鉄道における安全マネジメントに向けたヒューマンエラーのリスク評価とその支援手法

摘要

In recent years, generally in various industrial fields, various kinds oftroubles resulting from human error are called " organization accident" , andare regarded as questionable. Even if an individual work result becomes atrigger and the accident resulting from human error, the problem of theorganization, i. e.; system and climate for safety are considered to exist asthe background of accidents resulting from human error. According, inorder to prevent human error, systematic management is required. So, thispaper proposes the risk assessment technique (ex-ante analysis) of humanerror, and human factor analysis (ex-post analysis), as supporting methodfor safety management in railway.In the risk assessment of the human error using the existing evaluationstandard, it is hard to grasp the difference in a risk value of being based onthe difference among error conditions. The result of the assessment doesnot appear useful information to acquire in management. On other hand,it is difficult to utilize object data for risk assessment in the generalindustrial field. Then, in this paper, it could use the relative evaluationby specialists with experience of work and safety management. In orderto support their judgment, this paper proposed the valuation modelingreflecting the conditions of work or a system. The relative evaluationtechnique is AHP (Analysis Hierarchy Process) adapting a pair comparisonmethod. It showed examples corresponding to two evaluation levels.When there were requirement of evaluations about the whole task over oneworker, it proposed the rough evaluation by the model of six hierarchystructures. The example is about the operation handling work in a traindriver and “Level_1” of the model is “whose work is it?” “Level_2” isthe system conditions that influence the accident by a driveru27s human error.In the low rank, work mode (Leve_3), the work conditions (Level_4), thecontents of work (Leve_5), and human error actions (Level_6) areconsidered. On the other hand, when there were requirement ofevaluations about simplex work, it proposed the detail evaluation. Itanalyzed the situations which tasks generate in the work, found the pattern.And it analyzed the possibility of progressing to an accident by ask failure.Based on this analysis result, it created the model that structured the task,in the progress scenario for every situation pattern.Next, I proposed human factor analytical skills, as one of themeasures for acquiring the knowledge of the expert about work or its safetyin railway. In railway, the report form of accidents is specified. So itreported by choosing from the existing classification item about the featureand factor of a human error. However, about a regular classification item,it is a stage of collecting items and there is a problem that only an externalfactor cannot be extracted. In other industrial fields, the technique aboutqualitative case analysis is proposed in recent years, and to investigate toorganization / management factor is tried. Although the on-site person incharge is carrying out investigation and analysis also about the smallaccident of damage, such as a transportation obstacle, systematic analyticalskills are not used in particular in many cases, in present condition of themeasure of railway. In order to analyze an exact background factor, theprocedure that specifies the human error that should be analyzed isimportant. Then, this paper proposed the human factor analysis method asqualitative accident analytical skills that the on-site person can use inrailway. It is the technique of grasping simply the feature of thegenerating mechanism of human error (Existence of consciousness control,and existence of action expression), and delving into a background factorto organization / management factor. Moreover, this paper showedvalidity by the example, as a result of re-analyzing using the technique topropose. By using this technique, it became clear that it is inadequate ininvestigation, and it can work on a measure corresponding to the contentsof the human error. Furthermore, it investigated whether education aboutthe procedure could make a safety climate in trial. The result ofinvestigation proved that education about the procedure of human factoranalysis might activate communication in work place, and become onefactor of creation of a safety climate. Especially the effective acquisitionmethods are an exercise and individual guidance.Finally, the result of research proposed in this above paper wassummarized. Moreover, it made reference also about the presentcondition of result practical use. Furthermore, it summarized about thefuture expansive subject.