JPH0232656A - System for inferring mis-working of maintenance engineer for fault of exchange - Google Patents

Abstract

PURPOSE:To infer the cause of a fault due to mis-working in a short time by classifying work information executed in a recent constant time and execution date information for respective work in sequence of the execution of work, inferring the mis-working of a maintenance engineer based on a rule, and outputting a doubtful work list. CONSTITUTION:When fault information is inputted from an input means 11, it is held by a fault information input part 14 in an inference processing part 13. Next, when the content of the work and the execution date for respective work are inputted from an input means 12, the combination of respective content of the work and date is rearranged in sequence of time at a work content processing part 15 in the inference processing part 13 based on the date information. A doubtful work extraction part 16 selects doubtful work inferred as the cause of the fault for the fault information in the fault information input part 14 from the work processed by the work content processing part 15, and outputs it from an output means 18 as the doubtful work list. In such a way, it is possible to infer the cause of the fault due to the mis-working in a short time.

Description

[Detailed Description of the Invention] [) Already Required] This invention relates to a method for inferring a maintenance worker's work error that causes a fault from fault information in switching equipment fault diagnosis. The purpose of the present invention is to provide a method for inferring maintenance work errors in switching equipment failures that can be inferred based on time.
The inference processing section includes an input means, an inference processing section, a suspect work extraction rule storage section, and an output means, and the inference processing section stores fault information representing a fault state from the input means in the fault information input section, and stores recent information from the input means. Work content information and respective execution date and time information performed within a certain period of time are classified in order of work execution time in the work content processing unit, and based on the rules stored in the suspect work extraction rule storage unit, the information in the fault information human resources department is extracted. The suspect work extracting unit infers the maintenance worker's work error with respect to the work contents classified by the work content processing unit, and the meritorious argument processing unit is configured to output a list of suspected work that is the result of the inference processing from the output means. .

[Industrial Application Field] The present invention relates to a method for inferring a maintenance worker's work error that causes a failure from failure information in fault diagnosis of a switchboard.

If a fault occurs in an exchange and the switching function stops, not only will the subscribers accommodated in that exchange be unable to communicate, but communication with other regions connected via trunk lines will be interrupted, resulting in social problems. The impact is extremely large. For this reason, early repair of failures is essential, and an expert system that strongly supports maintenance personnel's diagnosis and repair work is required.

Furthermore, although some failures are caused by failures in replacement parts, many are due to simple operational errors by maintenance personnel. In other words, when operating a switching system, maintenance personnel perform periodic tests, switching between active and spare systems, setting data when changing equipment, etc. 9 different tasks are performed, and mistakes in these tasks can cause failures. There are many things. However, even if a failure is caused by such a relatively simple work error, even if the maintenance person is familiar with the technology of the switch, if the technical knowledge regarding the switch is poor (especially when the switch is exported and operated locally) etc.), or if the exchanger has not been installed for a long time, it is difficult to diagnose the cause of the failure in a short time.

Therefore, when a switch breaks down, it is desired to be able to quickly infer a failure caused by a maintenance worker's work error.

[Prior Art] A conventional procedure for dealing with a failure caused by a maintenance worker's work error will be explained.

As a specific example, if a problem occurs in which a particular subscriber cannot hear the dial tone from the exchange, Figure 6 shows a list of the tasks performed by a maintenance person just before the problem occurred. Suppose it had become like this. FIG. 6 shows an example of the maintenance person's work contents and work execution dates and times.

Conventionally, when such a failure occurs, maintenance personnel refer to a diagnostic manual or the like that describes how to deal with the failure when it occurs, and attempt to repair the failure. In that case, the maintenance person must check the various cards (printed circuit boards) that make up the exchange.
The maintenance worker had to extract items that could be the cause of failures from his own work, as shown in Figure 6, as well as the work he had performed in the past.

Specifically, there are various cases such as data setting errors (when adding subscribers, changing phone numbers, adding trunks, setting test data), when replacing the wrong card, or when the card is in the wrong position. There is a work error.

[Problem to be solved by the invention] As mentioned above, in the conventional method, maintenance personnel have to perform a wide variety of tasks, so when a fault occurs, it is difficult to detect which of the tasks performed by the maintenance worker caused the fault. This is time consuming and causes delays in troubleshooting.

SUMMARY OF THE INVENTION An object of the present invention is to provide a method for inferring a maintenance worker's work error in the case of a switch failure, which can quickly infer the cause of a fault caused by a maintenance worker's work.

[Means to solve the problem 1 A basic configuration diagram of the present invention is shown in FIG.

In FIG. 1, 11 is an input means for inputting failure information, 12 is an input means for inputting work contents and execution date and time, 13 is an inference processing section, 14 is a failure information input section, 15 is a work contents processing section, 1
6 is a suspect work extraction unit; 17 is a suspect work rule storage unit; 1
8 is a suspect work list output means.

Although the input means 11 and 12 in FIG. 1 are shown as two different devices to illustrate the principle, the functions of both can be realized by one common device.

The present invention performs inference by selecting the cause of a failure from among the works performed by a maintenance person based on failure information and a suspect work extraction rule.

[Operation] When a fault occurs in the exchange, a maintenance person inputs fault information from the input means 11 in FIG.

Next, the worker inputs through the input means 12 the details of the work performed within a predetermined period (for example, the work of the day) and the date and time of execution of each work. In this case, the work contents do not need to be input in descending order of execution date, and may be input at random. Then, inference processing unit 1
Based on the date and time information, the work content processing unit 15 of No. 3
The combinations of each work content and date/time are rearranged and classified (sorted) in chronological order (with the latest work at the top). The classified information and the information from the fault information input unit 14 are supplied to the suspect work extraction unit 16.

The suspected work extraction unit 16 stores the suspected work extraction rule storage unit 17.
Based on the rules stored in the fault information input unit 14, the work that is suspected of causing the fault is selected from among those processed by the work content processing unit 15, and the suspected work is selected from the output means 18. Output as a work list. When the suspect work extraction unit 16 performs extraction from the latest work to the oldest work, the entire suspected work list is displayed (including printing) on the output means 18.

The maintenance person can repair failures caused by work errors by looking at the list and checking again what was done in the work displayed as suspicious.

[Example] Fig. 2 shows a configuration diagram of an embodiment of the present invention, Fig. 3 shows a flowchart of the suspect work extraction process, Fig. 4 shows an example of the suspect work extraction rule, and Fig. 5 shows the suspect work extraction process. Made by.

Here is an example of a business list.

In FIG. 2, 20 represents a processor including a CPU and memory, 21 a keyboard, 22 a magnetic disk device in which suspect work extraction rules are stored, and 23 an output device (display or printer). When the exchange is stopped due to a failure, a maintenance person inputs information indicating the type of failure, such as ``dial tone listening addition'' from the keyboard 21, and the information is stored in the failure information area 202 of the memory in the processor 20. Next, using the keyboard 21, the maintenance person inputs the details of the work performed and the date and time of the work execution. An example of this will be explained using the work contents shown in FIG. 6 (used in the explanation of the conventional example). In this example, settings of various station data, card replacement, testing, etc. are input as the maintenance worker's work contents, along with the date and time when the work was performed. The contents are the work contents/execution date and time area 2 of the memory in the processor 20.
It is stored in 03. Note that the work content and work execution date and time may be configured such that the worker reads out and inputs the recorded work data from a recording device (not shown), which is input periodically by the worker.

The contents of the work content/execution date and time area 203 are sorted by the processing unit 201 based on the execution date and time of the work in descending order of the most recently executed.

The sorted work content/execution date/time information is returned to area 2.
03, the processing unit 201 stores the third
According to the process flow shown in the figure, the suspected work is extracted and output.

In the processing flow shown in FIG. 3, first, the magnetic disk device 22 is searched based on the input failure information (area 202), and the suspected work extraction rule shown in FIG. 4 is retrieved.
31) in the figure, then extract the suspected work according to the rules (32), If the failure information "r dial tone cannot be heard" is input, ■ Card exchange tA (subscriber), ■ Station data setting ( (subscriber) are extracted as suspect operations.

Next, it is checked whether the maintenance personnel actually performed the extracted suspect work (33 and below). That is, each of the work contents in the area 203 sorted in the order of execution is compared with the extracted suspect work, and if they are equal, a list of suspected work as shown in FIG. 5 is output to the output device 23, and all work contents are Once the investigation is complete, the process ends. As a result, a suspect list is displayed on the output device 23 in the order of the most recently executed tasks. In this example, both ■ and ■ are output to the suspect list.

The maintenance person can check the data and cards corresponding to the work content output on this list, and correct or replace them to repair failures caused by work errors. If the problem cannot be repaired even after doing so, it is determined that the failure is not due to a work error, and diagnostic processing can be performed to focus on other causes.

[Effects of the Invention] According to the present invention, when a failure occurs due to a maintenance worker's work error, the suspected work can be immediately extracted by simply inputting the failure information and the work details. It is possible to achieve early repair and to reduce the spread of the effects of replacement equipment failure.

[Brief explanation of the drawing]

FIG. 1 is a basic configuration diagram of the present invention, FIG. 2 is a configuration diagram of an embodiment of the present invention, FIG. 3 is a flowchart of a suspect work extraction process, FIG. 4 is a diagram showing an example of a suspect work extraction rule, FIG. 5 is a diagram showing an example of a suspected work list, and FIG. 6 is a diagram showing an example of the maintenance worker's work contents and work execution date and time. In Fig. 1, 11.12: Input means 13: Inference processing section 14: Fault information input section 15: Work content processing section 16: Suspected work extraction section 17: Suspected work extraction rule storage section 18: Output means 2 figure 337-

Claims (1)

[Claims] A system for inferring a maintenance worker's work error due to a switch failure, comprising: input means (11, 12), an inference processing section (13), a suspect work extraction rule storage section (17), and an output means (18). , the inference processing unit (13) stores failure information representing the failure state from the input means (11) in the failure information input unit (14), and stores the failure information representing the failure state from the input means (11), and stores the failure information representing the failure state from the input means (12), and stores the failure information representing the failure state from the input means (12). The work content information and the respective execution date and time information are classified in order of work execution time in the work content processing unit (15), and the failure information input unit (14) The suspect work extraction unit (16) infers the maintenance worker's work error based on the information and the work contents classified by the work content processing unit (15), and the inference processing unit (13) generates a suspect work list that is the result of the inference processing. A method for inferring a maintenance worker's work error in a switching equipment failure, characterized in that the following is output from an output means (18).