit 32 diagnoses a sign of failure and a pseudo-failure using this. As the recent data, it is preferable to obtain the time from the present or a time close to the present of the refrigerator 1 to some extent in the past. As a result, when the diagnosis is executed, it is possible to provide the user with highly fresh information. The operation history that is old to some extent (for example, 24 hours or more or 48 hours or more) from the present may be deleted from the storage unit 31. As a result, it is possible to prevent the storage capacity required for storing recent data from becoming excessive.

The timing of the diagnosis by the server 3 may be the timing when a new operation history is received from the refrigerator 1, the timing according to the algorithm determined by the service provider, or the timing specified by the service provider. Since the server 3 of the present embodiment diagnoses a sign of failure that is normally impossible for the user to notice, it is preferable to start the diagnosis in a manner other than the user's request. For example, the number of operation history data received from each refrigerator 1 may be set to a predetermined number or more from the previous diagnosis, or the load on the server 3 may not be excessive, and the data may be distributed and assigned to each identification number of the refrigerator 1. It may be timing.

[Diagnosis data]
FIG. 4A is a diagram showing the device to be diagnosed stored in the server 3 of the present embodiment. The server 3 acquires information including an operation history from each of the refrigerators 1 through the communication unit 33 and stores the information in the storage unit 31.

The storage unit 31 stores, for each device belonging to the refrigerator system 4, the device type (refrigerator in this embodiment), similar group code, lot number, identification number, elapsed time from the previous diagnosis, and recent data. ing. For convenience, these are collectively referred to as diagnosed data. The diagnostic data may include, for example, at least an identification number and recent data. If it is possible to identify the device to be diagnosed and notify the diagnosis result, the identification number may not be included.

The similar group code can be defined in advance by a person who provides the service of the refrigerator system 4 among a certain device type (refrigerator in the present embodiment). In the present embodiment, it is a group number when the service provider pre-groups by the model number of the refrigerator. For example, refrigerators of the same model number (same type) are grouped so as to belong to the same group. Further, even if the refrigerators have different model numbers (variant types), there may be refrigerators having the same or similar control software, refrigerating cycle configuration, and compressor model number. If they are not the same but similar, it is expected that the methods of finding possible failures and their signs will be similar. Therefore, the service provider can assign the same similarity group code to the model numbers of refrigerators that are determined to be similar to each other in advance. As a result, it is possible to perform a grouped analysis to some extent even for atypical refrigerators, so that the number of samples in each group can be increased and accuracy can be expected to be improved.

The lot number is a number assigned according to, for example, when and in which factory a device was manufactured. Refrigerators of the same lot or similar lots may have certain signs of failure rate and type of failure. Therefore, by storing this in the storage unit 31, it can be used as a reference when determining the failure sign. be able to.

The identification number is a unique number for each refrigerator. This makes it possible to distinguish each of the refrigerators from each other. Since the device type, similar group code, and lot number can be associated with the identification number, if sufficient calculation and communication speed can be secured, store these in another device and use the identification number to store them. You may call it as appropriate.

The elapsed time from the previous diagnosis indicates the elapsed time from the time when the device (refrigerator 1) having the corresponding identification number executed the previous diagnosis. This information may be used to perform the diagnosis, for example, every 12 hours or 24 hours.

The recent data address indicates an address in which the operation history received from the device having the corresponding identification number is stored for at least a certain period of time.
Using the data to be diagnosed including the recent data, the arithmetic unit 32 executes a diagnosis for determining the presence or absence of a sign of failure or a pseudo-failure.

[Abnormal teacher data]
FIG. 4B is a diagram showing abnormal teacher data including operation history (teacher data) of past devices, including signs of failures or pseudo-failures accumulated in the server 3 of the present embodiment. The abnormal teacher data may be stored in the server 3 as in the present embodiment, but it may be stored in another device so that the arithmetic unit 32 can read and learn or refer to the data as appropriate. When the calculation unit 32 determines that the abnormal teacher data is similar to the diagnosed data as a result of learning or referring to the abnormal teacher data, it can determine whether a sign of failure has been detected or a pseudo failure has been detected. That is, the abnormal teacher data can be used to classify the diagnosed data.

The abnormal teacher data includes the abnormal type and the teacher data address in addition to the similar group code, lot number, and identification number. Abnormal teacher data can be collected and created by, for example, a service person.

The type of abnormality is a device failure or pseudo-failure (abnormality) that the service person (call center 7) or server 3 or the like in charge of the diagnosis made when diagnosing a sign or pseudo-failure of the refrigerator 1 in the past. It is a type.

For example, in the case of a refrigerator as in the present embodiment, "door open continuation" (pseudo-failure) in which cold air leaks more than usual because the door of a certain storage room is not closed, and the refrigerant enclosed in the refrigeration cycle. "Refrigerant leak" (failure) where gas leaks for some reason, "valve abnormality" (failure) where the operation of the valve that switches the loop of the refrigerant flow in the refrigeration cycle becomes abnormal, hot food etc. is thrown into the storage room by the user etc. As a result, the storage chamber temperature is less likely to drop, such as "large heat load" (pseudo-failure).

The teacher data address indicates the address of the storage unit 31 in which information related to the specific operation history (teacher data) of the refrigerator (equipment) determined to be a failure or a pseudo failure is stored.

The calculation unit 31 can learn at least the device type, the abnormality type, and the teacher data, preferably further including the similar group code, the lot number, and the identification number as the abnormality teacher data. At that time, learning may be distinguished for each similar group code, and the criteria may be set to be used properly so that the learning related to a certain similar group code does not affect the learning of different similar group codes.

The failure usually means a device that cannot or is difficult for the user of the device to repair by himself / herself. For example, when the refrigerator 1 is mentioned as the device, the compressor 161 and the decompressor, the evaporator 162, 163, Fans 164, 165, heaters 153, 166, 167, other sensor failures, pipe breakage, disconnection of lead wires of circuits, etc. In the present embodiment, the server 3 autonomously diagnoses whether or not the sign is occurring, and when it detects that the sign is occurring, it notifies the user to that effect and prompts the user to take an early action. be able to.

On the other hand, the pseudo-failure means a device that may adversely affect the present or future due to an operation error of the device user or an excessive input. For example, when the refrigerator 1 is mentioned as a device, the storage room is used. Input of hot food and drink, open the door of the storage room for a long time, or make it a so-called half door. In the present embodiment, when the server 3 autonomously diagnoses the presence or absence of a pseudo-failure and detects that it has occurred, the user is notified to that effect, so that the user can be made aware of the defect and urge the response. ..

(Example of teacher data)
(Teacher data of the abnormality type "Door open continuation" stored in the teacher data address "TAA1001")
FIG. 5 is a diagram showing teacher data (operation history) related to the abnormality type “door open continuation” stored in the storage unit 31 of the present embodiment. The teacher data includes the refrigerating room temperature 40, the cooler temperature 43, and the refrigerating room door opening time 46. The horizontal axis is represented as a discrete time, and the door opening time 46 indicates, for example, an hourly opening time subtotal.

This teacher data may include at least a part including an abnormality type "door open continuation" or a sign thereof, and may also include an operation history in a "normal" time zone. In this teacher data, before the time t1, it is the operation history of the normal state that has not reached the "door open continuation". During this period, the refrigerating chamber temperature is less than 10 ° C., the cooler temperature 43 is generally −10 ° C. or lower, and the door opening time 46 is short. The door opening time 46 keeps a long time with the time t1 as a boundary, and in particular, in this teacher data, the door opening is detected for almost the entire 1 hour during the discrete time 1 hour. Further, after time t1, the refrigerator temperature 40 tends to increase, while the cooler temperature 43 tends to decrease.

In fact, when the refrigerating room door 111 "continues to open the door", the refrigerating room door sensor continues to detect the opening, and the cold air in the refrigerating room 11 leaks, so that the refrigerating room temperature 40 rises. On the other hand, in the refrigerator of the similar group code "001" giving the present abnormal teacher data, at least, the cooler temperature 43 shows a sign that the cooler temperature 43 is lowered in an attempt to generate cold air in order to lower the refrigerating room temperature 40.

Such a feature can be found by, for example, a service person of the call center 7, stored in the server 3 together with the operation history, or can be taught to artificial intelligence as a feature amount. Alternatively, the artificial intelligence may be made to learn a large amount of abnormal teacher data as it is without any particular instruction, and the actual operation history may be classified. When teaching the feature amount, for example, using this teacher data as an example, in addition to the continuous detection of the door opening, the approximate straight lines 41, 42 and the cooler that approximate the upward tendency of the refrigerating room temperature 40 and the cooler. The service person can draw approximate straight lines 44 and 45 that approximate the downward tendency of the temperature 43, and add this to the abnormal teacher data. Further, it may be learned with the information that it is normal until the time t1. If the refrigerating room temperature 40 is maintained even though the door opening detection is continued, there is a possibility that the refrigerating room door sensor may fail or indicate a sign thereof.

(Teacher data of the abnormal type "refrigerant leak" stored in the teacher data address "TAA1002")
FIG. 6 is a diagram showing teacher data (operation history) related to the abnormality type “refrigerant leakage” stored in the storage unit 31 of the present embodiment. The teacher data includes the refrigerating room temperature 50, the cooler temperature 53, and the refrigerating room door opening time 56. The horizontal axis is represented as a discrete time, and the door opening time 56 indicates, for example, an hourly opening time subtotal.

Before the time tn-2 in this teacher data, it is an operation history in a normal state in which no sign of "refrigerant leakage" has occurred. During this period, the refrigerating chamber temperature 50 is less than 10 ° C., and the cooler temperature 53 is about −20 ° C. The door opening time 56 is short.

However, in a short time from immediately after time tn-2 to tn-1, the cooler temperature 53 rose sharply, which was larger than the changes in the refrigerating room temperature 50 and the door opening time 56. This sudden rise was subsequently resolved, but the cooler temperature 53 rose again in a short time from immediately after time t1 to t2. Even though the cooler temperature 53 rises sharply and the door opening time 56 is short, the fluctuation of the refrigerator compartment temperature 50 is small.

In fact, when the refrigerant leaks from the refrigeration cycle, the low temperature refrigerant starts to leak out of the cycle, and the cooler temperature 53 rises. Normally, the refrigerating chamber is cooled by the low-temperature refrigerant. Therefore, if the cooler temperature 53 rises, the refrigerating chamber temperature 50 may show a downward tendency, but this relationship is irrelevant, and the refrigerating chamber temperature 50 rises. The opening of the door, which is a factor, has not been detected. The fact that such a sudden rise in the cooler temperature 53 is observed once, preferably multiple times, suggests a refrigerant leak.

In the present teacher data, preferably, the approximate straight lines 54 and 55 of the rapid rise of the cooler temperature 53 drawn by the service person may be input as the feature amount. Further, it may be learned with the information that it is normal until the time tun-2.

The calculation unit 32 has already learned about various other abnormal types using one or more abnormal teacher data.

[Notification of failure sign or diagnosis result of pseudo failure]
FIG. 7 is a flowchart showing the flow of the diagnostic process of the server 3 of the present embodiment. When, for example, there is unlearned abnormal teacher data prepared by the call center 7 (step S100, Yes), the server 3 causes, for example, artificial intelligence of the arithmetic unit 32 to learn it (step S110). In the present embodiment, it is expressed that the processing is performed after the unlearned abnormal teacher data disappears, but the present invention is not limited to this, and if a certain degree of learning is completed, the rest. Learning and subsequent processing (diagnosis processing) may be executed in parallel or alternately. Further, when the server 3 does not learn the abnormal teacher data but calculates the similarity with the diagnosed data by referring to each abnormal teacher data, steps S100 and S110 are not necessary.

While the server 3 sequentially receives the operation history transmitted from the refrigerator 1 (step S120) and determines that the diagnosis time of the refrigerator 1 has arrived (step S200, Yes), the server 3 stores the diagnostic data stored in the storage unit 31. By using this, it is diagnosed whether or not there is a pseudo-failure of the refrigerator 1 to be diagnosed or a sign of failure (step S210). In this diagnosis, the data to be diagnosed may be classified by the learned calculation unit 32, or the calculation unit 32 compares the data to be diagnosed with the respective abnormal teacher data to calculate the similarity, and the degree of similarity is high. If there is abnormal teacher data, it may be determined that the abnormal type has occurred. Various known methods can be adopted for calculating the degree of similarity.

When a pseudo failure is detected (step S300, Yes), the user of the refrigerator 1 to be diagnosed is notified of a confirmation request as shown in FIG. 8 described later. (Step S310).

When a sign of failure is detected (step S400, Yes), the user of the refrigerator 1 to be diagnosed is notified of a request for an inspection request as shown in FIG. 9 described later (step S410). Further, in this case, the call center 7 is notified of the identification number of the refrigerator 1 indicating the sign of failure (step S420). It is clear to those skilled in the art that a user can be contacted by making it possible to refer to the user's contact information (such as the e-mail address of the terminal 5 or the IP address of the refrigerator 1) in association with the identification number.

[Request output 70 for confirmation of pseudo-failure]
FIG. 8 is a diagram showing a confirmation request output 70 notified when a pseudo failure is detected in the refrigerator 1 of the present embodiment. When the server 3 detects a pseudo failure, it notifies the terminal 5 of the user of the refrigerator 1 to be diagnosed and the output unit of the input unit 101.

The output unit may output the content so that the user can perceive it, and may not only display an image or a moving image but also pronounce a voice. In addition, it is said to "notify the user" that the information is perceptibly transmitted to the user through the output unit.

The confirmation request output 70 includes a confirmation request item output 71, an instruction output 72, an abnormality remaining time instruction 73, and a request deletion output 74.

The confirmation request item output 71 displays the type and location of the pseudo-fault to the user according to the type of the detected pseudo-fault. When it is determined that the refrigerating room with the identification number XXXX continues to open as in the present embodiment, for example, "Isn't the door of the refrigerating room of the identification number XXX refrigerator kept open?" The question text is displayed.

The instruction output 72 requests the user for a specific action according to the type of the detected pseudo-failure. In the present embodiment, it is displayed that the user wants to confirm whether the door of the refrigerating room is closed. The instruction output 72 may include an image or a moving image that visually displays the confirmation target portion of the device. In this embodiment, an image highlighting the refrigerator door of the refrigerator 1 is output.

The abnormality remaining time instruction 73 assists in contacting the call center 7 when the pseudo-fault cannot be resolved by the action obtained by the instruction output 72 or the user does not know what to do. In the present embodiment, it is possible to transition to the contact screen (not shown) to the call center 7 by performing an operation through the instruction 73 when an abnormality remains.

The request deletion output 74 can delete the confirmation request output 70. When the pseudo-fault is resolved, the recent data of the refrigerator 1 is subsequently transmitted to the server 3, and if the diagnostic process is executed again, the server 3 can detect the resolution of the pseudo-fault. Can be processed to delete. However, in this case, since this notification remains on the user side for a certain period of time, the user can delete the request output 70 by performing an operation through the request deletion output 74.

In addition, the request output 70 can include a pseudo-fault authenticity determination output (not shown). The pseudo-failure authenticity determination output is an output for requesting an input as to whether or not the type of pseudo-failure notified by the server 3 actually occurred in the refrigerator 1 to be diagnosed. According to this output, feedback of the diagnosis result can be received from the user. When it is input that it is "true" or "false", the server 3 treats the recent data of the refrigerator 1 to be diagnosed as new abnormal teacher data (or normal teacher data indicating normal operation). Can be done. It is preferable that the arithmetic unit 32 also learns the normal teacher data.

[Request for inspection output 80]
FIG. 9 is a diagram showing a request output 80 for an inspection request notified when a sign of failure is detected in the refrigerator 1 of the present embodiment. When the server 3 detects a failure sign, it notifies the display unit of the user of the refrigerator 1 to be diagnosed.

The inspection request request output 80 includes a failure sign output 81, a failure impact output 82, and an inspection request auxiliary output 83. Request for inspection The output 80 may or may not display a specific failure type (for example, refrigerant leakage). The display is preferable in that more detailed information can be provided to the user. If it is not displayed, it is usually impossible for the user to repair the failure by himself / herself, and it is preferable in that it is possible to reduce the risk of causing unnecessary worries or making the user think that he / she should do something by himself / herself when notifying the user of the type.

The failure sign output 81 outputs that the refrigerator 1 has detected a sign of failure in the future.

The failure effect output 82 outputs the effect on the refrigerator 1 when a predicted failure occurs. When a refrigerant leak is predicted as in the present embodiment, the effect of "stop cooling of all storage chambers" is output. The range of influence may be visually expressed in the failure influence output 82. In this embodiment, the screen of the refrigerator 1 in which all the storage rooms are highlighted is displayed.

The inspection request auxiliary output 83 assists or promotes communication to the call center 7. In the present embodiment, the inspection request screen to the call center 7 can be transitioned to by operating through the inspection request auxiliary output 83 (for example, the call processing to the call center 7 or the transition to the inspection reservation form). However, it may simply display the telephone number, e-mail address, or the like of the call center 7.

The information to be notified to the user in this embodiment may be notified to the call center 7.

As described above, according to the present embodiment, the device can be diagnosed by using the recent data including the operation history, and when a sign or a pseudo failure of the device is detected, the user or the call center can be notified.

[Contrast with related technologies]
According to JP-A-2015-288760, when the keys 28 to 30 displayed by the personal computer 1 are operated by the user, the personal computer 1 has the current state of the refrigerator 7 with respect to the control communication circuit 23 of the refrigerator 7. The "first operation information" corresponding to the information corresponding to the above is determined and requested to be transmitted to the personal computer 1 via the server 5 (0025). For example, when the key 28 "not cold / not cold" is operated by the user, the communication control circuit 23 determines the current "power information (indicating whether or not power is supplied)" of the refrigerator 7, and as a result. The "first operation information" is transmitted to the personal computer 1 (0024,0025).
Other first operation information includes "time information" indicating whether or not the time has elapsed since the start of power supply, "weak operation information", "power saving information", "strong operation information", and the like, and the current state of the refrigerator 7. Information according to the state is given (0021). If the above "power supply information" is not used as the first operation information, these are determined in order (0026-0028).
Further, it is disclosed that the "unique advice" determined according to the selected keys 28-30 and the first operation information is displayed on the personal computer 1 (0031,0033, FIG. 6).

As described above, Japanese Patent Application Laid-Open No. 2015-288760 provides advice based on the user's operation. In addition, advice is created using the settings of the refrigerator 7 (currently) when the user selects keys 28 to 30, and advice is given unless the cause such as "not cold" can be confirmed even now. Cannot be created. Therefore, Japanese Patent Application Laid-Open No. 2015-288760 cannot create advice unless the user notices the abnormality. Further, in Japanese Unexamined Patent Publication No. 2015-288, when the refrigerator 7 is normal, but the state is temporarily disturbed due to past operations and the like and is currently recovering, the disturbed state was perceived. Even if the user suspects that it is a malfunction, he cannot give any advice.

In addition, it is also described and suggested about an event whose cause is difficult to estimate only from the current state of the refrigerator 7 at that moment, for example, a failure sign or a pseudo-failure determination using recent data over a certain period of time. Not.

1 Equipment (can include an input unit and an output unit) (for example, a refrigerator)
2 Wide area network 3 Server 31 Storage unit 32 Calculation unit (artificial intelligence)
33 Communication unit 4 Equipment system (refrigerator system)
5 Terminals (including input and output)
7 Call center (service person)
70 Confirmation request output 71 Confirmation request item output 72 Instruction output 73 Abnormal remaining instruction 74 Request deletion output 80 Inspection request request output 81 Failure sign output 82 Failure effect output 83 Inspection request auxiliary output

Claims (11)

A server having a calculation unit, a device, a network, an output unit of a user of the device, and a storage unit that acquires the operation history of the device via a communication unit that acquires information via the network and stores it as teacher data. And is a device system that includes
When the arithmetic unit diagnoses the device using recent data including the operation history and detects a sign of failure and / or a pseudo-failure of the device, the detected failure sign / or pseudo-failure. Notify the output unit of the user of the device, including the output of the abnormality type and the output requesting the user to input whether or not the abnormality type actually occurred .
A device system characterized in that feedback from a user of a device returned in response to an output for which input is requested is reflected in the teacher data in the storage unit . When it is input that the diagnosis result of the device is true or false as the feedback, the request is characterized in that the recent data is reflected as abnormal teacher data indicating abnormal operation or normal teacher data indicating normal operation. Item 1. The equipment system according to Item 1. The arithmetic unit diagnoses the presence or absence of signs of failure of the device, and determines the presence or absence of signs of failure.
The device system according to claim 1 or 2 , wherein the diagnosis is started by a trigger other than the request of the user. The arithmetic unit
It is possible to learn or refer to abnormal teacher data including teacher data as an operation history including an abnormal type of a device of the same or similar type as the device and a time zone in which the abnormal type has occurred or a sign thereof has occurred.
The data to be diagnosed , which is the recent data including the operation history used when diagnosing the device by using the result of the learning or the reference, classifies the sign of the failure of the device and / or the presence / absence of the pseudo-failure. The device system according to any one of claims 1 to 3, wherein the device system is characterized by the above. When a sign of failure of the equipment is detected, the request output for inspection request is output to the output unit.
The request output of the inspection request is
Failure sign output indicating that the device may fail, and / or
The device system according to any one of claims 1 to 4, wherein the device system includes a failure effect output that displays the effect when the device fails. The equipment system according to claim 5 , wherein the inspection request request output includes an inspection request auxiliary output that assists and / or prompts the inspection request to the call center . It is possible to start the diagnosis triggered by something other than the request from the user.
When a pseudo-failure of the device is detected, a confirmation request output is output to the output unit.
The confirmation request output is
The device system according to any one of claims 1 to 6, which is determined according to the type of pseudo-failure detected by the server and includes an instruction output for requesting a specific action from the user. The above confirmation request output is
Instructions for remaining abnormalities to assist in contacting the call center , or
The device system according to claim 7 , wherein the confirmation request output is deleted. The request deletion output is included. Communicate with multiple devices
The arithmetic unit uses the similar group code and / or the lot number of the device preset for each model number of the device to properly use the criteria used for the diagnosis or to learn by dividing the criteria.
The device system according to any one of claims 1 to 8, wherein the device is distinguished by using an identification number unique to each device . A program for making a computer as a server function as a computer of the device system according to any one of claims 1 to 9. A server that has a calculation unit
Acquire information via the network Acquire the operation history of the device via the communication unit and store it in the storage unit as teacher data .
When the arithmetic unit diagnoses the device using recent data including the operation history and detects a sign of failure and / or a pseudo-failure of the device, the detected failure sign / or pseudo-failure. Notify the output unit of the user of the device, including the output of the abnormality type and the output requesting the user to input whether or not the abnormality type actually occurred .
A server characterized in that feedback from a user of a device returned in response to an output for which input is requested is reflected in the teacher data in the storage unit .

Images