Model generation device, determination device, and determination method

Abstract

To accurately determine whether a work machine is to be inspected, from error codes, on the basis of past results of inspecting the work machine.SOLUTION: A model generation device executes an inspection information acquisition step, an error acquisition step, and a generation step. The inspection information acquisition step acquires, with respect to inspection of a work machine, the time at which the inspection was executed and the presence / absence of a failure detected according to the result of the inspection. The error acquisition step acquires a combination of error codes output earlier than the time at which the inspection was executed, from a control device of the work machine. The control device of the work machine outputs an error code representing the type of the error when the state of the work machine is not a desired state. The generation step generates a prediction model for predicting the presence / absence of a failure from the combination of error codes, on the basis of the presence / absence of a failure and the combination of error codes output earlier than the time at which the inspection was executed.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present disclosure relates to a model generation device, a determination device, and a determination method. [Background technology]

[0002] A control device mounted on a work machine monitors the state of the work machine via sensors and the like, and issues an error notification if the state of the work machine is not the desired state. The error notification is accompanied by an error code indicating the content of the error, and a maintenance person reads the error code to consider whether or not an inspection of the work machine is necessary. Patent Document 1 discloses a technology for determining the degree of abnormality of a work machine from the error code. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2002-180502 A Summary of the Invention [Problem to be solved by the invention]

[0004] An error is output when a sensor measurement value exceeds a specified range or when control such as communication fails. Therefore, the occurrence of an error does not necessarily indicate a breakdown in the work machine. This makes it difficult for a maintenance person to determine whether or not an inspection of the work machine is necessary from the error code. In addition, the accuracy of the determination may vary depending on the experience of the maintenance person. An object of the present disclosure is to provide a model generation device, a determination device, and a determination method for accurately determining whether or not a work machine needs to be inspected from an error code, based on the work machine's past inspection results and the occurrence of error codes. [Means for solving the problem]

[0005] According to one aspect of the present invention, a model generation device includes a processor, and the processor executes the following steps for acquiring, for an inspection of a work machine, the time at which the inspection was performed and the presence or absence of a fault revealed by the results of the inspection; an error acquisition step for acquiring a combination of error codes output at a time prior to the time at which the inspection was performed from a control device of the work machine that outputs an error code indicating a type of error when the state of the work machine is not in a desired state; and a generation step for generating a predictive model for predicting the presence or absence of a fault from the combination of error codes, based on the presence or absence of the fault and the combination of error codes output at a time prior to the time at which the inspection was performed. Effect of the Invention

[0006] According to the above aspect, by using a prediction model, it is possible to accurately determine whether or not an inspection of the work machine is required from the error code based on past inspection results of the work machine. [Brief description of the drawings]

[0007] [Figure 1] 1 is a schematic diagram showing a configuration of a trigger notification system according to a first embodiment. [Diagram 2] FIG. 2 is a schematic block diagram showing the configuration of a control device according to the first embodiment. [Diagram 3] FIG. 2 is a schematic block diagram showing a configuration of a management server according to the first embodiment. [Figure 4] 10 is a flowchart showing a model generation process by the management server according to the first embodiment. [Diagram 5] FIG. 2 is a diagram illustrating an example of a prediction model according to the first embodiment. [Figure 6] 10 is a flowchart showing a trigger generation process by the management server according to the first embodiment. [Figure 7] 13 is a flowchart showing a model generation process by a management server according to a second embodiment. [Figure 8] FIG. 11 is a diagram illustrating an example of a prediction model according to the second embodiment. [Figure 9] 10 is a flowchart showing a trigger generation process by the management server according to the first embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0008] First Embodiment Configuration of trigger notification system 1 Hereinafter, the embodiments will be described in detail with reference to the drawings. Fig. 1 is a schematic diagram showing the configuration of a trigger notification system 1 according to a first embodiment. The trigger notification system 1 determines whether or not inspection of a work machine 100 is necessary, and presents to a maintenance person those work machines 100 for which it has been determined that inspection is necessary. In other words, the trigger notification system 1 notifies the maintenance person of a trigger for the inspection of the work machine 100. The maintenance person can recognize the appropriate timing for inspecting the work machine 100 by the trigger notified from the trigger notification system 1.

[0009] The trigger notification system 1 includes a work machine 100, a management server 300, and a maintenance person terminal 500. The work machine 100, the management server 300, and the maintenance person terminal 500 are communicatively connected via a network.

[0010] For example, if the work machine 100 is a hydraulic excavator, it operates at a construction site and performs work such as excavating earth and sand. The work machine 100 shown in FIG. 1 is a hydraulic excavator, but in other embodiments, it may be another work machine. Examples of the work machine 100 include a bulldozer, a dump truck, a forklift, a wheel loader, and a motor grader. The work machine 100 monitors the state of its own vehicle, and transmits error data to the management server 300 if the state is not the desired state.

[0011] The management server 300 determines the work machines 100 to be inspected based on the error data of the work machines 100 . The management server 300 stores error data of the work machine 100 and a history of inspection information of the work machine 100. The management server 300 generates a prediction model that predicts the presence or absence of a malfunction of the work machine 100 based on the error data and the history of the inspection information. The management server 300 determines the work machine 100 to be inspected based on the prediction model, and transmits trigger information indicating the work machine 100 to be inspected to the maintenance operator terminal 500. In other words, the management server 300 is an example of a model generating device, and the management server 300 is an example of a determination device.

[0012] The maintenance person terminal 500 displays or prints the trigger information generated by the management server 300. The maintenance person terminal 500 also accepts input of inspection result data indicating the results of an inspection of the work machine 100 by a maintenance person, and transmits it to the management server 300. The inspection result data includes the date and time of the inspection, the presence or absence of a malfunction, and the details of the inspection and the malfunction.

[0013] Configuration of the work machine 100 The work machine 100 includes a running body 110, a rotating body 120, a work implement 130, and a control device 140. The running body 110 supports the working machine 100 so that the working machine 100 can run. The running body 110 is, for example, a pair of left and right caterpillar tracks. The rotating body 120 is supported by the running body 110 so as to be capable of rotating about a rotation center. The working machine 130 is supported on the front part of the revolving body 120 so as to be drivable in the vertical direction. The working machine 130 is hydraulically driven. Here, the part of the revolving body 120 to which the working machine 130 is attached is referred to as the front part. Furthermore, with respect to the revolving body 120, the part opposite the front part is referred to as the rear part, the left part as the left part, and the right part as the right part. The control device 140 controls the traveling body 110, the rotating body 120, and the work machine 130 based on the operation of the operator. The control device 140 is provided inside the driver's cab, for example. The control device 140 is an example of a fall risk presentation device.

[0014] The work machine 100 is equipped with a plurality of sensors 141 for detecting the work state of the work machine 100. Specifically, the work machine 100 is equipped with a service meter, an engine speed sensor, an exhaust temperature sensor, an oil temperature sensor, an oil pressure sensor, an inclination sensor, and the like.

[0015] FIG. 2 is a schematic block diagram showing the configuration of the control device 140 according to the first embodiment. The control device 140 is a computer that includes a processor 210, a main memory 230, a storage 250, and an interface 270.

[0016] The storage 250 is a non-transitory tangible storage medium. Examples of the storage 250 include a magnetic disk, an optical disk, a magneto-optical disk, and a semiconductor memory. The storage 250 may be an internal medium directly connected to the bus of the control device 140, or may be an external medium connected to the control device 140 via the interface 270 or a communication line. The storage 250 stores a program for controlling the work machine 100.

[0017] The program may be for realizing a part of the functions to be performed by the control device 140. For example, the program may be for realizing the functions by combining with other programs already stored in the storage 250 or by combining with other programs implemented in other devices. In other embodiments, the control device 140 may include a custom LSI (Large Scale Integrated Circuit) such as a PLD (Programmable Logic Device) in addition to or instead of the above configuration. Examples of PLDs include PAL (Programmable Array Logic), GAL (Generic Array Logic), CPLD (Complex Programmable Logic Device), and FPGA (Field Programmable Gate Array). In this case, some or all of the functions realized by the processor may be realized by the integrated circuit.

[0018] The processor 210 executes a program to function as an acquisition unit 211, a determination unit 212, and a transmission unit 213.

[0019] The acquisition unit 211 acquires measurement values ​​from each of the multiple sensors 141 . The determination unit 212 determines whether an error exists based on the measurement value acquired by the acquisition unit 211. The determination unit 212 determines that an error exists when the measurement value exceeds a predetermined range. The determination unit 212 also determines that an error exists when the acquisition unit 211 cannot acquire a measurement value from the sensor 141. The determination unit 212 identifies an error code that indicates the content of the error.

[0020] The transmission unit 213 transmits error data indicating the error detected by the determination unit 212 to the management server 300. The error data includes identification information of the work machine 100, the time the error occurred, and an error code. The transmission unit 213 transmits the error data to the management server 300 at a predetermined transmission timing. The transmission unit 213 may transmit the error data by batch processing, or may transmit the error data to the management server 300 in real time. When transmitting error data by batch processing, the acquisition unit 211 records the error data in the storage 250, and the transmission unit 213 transmits this to the management server 300. Note that in order to reduce the amount of communication, the transmission unit 213 may compress and transmit this history data as necessary.

[0021] In addition, in order to reduce the amount of communication, even if an error related to the same error code occurs multiple times within a predetermined notification suppression time, the transmission unit 213 may not transmit error data for errors other than the first error. Specifically, when an error with the same error code occurs within a predetermined notification suppression time (e.g., 20 hours) from the detection time of the error for which the previous error data was transmitted, the transmission unit 213 does not notify the error. For example, a case will be described in which errors related to tilt occur at 0:00, 10:00, and 22:00. The error detected at 0:00 is the first error. Therefore, the transmission unit 213 transmits error data related to the error detected at 0:00 to the management server 300. The error detected at 10:00 is an error related to the same error code as the error detected at 0:00, and 20 hours, which is the notification suppression time, have not elapsed since the detection. Therefore, the transmission unit 213 does not transmit error data related to the error detected at 10:00. The error detected at 22:00 is an error related to the same error code as the error detected at 0:00, but 20 hours, which is the notification suppression time, have elapsed since the detection. Therefore, the transmission unit 213 transmits to the management server 300 the error data related to the error detected at 22:00.

[0022] Configuration of Management Server 300 FIG. 3 is a schematic block diagram showing the configuration of the management server 300 according to the first embodiment. The management server 300 is a computer that includes a processor 310 , a main memory 330 , a storage 350 , and an interface 370 .

[0023] The storage 350 is a non-transitory tangible storage medium. Examples of the storage 350 include a magnetic disk, an optical disk, a magneto-optical disk, and a semiconductor memory. The storage 350 may be an internal medium directly connected to the bus of the management server 300, or an external medium connected to the management server 300 via the interface 370 or a communication line. The storage 350 stores a program for generating an incident report.

[0024] The program may be for realizing a part of the functions to be performed by the management server 300. For example, the program may be for realizing the functions by combining with other programs already stored in the storage 350 or by combining with other programs implemented in other devices. In other embodiments, the management server 300 may include a custom LSI in addition to or instead of the above configuration. In this case, some or all of the functions realized by the processor 310 may be realized by the integrated circuit. In other embodiments, the management server 300 may be realized by a computer virtualized on one or more computers.

[0025] The processor 310 executes a program to function as a receiving unit 311, a model generating unit 312, a determining unit 313, and a transmitting unit 314.

[0026] The receiving unit 311 receives error data from the work machine 100. The receiving unit 311 records the received error data in the storage 350. The receiving unit 311 receives the inspection result data from the maintenance person terminal 500. The receiving unit 311 records the received inspection result data in the storage 350. The maintenance person performs a direct inspection of the work machine 100 to determine the presence or absence of a malfunction, and generates the inspection result data by inputting the determination result to the maintenance person terminal 500. Note that the direct inspection is not necessarily performed when an error code occurs, but is performed in response to some trigger, such as when the maintenance person receives a request from the site supervisor. The receiving unit 311 receives error data that occurred immediately before the maintenance person performed the direct inspection, and inspection result data that indicates the result of the direct inspection.

[0027] The model generation unit 312 generates a prediction model for predicting the presence or absence of a malfunction of the work machine 100 from a combination of error codes output by the work machine 100, based on the past error data and inspection result data recorded in the storage 350. Specifically, the model generation unit 312 generates a prediction model which is a table indicating the failure rate and P value for each combination pattern of error codes by cross-tabulating the error codes and the presence or absence of malfunctions. The P value is the probability that a certain combination pattern of error codes will have a failure rate in a probability density distribution which indicates the failure rate when it is assumed that a malfunction occurs regardless of the combination of error codes. The model generation unit 312 records the created prediction model in the storage 350.

[0028] The determination unit 313 determines whether or not an inspection of the work machine 100 is necessary based on error data received from the work machine 100 within the most recent target period (e.g., the most recent 30 hours) and the prediction model recorded in the storage 350. The target period is a period of a predetermined time in the past starting from the present, and is at least a period longer than the notification suppression time of the same error by the work machine 100. For example, when the determination unit 313 receives error data related to error codes A, B, and C during the target period, it reads out the failure rate and P value associated with the combination of error codes A, B, and C. The failure rate is a value in which the number of inspections that occurred when the combination of error codes appeared in the past is used as the denominator and the number of times a failure was found is used as the numerator for the combination of error codes. The determination unit 313 determines that an inspection is necessary when the read failure rate is equal to or greater than a predetermined threshold value (e.g., 70%) and the read P value satisfies a predetermined significance level (e.g., 5%). On the other hand, when the read failure rate is less than a predetermined threshold value, or when the read P value does not meet a predetermined significance level, the determination unit 313 determines that an inspection is unnecessary. The fact that the P value does not meet the significance level indicates that the combination of error codes is highly likely not to be related to a failure. An example of a combination of error codes A, B, and C that can occur in parallel is error code A being an oil temperature abnormality, error code B being a fuel injector abnormality, and error code C being a coolant noise abnormality.

[0029] The transmission unit 314 transmits a list indicating the work machines 100 that require inspection among the multiple work machines 100 as trigger data to the maintenance terminal 500 based on the determination result of the determination unit 313. The list includes the necessity of inspection as well as the failure rate and P value read out by the determination unit 313. The maintenance person can recognize the work machines 100 that require inspection by the trigger data displayed on the maintenance terminal 500.

[0030] <<Processing of Management Server 300>> The receiving unit 311 of the management server 300 records the error data in the storage 350 every time it receives error data from the work machine 100, and records the inspection result data in the storage 350 every time it receives inspection result data from the maintenance terminal 500.

[0031] 4 is a flowchart showing a model generation process by the management server 300 according to the first embodiment. Upon receiving an instruction to generate a prediction model from an administrator, the management server 300 executes the process shown in FIG.

[0032] First, the model generation unit 312 of the management server 300 selects each piece of inspection result data recorded in the storage 350 (step S1), and executes the processes of steps S2 to S7 for each piece of inspection result data. The model generation unit 312 identifies the work machine 100 to be inspected and the date and time of the inspection from the selected inspection result data (step S2). The model generation unit 312 reads out from the storage 350 error data related to the identified work machine 100, which occurred during a target period ending at the identified date and time (for example, a period from 30 hours before the identified date and time to the identified date and time) (step S3).

[0033] The model generation unit 312 identifies a combination of error codes that occurred during the target period from the read error data (step S4). The model generation unit 312 identifies all error code patterns that can be configured with the error codes related to the identified combination (step S5). For example, when the model generation unit 312 identifies the error code combination {A, B, C} in step S4, it identifies the error code patterns {A}, {B}, {C}, {A, B}, {B, C}, {A, C}, {A, B, C} in step S5. The model generation unit 312 updates the failure rate of each identified error code pattern based on the presence or absence of a failure indicated by the inspection result data selected in step S1 (step S6). That is, the model generation unit 312 reads the error occurrence status for each inspection indicated by the inspection result data selected in step S1, and recalculates and updates the failure rate for each error code pattern corresponding to the error occurrence status. The failure rate is calculated by using the number of inspections performed during the period when a combination of error codes related to an error code pattern appears as the denominator and the number of failures found as the numerator. Specifically, for the failure rate of each error code pattern, the model generation unit 312 adds 1 to the denominator and numerator when the inspection result data indicates that there is a failure, and adds 1 to the denominator when the inspection result data indicates that there is no failure. In addition, the model generation unit 312 updates the P value for each error code pattern identified in step S5 (step S7). The failure rate and P value are calculated by forming a population for each model of the work machine 100. The failure rate and P value may be calculated by forming a population not only for each model, but also for each classification such as vehicle rank or region.

[0034] The model generating unit 312 generates a prediction model, which is a table showing the failure rate and P value for each error code pattern, based on the calculation results from steps S1 to S7 (step S8). The model generating unit 312 records the generated prediction model in the storage 350 (step S9). FIG. 5 is a diagram showing an example of a prediction model according to the first embodiment. FIG. 5 shows an example in which there are three types of error codes, A, B, and C. Note that the prediction model may be obtained by extracting only those combinations of all error codes whose failure rate is equal to or exceeds a predetermined threshold value (e.g., 70%) and whose read P value satisfies a predetermined significance level (e.g., 5%).

[0035] Fig. 6 is a flowchart showing a trigger generation process by the management server 300 according to the first embodiment. The management server 300 executes the process shown in Fig. 6 at predetermined intervals.

[0036] First, the determination unit 313 of the management server 300 selects the multiple work machines 100 one by one (step S21), and executes the processes of steps S22 to S27 for each work machine 100. The determination unit 313 reads out from the storage 350 error data related to the selected work machine 100, which error data occurred during a target period ending at the current time (for example, the period from 30 hours before the current time to the current time) (step S22).

[0037] The determination unit 313 identifies a combination of error codes that occurred during the target period from the read error data (step S23). The determination unit 313 reads out a failure rate and a P value associated with the combination of error codes identified in step S23 (step S24). The determination unit 313 determines whether the read failure rate is equal to or greater than a predetermined threshold (e.g., 70%) and whether the read P value is equal to or greater than a predetermined significance level (e.g., 5%) (step S25). If the failure rate is equal to or greater than the threshold and the P value is equal to or greater than the significance level (step S25: YES), the determination unit 313 determines that an inspection is required for the work machine 100 selected in step S21 (step S26). On the other hand, if the failure rate is less than the threshold or the P value is less than the significance level (step S25: NO), the determination unit 313 determines that an inspection is not required for the work machine 100 selected in step S21 (step S27).

[0038] The transmission unit 314 generates, as trigger data, a list indicating the work machines 100 that require inspection among the multiple work machines 100 based on the determination results from steps S21 to S27 (step S28). The trigger data includes the failure rate and P value read out by the determination unit 313 in addition to whether or not inspection is necessary. The transmission unit 314 transmits the trigger data to the maintenance person terminal 500 (step S29). The transmission unit 314 transmits the trigger data to, for example, the terminal of a maintenance person in charge of maintenance service of the work machines 100 that require inspection. Note that in other embodiments, the transmission unit 314 may transmit the trigger data to the owner of the work machine 100, or may transmit the trigger data to a monitor provided on the work machine 100 to notify the operator.

[0039] Actions and Effects In this way, the management server 300 according to the first embodiment executes, for the inspection of the work machine 100, an inspection information acquisition step of acquiring the time the inspection was performed and the presence or absence of a malfunction revealed by the inspection results, an error acquisition step of acquiring a combination of error codes output from the control device 140 of the work machine 100, and a generation step of generating a prediction model for predicting the presence or absence of a malfunction from the combination of error codes based on the time the inspection was performed, the presence or absence of a malfunction, and the combination of error codes. This enables the management server 300 to generate a prediction model that predicts the presence or absence of a malfunction based on the history of past error data and inspection results. In other words, the management server 300 can generate a prediction model for accurately determining whether or not a work machine inspection is required from an error code based on past inspection results of the work machine 100.

[0040] Furthermore, the management server 300 according to the first embodiment executes an error acquisition step of acquiring a combination of error codes output from the control device 140 of the work machine 100, and an output step of outputting information relating to the necessity of inspection of the work machine 100 based on the combination of the prediction model and the error code. This enables the management server 300 to accurately determine whether or not inspection of the work machine 100 is necessary from the error code, based on past inspection results of the work machine 100.

[0041] Second Embodiment The management server 300 according to the first embodiment generates a prediction model by cross-tabulation based on the inspection result data. In contrast, the management server 300 according to the second embodiment generates a prediction model by cross-tabulation based on the error codes of the work machines 100.

[0042] The configuration of the trigger notification system 1 according to the second embodiment is similar to that of the first embodiment. On the other hand, the second embodiment differs from the first embodiment in the method of generating a prediction model by the management server 300.

[0043] <<Processing of Management Server 300>> The receiving unit 311 of the management server 300 records the error data in the storage 350 every time it receives error data from the work machine 100, and records the inspection result data in the storage 350 every time it receives inspection result data from the maintenance terminal 500.

[0044] Fig. 7 is a flowchart showing a model generation process by the management server 300 according to the second embodiment. Upon receiving an instruction to generate a prediction model from an administrator, the management server 300 executes the process shown in Fig. 7.

[0045] First, the model generation unit 312 of the management server 300 selects the work machines 100 one by one (step S41), and executes the processing of steps S42 to S48 for each work machine 100. The model generation unit 312 identifies the state of the work machine 100 represented by a combination of error codes detected in a target period ending at a plurality of times obtained by dividing the period from the time related to the oldest error data recorded in the storage 350 to the current time into predetermined time intervals (step S42). The state of the work machine 100 is represented as a vector representing a pattern of the combination of error codes. For example, if there are five types of error codes, A, B, C, D, and E, and error data related to error codes A, B, and E exists in the target period, the state of the work machine 100 is represented as {1, 1, 0, 0, 1}.

[0046] The model generation unit 312 selects, one by one, the periods in which the work state identified in step S42 does not change (step S43), and executes the processing of the following steps S44 to S48 for each period. For example, if the state of the work machine 100 is {1, 0, 0, 0, 0} at time t1, time t2, and time t3, the state of the work machine 100 is {1, 1, 0, 0, 0} at time t4 and time t5, and the state of the work machine 100 is {1, 1, 0, 0, 1} at time t6, time t7, and time t8, the model generation unit 312 selects the period from time t1 to time t3, the period from time t4 to time t5, and the period from time t6 to time t8, respectively.

[0047] The model generation unit 312 judges whether or not there is inspection result data related to the working machine 100 selected in step S41, which was generated during the period selected in step S43 (step S44). If there is inspection result data generated during the period selected in step S43 (step S44: YES), the model generation unit 312 adds 1 to the number of inspection results for the state related to the period (step S45). The model generation unit 312 acquires all of the inspection result data related to the working machine 100 selected in step S41, which was generated during the period selected in step S43 (step S46). The model generation unit 312 judges whether or not there is one or more pieces of inspection result data in which a fault has been detected in the acquired inspection result data (step S47). If there is one or more pieces of inspection result data in which a fault has been detected (step S47: YES), the model generation unit 312 adds 1 to the number of fault results for the state related to the period (step S48).

[0048] Incidentally, if there is no inspection result data related to the work machine 100 selected in step S41 that was generated during the period selected in step S43 (step S44: NO), the model generation unit 312 does not add the number of inspection records related to the condition of the work machine 100 related to that period. Furthermore, if no faults have been detected in any of the inspection result data related to the work machine 100 selected in step S41 that was generated during the period selected in step S43 (step S47: NO), the model generation unit 312 does not add up the number of actual faults related to the state of the work machine 100 related to that period.

[0049] The model generating unit 312 calculates the failure rate by dividing the number of actual failures by the number of actual inspections for each combination pattern of error codes (step S49). Then, the model generating unit 312 generates a prediction model that associates at least the failure rate and the number of actual inspections for each combination pattern of error codes (step S50). The model generating unit 312 records the generated prediction model in the storage 350 (step S51). FIG. 8 is a diagram showing an example of a prediction model according to the second embodiment. FIG. 8 shows an example in which there are three types of error codes.

[0050] Fig. 9 is a flowchart showing a trigger generation process by the management server 300 according to the first embodiment. The management server 300 executes the process shown in Fig. 9 at predetermined intervals.

[0051] First, the determination unit 313 of the management server 300 selects the multiple work machines 100 one by one (step S61), and executes the processes of steps S62 to S69 for each work machine 100. The determination unit 313 reads out from the storage 350 error data related to the selected work machine 100, which error data occurred during a target period ending at the current time (for example, the period from 30 hours before the current time to the current time) (step S62).

[0052] The determination unit 313 identifies a combination of error codes that occurred during the target period from the read error data (step S63). The determination unit 313 reads out the failure rate and the number of inspections associated with the combination of error codes identified in step S63 (step S64). The determination unit 313 determines whether the read failure rate is equal to or greater than a predetermined threshold value (e.g., 70%) (step S65). If the failure rate is equal to or greater than the threshold value (step S65: YES), the determination unit 313 determines that an inspection is required for the work machine 100 selected in step S61 (step S66). Furthermore, the determination unit 313 determines whether the read number of inspection records is less than a predetermined threshold (e.g., 5 records) (step S67). If the number of inspection records is less than the threshold (step S67: YES), the determination unit 313 determines that it is preferable to inspect the work machine 100 selected in step S61 for data collection (step S68). On the other hand, if the failure rate is less than the threshold value and the number of inspection records is equal to or greater than the threshold value (step S67: NO), the determination unit 313 determines that inspection is not required for the working machine 100 selected in step S21 (step S69).

[0053] The transmission unit 314 generates, as trigger data (step S70), a list indicating the work machines 100 that require inspection among the multiple work machines 100, based on the determination results from steps S61 to S69. The trigger data includes, in addition to whether inspection is required, the failure rate and the number of inspection records read out by the determination unit 313. The transmission unit 314 transmits the trigger data to the maintenance terminal 500 (step S71).

[0054] Actions and Effects In this way, the management server 300 according to the second embodiment determines whether or not at least one fault has been found in each period in which the combination of error codes does not change, and generates a prediction model by statistical processing based on the number of periods in which at least one fault has been found. In other words, even if multiple inspections are performed in a period in which the combination of error codes does not change, the management server 300 counts the inspection record as 1. This makes it possible for the management server 300 to prevent bias in the weighting of the inspection response even if multiple inspections are performed in one response to a fault. In addition, the management server 300 specifies the combination of error codes as the state of the work machine 100, and therefore outputs one failure rate regardless of the number of error codes generated by the work machine 100.

[0055] Furthermore, the management server 300 according to the second embodiment executes a specifying step of specifying a combination of error codes with few inspection results. This makes it possible to prompt the maintenance person to collect data on the state of the work machine 100 where the accuracy of the prediction model is low due to a small amount of data.

[0056] Other Embodiments Although several embodiments have been described in detail above with reference to the drawings, the specific configuration is not limited to the above, and various design changes and the like are possible. That is, in other embodiments, the order of the above-mentioned processes may be changed as appropriate. Also, some of the processes may be executed in parallel. The management server 300 according to the above-described embodiment may be configured by a single computer, or may be configured by distributing the configuration of the management server 300 among multiple computers, and the multiple computers may cooperate with each other to function as the management server 300. For example, in another embodiment, the function of the management server 300 as a model generating device that generates a predictive model and the function as a determination device that determines the necessity of an inspection based on the predictive model may be implemented in separate computers.

[0057] The management server 300 according to the embodiment described above generates a prediction model by a statistical process called cross tabulation, but is not limited to this. For example, the management server 300 according to another embodiment may generate a prediction model by a statistical process using a machine learning model such as a neural network model. For example, the management server 300 generates a combination of the state of the work machine 100 (combination of error codes) at the inspection time and the presence or absence of a malfunction as a learning data set for each piece of inspection performance data. Then, the management server 300 trains parameters of an estimation model, which is a machine learning model, in the learning data set so that the state of the work machine 100 is used as an input and the malfunction rate is used as an output. By such statistical process, the management server 300 can generate a prediction model by a machine learning model.

[0058] The prediction model according to the above embodiment represents the relationship between the failure rate and the P value for each error code pattern as shown in FIG. 5, but is not limited to this in other embodiments. For example, the prediction model according to other embodiments may store the failure rate and the P value in addition to the failure rate and the P value in association with the content of the failure. Also, for example, the prediction model according to other embodiments may store the failure rate and the P value for each type of failure. [Explanation of symbols]

[0059] 1...Trigger notification system 100...Work machine 110...Traveling body 120...Rotating body 130...Work machine 140...Control device 141...Sensor 210...Processor 211...Acquisition unit 212...Determination unit 213...Transmission unit 230...Main memory 250...Storage 270...Interface 300...Management server 310...Processor 311...Receiver 312...Model generation unit 313...Determination unit 314...Transmission unit 330...Main memory 350...Storage 370...Interface 500...Maintenance person terminal

Claims

1. A processor is provided. The processor, an inspection information acquisition step of acquiring, for an inspection of a work machine, the time when the inspection was performed and the presence or absence of a malfunction found as a result of the inspection; an error acquisition step of acquiring a combination of error codes outputted at a time before the inspection was performed from a control device of the work machine that outputs an error code indicating a type of error when the state of the work machine is not a desired state; a generation step of generating a prediction model for predicting the presence or absence of a fault from a combination of error codes based on the presence or absence of the fault and a combination of error codes outputted at a time before the execution of the inspection; Run Model generation device.

2. In the generating step, the prediction model is generated by statistically processing the presence or absence of the fault and a combination of the error code outputted at a time before the execution of the test for each of a plurality of tests. The model generating device according to claim 1 .

3. In the generating step, the prediction model is generated by statistically processing a pattern of a combination of the error codes outputted at a time before the execution of the test and the presence or absence of the fault in each of the tests. The model generating device according to claim 1 .

4. In the generating step, for each period in which the combination of error codes does not change, it is determined whether or not at least one fault has been found in the period, and the prediction model is generated by performing statistical processing based on the number of periods in which the at least one fault has been found.

4. The model generating device of claim 3.

5. The processor, A step of identifying a combination of the error codes with a low inspection result is further performed.

5. The model generating device according to claim 3 or 4.

6. A processor is provided. The processor, an error acquisition step of acquiring a combination of the error codes outputted at a time before the current time from a control device of the work machine which outputs an error code indicating a type of error when the state of the work machine is not a desired state; an output step of outputting information regarding the necessity of inspection of the work machine based on a prediction model for predicting the presence or absence of a fault from the combination of error codes, the prediction model being generated based on the time at which the inspection of the work machine was carried out, the presence or absence of a fault found as a result of the inspection, and the combination of the error codes output at a time prior to the time at which the inspection was carried out, and based on the combination of the error codes output from a control device of the work machine at a time prior to the current time; A determination device that executes the above.

7. an error acquisition step of acquiring a combination of the error codes outputted at a time before the current time from a control device of the work machine which outputs an error code indicating a type of error when the state of the work machine is not a desired state; an output step of outputting information regarding the necessity of inspection of the work machine based on a prediction model for predicting the presence or absence of a fault from the combination of error codes, the prediction model being generated based on the time at which the inspection of the work machine was carried out, the presence or absence of a fault found by the results of the inspection, and the combination of the error codes output at a time prior to the time at which the inspection was carried out, and based on the combination of the error codes output from a control device of the work machine at a time prior to the current time; The determination method includes:

Images