JPH08221295A - Fault supporting device - Google Patents

Abstract

PURPOSE: To lighten the burden of a worker, and also to generate a correct work report by recording automatically work history, and generating automatically the work report on the basis of the work history. CONSTITUTION: In the case where a monitoring device 2 detects a fault, when the worker selects 'dealing with fault', a supporting device 1 for dealing with fault displays an operation picture for dealing with fault to present an operation item or the hypothesis of the cause of the fault and display a restoring operation item. When in response to this, the worker menu-selects and inputs the result of confirming operation or the result of restoring operation based on the cause hypothesis on the operation picture for dealing with fault, the cause hypothesis definition of the pattern knowledge base 4b of the result of the confirming operation or the result of the restoring operation at present is searched, and the new confirming operation or the fault cause hypothesis is displayed on the operation picture for dealing with fault. Then, when a finish button is clicked, the operation history in the course is written out to a database, and after executing the learning of knowledge, a fault processing history generating part 1g executes the generation processing of a fault record, that is, the generation of a fault record document.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a failure support device for dynamically instructing a cause of failure and a recovery support method based on accumulated failure support knowledge.

[0002]

2. Description of the Related Art Conventionally, a maintenance system of this kind receives a fault information from a diagnostic device which has detected a fault when a fault occurs and applies a fault analysis procedure to analyze the fault (for example, Japanese Patent Laid-Open No. 3-278240). There is a method of inferring a fault and determining a countermeasure (for example, Japanese Patent Laid-Open No. 4-150641) based on the fault handling knowledge of the maintenance system. However, regarding failure analysis and troubleshooting methods based on analysis results, a simple method of repeatedly searching not only test / operation results but also recovery operation results to find the cause and location of the failure, and the failure response to be applied There is no system that has a method to automatically acquire knowledge of fault handling from the result of operation history (work report) entered by a worker with simple menu selection when he / she does not have knowledge. Further, even in the failure analysis method itself, there is no mechanism for positively utilizing the target device configuration information when a failure occurs to select a failure caused by the same cause.

[0003]

In the conventional system,
For failures for which failure analysis procedures and failure handling methods have not been defined, system experts and failure support experts work together, and regularly review and select the failure handling procedures.
There is a drawback that the maintenance work load in operation is large and it is difficult to be cost-effective as a system, and it is an issue to reduce the maintenance work load in operation.

The present invention is intended to solve the above-mentioned problems, and in order to facilitate operation, it is easy to search for the cause of failure and the location of failure by repeatedly using the result of recovery operation and the result of test / operation. The purpose is to enable quick and effective response in the event of a failure by determining the failure caused by different methods and the same cause. It also aims to automatically create a work report from the operations performed by the worker.

[0005]

According to a first aspect of the present invention, there is provided a failure support device including a database having information on devices constituting a system, a knowledge base having a hypothesis for estimating an operation for a failure and its cause, and a failure. A fault information collection unit that receives information, a cause hypothesis generation unit that searches the database and the knowledge base to provide a worker with an operation and a causal hypothesis for the fault, and a worker that operates information presented by the causal hypothesis generation unit. And a means for automatically creating a work report from the operation history and the database in the failure treatment history creation section.

A failure support device according to a second aspect of the present invention includes a database having information on devices constituting the system,
A knowledge base having a hypothesis for estimating a failure operation and its cause, and a failure information collecting unit for receiving failure information,
A cause hypothesis generator that searches the database and the knowledge base to provide an operator with an operation and a cause hypothesis for the failure, and a failure that stores the operation history of the operator for the operation information presented by the cause hypothesis generator in the database A treatment history creation unit and an automatic learning processing unit that holds the number of times of the same operation performed by the worker in accordance with the operation provided by the cause hypothesis generation unit are provided. It is designed such that a large number of operations are preferentially instructed.

In the fault support apparatus according to the third aspect of the present invention, the fault treatment history creating section is provided with operation item adding means for adding an operation item, and the operation item adding means automatically stores the added item as an operation history in a database. It is designed to be registered as a causal hypothesis of the knowledge base as well as registered in.

In the fault support apparatus according to the fourth aspect of the present invention, the fault treatment history creating section is provided with a restoration action item means for adding a restoration action item, and the added restoration action item is automatically registered in the database. It is a thing.

The fault support apparatus according to the fifth aspect of the present invention is provided with a fault confirmation unit caused by the same cause for judging and grouping faults caused by the same cause with respect to a plurality of faults that have occurred at substantially the same time. It is a thing.

In the fault support apparatus according to the sixth aspect of the present invention, the fault checking unit caused by the same cause divides the same cause into a parent-child relationship and a parallel relationship.

[0011]

In the obstacle support device according to the first invention,
The fault information display unit displays the information collected by the fault information collection unit on the dialogue device, and the cause hypothesis generation unit sequentially refers to the database and the knowledge base for the fault selected by the operator for the displayed fault. The fault history creating unit automatically creates an operation history as the operator operates according to the presented contents, and creates a work report at the end of the operation.

The fault support apparatus according to the second invention is characterized by being provided with an automatic learning processing section for holding how many times the same operation has been used in the past, and the cause hypothesis generating section has been used most frequently in the past. The performed operation is preferentially presented to the worker.

The fault support apparatus according to the third invention is characterized in that the fault treatment history creating section is provided with means for adding operation items. The added items are registered in the database as an operation history and a knowledge base is added. It is also registered as the cause hypothesis of.

The failure support apparatus according to the fourth aspect is characterized in that a means for adding a recovery action item is provided in the failure action history creating section, and the added recovery action item is automatically stored in the database as an operation history. It is also registered as a causal hypothesis of the knowledge base.

The fault support apparatus according to the fifth aspect of the present invention is characterized in that a fault confirmation unit for checking the causal relationship between a plurality of faults which occur at the same time and which is caused by the same cause is provided.
The fault confirmation unit caused by the same cause extracts, from a plurality of faults, faults caused by the same cause, groups them, and reports them to the fault display unit.

In the fault support apparatus according to the sixth aspect of the present invention, the fault confirmation unit caused by the same cause reports the same cause by classifying it into a parent-child relationship and a parallel relationship.

[0017]

【Example】

Example 1. FIG. 1 is a diagram showing the configuration of an embodiment of the present invention. In the figure, reference numeral 1 denotes a failure support device, and this failure support device 1 is provided with a failure via a communication interface 3 from a plurality of external monitoring devices 2 for monitoring line network components such as a circuit switch and a modem. A failure information collection unit 1a that receives information, a plurality of collected failure information items are checked to correlate a failure caused by the same cause, and a failure confirmation unit 1b caused by the same cause for grouping is performed. The operation hypothesis generation unit 1c that generates an operation (test) to be performed and a fault cause hypothesis, and when a new causal relationship occurs due to the fault response performed by the worker, it is extracted from the operation history of the database described later and organized, and knowledge is obtained. Automatic learning processing unit 1 added to the base operation priority / hypothesis definition
d, a fault information display unit 1e that displays fault information, hypotheses, and operation histories that have been grouped, and a result input by the operator through the fault-corresponding operation screen of the interactive device 1h; and a worker is displayed on the interactive device 1h. When the required failure information item is selected (input) from the function selection menu bar, the information related to the selected failure is displayed on the related information display screen of the interactive apparatus 1h from the related information generation unit 1f and the operation performed by the operator. It is composed of a failure treatment history creation unit 1g that creates an operation history and a failure record.

Reference numeral 4 denotes a storage device which is a part of the fault support device 1. Various data is stored in the storage device 4, and device information, user information, line carrier information, line connection information, and fault record are stored. Knowledge base of operation priority / hypothesis definition, operation item definition format, hypothesis definition format, etc. for giving priority to operation instructions to the operator and a database 4a such as manuals and operation history and presenting hypotheses for the cause of failure 4b is stored.

FIG. 2 is an example of a display screen used in this embodiment, which is an example of a screen on a computer with which a worker interacts.
In this case, it consists of 5 screens. The display screen 5 is
A trouble list screen 5a that is automatically updated every time a trouble occurs, related network information and device information for an operator to take measures against the trouble, search and display / printing of users affected by the trouble, and automatic communication of various communications Function selection menu bar 5b for selecting FAX transmission, automatic creation of work report, failure response screen, screen for adding operation history, failure response operation screen 5c for the operator to operate when a failure occurs, On the screen displaying the work history, the contents of this screen are reflected in the fault record, and the operation history screen 5d and faults used for generating new fault handling knowledge and changing the usage frequency of existing fault handling knowledge are generated. It has a related information display screen 5e for displaying information about the failure when the user does.
Of these screens, the failure list screen 5a and the function selection menu bar 5b are screens that are always displayed. FIG.
4 and FIG. 4 are flowcharts for explaining the operation of this embodiment. Further, FIG. 5 is an example of a screen displayed as a failure information list.

The operation of this embodiment will be described below with reference to the drawings. [Overall Processing] When the monitoring device 2 detects a failure and sends information about the failure via the communication interface 3, the processing flow shown in FIG. 3 is activated. Upon receiving the failure information in step S12, the failure information collection unit 1a sends the information to the failure confirmation unit 1c caused by the same cause. The fault confirming unit 1c caused by the same cause checks in step S13 whether or not there is a fault caused by the same cause, and sends the result to the fault information display unit 1e together with the fault information. The failure information display unit 1e displays a failure list as shown in FIG. 5 on the failure list screen 5a in step S14.

The operator selects a fault to be processed from the fault list shown in FIG. 5 by double-clicking the mouse, for example (step S32). By this processing, the related information display screen 5e is opened (step S33), and device information and user information related to the designated failure are displayed. At the same time, the operation history screen 5d is opened, and the failure name and its occurrence time are displayed on the screen. At this stage, the operator selects items such as “device information” and “fault affected user” on the function selection menu bar by clicking a mouse button or the like, and selects information different from the information already displayed (for example, It is also possible to display the relevant information display screen 5e on the connection information of the devices constituting the circuit network, the configuration devices, the information on the user who is affected by the failure, etc.) and use this information to handle the failure. In the example, a case will be described in which "fault handling" is selected from the function selection menu bar 5b (step S34) and the action for the fault is taken.

By the operator selecting "handling a fault",
The failure handling support device 1 displays a failure handling operation screen 5c displaying operation (test) items, a hypothesis of a cause of failure, and a recovery operation item. On the other hand, the operator displays the confirmation operation result and the recovery operation result based on the cause hypothesis on the failure response operation screen 5c.
When the menu selection is input with (step S37), it is confirmed whether or not the operation end button on the failure handling operation screen 5c is pressed (step S39), and if the end button is not pressed, the current confirmation operation result or The cause hypothesis definition of the knowledge base 4b is searched for in the pattern of the recovery operation result (step S311), and a new confirmation operation or failure cause hypothesis is displayed again on the failure handling operation screen 5c (step S36).
This process is continued until the worker finishes the handling of the fault or the restoration operation result of the fault cause hypothesis becomes "recovery". When the end button is clicked, the operation history in the middle is written in the database 5a (step S3A), and after the knowledge is learned (step S3C), the failure treatment history creation unit 1
g performs a trouble record generation process (step S3D), that is, a trouble record document is created. It should be noted that in the course of the above operation, the fault treatment history creating section 1g is operated by the operator and the operation history screen 5 is displayed.
The operation history displayed in d is automatically collected and added to the operation history of the database 4a (steps S35, S
38, S3A).

[Creation of Operation History and Trouble Record] The operation history and trouble record generated in the course of the above operation will be described. FIG. 6 is an example of the operation history screen, and FIG. 7 is an example of the trouble record document. The operation history screen is displayed when the operator designates a specific failure on the failure list screen 5a as described above, and in the example of FIG.
The case where the line disconnection that occurred in the minute is selected is shown. This screen is updated each time the above operation is performed, and in this example, it is recorded that the line re-arrangement was instructed at 15:00 and the line was restored at 16:00. FIG. 7 is an example of the above-mentioned trouble record document created by the trouble handling history creating unit 1g in step S3D of FIG. In the fault record number, the faulty device address of FIG. 5 is embedded in the faulty device number, the alarm column of FIG. 5 is similarly embedded in the content column, and the content of the operation history screen is embedded in the response progress column. Is automatically embedded with the summary (keyword) of the correspondence progress column. This trouble record is used as a work report of the worker.

[Learning of Knowledge] Learning of knowledge working in the course of the above-mentioned failure handling processing will be described below. FIG. 8 is a diagram showing the flow of knowledge learning processing (step S3C), and FIG. 9 is an example of the operation priority / hypothesis definition knowledge base in the knowledge base 4b. In the example of FIG. 9, the operations for disconnecting the line include (1) switch port return test,
(2) Switch card exchange test and (3) Modem loopback test are registered. In addition, hypothetical item (cause of failure)
And as recovery operations, (a) modem card exchange, (b)
Circuit switching is registered. Further, the priority order of the operation for each operation is registered, and in the (a) column of FIG. 9, (1)
Switch port wrapping test is the highest priority.
Also, in the column (b), the exchange card exchange test of (2) has the highest priority.

As the knowledge learning (step S3C), the automatic learning processing unit 1d obtains the above-mentioned operation history obtained by the worker performing the failure handling processing as a daily work (FIG. 6).
The operation result and the recovery operation result are automatically taken out from, the operation result / recovery operation result pattern is newly generated, and the operation priority / hypothesis definition shown in FIG. 9 is updated. If the corresponding operation / recovery operation result pattern has already been defined in the operation priority / hypothesis definition of the knowledge base 4b, the operation item selected in the operation selection (step S37) on the failure handling operation screen described later or The frequency of use of the recovery operation item is updated (step S42), and if there is no corresponding pattern, the pattern of the operation / recovery operation result and the failure cause hypothesis / recovery operation are automatically retrieved from the operation history of the database 4a to obtain the knowledge base 4b. A new pattern is newly registered (step S43) with the weight "1" (the number of times of use 1) in the operation priority order / hypothesis definition of.

By performing this weighting process, the weight of the operation frequently appearing in each failure and the recovery operation increases as the operation is performed, and the priority of the operation items presented as a result becomes an appropriate order for each failure. For example, looking at the column (a) in FIG. 9, the operation priority is (1) the weight of the switch port loopback test is "3", which is the highest priority. Therefore, in FIG. 10 of the failure handling operation screen, which will be described later, the port return test of the exchange is displayed as the highest priority at the top. The operation priority / hypothesis definition (FIG. 9) of the knowledge base 4b specifies the next operation priority and the cause hypothesis from the operation status (operation result definition) and the pattern of the recovery operation result, and gives priority based on the weight (number of times of use). The following operation items and recovery operation items are presented on the screen in order.

[Fault Handling Operation] Next, a fault handling operation will be described. 10 to 12 are diagrams showing the transition of the failure handling operation screen. As shown in the figure, this screen has a test operation input / output area in the upper part and an operation input / output area for the failure cause hypothesis in the lower part, and displays the operation status at the present time. For example, in the case of "line disconnection" failure, the failure handling operation screen of FIG. 10 appears. In the operation item column, the highest priority "switch return test of exchange" is displayed at the top. When the worker instructs the department in charge of this operation and the operation result is a failure, the worker inputs “failure” in the operation result column of the failure handling operation screen by selecting the menu. As a result, the use frequency of the operation item selected at this time in the operation priority / hypothesis definition of the knowledge base 4d is counted up. That is, in this case, the column (1) of the item of the operation priority order in the column (a) of FIG.
4 ". In learning of knowledge, if the operation priority / hypothesis definition of the knowledge base 4b is searched for in the pattern of the operation status including the operation result and the recovery operation result (not in this case), the same pattern (Fig. As in (9) (b), the operation status is (1) the exchange port return test fails,
Since (2) switch card exchange test and (3) modem loopback test have not been performed), the new operation priority and the cause hypothesis / restoration procedure are displayed as FIG. Figure 11
Then, the "exchange card exchange test" is instructed and the operation result "failure" is input (at this time, the weight is changed by using the operation item is shown in FIG. 9B). Further, when searching the operation priority / hypothesis definition (FIG. 9) of the knowledge base, there are the same patterns as shown in (c) of FIG. It When reaching this stage, the estimated failure location and instructions for restoration measures are issued, so when the restoration operation result is "restore",
The failure handling process will end. The recovery operation result is "
If it is not "recovery", the process is repeated.
Display items such as the operation items and the failure cause hypothesis of the failure handling operation screen shown in FIGS.
Set the operation item definition format and the hypothesis definition format of in advance.

[Operation History] Next, utilization of the operation history will be described. The operation items automatically obtained from the monitoring device 2 and the FAX modem are automatically added / updated to the operation history as described in the explanation of FIG. 4 (steps S35, S38,
In S3A), the operation history that cannot be automatically fetched, such as telephone contact or operation result, can be obtained by the operator selecting "Add operation history" from the function selection menu bar 5b based on the information displayed on the interactive device 1h. Operation history can be set. This mechanism will be described with reference to FIGS. 13 to 18.
FIG. 13 shows an example of a menu, in which various necessary definitions can be made so that various operation items and results can be input. For example, select “Contact / Report”
By selecting "Worker → User" in the direction selection and "Status report" in the standard operation item selection, the operation history of the database 4a is recorded along with the input time as "Status report from worker to user". Remains. Using the same menu, you can also enter operation items, results, restoration operations and results for the causal hypothesis. This operation history is used to create a failure record (report), but is also used to create a pattern of failure response operation / recovery operation results and to create / register failure response knowledge. The items selectable in the operation history setting menu shown in FIG. 13 are defined by the item definition interface as shown in FIGS. 14 to 18, so that the menu items themselves can be easily set.

Note that the items that can be selected when selecting and registering the operation item, operation result, restoration procedure, and restoration operation result in the category selection in FIG. 13 are displayed with hatched portions in FIGS. 19 and 20 to be described later. Item. For example, as the operation items of FIG. 17, the standard operation items of the worker are shown in FIG.
The device name and operation item of are automatically displayed. In addition, as the standard work items of the operator of the recovery procedure in FIG. 18, the device name, the estimated cause of failure, and the recovery procedure items shown by the shaded areas in FIG. 20 are automatically displayed.

[Operation item / hypothesis definition format] FIGS. 19 and 20 are diagrams showing the formats of operation items and hypothesis definitions, and are formats for previously setting an operation for a failure and a failure cause (hypothesis) for the failure. Is shown. When the operation priority / hypothesis definition of the database 4b shown in FIG. 9 is defined based on the information set here, but an undefined operation result / recovery operation result pattern occurs during failure handling. The above-mentioned automatic learning processing function of the automatic learning processing unit 1d operates to add a definition pattern to the operation priority / hypothesis definition of the knowledge base 4b. For example, in FIG. 12, if a result that the operation item "modem loopback test" has failed is entered, there is no corresponding pattern in the operation priority / hypothesis definition of the knowledge base 4b shown in FIG. Operation screen 5c (Fig. 12)
Is not updated. At this time, if the worker judges that the line is defective and the fault is recovered by switching the line, the worker uses the operation history selection menu shown in FIG. 13 to perform the “line switching” and “recovery”. Select and enter in order. Also, the operator inputs a hypothesis in the hypothesis definition format of FIG. 20 prior to this operation. As a result, the learning of knowledge (the automatic learning processing unit 1d) searches the column (a) of FIG. 20 to acquire the hypothesis, and the new operation priority / hypothesis definition knowledge is acquired as shown in (e) of FIG. It is automatically added (if the operation result is also added to the operation history at this time, this is also added to the operation priority / hypothesis definition pattern).

The items of the operation item definition format shown in FIG. 19 and the hypothesis definition format shown in FIG. 20 can be set almost comprehensively, but in rare cases, items that do not appear in the operation history selection menu because they could not be assumed in advance. Suppose there was. In this case, since the operation item does not appear in the selection menu, the operator sets the operation item and the recovery processing result by manually inputting arbitrary characters, and selects and sets only the result from the operation history selection menu. (Example: "Power cord is not soldered properly and soldered to recover"). Such trouble records are accumulated, and “restoration” is searched from the operation history of trouble records by batch processing, and the operation item / recovery operation item is the operation item definition format (FIG. 19) and the hypothesis definition format (FIG. 20). It is also possible to prevent knowledge acquisition omission by displaying or printing as a list even those not defined in. In this case, the knowledge is added manually, but this work is rarely necessary.

[Measures Caused by the Same Cause] Since a plurality of faults caused by the same cause confuses the operator and reduces the support efficiency of fault handling, the fault handling support apparatus 1 is checked for faults caused by the same cause. A unit 1c is provided, and whenever a failure occurs, it is automatically determined whether or not a plurality of failures are caused by the same cause, information is displayed to the operator, and a failure caused by the same cause is identified as one failure among them. Grouping of obstacles is performed so that assistance for obstacles should be provided. There are two types of failures due to the same cause, parent-child relationship and parallel relationship, and are failures that occur from the same or different devices within a certain time width. 21 and 22 show conditions for making this determination.

FIG. 21 is a diagram for explaining the parent-child relationship.
0 and the device X1 are connected in series, and a failure is simultaneously reported from the devices X0 and X1. Further, FIG. 22 is a diagram for explaining the parallel relationship, and is a case where the devices X0 and X1 are connected in parallel and a failure is simultaneously reported from the devices X0 and X1. There is no limit to the number of parent-child relationships and parallel relationships. This example is shown in the failure list shown in FIG.

In the above embodiment, an example of a line failure has been described, but it goes without saying that the present invention is not limited to a line failure and can be applied to a general data processor failure. .

[0035]

As described above, according to the present invention, the operation history of the worker is automatically recorded, and the work report is automatically created based on this operation history. However, it is not necessary to create a report later, the burden on the worker can be reduced, and an accurate and reliable work report can be created.

Further, since the automatic learning processing unit is provided to instruct the operation by utilizing the past knowledge for the obstacle, it is possible to give an appropriate work instruction to the worker.

Since means for adding operation items can be provided, a richer knowledge base can be generated and work instructions can be given with higher accuracy.

Further, since the means for adding the recovery action item is provided, the fault handling knowledge can be accumulated more accurately.

Further, since the failures caused by the same cause are grouped, the failures can be dealt with more quickly.

Further, since the same cause is divided into the parent-child relationship and the parallel relationship, it becomes easier to deal with the failure.

[Brief description of drawings]

FIG. 1 is a diagram showing a configuration of an embodiment of the present invention.

FIG. 2 is an example of an operation screen on which an operator operates.

FIG. 3 is a flow chart of the processing of the device according to the present invention.

FIG. 4 is a flowchart of a failure handling processing function.

FIG. 5 is an example of a failure list display.

FIG. 6 is an example of an operation history screen.

FIG. 7 is an example of a fault record document.

FIG. 8 is a flow chart of an automatic knowledge learning function.

FIG. 9 is a knowledge base operation priority / hypothesis definition.

FIG. 10 is an example of an initial screen of a failure handling operation screen.

FIG. 11 is an example of a screen next to the failure handling operation screen.

FIG. 12 is an example of a screen next to the failure handling operation screen.

FIG. 13 is an operation selection menu for setting an operation history.

FIG. 14 is an item definition format for setting operation history (contact / report).

FIG. 15 is an item definition format for setting an operation history (answer).

FIG. 16 is an item definition format for setting an operation history (request / instruction).

FIG. 17 is an item definition format for setting an operation history (operation item).

FIG. 18 is an item definition format for operation history setting (restoration procedure).

FIG. 19 is a definition format of an operation item.

FIG. 20 is a restoration procedure definition format.

FIG. 21 is a definition of a method of determining a failure caused by the same cause having a parent-child relationship.

FIG. 22 is a definition of a method of determining a failure caused by the same cause in parallel relationship.

[Explanation of symbols]

1 Failure response support device, 1a Failure information collection unit, 1b
Fault confirmation part caused by the same cause, 1c Cause hypothesis generation part, 1d Automatic learning processing part, 1d Failure processing history creation part, 2 Monitoring device, 4a database, 4b Knowledge base, 5a Failure list screen, 5b Function selection menu bar 5c Fault response operation screen, 5d Operation history screen,
5e Related information display screen.

Claims (6)

[Claims] 1. A database having information of devices constituting a system, a knowledge base having a hypothesis for estimating an operation for a failure and its cause, a failure information collecting unit for receiving failure information, the database and the knowledge base. A cause hypothesis generation unit that provides an operator with an operation and a causal hypothesis for a fault by searching for a fault, and a fault treatment history generation unit that stores the operation history of the worker for the operation information presented by the cause hypothesis generation unit in the database. The failure support history creating unit is configured to automatically create a work report from the operation history and the database. 2. A database having information on devices constituting the system, a knowledge base having a hypothesis for estimating an operation for a failure and its cause, a failure information collecting unit for receiving failure information, the database and the knowledge base. A cause hypothesis generation unit that provides an operator with an operation and a causal hypothesis for a fault by searching for a fault, and a fault treatment history generation unit that stores the operation history of the worker for the operation information presented by the cause hypothesis generation unit in the database. And an automatic learning processing unit that retains the number of times of the same operation performed by the worker according to the operation provided by the cause hypothesis generation unit, and the cause hypothesis generation unit gives priority to an operation that has been used frequently. Disability support device characterized by giving instructions. 3. The fault action history creating section is provided with operation item adding means for adding an operation item, and the operation item adding means automatically registers the added item as an operation history in a database and causes the knowledge base to be hypothesized. The failure support device according to claim 1 or 2, wherein the failure support device is also registered. 4. A recovery action item means for adding a recovery action item to the failure action history creation unit is provided, and the added recovery action item is automatically registered in the database and also registered as a cause hypothesis of the knowledge base. The failure support device according to any one of claims 1 to 3, which is characterized. 5. The fault confirmation section for determining the faults caused by the same cause for a plurality of faults occurring at substantially the same time and grouping the faults by the same cause is provided. The fault support device according to claim 4. 6. The failure support apparatus according to claim 5, wherein the failure confirmation unit caused by the same cause divides the same cause into a parent-child relationship and a parallel relationship.