JP4528492B2 - Accident analysis support system - Google Patents

Description

[0001]
The present invention relates to an accident analysis system for analyzing the cause of an accident occurring at a work site in various operations, and in particular, an accident analysis for supporting cause analysis of an accident involving human factors. support About the system.
[0002]
[Prior art]
When an accident occurs, it is necessary to determine the cause and take measures to prevent recurrence. For this reason, various accident analysis systems have been proposed.
[0003]
Such an accident analysis is considered to be relatively simple if it is an accident caused by a failure or defect of the apparatus, that is, an accident caused by a physical factor. However, many accidents involving human factors involve many complicated causes, and the way in which human factors contribute to the accident varies greatly depending on the environmental factors at the site and the person's own condition. In many cases, it is difficult to analyze the cause of an accident and to plan countermeasures.
[0004]
Conventionally, as a system for analyzing an accident involving human factors, a system disclosed in Patent Document 1 below is known.
[0005]
This accident analysis system is a system using a variation tree method, which is one of the techniques for analyzing an accident. The variation tree method is based on the idea that an accident occurs due to an event (deviation event) that is different from the flow of events that should be.
[0006]
In the analysis, first, in the field, the events that occurred before the accident occurred are input to the system in time series. After input, the system expands the event in a tree shape and displays the variation tree on the display screen. The user can analyze the cause of the accident and devise a countermeasure while referring to the variation tree.
[0007]
[Patent Document 1]
Japanese Patent Laid-Open No. 9-22313
[0008]
[Problems to be solved by the invention]
However, the accident analysis system is a system for assisting even a beginner who does not have knowledge of the variation tree so that the variation tree for analyzing the cause of the accident can be efficiently created. That is, this system only supports the creation of a variation tree, and therefore the analyst needs to analyze the cause of the accident himself while looking at the variation tree.
[0009]
However, there are many cases where it is difficult to analyze and plan countermeasures for accidents involving human causes as described above. Therefore, in order to conduct accident analysis using the above accident analysis system, it is not accurate and effective unless it is not an analyst or beginner who lacks specialized knowledge but an expert with psychology or ergonomic knowledge. It is difficult to perform cause analysis.
[0010]
The present invention has been made in view of such problems, and accident analysis capable of performing advanced cause analysis even by an analyst lacking specialized knowledge. support The purpose is to provide a system.
[0011]
In order to solve the above problems, the accident analysis support system according to the present invention is an accident analysis support system that supports the cause analysis of an accident involving human factors, and the analyst inputs the cause of the accident involving human factors. Input means, output means for displaying support information for assisting the cause analysis by the analyst, a plurality of causes, n-order cause (n: natural number), and n + 1-order that is the cause of the occurrence of the n-order cause The storage means for storing the information in association with the cause, and when the analyst inputs the n-order cause of the accident via the input means, the recording means is referred to and n + 1 stored in association with the n-order cause. Control means for performing support information display control so as to display a list of secondary causes as the support information on the output means, and the control means, when an analysis is performed on a certain n-order cause, N When a new n + 1-order cause other than the n + 1-order cause stored in the storage means in association with the cause is input by the analyst through the input means, the n + 1-order cause is associated with the n-order cause. The accumulation control is performed so as to be stored in the storage means, and this new n + 1-order cause is continuously set as the n-order cause input by the analyst. Referring to the recording means, if there are n + 1-order causes stored in association with the n-order cause, they are extracted and displayed as a list on the output means as the support information The support information display control keep working on it It is characterized by performing.
[0012]
In the accident analysis support method according to the present invention, the input means, the output means, and a plurality of causes are associated with the n-order cause (n: natural number) and the n + 1-order cause that causes the n-order cause. In an accident analysis support method for supporting cause analysis of accidents involving human factors by means of an accident analysis support system comprising a storage means for storing an Is input to the accident analysis support system, the recording means is referred to extract an n + 1-order cause stored in association with the n-order cause, and the n + 1-order cause extracted in the extraction step is supported. A display step of displaying a list on the output means as information, and when an analysis is performed on a certain n-order cause, n + stored in the storage means in association with the n-order cause When a new n + 1-order cause other than the next cause is input by the analyst via the input means, an update process is performed in which the n + 1-order cause is stored in the storage means in association with the n-order cause. This new n + 1 order cause is entered by the analyst as the nth order cause. Referring to the recording means, if there are n + 1-order causes stored in association with the n-order cause, they are extracted and displayed as a list on the output means as the support information The extraction step and the display step keep working on it It is characterized by performing.
[0013]
The accident analysis support program according to the present invention includes an input device, an output display device, a plurality of causes, an n-order cause (n: natural number), and an n + 1-order cause that causes the n-order cause. In an accident analysis support program for causing a computer comprising a storage device to store information in association to execute support for analysis of the cause of an accident involving human factors, the analyst uses the input device to cause the n-th cause of the accident. To the computer, an extraction step of extracting the n + 1-order cause stored in association with the n-order cause and the output of the n + 1-order cause extracted in the extraction step as the support information A display step of displaying a list on a display device, causing a computer to execute the analysis, and when an analysis is performed on a certain n-order cause, the display step relates to the n-order cause When a new n + 1-order cause other than the n + 1-order cause stored in the storage means is input by the analyst via the input device, the n + 1-order cause is associated with the n-order cause in the storage device. Let the computer execute the update step to be stored and continue to use this new n + 1-order cause as the n-order cause entered by the analyst. Referring to the recording means, if there are n + 1-order causes stored in association with the n-order cause, they are extracted and displayed as a list on the output means as the support information The extraction step and the display step keep working on it The computer is executed.
[0014]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, embodiments of the present invention will be described with reference to the drawings.
[0015]
FIG. 1 is a flowchart showing an outline of an accident analysis system 1 according to an embodiment of the present invention. As shown in the figure, an accident analysis system 1 according to the present embodiment includes a personal computer main body (hereinafter referred to as “PC main body”) 10, a display 20 as an output display device, and a keyboard 30 as an input device. It is composed of The PC main body 10 includes a CPU 11 as various arithmetic devices and a storage device 12 such as a hard disk or a flash memory. The storage device 12 includes a human error tree diagram database (hereinafter, “database” is referred to as “DB”) 13. The human error tree diagram DB 13 stores information to be referred to when performing a reason analysis as will be described later.
[0016]
Next, an outline of a process flow when the accident analysis is performed by the accident analysis system 1 according to the present embodiment will be described with reference to FIG. FIG. 2 is a flowchart showing the outline. As shown in the figure, the accident analysis according to the present embodiment is roughly divided into “extraction of deviation event” by time series analysis (step 1 (hereinafter, “step” is referred to as “S”)), why It is composed of three steps: “Analysis of the cause of the departure event” (S2) by the analysis and “Sorting the problems of the accident” (S3) by the m-SHEL analysis.
[0017]
Here, the “deviation event” is a time series comparison of the event at the time of the accident and the event that should have occurred when performing a certain work, and is different from the event that should have occurred at the time of the accident. Means an event. If work is carried out according to the flow of events that should be originally intended, it is considered that an accident will not occur. Therefore, this deviating event is considered as a direct cause of the occurrence of the accident. In the present embodiment, only deviation events related to human factors are extracted.
[0018]
Also, why and why analysis is an analysis method that does not consider the cause of occurrence of a certain event, but regularly and in order. For example, if A and B are considered as causes that cause a certain event, then the causes that A and B are generated are respectively A-1 and A-2, B-1 and B-. This is a method of thinking about the cause over a plurality of layers, such as 2 and further considering the cause of each occurrence. In addition, why and why analysis is explained in detail in Hitoshi Ogura, “Why Why and Thorough Analysis” (published by the Japan Plant Maintenance Association, July 15, 1997).
[0019]
The m-SHEL analysis is based on six viewpoints: a) individuals, b) individuals and others, c) procedures, d) equipment, e) environment, and f) management to take measures against accidents. This is a method for estimating the problem that a departure event has occurred.
[0020]
Next, the details of the “extraction event extraction” (S1) described above will be described with reference to FIGS.
[0021]
FIG. 3 is a flowchart for explaining the flow of processing when a departure event is extracted by time series analysis. First, in S10, the analyst inputs, through the keyboard 30, events that should originally be in time series until an accident event is reached. When entering, the event Time to occur Also enter together. Next, in S11, the analyst similarly inputs the event when the accident occurs. These events input by the analyst and their occurrence times are stored in the storage device 12 in the PC main body 10.
[0022]
Subsequently, in S12, the CPU 11 in the PC main body 10 compares the event that should be originally stored in the storage device 12 and the event at the time of the accident for each time, and extracts different items, that is, deviation events. The different items in this case are extracted as deviation events only for events related to the behavior of the worker, and physical events such as devices are not extracted even if there are different items. In S13, the extracted departure event is stored in the storage device 12 and recorded.
[0023]
FIG. 4 is a diagram showing an example of the result of the departure event extraction in S1, and the contents shown in FIG. 4 are displayed on the display 20 and presented to the analyst. This figure shows the extraction results for the case where the machine exploded because the operator pulled the wrong lever. As events that should be, “signal blinks” at time 0:00, “worker confirms blinking of signal” at time 0:01, “worker pulls lever I” at time 0:02, The operator inputs “the meter needle swings to the left” at time 0:03 and “the machine stops” at time 0:04.
[0024]
On the other hand, the flow of events at the time of the accident was as follows: “Signal blinked” at time 0:00, “Worker confirmed blinking of signal” at time 0:01, “Worker at time 0:02” ”Pulls lever II”, nothing happens at time 0:03, and there is an accident that “the machine exploded” at time 0:04. In this case, since the deviation event related to the behavior of the worker is an event “time when the worker pulls lever II” at time 0:02, an indication that there is a deviation event {circle around (1)} at the time 0:02. Has been made.
[0025]
Next, details of the above-described “analysis of the cause of the departure event” (S2) will be described with reference to FIGS.
[0026]
FIG. 5 is a flowchart showing the flow of processing when analyzing the cause of the departure event by the reason why analysis. FIG. 6 is a diagram conceptually showing information stored in the human error tree diagram DB 13.
[0027]
As shown in FIG. 5, first, in S <b> 200, the departure event extracted in S <b> 1 and stored in the storage device 12 is displayed on the display 20. Subsequently, in S <b> 201, the analyst estimates the primary cause of the departure event displayed in S <b> 1 and inputs it via the keyboard 30. The input primary cause is stored in the storage device 12 and displayed on the display 20.
[0028]
In step S202, the CPU 11 in the PC main body 10 searches the human error tree diagram DB 13 in the storage device 12 in order to extract the cause of the primary cause (secondary cause). In S203, if there is a matching cause as the cause of the primary cause, the process proceeds to S204, and if there is no cause that matches, the process proceeds to S207.
[0029]
In S204, the cause extracted as the secondary cause in S203 is displayed on the display 20. Here, as shown in FIG. 6, various causes are stored in the human error tree diagram DB 13 in association with other causes. For example, looking at a certain x-order cause, it is stored in association with the cause of this x-order cause, that is, x + 1 cause. Therefore, CPU11 extracts the secondary cause linked | related with the primary cause input by the analyzer as an option.
[0030]
Next, in S205, the analyst selects an appropriate cause as the cause of the primary cause (secondary cause) from the causes displayed as the options. In this selection, one cause may be selected, and when there are a plurality of appropriate causes, a plurality of causes may be selected as secondary causes. If there is no appropriate secondary cause, the process proceeds to S207 without selection.
[0031]
If the analyst selects an appropriate cause in S205, the option is determined as a secondary cause (S206). The determined secondary cause is stored in the storage device 12 and displayed on the display 20 as the secondary cause. If there is no cause for matching in S203 or there is no appropriate option in S205, the process proceeds to S207, where the analyst estimates the cause of the primary cause (secondary cause) through the keyboard 30. Enter. Since the input association between the secondary cause and the primary cause is information not stored in the human error tree diagram DB 13, it is stored in the human error tree diagram DB 13 in the storage device 12 in S 208. . Then, this input is determined as a secondary cause (S206), and is stored in the storage device 12 and displayed on the display 20 as in the above case.
[0032]
Next, proceeding to S209, the determination of the tertiary cause that is the cause of the secondary cause is performed in the same manner as in the case of the determination of the secondary cause. In this way, the cause is searched for one after another by the reason why analysis, and when the last n-th cause is determined in S220 to S224, the analysis of the cause of the occurrence of the deviation event of S2 is completed. As far as the order of analysis should be performed, there are reasons why the cause of the deviating event has been completely exhausted, and a reliable and effective preventive measure can be taken in organizing the problems of the accident in S3. If the analyst determines that it has been found, the cause analysis can be ended there.
[0033]
As described above, in the present embodiment, when analyzing the cause of the departure event by the reason why analysis, the cause option is displayed on the display 20 using the human error tree diagram DB 13 and presented to the analyst. Therefore, it is possible to easily analyze the cause of the deviation event. In particular, accidents involving human factors that are the subject of analysis in this embodiment are difficult to perform accurate and effective cause analysis unless they are specialists such as psychologists. Even if it is not such an expert, it is possible to perform cause analysis relatively easily.
[0034]
Further, in the present embodiment, as shown in S208 and S226, associations between new causes are additionally accumulated in the human error tree diagram DB 13, and the updated human error is analyzed in the subsequent why-why analysis. A tree diagram DB 13 can be used. Therefore, a more effective accident analysis system can be constructed by continuing to use the accident analysis system 1 according to the present embodiment. In particular, if you continue to use it at a specific site, you can build an effective accident analysis system suitable for that site.
[0035]
Next, details of the “arrangement of accident problems” (S3) using the m-SHEL analysis described above will be described with reference to FIGS. FIG. 7 is a diagram showing six viewpoints in the m-SHEL analysis used in “sorting of accident problems”.
As shown in Fig. 7, the m-SHEL analysis shows a) “individuals” regarding the personality and attitude of the person, b) “individuals and others” regarding communication and dependence between the person and the other person, c) "Procedure" related to the person and the procedure, manual, etc. d) "Equipment" related to the person and the equipment, tools, etc. e) "Environment" related to the person and the working environment, f) "Management" related to the person and education, supervision, measures, etc. This is a technique for analyzing the factor (problem) behind the cause of the departure event based on the above six viewpoints.
[0036]
In organizing the problems of S3, the analyst uses the m-SHEL analysis method while viewing the analysis result of the cause of the departure event by S2 displayed on the display 20, and causes the occurrence of the departure event. The factor behind the problem (problem) is analyzed and input via the keyboard 30. This problem is stored in the storage device 12 and displayed on the display 20. Then, the analyst plans measures for improving these problems while looking at the list of problems.
[0037]
As described above, in the “arrangement of accident problems” according to the present embodiment, the problems are extracted from various viewpoints by the m-SHEL analysis, so the analyst takes effective countermeasures. be able to. In addition, by separating human factors into individuals (persons) and others (others), individual factors and interpersonal factors can be analyzed more clearly. In the field where work is performed, there are many tasks in which multiple people work together rather than single work, so if you analyze the problems from this classified viewpoint, you will point out more accurate problems and thus more effective Can devise various countermeasures.
[0038]
FIG. 8 is a diagram showing a result of an analysis by m-SHEL analysis using a model similar to the above-described departure event extraction (S1) as a model. As shown in the figure, from the perspective of “individuals”, the problem of “being incapable of seeing around when you concentrate on something”, and from the perspective of “individuals and others” From the viewpoint of “procedures”, there are problems such as “there is a tendency to place too much emphasis on efficiency as a whole”, and from the viewpoint of “procedures”, problems such as “the confirmation procedure of work different from usual has not been decided” are found by analysis . Furthermore, from the viewpoint of "equipment", from the viewpoint of "Lever I and Lever II are adjacent and similar in shape and easy to make mistakes", and from the viewpoint of "Environment", "The operation room is dark. From the point of view of “management”, there are problems such as “There is not enough work plan and work time is often insufficient due to train delays”. It is.
[0039]
In this model, problems are pointed out for all six viewpoints. However, it is not always necessary to point out all the viewpoints, and only the viewpoints where appropriate problems are found may be used.
[0040]
Next, analysis results by the accident analysis system 1 according to the present embodiment will be described along specific examples. In this example, “The conductor of a 999 train operating from A station to B station opens the door without confirming the stop position of the train that has stopped 100 meters past the predetermined stop position of B station. After that, the conductor confirmed the passenger's safety, met with the driver, moved the train backward, stopped at the predetermined stop position, and then got on and off the passenger. "
[0041]
FIG. 9 is a diagram illustrating a departure event extraction result (S1) by time-series analysis in this specific example. As described above, in S1, the accident analysis system 1 extracts based on the information of “time”, “event flow that should be”, and “event flow at the time of accident” input by the analyst. is there. The extraction result is displayed on the display 20.
[0042]
As shown in the figure, the departure event (1) is “accident” even though the conductor should return from the in-vehicle tour at “time” 10:22:00, The “event flow” is that the conductor has not returned in response to passenger complaints. Deviation event (2) is “time of event” at 10:22:30 as “the flow of events that should be,” while the conductor should be aware of the train state, Since he has not yet returned from the in-car patrol, he has not taken any cautionary action. The departure event {circle over (3)} is “11.22: 40” at the time of “11:22:40”, where the conductor should confirm the passenger status on the platform as “the flow of the event that should be,” "It has just returned from the in-car patrol and has not taken any confirmation actions." The departure event (4) is “accidental accident” at the time “11:22:55” although the conductor should point to the stop position of the train and confirm the call as “the flow of the event that should be.” The “event flow at the time” is that no finger pointing confirmation action is taken.
[0043]
Next, the analysis result (S2) of the cause of the departure event in this specific example will be described. FIG. 10 is a diagram showing the analysis result of the cause of the departure event (1), and FIG. 11 is a diagram showing the analysis result of the cause of the departure event (4). Here, in this specific example, analysis results are shown for the fundamental departure events (1) and (4) of the accident among the departure events (1) to (4). As described above, these analysis results are obtained when the analyst analyzes why and why using the accident analysis system 1.
[0044]
In the case of departure event (1) “The conductor has not returned from the in-vehicle circuit” shown in FIG. 10, the primary cause is “because the conductor has forgotten to return from the in-vehicle circuit”, and the secondary cause is “ The reason for the third cause is that the conductor was not able to round up the passenger response, and the fourth cause was that the conductor had not received difficult passenger response training. Analysis has been done.
[0045]
Further, in the case of the deviation event (4) “Confirmed that the conductor has not pointed to the stop position” shown in FIG. 11, the primary cause is “because the conductor forgot to confirm the stop position”, 2 The next cause is "because the conductor's attention has shifted to opening the door", and the third cause is "the conductor has not returned from the in-vehicle patrol when entering the station" and "the conductor's awareness is more suitable for regular driving than safety." “Because there is a tendency,” cause analysis is done.
[0046]
Here, by taking the case of the deviating event (4) as an example, the details of why and why will be described. FIG. 12 is a diagram showing a part of information in the human error tree diagram DB 13 shown in FIG. 6 in this specific example. In the description of the present embodiment described above, when analyzing the “primary cause” from the “deviation event”, the analyst must manually input, and from the “n-order cause” to the “n + 1-order cause”. The human error tree diagram DB 13 is used only when analyzing, but in this specific example, the former is also analyzed using the human error tree diagram DB 13. This is because, depending on the content of the “deviation event”, the association between a cause and another cause may be applied as it is. In this way, if the human error tree diagram DB 13 is used from the stage of analyzing the “departure event” to the “primary cause”, it is not necessary to divide the processing as in S201 of FIG. 5 and S202 to S206. The system configuration can be simplified.
[0047]
First, in the analysis from “deviation event” to “primary cause”, “deviation event” is “not done”, and referring to FIG. 12, “forgot” and “no need to do”. Since “I thought” matches as the primary cause (corresponding to S203), these are displayed as options on the display 20 and presented to the analyst (corresponding to S204). As shown in FIG. 12, in this specific example, the analyst selects “Forgot” (corresponding to S205). It should be noted that the analyst may appropriately manually input, for example, “the conductor confirms the stop position”, which is the primary cause object of what is “forgotten”. In the why-why analysis according to this embodiment, it is important to present n + 1 cause choices for the n-th cause in consideration of psychological factors. Even so, it can be easily applied.
[0048]
Subsequently, in analyzing the secondary cause from the primary cause "because the conductor forgot to confirm the stop position" FIG. The analysts are given four options: “Because attention has shifted to”, “Because of being vague”, “Because there was often“ do not ””, and “Because of“ doing ”work immediately before” Presented (S222). In this specific example, the analyst has selected “Because attention has shifted to” (S223). Next, since there is no matching information in the human error dendrogram DB13 in the analysis of the tertiary cause from the secondary cause "because the conductor's attention has shifted to opening the door" (in the case of "No" in S221) ), The analyst manually entered that the conductor did not return from the in-vehicle patrol when entering the station and that the conductor's consciousness tends to be more suitable for driving than safety. The former third cause is the same as the deviating event (4), and in the analysis up to this point, the analyst has determined that the cause has been issued without omission, so the analysis is terminated here.
[0049]
Next, the m-SHEL analysis result (S3) of this specific example will be described. FIG. 13 is a diagram showing the m-SHEL analysis result of the departure event (1), and FIG. 14 is a diagram showing the m-SHEL analysis result of the departure event (4).
[0050]
The analyst points out the problems of each viewpoint through m-SHEL analysis while looking at the analysis results of the cause of departure event (1) shown in FIG. The analysis results point out problems related to two viewpoints as shown in FIG. One is c) the “procedure” viewpoint, “the passenger handling manual was not clearly defined”, and the other, f) “the passenger who is difficult to understand” It was a problem such as “I didn't provide training or training for the response”.
[0051]
Further, as shown in FIG. 14, the analysis result of the departure event (4) shows that b) “Consider awareness tends to place more importance on regular driving than safety” from the viewpoint of “individuals and others”. And d) problems such as “a structure in which the door opens even if the vehicle is not stopped at the stop position” from the viewpoint of “equipment” are pointed out. The m-SHEL analysis results shown in FIGS. 13 and 14 are input by the analyst via the keyboard 30 and stored in the storage device 12.
[0052]
FIG. 15 is a diagram that collectively displays the problems of the subject according to the specific example shown in FIGS. 13 and 14. The contents shown in FIG. 15 are created by the CPU 11 summing up the problems of the departure event (1) and the departure event (4) stored in the storage device 12 and displayed on the display 20. The analyst makes a preventive measure against the accident of this specific example with reference to FIG.
[0053]
FIG. 16 shows countermeasures against the accident of this specific example, which was made by the analyst with reference to FIG. As shown in the figure, two measures are proposed for the problem of viewpoint b) “individual and others”. As a first measure, “The guidance of the meaning of the confirmation operation at the time of a call and guidance was given so that the confirmation operation should be performed more accurately when the person is impatient.” The purpose is “so that the confirmation operation can be performed even when impatient. The person responsible is “Chief”, the enforcer is “Director of duty”, the subject is “Conductor”, and the deadline / period is “2002/8 / 20-2002 / 9 / 12 ”, the audit method is“ Confirmation of precautions for call ”, and the audit period is“ 3 months later ”.
[0054]
In addition, as a second countermeasure against the problem of viewpoint b), “the confirmation of the stop position should be pointed out and described in the manual, and the change point was posted”. To raise awareness of stopping position confirmation, responsible person is "Director", enforcer is "Supervisor", subject is "Conductor", deadline / period is "Revised 2002/8/20 Manual" , 2002/8 / 12-2002 / 8/31 ", the audit method is" Confirmation of manuals and presentation documents ", and the audit period is" 3 months later ".
[0055]
And as a measure against the problems of viewpoint c) procedure and f) management, “To round up the response to passengers is made manual for each common pattern, simulation training is carried out under each assumption, The purpose of the simulation was to introduce simulation education to the education. The purpose is to enable the conductor to properly round up the passenger response. "Chief", subject is "conductor", deadline / period is "2002/8 / 20-2002 / 8/31, once a year, new employee education", audit method is "manual, confirmation of educational results", audit The plan is to formulate a measure “after 3 months”.
[0056]
Here, the functions of the accident analysis system 1 according to the present embodiment are realized by installing a program for realizing these functions in the PC body. Distribution of this program may be performed via a recording medium or may be performed via a network.
[0057]
As described above, the embodiment of the present invention has been described in detail. However, according to the accident analysis system according to the present embodiment, first, the departure event that caused the accident is extracted, and the cause of the occurrence of the departure event is analyzed. From this, problems that cause deviating events are extracted, and measures are taken to prevent these problems, thereby preventing accidents. Therefore, it is possible not to grasp the prevention of the whole accident vaguely as in the past, but to make an effective measure after clarifying the target of the accident prevention measure.
[0058]
In the present embodiment, in analyzing the cause of the departure event, the accident analysis system uses the human error tree diagram DB to present the analysis options to the analyst. Therefore, even an analyst lacking specialized knowledge such as psychology or ergonomic knowledge can perform advanced cause analysis in consideration of human factors. In addition, if the analyst determines that there is no cause for matching in the cause associations prepared in advance in the human error tree diagram DB during the reason why analysis, the cause input by the analyst Is stored in the human error tree diagram DB as a new association. In this way, by repeating the analysis at a specific site using this accident analysis system, this accident analysis system learns the correlation of causes according to each site, and as a system more suitable for each site Can develop.
[0059]
In this embodiment, since the CPU automatically performs the extraction of the deviation event from the “event flow that should be originally” and the “event flow at the time of the accident” input by the analyst. The burden on the analyst can be greatly reduced.
The present invention is not limited to this embodiment, and various modifications can be made without departing from the spirit of the present invention. For example, in the accident analysis system according to the present embodiment, in the analysis of the cause of the departure event, only the departure event related to the human factor is extracted, but the departure event related to the physical factor is included together. You may make it extract. In this case, it is only necessary to store association information about physical factors in the human error tree diagram DB and analyze why and why as in the case of human factors.
[0060]
【The invention's effect】
As described above in detail, according to the present invention, even if an analyst lacks expertise in psychology or ergonomic knowledge, accident analysis that can perform advanced cause analysis support A system can be provided.
[Brief description of the drawings]
FIG. 1 is a diagram showing a schematic configuration of an accident analysis system according to an embodiment of the present invention.
FIG. 2 is a flowchart showing an outline of a process flow when an accident analysis is performed by the accident analysis system according to the present embodiment.
FIG. 3 is a flowchart for explaining a flow of processing when extracting a departure event according to the embodiment of the present invention.
FIG. 4 is a diagram showing an example of a departure event extraction result according to the embodiment of the present invention.
FIG. 5 is a flowchart showing a flow of processing when analyzing the cause of the occurrence of the departure event according to the embodiment of the present invention.
FIG. 6 is a diagram conceptually stored in a human error tree diagram DB according to the embodiment of the present invention.
FIG. 7 is a diagram showing six viewpoints in the m-SHEL analysis according to the embodiment of the present invention.
FIG. 8 is a diagram showing an analysis result of m-SHEL analysis according to the embodiment of the present invention.
FIG. 9 is a diagram showing an extraction result of a departure event of a specific example according to the embodiment of the present invention.
FIG. 10 is a diagram showing an analysis result of the cause of the departure event (1) of the specific example according to the embodiment of the present invention.
FIG. 11 is a diagram showing the analysis result of the cause of the departure event (4) of the specific example according to the embodiment of the present invention.
FIG. 12 is a diagram showing a part of information stored in a human error tree diagram DB in a specific example according to the embodiment of the present invention;
FIG. 13 is a diagram showing an m-SHEL analysis result of a departure event (1) of a specific example according to the embodiment of the present invention.
FIG. 14 is a diagram showing an m-SHEL analysis result of a departure event (4) of a specific example according to the embodiment of the present invention.
FIG. 15 is a diagram collectively showing m-SHEL analysis results of deviation events (1) and (4) of a specific example according to the embodiment of the present invention.
FIG. 16 is a diagram showing countermeasures against accidents in a specific example according to the embodiment of the present invention.
[Explanation of symbols]
1 Accident analysis system
10 PC body
11 CPU
12 Storage device
13 Human error tree diagram DB
20 display
30 keyboard

Claims (3)

An input means for the analyst to input the cause of the accident involving human factors,
An output means for displaying support information to support the cause analysis by the analyst;
Storage means for storing a plurality of causes in association with an n-order cause (n: natural number) and an n + 1-order cause that is the cause of the occurrence of the n-order cause;
When the analyst inputs the n-th cause of the accident via the input means, the recording means is referred to, and the n + 1-order causes stored in association with the n-order cause are listed in the output means as the support information. Control means for performing support information display control to display,
When the control means is analyzing a certain n-order cause, a new n + 1-order cause other than the n + 1-order cause stored in the storage means in association with the n-order cause is input to the control means. When input by the analyst through the means, the accumulation control is performed so as to store the n + 1-order cause in the storage means in association with the n-order cause, and the analyst continuously inputs the new n + 1-order cause. The support information that refers to the recording means as the n-th order cause and extracts and displays the list as the support information on the output means when there are n + 1-order causes stored in association with the n-order cause. An accident analysis support system that supports the cause analysis of accidents involving human factors, characterized by continuous display control. And an input means, an output means, and a storage means for storing a plurality of causes in association with each other as an n-order cause (n: natural number) and an n + 1-order cause that is the cause of the occurrence of the n-order cause. In the accident analysis support system that supports the cause analysis of accidents involving human factors by the accident analysis support system,
When an analyst inputs an n-order cause of an accident to the accident analysis support system via the input means, an extraction for extracting an n + 1-order cause stored in association with the n-order cause with reference to the recording means Process,
A display step of displaying a list of the n + 1 primary causes extracted in the extraction step as support information on the output means,
When an analysis is performed on a certain n-order cause, a new n + 1-order cause other than the n + 1-order cause stored in the storage means in association with the n-order cause is analyzed via the input means. When the user inputs the n + 1-order cause in association with the n-order cause, an update step of storing the n + 1-order cause in the storage means is performed, and the new n + 1-order cause is continuously recorded as the n-order cause input by the analyst If there are n + 1-order causes stored in association with the n-order cause, the extraction step and the display step of extracting these and displaying them as the support information on the output means are continued. accident analysis support method, characterized in that it carried out. An input device, an output display device, and a storage device that stores a plurality of causes in association with each other as an n-order cause (n: natural number) and an n + 1-order cause that is the cause of the occurrence of the n-order cause. In an accident analysis support program that causes a computer to perform support for the cause analysis of accidents involving human factors,
When an analyst inputs an n-order cause of an accident to the computer via the input device, an extraction step of extracting an n + 1-order cause stored in association with the n-order cause with reference to the storage device;
Causing the computer to execute a display step of displaying a list of n + 1-order causes extracted in the extraction step as support information on the output display device;
When an analysis is performed for a certain n-order cause, a new n + 1-order cause other than the n + 1-order cause stored in the storage means in association with the n-order cause is analyzed via the input device. When the user inputs the n-th cause, the computer executes an update step of storing the n + 1-order cause in the storage device in association with the n-order cause, and the analyst continues to input the new n + 1-order cause. In the extraction step and the display step , when there are n + 1-order causes stored in association with the n-order cause, they are extracted and displayed as a list on the output means as the support information Accident analysis support program characterized in that the computer is continuously executed.

Images