Method and system for providing maintenance service for recording media in electronic devices

Abstract

This method executed by a calculation circuit of at least one computer comprises: a step for periodically collecting log information from an electronic device; a step for inputting, to a prediction model, log information collected from the electronic device earlier than the present time; a step for causing the prediction model to predict an abnormal or normal operation state of a recording medium at the present time or later; and a step for transmitting, to the electronic device, replacement notification information based on the prediction result. The prediction model is generated through machine learning using, as training data, training log information collected from a plurality of training electronic devices having training recording media in a predetermined period until a reference time, and state information indicating operation states of the training recording media, determined at the reference time or later. The log information is obtained by associating time information with information about operation states of the electronic devices and recording media.

Description

[Technical field]

[0001] The present disclosure relates to a method and system for providing a maintenance service for a recording medium included in an electronic device. [Background technology]

[0002] Patent Document 1 discloses a method for diagnosing the remaining life of a storage device in a data center. The method of Patent Document 1 collects operational attributes of a damaged storage device and a record of the time until the damage, and generates a trend model that generates the remaining life of the storage device from the operational attributes and the record of the time until the damage using an algorithm such as machine learning. When a set of operational attributes of a storage device in operation is input, the trend model calculates the remaining life of the storage device. The method of Patent Document 1 provides the calculated remaining life data to the data center administrator so that the administrator can decide to back up or replace the storage device and perform repairs. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2019-87217 A Summary of the Invention

[0004] A method and system for providing a maintenance service that predicts the operating state of a recording medium in an electronic device and enhances user convenience.

[0005] A method according to one aspect of the present disclosure is a method for providing a maintenance service for a recording medium included in an electronic device, the method being executed by one or more arithmetic circuits of a computer, the method including the steps of periodically collecting log information from the electronic device, inputting the log information collected from the electronic device before a current time into a prediction model to predict an operating status of the recording medium after the current time, and transmitting replacement notification information based on a prediction result to the electronic device. receiving replacement request information for the recording medium from a user of the maintenance service; comparing, when the replacement request information is received, the replacement request information with management information for the maintenance service provider to provide a replacement recording medium to the user; and updating the management information with information indicating a history of replacement of the recording medium with the replacement recording medium.The predictive model includes: a) a prediction model is generated by performing machine learning using, as training data, log information collected from a plurality of electronic devices, each having a recording medium, within a predetermined period until a reference time, and status information indicating an operating status of the recording medium possessed by each of the plurality of electronic devices, determined after the reference time; the log information is information that associates time information with information regarding the operating status of the electronic device and the recording medium possessed by the electronic device, and the operating status is abnormal or normal.

[0006] A system according to one aspect of the present disclosure is a system for providing a maintenance service for a recording medium held by an electronic device, the system comprising: an arithmetic circuit; a storage device that stores a prediction model for predicting an operating state of the recording medium; and a communication circuit that receives log information from the electronic device, the arithmetic circuit periodically collects log information from the electronic device, inputs log information collected from the electronic device before a current time into a prediction model to predict an operating state of the recording medium after the current time, and transmits replacement notification information based on the prediction result to the electronic device via the communication circuit; the arithmetic circuit receives replacement request information for the recording medium from a user of the maintenance service via the communication circuit, the storage device stores management information for the maintenance service provider to provide a replacement recording medium to the user, and when the communication circuit receives the replacement request information, the arithmetic circuit compares the replacement request information with the management information and updates the management information with information indicating a history of replacement of the recording medium with the replacement recording medium; The predictive model is generated by machine learning using as training data log information collected from multiple electronic devices, each of which has a recording medium, within a specified period up to a reference time, and status information indicating the operating status of the recording media possessed by each of the multiple electronic devices determined after the reference time, where the log information corresponds to time information and information regarding the operating status of the electronic device and the recording media possessed by the electronic device, and the operating status is abnormal or normal.

[0007] According to the method and system for providing a maintenance service according to the present disclosure, the operating state of a recording medium in an electronic device can be predicted, thereby improving convenience for the user. [Brief description of the drawings]

[0008] [Figure 1] FIG. 1 is a diagram illustrating an example of an overall configuration of a maintenance system according to a first embodiment; [Diagram 2] Diagram for explaining log information [Diagram 3] FIG. 1 is a diagram for explaining an example of the operation of the maintenance system according to the first embodiment; [Figure 4] Flowchart illustrating a process of collecting logs [Diagram 5] 1 is a flowchart illustrating an operation of a maintenance server according to the first embodiment. [Figure 6] FIG. 1 is a diagram for explaining an example of the operation of a prediction model. [Figure 7] FIG. 1 is a diagram for explaining a process for generating a prediction model using training data. [Figure 8] FIG. 11 is a diagram for explaining an example of the operation of the maintenance system according to the second embodiment. [Figure 9] 11 is a flowchart illustrating an operation of a maintenance server according to a second embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0009] Hereinafter, a method and system for providing maintenance service according to an embodiment will be described with reference to the accompanying drawings. In the following description, like components are given the same reference numerals. Duplicate descriptions thereof will be omitted as appropriate.

[0010] (Embodiment 1) In the first embodiment, as an example of executing the method for providing a maintenance service according to the present disclosure, a maintenance system is described in which a maintenance server predicts the operating state of a solid-state drive (SSD) of an electronic device and provides information based on the prediction result. In this specification, the SSD itself may be referred to as a "recording medium." It should be noted that the term "recording medium" may be used in a narrow sense to mean, for example, a memory chip in an SSD or a hard disk in a hard disk drive, but may also be used in a broad sense to mean a storage device in this specification.

[0011] 1. Configuration The configuration of the maintenance system according to the first embodiment will be described with reference to FIG.

[0012] 1 is a block diagram illustrating an example of the overall configuration of a maintenance system 100 according to the present embodiment. The maintenance system 100 includes, for example, one or more electronic devices 10 having an SSD 13, and a maintenance server 20 having a prediction model 27 generated by machine learning. In this system 100, the maintenance server 20 functions as a system (apparatus) for providing maintenance services for the SSD 13. The maintenance server 20 periodically collects log information from the electronic device 10 via a communication network (not shown) such as the Internet, and predicts the operating state of the SSD 13 using the prediction model 27. As an example of providing a maintenance service for the SSD 13, the maintenance server 20 transmits replacement notification information based on the prediction result of the operating state to the electronic device 10.

[0013] 1-1. Maintenance servers and electronic devices Hereinafter, a configuration example of the maintenance server 20, which is an example of a maintenance service providing system in the present system 100, and the electronic device 10 will be described with reference to FIG.

[0014] The maintenance server 20 is configured with an information processing device such as a computer. The maintenance server 20 includes a CPU 21 that performs calculation processing, a communication circuit 22 for communicating with other devices, and a storage device 23 that stores data and computer programs.

[0015] The CPU 21 is an example of an arithmetic circuit of the information processing device in this embodiment. The CPU 21 executes a control program 26 stored in the storage device 23 to realize predetermined functions including training and execution of a prediction model 27. The maintenance server 20 realizes the function of a maintenance service providing system in this embodiment by the CPU 21 executing the control program 26. The arithmetic circuit configured as the CPU 21 in this embodiment may be realized by various processors such as an MPU or a GPU, or may be configured by one or more processors.

[0016] The communication circuit 22 is a communication circuit that performs communication in compliance with a standard such as IEEE802.11, 4G, or 5G. The communication circuit 22 may perform wired communication in compliance with a standard such as Ethernet (registered trademark). The communication circuit 22 is connectable to a communication network such as the Internet. The maintenance server 20 may communicate directly with other devices via the communication circuit 22, or may communicate via an access point.

[0017] The storage device 23 is a storage medium that stores computer programs and data necessary to realize the functions of the maintenance system 100, and stores a control program 26 executed by the CPU 21, a prediction model 27, and various data. The prediction model 27 will be described in detail later. The storage device 23 is configured, for example, by a magnetic storage device such as a hard disk drive (HDD), an optical storage device such as an optical disk drive, or a semiconductor storage device such as an SSD. The storage device 23 may include a temporary storage element configured, for example, by a RAM such as a DRAM or an SRAM, and may function as an internal memory of the CPU 21.

[0018] The database 28 manages log information to be input to the prediction model 27, training data for the prediction model 27, and the like. The training data includes data in which log information collected from multiple electronic devices, each having an SSD as a recording medium, within a predetermined period until a reference time is associated with status information indicating the operating status of each recording medium determined after the reference time. Multiple times may be set as the reference time. The log information is information in which time information is associated with information regarding the operating status of the electronic device 10 and the SSD 13. Details of the log information will be described later.

[0019] The electronic device 10 is configured to be able to communicate with the maintenance server 20 via a communication network. The electronic device 10 is, for example, a personal computer (PC), but may also be a tablet terminal, etc. The electronic device 10 includes a CPU 11, a communication circuit 12, and an SSD 13.

[0020] The CPU 11 is an arithmetic circuit that controls the operation of the electronic device 10. The CPU 11 executes an application program 14 stored in the SSD 13, for example, to realize a predetermined function of the electronic device 10. The CPU 11 may realize the predetermined function by executing the application program 14 provided via a communication network.

[0021] The communication circuit 12 is a communication circuit for communicating with other devices in accordance with a wireless communication standard and / or a wired communication standard. Examples of wireless communication standards are IEEE802.11, 4G, and 5G. An example of a wired communication standard is Ethernet (registered trademark). The communication circuit 12 is connectable to a communication network.

[0022] The SSD 13 is an example of a recording medium included in the electronic device 10 of this embodiment. The SSD 13 has a Self-Monitoring, Analysis and Reporting Technology (SMART) function that generates data on the operating state at a predetermined interval, and stores SMART information 15 generated by the function. In the following, "SMART" will be simply referred to as "SMART". The SSD 13 may temporarily store application programs and data. For example, the SSD 13 temporarily stores log information to be transmitted to the maintenance server 20. The SSD 13 includes a memory chip including a flash memory, a controller chip for controlling the operation, and a board on which each chip is mounted.

[0023] Although not shown, the electronic device 10 also includes a battery that supplies power to the electronic device 10, a board on which the CPU 11 and the like are mounted, a fan used for cooling the inside of the electronic device 10, and a memory that functions as a working area for the CPU 11. The memory is realized by, for example, a dynamic random access memory (DRAM).

[0024] 1-2.Log information The log information collected from the electronic device 10 will be described with reference to Fig. 2. Fig. 2 is a diagram for explaining the log information D1 in the maintenance system 100.

[0025] 2 includes a time indicating time information, and various information such as the device model number of the electronic device 10, the serial number of the SSD 13, device usage information of the electronic device 10, and SMART information 15 of the SSD 13 as information related to the operating status of the electronic device 10 and the SSD 13. The log information D1 manages the time and the various information in association with each other. The time is, for example, the time when the electronic device 10 transmits the various information to the maintenance server 20.

[0026] In the example of Figure 2, log information D1 includes, as device usage information, the temperature measured at a specific location on the electronic device 10, application information related to the application program 14 being operated, the power consumption of the SSD 13, as well as the usage time, manufacturing date, and start date of use of the electronic device 10.

[0027] The temperature in the device usage information is measured at, for example, the CPU 11, memory, battery, front and back of the board, and fan of the electronic device 10. The temperature of the battery may be used as the outside air temperature for the electronic device 10.

[0028] Furthermore, the application information in the device usage information includes the name of the application program 14 executed by the electronic device 10, the number of writes of the application program 14 to the SSD 13, and the like. The number of writes is, for example, the cumulative number of writes, but may be the number obtained by calculating the difference from the number of writes acquired in the previous period for collecting the log information D1. Since the number of writes to the SSD is limited, the future operating state of the SSD 13 can be predicted with higher accuracy by including the number of writes to the SSD 13 in the periodically collected log information D1.

[0029] The maintenance server 20 of this embodiment periodically collects various log information D1 as described above from the electronic device 10. Note that the log information D1 is not limited to the above example, and may be, for example, a part or a combination of the above various information. Furthermore, the log information D1 may include other information in addition to or instead of a part of the above example. For example, the type of flash memory included in the memory chip of the SSD 13, the manufacturer, firmware version information, and the like may be included in the log information D1.

[0030] 2. Operation The operation of the maintenance system 100 configured as above will be described.

[0031] 2-1. Overview of operation An overview of the operation of the maintenance system 100 according to this embodiment to provide a maintenance service will be described with reference to Fig. 3. Fig. 3 is a diagram for explaining an example of the operation of the maintenance system 100 according to this embodiment.

[0032] In the maintenance system 100, the electronic device 10 transmits log information D1 to the maintenance server 20 at a predetermined cycle in association with the transmission time. The predetermined cycle is determined in advance, for example, depending on the type of information in the log information D1. For example, for device usage information, an hourly cycle may be used from the viewpoint of capturing detailed changes over time, and for other information, a daily cycle may be used from the viewpoint of capturing longer-term changes or because frequent changes over time are not expected.

[0033] The maintenance server 20 collects log information D1 from the electronic device 10, for example, at a period synchronized with the period of transmission by the electronic device 10 (S0). After collecting the log information D1 for a predetermined number of periods, the maintenance server 20 inputs the log information D1 collected before the current time into the prediction model 27 to predict the operating state of the SSD 13 of the electronic device 10 after the current time (S2). The predetermined number of periods is set to a sufficiently large number of times, for example, for the period for collecting device usage information of the log information D1, from the viewpoint of using the change over time of the log information D1 to predict the operating state. The predetermined number of periods may vary depending on the type of the SSD 13, for example, the model number of the controller chip of the SSD 13, the firmware version, whether it is a single / multi / triple level cell, etc., and therefore a specific numerical example is omitted.

[0034] In step S2, the prediction model 27 of the maintenance server 20 predicts the operating state of the SSD 13 as "normal", "failure", or "deterioration". Note that "failure" and "deterioration" predicted by the prediction model 27 of this embodiment are examples of the SSD 13 being "abnormal", and the operating state may be predicted as "abnormal" or "normal". Possible factors for predicting an abnormality in the SSD 13 include physical factors such as damage to a memory chip, damage to a controller chip, and damage to a board, as well as logical factors such as damage to a file system in the SSD 13. For example, the operating state of the SSD 13 (recording medium) may be determined to be "abnormal" due to the presence of such physical factors and / or logical factors, and may be determined to be "normal" when no physical or logical factors exist.

[0035] The maintenance server 20 of this embodiment transmits the prediction result obtained by predicting the operating state of the SSD 13 using the prediction model 27 to a user 1 of the maintenance service provided by the maintenance system 100. The user 1 may be, for example, a management department of a business entity that owns the electronic device 10, or may be a user of the electronic device 10. The user 1 receives the prediction result via an information processing terminal of the management department, the electronic device 10, or by means of e-mail or the like, and can check and manage the operating state of the SSD 13 of the electronic device 10.

[0036] The maintenance server 20 judges whether the predicted result of the operating state is "deterioration" or "failure" (i.e., "abnormal") (S4). If the predicted result is deterioration or failure, the maintenance server 20 transmits replacement notification information regarding replacement of the SSD 13 to the electronic device 10 (S5). Based on the replacement notification information, the user 1 can request replacement of the SSD 13 from a business that performs SSD replacement (e.g., a maintenance service provider).

[0037] In this embodiment, in the maintenance system 100 as described above, the maintenance server 20 predicts the operating state of the SSD 13 of the electronic device 10 after the current time (S2), and if the operating state is abnormal, transmits replacement notification information to the electronic device 10 (S5). With such maintenance system 100, it is possible to take action before an abnormality occurs in the SSD 13 and repair becomes difficult, for example, and it is possible to reduce data loss in the SSD 13 and the time when the electronic device 10 cannot be used. This makes it possible to provide a proactive maintenance service that is highly convenient for the user 1.

[0038] 2-2. Maintenance server operation The operation of the maintenance server 20 in the system 100 will be described below with reference to FIGS.

[0039] 4 is a flowchart illustrating a log collection process (S0) for collecting log information D1 from the electronic device 10. This flowchart is started, for example, when log information D1 is received for the first time from the electronic device 10, and each process is executed by the CPU 21 of the maintenance server 20.

[0040] The CPU 21 acquires the log information D1 transmitted from the electronic device 10, for example, via the communication circuit 22 (S01). The CPU 21 stores the acquired log information D1, for example, in the database 28 of the storage device 23 (S02).

[0041] The CPU 21 repeatedly executes the above process at the above-mentioned predetermined period, for example, to periodically collect the log information D1 (S0).

[0042] Fig. 5 is a flowchart illustrating the operation of the maintenance server 20 according to this embodiment. This flowchart corresponds to the process after S0 in Fig. 3, and is started every time the log collection process (S0) is repeated a predetermined number of times. Each process in this flowchart is executed by the CPU 21 of the maintenance server 20, for example.

[0043] First, the CPU 21 acquires the log information D1 collected up to the current time from, for example, the database 28 (S1).

[0044] The CPU 21 inputs the acquired log information D1 to the prediction model 27 to predict the operating state of the SSD 13 after the current time (S2). At this time, data that has been preprocessed in the log information D1, for example, by calculating the difference between two consecutive times, may be input to the prediction model 27. After predicting the operating state (S2), the CPU 21 transmits the prediction result to the user 1, for example, via the communication circuit 22 (S3).

[0045] The CPU 21 determines whether the predicted operating condition is normal or not (S4), and if the operating condition is other than normal (i.e., an abnormality including deterioration or failure) (NO in S4), it transmits replacement notification information to the electronic device 10 (S5).

[0046] The replacement notification information is used, for example, by the maintenance service provider to provide the user 1 with a service for replacing the SSD 13, and includes a maintenance service contract number, a replacement code indicating identification information for the replacement, a shipping destination and date and time of the SSD 13, etc. The contract number is, for example, a number associated with the device number of the electronic device 10.

[0047] After transmitting the replacement notification information (S5), or if the operating state is normal (YES in S4), the CPU 21 ends the process shown in this flowchart. After that, when the log collection process (S0) is executed for a predetermined number of cycles, the process of this flowchart is started again.

[0048] As described above, when the result of predicting the operating state of SSD 13 (S2) using log information D1 collected from electronic device 10 having SSD 13 is other than normal (NO in S4), CPU 21 of maintenance server 20 transmits replacement notification information to electronic device 10 (S5). This makes it easier for user 1 of the maintenance service to carry out procedures related to replacement of SSD 13 using the replacement notification information, for example, thereby improving convenience for user 1.

[0049] 2-3. Operation of predictive model The operation of the prediction model 27 executed in step S2 in FIG. 3 and FIG. 5 will be described with reference to FIG.

[0050] FIG. 6 is a diagram for explaining an example of the operation of the prediction model 27. The prediction model 27 of this embodiment is configured as, for example, an RNN (Recurrent Neural Network). The prediction model 27 of FIG. 6 includes an input layer 27I, one or more hidden layers 27H, an output layer 27O, and a classifier 27C. Although FIG. 6 shows an example having two hidden layers 27H, the number of hidden layers 27H may be one, or may be three or more. Each of the layers 27I, 27H, and 27O has a plurality of units or nodes including parameters.

[0051] When log information D1 collected from electronic device 10 multiple times up to the current time is input to input layer 27I in sequence for each time corresponding to each time, prediction model 27 propagates the information to hidden layer 27H according to the weighting of the input value by the parameter. For each time, prediction model 27 applies weighting by the parameter to the input value at that time and the output value of hidden layer 27H at the previous time, and calculates the output value of hidden layer 27H at that time.

[0052] The prediction model 27 propagates the output value of the hidden layer 27H calculated corresponding to the last time of the input log information D1 to the output layer 27O via parameters, and calculates the probability distribution of the operating state of the SSD 13 after the current time by applying, for example, a softmax function. In the example of Fig. 6, the output layer 27O calculates the probabilities that the operating state is "normal", "fault", and "deterioration". Note that the probability of "abnormality", which is a combination of fault and deterioration, may also be calculated.

[0053] The prediction model 27 outputs, for example, by the classifier 27C, the operating state with the highest probability in the probability distribution output from the output layer 27O as a prediction result of the operating state.

[0054] By the operation of the prediction model 27 as described above, it is possible to predict the operating status of the SSD 13 of the electronic device 10 after the current time based on the log information D1 collected from the electronic device 10 up to the current time, while reflecting the history of changes in the log information D1.

[0055] 2-4.Generating predictive models The process of generating the prediction model 27 that operates as described above by machine learning will be described with reference to FIG.

[0056] 7 is a diagram for explaining a process for using training data to generate a prediction model 27. The process for generating the prediction model 27 is executed by the CPU 21 in the maintenance server 20, for example.

[0057] 7 illustrates training data D2 for the prediction model 27 stored in the database 28 in this embodiment. The training data D2 is collected from a plurality of electronic devices each having an SSD as information associating each of log information D1 (an example of training log information) collected at a plurality of times within a predetermined period from an electronic device (an example of a training electronic device) having an SSD (an example of a training storage medium) as a recording medium with status information that determines the operating status of the SSD at each time. The status information is correct answer information that indicates that the operating status is normal, deteriorated, or broken. The predetermined period is set as a period corresponding to a predetermined number of cycles in the log collection process (S0 in FIG. 3), for example.

[0058] In the example of FIG. 7, the CPU 21 inputs input data I1 for five time periods collected from an electronic device with device part number X1 up to time t5 as a reference time into the prediction model 27, and predicts the operation state of the electronic device one time period ahead from the reference time (i.e., time t6). In the prediction model 27 configured as an RNN as in FIG. 6, for example, cross entropy is used as an error function. The CPU 21 updates the parameters of the prediction model 27 so as to minimize the error between the correct state information O1 and the output of the operation state predicted by the prediction model 27. The CPU 21 trains the prediction model 27 by repeating such processing, for example, for data periodically collected from the electronic device with device part number X1 in the training data D2 and data collected from electronic devices with different device part numbers.

[0059] Through the above-described processing, a prediction model 27 for predicting the operating state of the SSD 13 can be generated using the training data D2. The information on the operating state of the SSD in the log information D1 of the training data D2 may include SMART information 15 obtained by recovering data from a failed SSD. This expands the training data D2, and for example, the prediction accuracy of the prediction model 27 can be improved. In addition, data that has been pre-processed, for example, to take the difference between two consecutive times, may be used to train the prediction model 27.

[0060] 3. Effects etc. As described above, in the maintenance system 100 according to the present embodiment, the CPU 21 of the maintenance server 20, which is an example of a computation circuit of a computer, executes a method for providing a maintenance service for the SSD 13 of the electronic device 10. This method includes a step (S0) of periodically collecting log information D1 from the electronic device 10, a step (S2) of inputting the log information D1 collected from the electronic device before the current time into the prediction model 27 to predict the operating state of the SSD 13 after the current time, and a step (S5) of transmitting replacement notification information based on the prediction result to the electronic device 10. The prediction model 27 is generated by performing machine learning using the log information D1 collected from a plurality of electronic devices each having a recording medium within a predetermined period until a reference time and status information indicating the operating state of the recording medium of each of the plurality of electronic devices determined after the reference time as training data D2. The log information D1 is information in which time information is associated with information on the operating state of the electronic device and the SSD of the electronic device. The operating state is abnormal or normal.

[0061] According to the above-described method for providing a maintenance service, log information D1 collected from the electronic device 10 up to the current time (S0) is input to the prediction model 27 to predict the operating state of the SSD 13 of the electronic device 10 after the current time (S2). As a result, replacement notification information based on the prediction result is transmitted to the electronic device 10 (S5), making it easier for the user 1 of the maintenance service to take measures related to maintenance, such as early replacement of the SSD 13, and improving convenience for the user 1.

[0062] In this embodiment, the abnormality in the operating state is degradation or failure. This allows the type of abnormality to be distinguished in the prediction result, improving the convenience for the user 1. Note that degradation and failure are examples of SSD abnormalities, and other classifications may be used.

[0063] In this embodiment, the method of providing the maintenance service further includes a step (S3) of transmitting the prediction result to the user 1 of the maintenance service. The replacement notification information is transmitted (S5) when the operating status of the prediction result is other than normal (NO in S4). This allows, for example, the user 1 to use the prediction result for managing the SSD 13 even when the operating status is normal.

[0064] In this embodiment, the SSD (Solid State Drive) 13 is an example of a recording medium included in the electronic device 10. Note that the present disclosure may be applied to provide a maintenance service for a recording medium other than an SSD, such as an HDD (Hard Disk Drive) included in the electronic device 10.

[0065] In this embodiment, the information on the operating state in the log information D1 includes usage information of the electronic device and the SSD's SMART information 15. The maintenance server 20 can predict the operating state of the SSD 13 by the prediction model 27 using such log information D1.

[0066] In this embodiment, the usage information includes the temperature measured in the electronic device, the application programs executed in the electronic device, the power consumption of the SSD in the electronic device, and the usage time, manufacturing date, and / or start date of the electronic device. Note that a combination of one or more of the above information may be used as the usage information.

[0067] In this embodiment, the training data D2 includes SMART information 15 obtained by recovering data from a failed SSD in the information on the operating state in the log information D1. The SMART information 15 may be provided by, for example, a company that performs data recovery for SSDs. Note that, although the use of the SMART information 15 can improve the prediction accuracy of the prediction model 27, the training data D2 may not include the SMART information 15 in some cases.

[0068] In this embodiment, the maintenance server 20 is an example of a system for providing a maintenance service for the SSD 13 (recording medium) of the electronic device 10. The maintenance server 20 includes a CPU 21 (arithmetic circuit), a storage device 23 that stores a prediction model 27 for predicting the operating state of the SSD 13, and a communication circuit 22 that receives log information D1 from the electronic device 10. The CPU 21 periodically collects the log information D1 from the electronic device 10 (S0), inputs the log information D1 collected from the electronic device 10 before the current time to the prediction model 27 to predict the operating state of the recording medium after the current time (S2), and transmits replacement notification information based on the prediction result to the electronic device 10 by the communication circuit 22 (S5). The prediction model 27 is generated by performing machine learning using the log information D1 collected from multiple electronic devices each having an SSD within a predetermined period until a reference time and status information indicating the operating state of the SSD of each of the multiple electronic devices determined after the reference time as training data D2. The log information D1 is information in which time information is associated with information on the operating status of the electronic device and the recording medium contained in the electronic device. The operating status is either abnormal or normal.

[0069] In this embodiment, the CPU 21 transmits the prediction result and the exchange notification information via the communication circuit 22 (S3, S5).

[0070] In this embodiment, the above-mentioned method for providing maintenance services is executed by the CPU 21 of the maintenance server 20, but the method may be executed by the maintenance server 20 and another information processing device operating in cooperation with each other. In this case, for example, each step of the method is executed by the arithmetic circuits of the maintenance server 20 and the other information processing device (computer).

[0071] (Embodiment 2) In the first embodiment, a maintenance system 100 in which the maintenance server 20 transmits replacement notification information to the electronic device 10 has been described as an example of a method for providing a maintenance service for the SSD 13 of the electronic device 10. In the second embodiment, in addition to the operations as in the first embodiment, a maintenance system 100 is described that can be used to provide a maintenance service that includes replacement of the SSD 13 of the electronic device 10 that has received the replacement notification information and, in some cases, data recovery of the SSD 13.

[0072] Hereinafter, the maintenance system 100 according to this embodiment will be described, while omitting descriptions of configurations and operations similar to those of the maintenance system 100 according to the first embodiment as appropriate.

[0073] In the maintenance system 100 of this embodiment, the database 28 in the storage device 23 of the maintenance server 20 stores, in addition to the log information D1 and training data, management information for the maintenance service and registration information related to the SSD 13 of the electronic device 10. The management information is information used, for example, by a maintenance service provider to provide a replacement SSD to the user 1, and manages contract information between the provider and the user, and a replacement history indicating whether the SSD 13 has been replaced in the past. The registration information is information including the product number of the SSD 13. Furthermore, the maintenance server 20 of this embodiment includes, in addition to the configuration of the first embodiment (FIG. 1), a display for displaying information and an input device such as a touch panel or a keyboard.

[0074] An overview of the operation of the maintenance system 100 according to this embodiment for providing a maintenance service will be described with reference to FIG.

[0075] 8 is a diagram for explaining an example of the operation of the maintenance system 100 according to the embodiment 2. In this embodiment, the maintenance server 20, like the operation of the embodiment 1, collects log information D1 from the electronic device 10 (S0), predicts the operating state of the SSD 13 of the electronic device 10 (S2), and transmits replacement notification information to the electronic device 10 if the operating state is abnormal (S5).

[0076] In the maintenance system 100 of this embodiment, after transmitting the replacement notification information (S5), the maintenance server 20 receives replacement request information transmitted from the electronic device 10 in response to the replacement notification information. The replacement request information includes, for example, the contract number and replacement code contained in the replacement notification information, and information on the operating state of the SSD 13 at the time of transmitting the replacement request information. In addition, for example, the SSD 13 whose operating state is predicted to be abnormal is shipped from the user 1 to a maintenance service provider.

[0077] The maintenance server 20 compares the management information with the received replacement request information (S7) and presents to the maintenance service provider information such as, for example, whether or not the contract number is present in the management information and the replacement history of the SSD 13. In addition to this information, the maintenance service provider can check the replacement code and the operating state at the time of transmission that are included in the replacement request information, and perform replacement work such as transferring the data of the SSD 13 to a replacement SSD.

[0078] Here, it is assumed that the SSD 13 does not operate during replacement work, such as when the operating state of the SSD 13 is broken when the replacement request information is sent, and the maintenance service provider cannot extract the data of the SSD 13. In this embodiment, the maintenance server 20 transmits registration information of the SSD 13 from the maintenance service provider to the recovery company 3 that recovers data recorded in the SSD. The maintenance service provider then ships the SSD 13 from which data cannot be extracted due to a breakdown or the like to the recovery company 3. As a result, even if preventive replacement based on a prediction of the operating state, such as a sudden breakdown of the SSD 13, cannot be performed, the maintenance service provider can receive the data of the SSD 13 recovered from the recovery company 3 and perform replacement work of the SSD 13.

[0079] After the replacement work, when the maintenance service provider ships a replacement SSD to which the data of SSD 13 has been transferred to user 1, for example, the replacement history related to the replacement work is entered into the maintenance server 20 and the management information is updated (S10).

[0080] According to the maintenance system 100 described above, it becomes easier to efficiently provide the user 1 with a replacement service for SSD 13 using replacement notification information as part of the maintenance service, thereby improving the convenience of the user 1. Also, for the maintenance service provider, it is possible to predict the demand for replacement SSDs based on the predicted operating state results, and reduce excess inventory while securing the inventory of SSDs necessary for providing the service.

[0081] Hereinafter, the operation of the maintenance server 20 in the maintenance system 100 of the second embodiment will be described with reference to FIG.

[0082] Fig. 9 is a flowchart illustrating the operation of the maintenance server 20 according to the second embodiment. In this flowchart, each process is executed by, for example, the CPU 21 of the maintenance server 20, corresponding to the processes after S0 in Fig. 8. The processes of steps S1 to S5 are executed in the same manner as the operation of the maintenance server 20 in the first embodiment (Fig. 5).

[0083] After transmitting the replacement notification information (S5), CPU 21 receives replacement request information from electronic device 10, for example, via communication circuit 22 (S6). CPU 21 compares the received request information with management information previously stored in database 28 (S7) and displays the comparison result on, for example, a display of maintenance server 20.

[0084] The CPU 21 judges whether or not the state of the SSD 13 meets a predetermined condition (S8). For example, the operating state of the SSD 13 at the time of transmitting the replacement request information is determined in advance as the predetermined condition. Alternatively, the predetermined condition may be, for example, that the replacement code in the replacement request information indicates a failure of the SSD 13, or that the replacement code indicates damage to data in the SSD 13. If the former applies, it is considered that the SSD 13 will no longer operate. If the latter applies, the SSD 13 may still be operable, although unstable. The replacement code in the replacement request information may be specified according to the state of the SSD 13.

[0085] If the state of the SSD 13 satisfies a predetermined condition (YES in S8), the CPU 21 transmits registration information related to the SSD 13 to the restoration operator 3 (S9). After that, the data of the SSD 13 restored by the restoration operator 3 may be transmitted to the maintenance server 20.

[0086] After transmitting the registration information (S9), the CPU 21 acquires, from an input device, information indicating the replacement history input by the maintenance service provider who performed the replacement work of the SSD 13, and updates the management information to reflect the replacement history (S10). If the state of the SSD 13 does not meet a predetermined condition (NO in S8), the CPU 21 does not execute the process of step S9 and updates the management information (S10).

[0087] As described above, in the maintenance system 100 according to this embodiment, the method for providing maintenance service for the SSD 13 of the electronic device 10 is executed by the CPU 21 of the maintenance server 20, and further includes a step (S6) of receiving replacement request information for the SSD 13 from the user 1. This makes it possible to provide a maintenance service including replacement of the SSD 13 using, for example, replacement notification information, thereby improving the convenience of the user 1.

[0088] In this embodiment, the method of providing the maintenance service further includes a step (S7) of collating the replacement request information with management information for the maintenance service provider to provide a replacement SSD to the user 1 when the replacement request information is received, and a step (S10) of updating the management information with information indicating the replacement history between the SSD 13 and the replacement SSD. This makes it possible to provide the maintenance service, including the replacement of the SSD 13, more efficiently.

[0089] In this embodiment, the method of providing the maintenance service further includes a step (S9) of transmitting registration information about the SSD 13 to a recovery company 3, which is an example of a company that recovers data from the SSD, when the operating state is abnormal and the state of the SSD 13 meets a predetermined condition (YES in S8). This makes it possible to prevent loss of data from the SSD 13 and to improve the convenience of the user 1, even if the SSD 13 suddenly breaks down.

[0090] In this embodiment, the CPU 21 receives replacement request information (S6) and transmits registration information (S9) via the communication circuit 22. Furthermore, the storage device 23 of the maintenance server 20 stores management information.

[0091] (Other embodiments) As described above, the first and second embodiments have been described as examples of the present disclosure. However, the present disclosure is not limited to these and can be applied to other embodiments. Other embodiments will be exemplified below.

[0092] In the first and second embodiments, the maintenance server 20 transmits the prediction result to the user 1 (S3), and transmits replacement notification information to the electronic device 10 (S5) according to the operating state of the prediction result (S4). In the present embodiment, the maintenance server 20 may transmit only the replacement notification information to the electronic device 10 without transmitting the prediction result to the user 1. Also, the judgment of step S4 may not be particularly executed, and the prediction result may be transmitted as the replacement notification information, for example.

[0093] In the first and second embodiments, the replacement notification information is used for the maintenance service provider to provide a service of replacing the SSD 13. The replacement notification information in this embodiment may be used to provide a replacement SSD to the user 1 so that the user 1 can replace the SSD 13. In this case, the maintenance service provider may periodically back up the data of the SSD 13 via a communication network or the like, and transfer the data of the SSD 13 to a replacement SSD and provide it to the user 1.

[0094] In this embodiment, the method for providing the maintenance service further includes a step of acquiring backup data of the SSD 13 from the electronic device 10, and the replacement SSD is the SSD to which the backup data has been transferred. In this case, for example, the CPU 21 of the maintenance server acquires the backup data of the SSD 13 from the electronic device 10 via the communication circuit 22.

[0095] In the first and second embodiments, the process of generating the prediction model 27 is executed in the maintenance server 20. In the present embodiment, the prediction model 27 generated in an information processing device other than the maintenance server 20 may be provided to the maintenance server 20 via the communication circuit 22.

[0096] In the first and second embodiments, the prediction model 27 configured as an RNN has been described. In the present embodiment, the prediction model 27 may be configured using not only an RNN but also a Convolution Neural Network (CNN), a Support Vector Machine (SVM), a Random Forest, or the like. In addition, a Long Short-Term Memory (LSTM), a Gated Recurrent Unit (GRU), or the like may be used for the prediction model 27.

[0097] In the first and second embodiments, the state information in the training data D2 is correct answer information that determines the operating state as either normal, deteriorated, or broken. In the present embodiment, the state information may correspond to either an operating state of normal or abnormal, including deteriorated and broken. In addition, the training data D2 may include information indicating a factor that determines the operating state as abnormal.

[0098] The present disclosure is not limited to the above-described embodiment, and various modifications are possible. In other words, embodiments obtained by combining technical means appropriately modified by a person skilled in the art are also within the scope of the present disclosure. [Industrial Applicability]

[0099] The present disclosure is applicable to a method and system for providing a maintenance service for a recording medium included in an electronic device. [Explanation of symbols]

[0100] 1 User 3. Restoration company 10 Electronic equipment 20 Maintenance Server 11, 21 CPU 12, 22 Communication circuits 13 SSD 14 Application Programs 15 SMART Information 23 Storage device 26 Control Program 27 Predictive Models 28 Database D1 Log information D2 Training data

Claims

1. A method for providing maintenance service for a recording medium included in an electronic device, the method being executed by an arithmetic circuit of one or more computers, the method comprising: periodically collecting log information from the electronic device; inputting the log information collected from the electronic device prior to a current time into a predictive model; A step of causing the prediction model to predict an operating state of the recording medium after the current time; transmitting replacement notification information based on the prediction result indicating the predicted operating state to the electronic device; receiving information requesting replacement of the recording medium from a user of the maintenance service; a step of, when receiving the replacement request information, comparing the replacement request information with management information for the maintenance service provider to provide the user with a replacement recording medium; updating the management information with information indicating a history of replacement of the recording medium with the replacement recording medium; Including, The prediction model is generated by performing machine learning using training log information collected from a plurality of training electronic devices each having a training recording medium within a predetermined period until a reference time and status information indicating an operating status of the training recording medium determined after the reference time as training data; the log information is information in which time information is associated with information regarding the operating states of the electronic device and the recording medium, The predicted operating condition is abnormal or normal. method.

2. The method of claim 1 , wherein the abnormality in the predicted operating condition is a degradation or a failure.

3. transmitting the prediction result to a user of the maintenance service; The replacement notification information is transmitted when the operation status indicated by the prediction result is other than normal. The method according to claim 1 or 2.

4. The method according to claim 1 , wherein the recording medium is a solid-state drive.

5. The method according to any one of claims 1 to 4, wherein the log information includes usage information of the electronic device and S.M.A.R.T. (Self-Monitoring, Analysis and Reporting Technology) information of the recording medium.

6. 6. The method of claim 5, wherein the usage information includes one or more of the following: a temperature measured in the electronic device, an application program executed in the electronic device, power consumption of the recording medium, a usage time of the electronic device, a manufacturing date of the electronic device, and a date on which the electronic device was first used.

7. The method according to claim 5 or 6, wherein the training log information includes S.M.A.R.T. information obtained by recovering data from a failed recording medium.

8. The method according to any one of claims 1 to 7, further comprising the step of: if the predicted operating status is abnormal and the status of the recording medium meets a predetermined condition, sending registration information regarding the recording medium to a business entity that recovers data on the recording medium.

9. The method further includes acquiring backup data of the recording medium from the electronic device, the replacement recording medium is the recording medium to which the backup data has been transferred; 8. The method according to any one of claims 1 to 7.

10. A system for providing maintenance services for a recording medium included in an electronic device, comprising: An arithmetic circuit; a storage device that stores a prediction model for predicting an operating state of the recording medium; a communication circuit for receiving log information from the electronic device; The arithmetic circuit includes: periodically collecting log information from the electronic device; inputting the log information collected from the electronic device prior to a current time into a predictive model; causing the prediction model to predict an operating state of the recording medium after the current time; transmitting replacement notification information based on the prediction result indicating the predicted operating state to the electronic device by the communication circuit; the arithmetic circuit receives, via the communication circuit, information requesting replacement of the recording medium from a user of the maintenance service; the storage device stores management information for the maintenance service provider to provide a replacement recording medium to the user; The arithmetic circuit includes: When the communication circuit receives the replacement request information, the communication circuit compares the replacement request information with the management information; updating the management information with information indicating a history of replacement of the recording medium with the replacement recording medium; The prediction model is generated by performing machine learning using training log information collected from a plurality of training electronic devices each having a training recording medium within a predetermined period until a reference time and status information indicating an operating status of the training recording medium determined after the reference time as training data; the log information is information in which time information is associated with information regarding the operating states of the electronic device and the recording medium, The predicted operating condition is abnormal or normal. system.

11. The system of claim 10 , wherein the abnormality in the predicted operating condition is a degradation or a failure.

12. The arithmetic circuit, by the communication circuit, Transmitting the prediction result to a user of the maintenance service; The system according to claim 10 or 11, wherein the replacement notification information is transmitted when the operating status indicated by the prediction result is other than normal.

13. The system according to claim 10 , wherein the recording medium is a solid-state drive.

14. The system according to any one of claims 10 to 13, wherein the log information includes usage information of the electronic device and S.M.A.R.T. (Self-Monitoring, Analysis and Reporting Technology) information of the recording medium.

15. 15. The system of claim 14, wherein the usage information includes one or more of a temperature measured at the electronic device, an application program executed at the electronic device, power consumption of the recording medium, a usage time of the electronic device, a manufacturing date of the electronic device, and a date on which the electronic device was first used.

16. 16. The system according to claim 14, wherein the training log information includes S.M.A.R.T. information obtained by recovering data from a failed recording medium.

17. 17. The system according to claim 10, wherein, when the predicted operating status is abnormal and the status of the recording medium meets a predetermined condition, the arithmetic circuit transmits, via the communication circuit, registration information regarding the recording medium to a business entity that recovers data from the recording medium.

18. the arithmetic circuit acquires backup data of the recording medium from the electronic device via the communication circuit; the replacement recording medium is the recording medium to which the backup data has been transferred; A system according to any one of claims 10 to 16.

Images