JPH0844700A - Complaint support method - Google Patents

Abstract

(57) [Abstract] [Purpose] It is an object of the present invention to maintain and improve customer satisfaction regarding a complaint handling support method used when processing a customer's complaint for mass-produced electronic equipment. [Structure] A first knowledge base in which a fault state is associated with a fault cause and a countermeasure using the knowledge of a designer, and a fault cause and a countermeasure corresponding to the fault state input using the first knowledge base. Designed to correspond to test items and test methods, test items and test results, and next test items, past fault conditions and their causes, and test methods The second that was related using the knowledge of the person
And a search / inference unit that searches the second knowledge base and infers the cause of failure, and the search / inference unit is
The past similar failure state and the cause candidate are extracted from the failure state sent by the inference unit, the cause is inferred by testing according to the corresponding test procedure, and the inference is repeated until the match between the inference cause and the cause candidate is confirmed. To configure.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a complaint handling support method used when processing customer complaints for mass-produced electronic equipment.

When handling complaints for mass-produced electronic devices, such as mobile phones, the support is not sufficient as compared with order-received devices because of the complexity of distribution channels and the high price / maintenance cost.

Therefore, it is necessary to systematize and support customer complaints for mass-produced electronic devices to maintain and improve customer satisfaction.

[0004]

2. Description of the Related Art FIG. 12 is a diagram illustrating a conventional example of a complaint handling support method. In the figure, a device shipped from a factory, for example, a mobile phone is sold to a customer through a sales agent, but if the purchased mobile phone fails, the customer brings the mobile phone to a sales agent point.

The sales agent point checks whether or not the condition as the customer's complaint can be reproduced with the mobile phone brought in, and estimates the approximate cost and period for repair by using the table possessed by the sales agent point. Present to the customer. Then, the customer support department (hereinafter referred to as CS
It is abbreviated as department, but it is the part that performs fault repair work)

[0006] The CS department diagnoses the mobile phone by looking at the sent slip, finds the faulty part, and repairs it. When the repair is completed, the mobile phone is sent back to the sales agent with a bill. Therefore, the sales agent hands the repaired mobile phone to the customer and receives the repair cost.

[0007]

With respect to the above-mentioned customer complaint processing, the problems at the sales agency and factory are as follows. (a) In response to customer complaints, the sales agent does not have the technical know-how, so it may return to the factory even those that are not obstacles such as incorrect usage or mishandling of the mobile phone. (b) The sales agent estimates and presents the estimated cost and period for repair to the customer, but the accuracy is low (the cost is impaired because it has only a low level of knowledge as compared to the factory level). Due to the reason, the period depends on the operating state of the factory, so it is difficult to judge if there is no knowledge of the factory level to some extent). (c) The CS department of the factory sends the slip with the customer complaint and the broken mobile phone together, which delays the work of arranging parts and preparing a repair system. (d) To quickly repair a mobile phone that has failed,
A skilled person who estimates and determines the cause of the failure is required. However, it takes time to gather people from other departments, educate them, and bring them up to be skilled. There are four issues.

An object of the present invention is to solve the above problems and to maintain and improve customer satisfaction.

[0009]

According to a first aspect of the present invention, in a first place, there is a correspondence between a fault condition, a fault cause, and a countermeasure,
A first knowledge base associated using the knowledge of the designer and an inference unit that infers a failure cause and a countermeasure corresponding to the failure state input using the first knowledge base are provided.

In addition, the type of the replacement parts and the weighted quantity are estimated by using the inference result and the repair record data from the inference section input in the second place, and the repair support for calculating the corresponding repair work man-hours. Section, test item and test method correspondence, test item and test result, and next test item correspondence,
A second knowledge base associated with the correspondence between past failure states and their causes and test methods, using a designer's knowledge, and a search / reason unit for searching the second knowledge base and inferring the failure cause And.

Then, the repair support unit arranges the members necessary for the repair work and prepares the system before the arrival of the troubled device, and the retrieval / inference unit sends the fault state sent by the inference unit. The similar failure state and the cause candidate in the past were extracted, the test was conducted according to the corresponding test procedure to infer the cause, and the inference was repeated until the match between the inferred cause and the cause candidate was confirmed.

According to a second aspect of the present invention, the similar fault state is obtained by decomposing the two fault states of the fault state sent by the inference unit and the past fault state into a condition unit, a subject, and a predicate. Comparison was made, and when it was detected that at least one decomposed portion was in agreement, it was determined to be similar.

[0013]

In order to solve the above-mentioned problem (a), the present invention relates to the first place, for example, the sales agent, by using the knowledge of the designer regarding the correspondence between the failure condition, the failure cause and the countermeasure. 1
And a reasoning unit that infers a cause of failure and a countermeasure corresponding to the input failure state by using the first knowledge base.

The sales agent uses the primary diagnosis function unit to
A plurality of causes of failure and countermeasures are inferred and sent to the factory side in real time with respect to the failure status, the number of the failed device, and the customer complaint. By providing the primary diagnosis function unit, there is no possibility of sending a device without any failure to the factory side.

To solve the problem (b), the repair support unit determines a plurality of causes of failure and countermeasures sent from the sales agent.
For example, since the individual cause is narrowed down, repair part data is used to extract replacement part information by product lot, failure frequency information for each used part lot, and replacement part information by season.

Then, a weighting coefficient is obtained from these pieces of information, the required number of replacement parts is estimated, and an additional man-hour is obtained, and from this result, a repair cost with accuracy information and a highly accurate repair completion time are calculated. And notify the sales agent.

For the problem (c), the replacement parts estimated as described above are arranged in advance, and the load of the process can be known from the calculation of the repair man-hours, so that the adjustment for the repair work can be performed in advance. .

With respect to the problem (d), the factory has accumulated know-how about what kind of test has been conducted to pursue the cause of failure. Therefore, in order to be able to utilize this know-how when diagnosing failures, it is the designer's responsibility to deal with test items and test methods, test items and test results, next test items, past failure conditions, their causes, and test methods. A second knowledge base associated using knowledge and a secondary diagnostic function unit consisting of a search / reason unit that searches the second knowledge base and infers the cause of a failure are used as servers, and test equipment and personal computers are used. Provided to the client.

As a result, even an inexperienced diagnostic operator can
The failure status from the primary diagnostic function unit can be sent to the secondary diagnostic function unit using a personal computer, and the cause can be specified by proceeding with the diagnostic in an interactive manner with the secondary diagnostic function unit.

Since the problems (a) to (d) can be solved by the above method, the customer satisfaction can be maintained and improved.

[0021]

DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1 is an explanatory view of a system of an embodiment of the present invention, FIG. 2 is an explanatory view of a primary diagnostic system of the present invention, FIG. 3 is an explanatory view of data processing for estimating replacement parts of the present invention, and FIG. 4 is a repair. FIG. 5 is a diagram showing an example of the result data, FIG. 5 is a functional explanatory diagram of scheduling / process management, and FIG. 6 is a secondary diagnostic system explanatory diagram of the present invention.

FIG. 7 is an example of the test procedure explanatory diagram of the present invention, FIG. 8 is an example of the fault example explanatory diagram of the present invention, FIG. 9 is an example of the diagnostic tree explanatory diagram of the present invention, and FIG. FIG. 11 is a schematic flow chart of fault diagnosis, and FIG. 11 is an explanatory view of a screen at the time of fault repair of the present invention.

The present invention will be described below with reference to FIGS. In the complaint handling support system of the present invention, as shown in FIG. 1, a sales agent having a primary diagnosis function unit 1 as a client, a repair support department, and a repair department, and a repair support function unit 2 and a secondary diagnosis function as a server 4. Part 3 is provided,
The server and the sales agent are connected online via a public line, and the server, the repair support department 5, and the repair department 6 are connected online via LAN. The details of the primary diagnostic function unit, the repair support functional unit, and the secondary diagnostic function unit will be described later.

When a sales agent receives a customer complaint about a sold device (for example, a mobile phone), a primary diagnosis of this complaint is made and the result of the primary diagnosis is predetermined online in a server via a public line. It is sent to the stored primary diagnosis result storage unit (not shown).

Therefore, the repair support department uses the primary diagnosis result stored in the server and the information necessary for repair support (stored in the repair support function section 2) to obtain the repair member and the repair process. In addition to formulating, estimate the accurate repair cost and period and send it to the sales agent.

Further, the repair department retrieves the primary diagnosis result and the information necessary for the secondary diagnosis (stored in the secondary diagnostic function unit 3), searches for the cause of the fault of the faulty device sent, and repairs it. , Send the repaired equipment back to the corresponding distributor.

The primary diagnosis function unit, repair support function unit,
The secondary diagnostic function unit will be sequentially described from the primary diagnostic function unit. (a) Primary diagnosis function unit The primary diagnosis function unit has a built-in database for primary diagnosis,
A person in charge can infer the cause of failure and countermeasures interactively by accessing a preset primary diagnosis database via a personal computer.

Here, as shown in FIG. 2, the primary diagnostic database 11 has a simple diagnostic tree for each device / function (in the figure, there are three types of diagnostic trees), and the vertical axis direction is The customer's complaint is the top layer, the question of the sales agent for this complaint, the probable cause of failure (probable cause) when YES to the question, the bottom layer is the countermeasure (what kind of countermeasure to the probable cause There is an item).

In the horizontal axis direction, check in each layer,
Although several tens of items are usually set as confirmation items, the items of each layer and the corresponding items of the lower layer are connected by a path. It should be noted that this path is provided using the knowledge of the designer.

Here, for example, it is assumed that the first diagnostic tree of the three types of diagnostic trees shown in FIG. 2 is collected for all sounds, and the claim is "noisy",
“Sound is cracked”, “Volume is low”, “Volume cannot be adjusted”, etc., but the questions include “Did you drop the device”, “Did you spill liquid”, “Did you ever adjust the volume”? Assumed to be shown.

Now, when there is a customer complaint that "the volume is low", the person in charge of the sales agency puts up this table on the screen and asks the question "I have not adjusted the volume. If the customer's reply is YES, the probable cause linked to this question is “unadjusted” and the corresponding measure is “volume adjustment”. Adjust the volume and give it to the customer.

However, when a plurality of probable causes and countermeasures can be inferred as a result of the diagnosis using the diagnosis tree, a primary diagnosis result storage unit (not shown) in the server 4 stores these and the mobile phone manufacturing number brought in by the customer. Send online to.

In other words, when there is a customer complaint, the presumed cause and countermeasure can be inferred by using the primary diagnosis system provided in the sales agent, so the accuracy of the primary diagnosis is improved and non-obstacles are detected. It will not be sent to the CS department on the factory side.

On the other hand, since the CS department is online connected to the above-described primary diagnosis system, the repair support function section on the factory side, the production management system, and the process management system, the primary diagnosis result from the sales agent (multiple Probable cause and countermeasure) and the serial number of the mobile phone that caused the failure, guess what kind of material is necessary and match it with the data of the estimated material and the repair material managed by the production management system. Check whether the estimated parts are in stock.

Further, by grasping the status of the process from the process management system, it is checked at what time it is possible to repair the failure, and the delivery date and the cost are paid before the defective mobile phone is sent to the CS department. Calculate (these are called repair support functions)
.

For this reason, the CS department uses the repair record data that the factory has. As shown in FIG. 4, the repair record data includes the device serial number, the cause of failure, countermeasures against the cause,
There are items such as replacement parts and date of repair, which are entered in the order in which they were repaired. This is the repair record data storage part (not shown) in the repair support function unit 2 provided in the server in FIG. ).

Here, it is assumed that the mobile phone with the manufacturing number of 1020 is returned by the customer with “sound crack”, and the manufacturing lot of the mobile phone at one time is 1000 units. Now, the person in charge of the repair support department uses the repair support terminal 5 in FIG. 1 to extract the manufacturing numbers 1000 to 1999 in the repair record data, and the cause of the failure is “sound crack”. Are further extracted. And
The items of the replacement parts of which the cause of the trouble is “sound crack” are arranged and taken out in the order of the number of replacement parts (see FIG. 3).

At this time, the number of replacement parts is 50, of which 30 are amplifiers, 10 are speakers, 5 are resistors, and 5 are others.
Set to 50. Then, when normalized by N and weighted,
The amplifier is 30 × N ₁ , the speaker is 10 × N ₂ , the resistance is 5 × N ₃ , and the others are 5 × N ₄ (N ₁ > N ₂ > N ₃ > N ₄ ), which is the member with the largest value. Is likely to be the cause of the failure. As shown in FIG. 3, when there are a plurality of weighting types for each part lot and each season, each weighting (M,
The value obtained by multiplying all P) is, for example, the largest value as the ordering target member.

Then, the production management system 71 is used to check whether or not the stock of the material estimated as described above is available. In addition, the process management system grasps the status of factory processes, estimates delivery times and costs, and sends them to sales agents. To this estimation result, past cost calculation results and estimated cost accuracy information are also added.

As a result, the customer can easily determine whether or not the repair can be performed depending on the amount of the repair cost, and since the repair turn is simulated from the man-hours of the incidental work, the highly accurate repair completion time is indicated.

Further, regarding the repair process in the CS department, since the necessary members are estimated from the repair result data as described above, it is possible to estimate what the repair work should be done with the estimated members, From this, the estimated work man-hours can be known (standard work man-hours for various works are stored in the database).

As a result, the estimated completion date can be known by comparing the estimated man-hours with the actual factory schedule (process control) as shown in FIG. On the other hand, if there is a target completion date and the expected completion date is later than the target completion date, an alarm is sent to the relevant departments because there is a shortage of measuring instruments and personnel from the viewpoint of process control.

That is, since the troublesome mobile phone can know the load of the members and processes required before the mobile phone arrives at the CS department, the adjustment can be made in advance, and the delay of the repair work is almost eliminated. By the way, in order to promptly repair a mobile phone having an obstacle, an expert who estimates and determines the cause of the obstacle is required. However, it takes time to gather people from other departments, educate them, and bring them up to be skilled.

However, since the factory has accumulated know-how as to what kind of tests were conducted in pursuit of the cause of failure, in order to make use of this know-how in failure diagnosis, as shown in FIG. Search section 33, inference section 34 and test procedure database 313, diagnostic tree 312, failure case base 31
The dotted line portion including the knowledge base 31 consisting of 1 and the diagnostic subtree 32 is stored in the secondary diagnostic function unit in the server of FIG. 1, and the dashed line portion is provided in the repair department 6 of FIG.

Here, the test procedure database in the knowledge base shown in FIG. 6 is divided into device names, functional block names, test items, test methods, etc. as shown in FIG. , Test standards, etc. are shown.

As shown in FIG. 8, the fault case base is a collection of data relating to faults that actually occurred. As shown in the diagram, the fault case base is divided into device names, functional block names, phenomena, causes, test methods, etc. , The cause is stored for each phenomenon. It also shows the test method used when estimating the cause and is linked to the test procedure above.

As shown in FIG. 9, the diagnosis tree shows the correspondence between the test item and the test result and the next test item. For example, if the transmission output is out of the standard after being tested according to the test method shown in the test procedure, another specified test item is tested. If this test also fails the specifications, the cause of the failure can be diagnosed by repeating the test with another test item.

The diagnostic subtree 32 is one of the diagnostic trees.
Only one of the functional blocks (for example, the transmission unit) is taken out, and the diagnostic-to-cutout unit 332 in FIG. 6 cuts out the functional blocks required for the diagnostic tree color.

The correspondence between the above test items and the test method, the correspondence between the test item and the test result and the next test item, and the correspondence between the past failure state and its cause and the test method are based on the knowledge of the designer. Are associated with.

The case retrieval unit 331, the test procedure retrieval unit 333, and the diagnosis tree cutout unit 332 in the retrieval unit 33 follow the fault case base and the test procedure database according to the instructions of the diagnostic operator input via the personal computer. , Access the diagnostic tree and retrieve the required information, for example.

Furthermore, the diagnostic tree collating unit 341 in the inference unit
Repeats reading the test item to be performed next from the diagnostic subtree according to the instruction of the diagnostic operator. And
When the cause is reached, the cause identifying unit 342 displays the cause determination to the diagnostic operator on the personal computer screen. At this time, if the cause is new, it is displayed and stored in the failure case base.

Now referring to FIGS. 1, 6 and 11, FIG.
A schematic flow of the fault diagnosis shown in FIG. First, the diagnostic operator uses the tester console 62 to send a customer complaint, for example, that the device name is a mobile phone, the functional block is a transmission part, and the failure phenomenon is from a sales agency via the personal computer 61 to the server 4 in FIG. It is sent to the secondary diagnostic function unit 3 stored in.
Therefore, the case retrieval unit 331 extracts similar phenomena and cause candidates from the fault case base 311, sends them out, and displays them on the personal computer screen as shown in FIG. 11-. At this time, the similar phenomena and the display of the cause candidates are arranged in descending order of probability (see Figure 10-S1 and S2).

In order to extract the similar phenomenon, the failure phenomenon sent from the sales agent and the failure phenomenon in the failure case base are decomposed into a condition part, a subject, and a predicate.
Enter "when dialing", "dial tone" as the subject, and "inaudible" as the predicate.

Further, in order to search for a similar phenomenon, all the logical sums of the coincident ones are extracted and displayed. At this time, the certainty factor is calculated by the following formula and displayed in descending order of certainty factor. The higher the certainty factor is, the higher the probability of recurrence is.

Confidence = (Number of occurrences of each phenomenon) / (Number of all phenomena of the relevant functional block) Now, when the diagnostic operator looks at the screen and selects a cause candidate with a high probability, the test procedure search unit 333 causes the test procedure search unit 333 to perform a test procedure. The database
Search for 313 and display the corresponding test procedure (test item and test method) on the PC screen as shown in Figure 11-.
Therefore, the diagnostic operator tests the DUT according to the test procedure displayed using the tester console 62 (Fig. 10-S3,
(See S4, S5).

When the diagnostic operator inputs the test result to the personal computer, the test result corresponding to the test procedure is displayed as, for example, OK / NG as shown in FIG. 11-. (Figure 11-S6, S7
See).

It is checked whether the inferred cause matches the cause (candidate cause in the failure case base) given as a candidate in FIG. 11-S2. -The cause is identified as shown in and completed (
(See Figure 11-S8, S17-S19).

However, if no, the next cause candidate is selected and the processing shown in FIGS. 11-S3 to S9 is repeated, but it is performed for all candidates until they match. Here, when the inferred cause and the actual cause do not match even after checking all the candidates, the untested procedure is extracted from the test procedure and the diagnostic tree. Then, the untested procedure is displayed and selected, and the device under test is tested (see FIGS. 11-S10 and 11).

When the diagnostic operator inputs the test result to the personal computer 61, the secondary diagnostic system makes an inference using the diagnostic tree. Check whether the inference result matches the actual cause,
If it is confirmed that they match, a case is added to the failure case base for the new cause and completed (Fig. 11-S12 to S14, S20 to S2
2).

If not, the procedure returns to FIG. 11-S11, another untested procedure is extracted, and the above-mentioned processing is repeated. However, if the untested procedures do not match, the diagnosis fails (FIG. 11- S1
(See 5, 16).

That is, the sales agent improves the accuracy of the primary diagnosis, and there is no possibility of sending non-obstructive equipment to the factory side. On the other hand, the CS department sends information such as the primary diagnosis result of the sales agent, failure status, and device serial number to the factory side online.Therefore, these information are used to determine the type and number of parts to be arranged. In addition to making advance arrangements, we grasp the process of the factory and estimate highly accurate costs and delivery dates and send it to the sales agent. If the expected completion date is later than the target completion date, an alarm will be issued.

Further, a secondary diagnostic function unit is provided so that even a non-skilled diagnostic operator can identify the cause of failure.

[0063]

As described above in detail, according to the present invention, since the system for supporting the customer complaint is systematically supported, there is an effect that the customer satisfaction can be maintained and improved.

[Brief description of drawings]

FIG. 1 is a system explanatory diagram of an embodiment of the present invention.

FIG. 2 is an explanatory diagram of a primary diagnostic system of the present invention.

FIG. 3 is an explanatory diagram of data processing for estimating replacement parts of the present invention.

FIG. 4 is a diagram showing an example of repair record data.

FIG. 5 is a functional explanatory diagram of scheduling / process management.

FIG. 6 is an explanatory diagram of a secondary diagnostic system of the present invention.

FIG. 7 is an example of a test procedure explanatory diagram of the present invention.

FIG. 8 is an example of a failure case explanatory diagram of the present invention.

FIG. 9 is an example of a diagnostic tree explanatory diagram of the present invention.

FIG. 10 is a schematic flowchart of fault diagnosis according to the present invention.

FIG. 11 is an explanatory diagram of a screen when repairing a fault according to the present invention.

FIG. 12 is a diagram illustrating a conventional example of a complaint handling support method.

[Explanation of symbols]

1 Primary diagnosis function unit 2 Repair support function unit 3 Secondary diagnosis function unit 4 Server 5 Repair support department 6 Repair department

Claims (2)

[Claims] 1. A first knowledge base, which is associated with a failure state, a cause of the failure, and a countermeasure by using the knowledge of the designer, is input to the first place by using the first knowledge base. An inference unit that infers a cause of failure and a countermeasure for the failure state is estimated in the second place by using the input inference result and repair record data from the inference unit and estimating the type and weighted quantity of replacement parts. , Correspondence between the repair support section that calculates the corresponding repair work man-hours, the test item and the test method, the test item and the test result and the next test item, the past fault condition and its cause, and the test method With respect to the above, a second knowledge base associated by using the knowledge of the designer and a search / inference unit that searches the second knowledge base and infers the cause of the failure are provided, and the repair support unit causes the failure. Required for repair work before the arrival of the equipment In order to infer the cause, the search / inference unit extracts past similar fault states and cause candidates from the fault state sent by the inference unit and tests them according to the corresponding test procedure. Then, the inference reason is repeated until the inference cause and the cause candidate are confirmed to match each other. 2. The similar failure state is divided into at least one of a condition section, a subject, and a predicate by comparing the two failure states transmitted by the inference unit and the past failure state, and at least 1
2. The complaint handling support method according to claim 1, wherein it is determined to be similar when it is detected that the two decomposed portions match.