JP7822191B2 - Maintenance Management Support System - Google Patents

Description

The present invention relates to a maintenance management support system, and in particular to a maintenance management support system for operating equipment.

For example, when an abnormality or malfunction (collectively referred to as "malfunction") occurs in industrial equipment operating in factories or offices, such as inkjet printers that print the manufacturing date and serial number for each product manufactured in a factory, printers that print documents created on office information equipment, and copiers that copy documents (hereinafter referred to as "operating equipment"), maintenance management such as repair or replacement of the malfunctioning part (hereinafter referred to as the malfunctioning part) is required.

This maintenance management involves service personnel visiting the location of operating equipment based on failure data to repair or replace faulty parts. However, with this maintenance management by service personnel, due to a decrease in experienced service personnel in maintenance management work and a lack of technical skills among new service personnel, there are often errors in estimation when identifying faulty parts, and the unnecessary issuance or replacement of parts.

This has led to a variety of issues, including increased time spent responding to customer inquiries, increased costs for replacement parts and unnecessary replacement work, increased calls for maintenance, a reduction in the number of operating devices each service technician is responsible for, and reduced efficiency in maintenance work.

In response to these issues, for example, Japanese Patent Application Laid-Open No. 2004-265159 (Patent Document 1) proposes technology related to a diagnostic function that collects operational information from facility equipment such as air conditioners and searches for the cause of a failure when one occurs, as well as a predictive function that predicts the location of a failure based on alarms and failure information output by the facility equipment.

Japanese Patent Application Laid-Open No. 2004-265159

In order to improve the accuracy of diagnosing faults in operating equipment and predicting where a fault will occur, it is necessary to obtain detailed operational data (driving data) from the equipment. However, limitations such as communication speed and data storage capacity make it difficult to obtain such a large amount of operational data from operating equipment. This creates a new challenge: it is not possible to predict where a fault will occur with sufficient accuracy. This also hinders the identification of failed parts.

The object of the present invention is to provide a maintenance management support system that can accurately predict fault locations and identify faulty parts, even when it is difficult to obtain a large amount of operational data from operating equipment.

The present invention is a maintenance management support system that includes a failure information input unit into which failure information is input, a parts database that stores data correlating replacement parts with the number of times failure events have occurred, a failure knowledge database that stores failure knowledge data correlating failure information with replacement parts, a failing part inference unit that infers the failing part based on the failure knowledge data, and an inferred failing part display unit that displays the inferred failing part.The failing part inference unit infers multiple failing parts associated with the failure information from the failure knowledge data based on the failure information input from the failure information input unit, and further extracts replacement parts corresponding to the inferred failing parts from the parts database and displays the extracted replacement parts in the inferred failing part display unit in descending order of the number of times failure events have occurred.

According to the present invention, even when it is difficult to obtain a large amount of operational data from operating equipment, it is possible to improve the accuracy of predicting the location of a failure and accurately identify the failed part.

1 is a configuration diagram showing a configuration of a maintenance management support system according to an embodiment of the present invention; FIG. 2 is an explanatory diagram showing fault knowledge data stored in a fault knowledge database; 10 is a flowchart illustrating a processing flow of a first example for estimating and displaying a faulty part. 10 is a flowchart illustrating a processing flow of a second example for estimating and displaying a faulty part. FIG. 10 is an explanatory diagram illustrating clustering in event analysis. 10 is a flowchart illustrating a first example of a processing flow for updating fault knowledge data. 10 is a flowchart illustrating a second example of a processing flow for updating fault knowledge data.

Embodiments of the present invention will be described in detail below with reference to the accompanying drawings. However, the present invention is not limited to the following embodiments, and its scope also includes various modifications and applications within the technical concept of the present invention.

Figure 1 shows the configuration of a maintenance management support system that helps locate faults in monitored operating equipment and identify faulty parts.

In Figure 1, the maintenance management support system 10 includes a monitoring and diagnosis database 12 that stores various data including sensor data, operation sequence data, and alarm data obtained from operating equipment 11; a failure knowledge database 13 that stores failure data on operating equipment 11; a maintenance management results database 14 that stores maintenance management work result data; and a parts database 15 that stores parts data on operating equipment and data on newly replaced parts (e.g., the number of parts issued). These databases are connected to each other so that they can exchange information.

The maintenance management system 10 also includes a failure information input unit 19 into which service personnel or the like input customer data, failure data for operating equipment, and operation data for operating equipment. Here, the failure data is primarily used by the first failed part estimation unit (A), described below, to identify the failed part, and the operation data is primarily used by the second failed part estimation unit (B), described below, to identify the failed part.

The maintenance management system 10 also includes a first failing part estimation unit (A) 16 that estimates the failing part from the failure data input by the failure information input unit 19 and the data in each database 12-15, a first estimated failing part display unit (A) 20 that displays the estimation results of the first failing part estimation unit (A) 16, and a first failing part selection unit (A) 21 that selects the failing part from the estimation results of the first failing part estimation unit (A) 16.

The maintenance management system 10 also includes a second faulty part estimation unit (B) 17 that estimates the faulty part from the operating data input by the fault information input unit 19 and the data in each database 12-15, a second estimated faulty part display unit (B) 22 that displays the estimation results of the second faulty part estimation unit (B) 17, and a second faulty part selection unit (B) 23 that selects the faulty part from the estimation results of the second faulty part estimation unit (B) 17.

Furthermore, the maintenance management system 10 includes a maintenance work result input unit 24 for inputting detailed maintenance work result data, a maintenance work report output unit 25 for outputting work reports to be submitted to customers, and a database update unit 18 for updating the data stored in each of the databases 13-15. Furthermore, each of the databases 12-15 and each of the functional configuration units 16-25 are interconnected to enable mutual information exchange.

Each of the databases 12-15 can be realized using a storage device such as a hard disk drive (HDD) or random access memory (RAM). Furthermore, the first faulty part estimation unit (A) 16, second faulty part estimation unit (B) 17, and database update unit 18 can be realized using a computing device such as a central processing unit (CPU) operated by a program stored in read-only memory (ROM) or RAM.

The failure information input unit 19, first presumed failed part display unit (A) 20, first presumed failed part selection unit (A) 21, second presumed failed part display unit (B) 22, second presumed failed part selection unit (B) 23, maintenance work result input unit 24, and maintenance work report output unit 25 are each realized by an input/output device such as a touch panel, a device combining a display and keyboard (e.g., a personal computer, tablet terminal, smartphone, etc.), or a printer, but are not limited to these.

Next, we will explain the functions of each component. Note that the present invention is not limited to any specific operating device, but the following explanation will use an inkjet printer as an example.

Fault data and operation data are uploaded to the monitoring diagnostic database 12 from the operating device 11 to be monitored, in this case an inkjet printer, via a network such as the Internet. Because there are limits to the connection speed of the Internet line and the storage capacity of the monitoring diagnostic database 12, the fault data and operation data constantly uploaded here is data that satisfies the limitations of the connection speed of the Internet line and the capacity of the monitoring diagnostic database 12. Note that operation data includes data stored by the operating device, such as the operating sequence of the operating device, sensor measurements, alerts, and errors.

The fault knowledge database 13 records the structure of the inkjet printer, the relationship between the major components that make up the inkjet printer and functional failures, the relationship between functional failures and failure effects, the relationship between functional failures and failure modes, and the relationship between functional failures and event analysis results as fault knowledge data 30 (see Figure 2). Furthermore, the fault knowledge data 30 does not necessarily contain the overall fault knowledge of the inkjet printer 11 and its relationships; it may also store fault knowledge for each target structure or functional part. Overall structural expansion data for the inkjet printer 11 is also recorded.

Figure 2 shows an example of fault knowledge data 30. Here, an inkjet printer is used as an example, and some of the main components of the ink circulation unit are particularly shown. The fault knowledge data 30 contains data relating to partial knowledge ID 31, main components 32, functional failure 33, failure effect 34, failure mode 35, and event analysis results 36, which are associated with each other on each line.

The partial knowledge ID 31 represents a name, and associated data is arranged in each row accordingly. For main components 32, the component ID and component name are associated as a pair. For functional failures 33, the failure ID of the functional failure and the physical state at the time the functional failure occurred are associated as a pair. As described above, each piece of data is associated based on the functional failure 33.

Failure impact 34 associates a pair of a failure ID with the impact that occurs in response to the failure. It is also possible to associate the degree of impact on safety and operational aspects. Failure mode 35 associates a pair of a failure mode ID with the failure mode that corresponds to the cause of the failure.

The event analysis results correspond to the operational data log, and for example, for past functional failures 33, weighting values are recorded according to the number of times an event recorded in the operational data occurred. An event is the smallest unit of operation sequence, sensor measurement value, alert, error, etc. saved as operational data. The event analysis results (e.g., weighting values) are updated each time. Here, the event analysis results are calculated by calculating the number of failure events that occur each time the inkjet printer is operated or per unit time, extracting characteristic events that occur frequently, assigning weighting values according to the number of occurrences, and storing them in a database.

The fault knowledge data 30 does not need to be created in its entirety at once; it may be created by an expert with specialized knowledge, such as an inkjet printer service technician or designer, or it may be created in parts due to additional data added in a version upgrade. The status of the knowledge data associated with its creation is recorded as a partial knowledge ID 31.

The fault knowledge database 13 also stores data showing the structural expansion of the operating equipment 11, such as the main components 32 in Figure 2. Here, the structural expansion is shown using an inkjet printer as an example, and only some of the main components 32 in the ink circulation section are shown.

Each row of the fault knowledge data 30 in Figure 2 (indicated by numbers 1 to 14 in Figure 2) represents a combination of a major component 32, functional failure 33, fault effect 34, failure mode 35, and event analysis result 36, which are treated as the smallest unit of fault knowledge.

In the fault knowledge database 13, the relationships between each piece of data recorded in the fault knowledge data 30 describe, as relationships between each data ID, which major component 32's functional failure 33 is the cause of the fault effect 34.

Next, a method for estimating a faulty part using the maintenance management support system 10 will be explained using Figures 3 and 4. Figure 3 shows an example of estimating a faulty part using the first faulty part estimation unit (A) 16, and Figure 4 shows an example of estimating a faulty part using the second faulty part estimation unit (B) 17.

In Figure 3, the service technician first interviews the customer to obtain failure information. This failure information can be viewed as general information about the actual phenomenon that occurred due to the failure. Note that the failure information may not be just one piece of information, but may include multiple different pieces of information.

Then, based on the failure information heard, the service technician enters customer information and failure information (for example, failure information such as "unable to print," "unstable printing," or "ink leaking," as shown in Figure 2) into the failure information input unit 9 in "Step S10." Once this failure information is entered, the first failed part estimation unit (A) 16 uses data from the failure knowledge database 13 to estimate the failed part.

Specifically, in "Step S11", the fault knowledge data 30 (see Figure 2) in the fault knowledge database 13 is used to select multiple fault effects 34 that are estimated to be highly likely to be the cause of the fault "unable to print" based on the fault information entered in "Step S10", such as "unable to print". Of course, the same applies to fault information such as "unstable printing" and "ink leakage".

Once the failure effect 34 is selected, in step S12, multiple functional failures 33 that are estimated to have a high probability of causing the failure "unable to print" are selected from the functional failures 33 in the failure knowledge data 30. At this time, if a functional failure 33 was input in step S10, the input functional failure is selected.

Once a functional failure 33 is selected, in step S13, multiple major components 32 that are highly related to the multiple functional failures 33 can be narrowed down as candidate faulty components. In this way, it is possible to estimate multiple faulty components related to the input general fault information. In this way, through the steps up to this point, multiple major components 32 that may be experiencing a fault can be extracted as faulty components.

Next, in step S14, the first faulty part selection unit (A) selects main parts 32 corresponding to the narrowed-down multiple faulty parts from the main parts 32 associated with the number of times that a fault has occurred and stored in the parts database 15, and the first suspected faulty part display unit (A) 20 displays the selected multiple faulty parts in descending order of the number of times that a fault has occurred.

In other words, the parts database 15 stores a list of the main parts that make up the inkjet printer, and each main part is associated with a number of parts issued. The number of parts issued is the number of replacement parts that are replaced due to a malfunction, and can therefore be considered the number of malfunctions. In this specification, the number of malfunction events is defined to include events that can be considered the number of malfunctions (for example, number of parts issued, frequency of malfunction, etc.).

For this reason, when a major part 32 corresponding to the faulty part extracted in step S13 from the parts database 15 is selected, the number of issues (which can be considered the number of times the fault has occurred) is linked, making it possible to display these in order of the most issued number.

In step S15, major parts corresponding to the narrowed-down multiple faulty parts are selected from the major parts associated with the number of issues stored in the maintenance result database 14, and the multiple faulty parts that have been read out are displayed in the first suspected faulty part display unit (A) 20 in descending order of the number of faults that have occurred.

Here, the display in "Step S14" and the display in "Step S15" may be displayed by dividing the display screen of the first suspected faulty part display unit (A) 20 into two, or may be switched by a switching input from the fault information input unit 19.

This method allows service personnel to identify and dispatch multiple suitable replacement parts from the faulty parts displayed in descending order of the number of times they have failed, thereby improving the accuracy of predicting the location of the failure and enabling accurate identification of the faulty part.

Next, we will explain an example of narrowing down the faulty parts using the event analysis results by the second faulty part estimation unit (B) 17.

In Figure 4, the service technician first interviews the customer to obtain fault information. Note that this fault information may include not just one piece of information, but multiple different pieces of information.

Then, based on the malfunction information heard, the service technician enters customer information and malfunction information (e.g., malfunction information such as "unable to print," "unstable printing," or "ink leaking") into the malfunction information input unit 9 in "Step S20."

Next, in step S21, the fault knowledge data 30 in the fault knowledge database 13 is used to select multiple fault effects 34 that are estimated to be highly likely to be the cause of the fault "unable to print" based on the fault information entered in step S20, such as "unable to print." Of course, the same applies to fault information such as "unstable printing" and "ink leakage."

Next, in step S22, from among the functional faults 33 in the fault knowledge data 30, multiple functional faults 33 that are estimated to have a high probability of causing the fault "unable to print" are selected. At this time, if a functional fault 33 was input in step S10, the input functional fault is selected.

Once functional failure 33 is selected, in step S23, operation log data from the inkjet printer and other data are input and an event analysis is performed. Event analysis calculates the number of failure events that occur for each operation of the operating equipment or per unit time from the operation log data, extracts characteristic events with a high number of failure events, and assigns a weighting value according to the number of failure events. This assigned weighting value is compared with the event analysis results in the failure knowledge database 30 as a reference value, and the closer it is to the weighting value of the current event analysis (the smaller the deviation), the more likely it is that a failure has occurred in that major component. Therefore, the closer this reference value is to the current weighting value (the smaller the deviation), the more likely it is that a failure has occurred.

Once the event analysis is complete, in step S24, the event analysis results 36 stored in the fault knowledge data 30 in the fault knowledge database 13 are matched with the current event analysis results 36, and multiple event analysis results 36 with small deviations in the weighting values of both are selected.

Next, in step S25, multiple faulty parts are narrowed down based on the AND condition of the failure impact 34 obtained in step S21 and/or the functional failure 32 obtained in step S22, and the event analysis result 36 obtained in step S24. This improves the accuracy of estimating the faulty part.

Once multiple faulty parts have been narrowed down, in step S26, the narrowed down faulty parts are sorted in order of smallest deviation of the weighting values, which are the matching results of the event analysis results 36, and displayed in the second suspected faulty part display section (B) 22.

Furthermore, similar to "Step S14" in Figure 3, in "Step S27", major parts corresponding to the narrowed-down multiple faulty parts are selected from the major parts stored in the parts database 15 and linked with the number of fault occurrences, and the selected multiple faulty parts are displayed in the second suspected faulty part display unit (B) 22 in descending order of the number of fault occurrences.

Similarly, in "Step S28", major parts corresponding to the narrowed-down multiple faulty parts are selected from the major parts associated with the number of fault occurrences stored in the maintenance result database 14, and the selected multiple faulty parts are displayed in the second suspected faulty part display section (B) 22 in descending order of the number of fault occurrences.

It is also possible to omit steps S21 and S22 and identify the faulty part through the event analysis in step S23.

The display in "Step S26", the display in "Step S27", and the display in "Step S28" may be displayed by dividing the display screen of the second suspected faulty part display unit (B) into three parts, or may be switched by switching input from the fault information input unit 19.

The event analysis in "Step S23" can be performed by clustering. Now, with the maintenance management support system 10 in operation, multiple event analysis results 36 may be obtained as a result of event analysis for a failure with the same main component 32, functional failure 33, and failure effect 34 recorded in the failure knowledge data 30. Therefore, even for the same failure, the weighting values of the elements Event 1 to Event n in the event analysis result 36 may differ.

As shown in Figure 5, for multiple failures with the same main component 32, functional failure 33, and failure effect 34, event elements Even 1 to Event n or several event elements Event are grouped to form a cluster CL, which is then used as classified failure information CL-1 to CL-n. By comparing the event analysis results 36 obtained by the event analysis of the second failure component estimation unit (B) 17 with this clustered failure information CL and narrowing down the main components 32 that are highly related to the event analysis results, it is possible to more accurately estimate the failed component.

Once multiple faulty components are identified using the methods shown in Figures 3 and 4, the service technician will perform the following tasks based on this information.

The service technician selects several faulty parts from those displayed in the first presumed faulty part display area (A) 20 or the second presumed faulty part display area (B) 22, and dispatches them to the customer site.

Even if operation data cannot be obtained, the service technician can select several faulty parts from those displayed in the first presumed faulty part display section (A) 20, have them delivered, and be dispatched to the customer site. At the customer site, the service technician can operate the second faulty part estimation section (B) by extracting operation data from the operating equipment and inputting it into the maintenance management support system 10. This allows the service technician to check the faulty parts displayed in the second presumed faulty part display section (B) 22 and select the most appropriate faulty part from the delivered faulty parts.

Furthermore, at the customer site, by investigating the parts of the actual inkjet printer in the order of the faulty parts displayed in the suspected faulty part display area (A) 10 or the suspected faulty part display area (B) 12, the faulty part can be efficiently identified, thereby shortening the maintenance work time.

After the maintenance work is completed, the service technician enters the details of the maintenance work and the parts used using the maintenance work result input unit 24, which automatically creates a maintenance work report and outputs it from the maintenance work report output unit 25. This eliminates the need for the service technician to return to their office to create a maintenance work report, improving work efficiency.

Here, a first example of updating the fault knowledge data 30 by a service technician will be explained based on Figure 6.

When the maintenance management support system 10 is in operation, it may happen that functional failures 33 or failure effects 34 for the same failure of a major component 32 recorded in the failure knowledge data 30 do not exist in the failure knowledge data 30. In this case, it is important to update the failure knowledge data 30 by adding these new data to the data update unit 18.

Therefore, as shown in Figure 6, the data update unit 18 is used to input failure information in "Step S30." If there is no failure information entered in the functional failure 33 or failure effect 34 at this time, "Step S31" and/or "Step S32" are executed.

If there is no failure corresponding to failure effect 34, in "Step S31", failure effect 34 is extracted from the failure knowledge database 13 and the pre-set item "Other" is selected. Similarly, if there is no failure corresponding to functional failure 33, in "Step S32", functional failure 33 is extracted from the failure knowledge database 13 and the pre-set item "Other" is selected.

When "Other" is selected for the failure effect 33 or functional failure 34, in "Step S33," functional failure data, failure effect data, failure mode data, and event analysis result data corresponding to the failure effect 33 and functional failure 34 at that time are extracted. Furthermore, in "Step S34," each extracted data is assigned a new failure number (No. 15 in Figure 2), and added to and updated in the failure knowledge data 30.

Next, a second example of updating the fault knowledge data 30 by a service technician will be explained with reference to Figure 7.

After the maintenance work is completed, the service technician enters the details of the maintenance work and the replacement parts into the maintenance work result input unit 14. At this time, the major parts 32 presented as faulty parts by the maintenance management support system 10 may differ from the major parts 32 entered by the service technician. In this case, the data update unit 18 must add these to the fault knowledge data 30 and update it.

Therefore, as shown in Figure 7, the data update unit 18 is used to input work result information for the maintenance management work results in "Step S40." Next, in "Step S41," if the major part used (replaced) is the same as the estimated major part, the process skips to "END" and ends. On the other hand, if the major part is different from the estimated major part, "Step S42" is executed.

Next, in "Step S42," if the input functional failure 33 is the same as the functional failure 33 in the fault knowledge data 30, the process exits to "END" and ends. On the other hand, if the functional failure 33 is different from the functional failure 33 in the fault knowledge data 30, "Step S43" is executed.

Next, in "Step S43," if the input failure effect 34 is the same as the failure effect 34 in the failure knowledge data 30, the process exits to "END" and terminates. On the other hand, if the failure effect 34 is different from the failure effect 34 in the failure knowledge data 30, "Step S44" and/or "Step S45" is executed.

If "Yes" is returned in steps S41 to S43 and the input failure effect 34 does not exist in the failure knowledge data 30, then in step S44 the failure effect 34 is extracted from the failure knowledge database 13 and the pre-set "Other" item is selected. Similarly, if the input functional failure 33 does not exist in the failure knowledge data 30, then in step S45 the functional failure 33 is extracted from the fault knowledge database 13 and the pre-set "Other" item is selected.

When "Other" is selected for the failure effect 33 or functional failure 34, in "Step S46," functional failure data, failure effect data, failure mode data, and event analysis result data corresponding to the failure effect 33 and functional failure 34 at that time are extracted. Furthermore, in "Step S47," each extracted data is assigned a new failure number (No. 15 in Figure 2), and added to and updated in the failure knowledge data 30.

As described above, the present invention is a maintenance management support system comprising a failure information input unit into which failure information is input, a parts database that stores data correlating replacement parts with the number of times failure events have occurred, a failure knowledge database that stores failure knowledge data correlating failure information with replacement parts, a failing part inference unit that infers a failing part based on the failure knowledge data, and an inferred failing part display unit that displays the inferred failing parts.The failing part inference unit infers multiple failing parts associated with the failure information from the failure knowledge data based on the failure information input from the failure information input unit, and further extracts replacement parts corresponding to the inferred failing parts from the parts database and displays the extracted replacement parts in the inferred failing part display unit in descending order of the number of times failure events have occurred.

This makes it possible to improve the accuracy of predicting fault locations and accurately identify faulty parts, even when it is difficult to obtain a large amount of operational data from operating equipment.

The present invention is not limited to the several embodiments described above, and includes various modifications. The above embodiments have been described in detail to clearly explain the present invention, and are not necessarily limited to those including all of the configurations described. Furthermore, it is possible to replace part of the configuration of one embodiment with the configuration of another embodiment, and it is also possible to add the configuration of another embodiment to the configuration of one embodiment. It is also possible to add, delete, or replace other configurations with respect to the configuration of each embodiment.

10...Maintenance management support system, 11...Monitored equipment, 12...Monitoring diagnosis database, 13...Failure knowledge database, 14...Maintenance result database, 15...Component database, 16...First failing component estimation unit (A), 17...Second failing component estimation unit (B), 18...Database update unit, 19...Failure information input unit, 20...First estimated failing component display unit (A), 21...First failing component selection unit (A), 22...Second estimated failing component display unit (B), 23...Second failing component selection unit (B), 24...Maintenance work result input unit, 25...Maintenance work report output unit, 30...Failure knowledge data, 31...Partial knowledge ID, 32...Major component, 33...Functional failure, 34...Failure effect, 35...Failure mode, 36, Event analysis results.

Claims (1)

a failure information input unit to which failure information of an operating device to be monitored is input, a parts database storing data correlating a failed part with the number of times a failure event has occurred, a failure knowledge database storing failure knowledge data correlating the failure information with the failed part, a failed part inferring unit (A) for inferring the failed part using the failure knowledge database, a failed part selecting unit (A) for extracting the number of times a failure event has occurred of the failed part using the parts database, an inferred failed part display unit (A) for displaying the inferred failed part, and a database updating unit for updating the failure knowledge data,
the faulty part estimation unit (A) estimates, based on the fault information input from the fault information input unit, a plurality of faulty parts associated with the fault information from the fault knowledge database;
the faulty part selection unit (A) extracts, from the part database, the fault event occurrence counts corresponding to the plurality of faulty parts estimated by the faulty part estimation unit (A),
a maintenance management support system characterized in that the estimated faulty part display unit (A) uses the number of times the fault event has occurred extracted by the faulty part selection unit (A) to display the estimated faulty parts in descending order of the number of times the fault event has occurred.