There has been a significant decline in the number of rail accidents in Korea since system safety management activities were introduced. Nonetheless, analyzing and preventing human error-related accidents is still an important issue in railway industry. As a railway system is increasingly automated and intelligent, the mechanism and process of an accident occurrence are more and more complicated. It is now essential to consider a variety of factors and their intricate interactions in the analysis of rail accidents. However, it has proved that traditional accident models and methods based on a linear cause-effect relationship are inadequate to analyze and to assess accidents in complex systems such as railway systems. In order to supplement the limitations of traditional safety methods, recently some systemic safety models and methods have been developed. Of those, FRAM(Functional Resonance Analysis Method) has been recognized as one of the most useful methods for analyzing accidents in complex systems. It reflects the concepts of performance adjustment and performance variability in a system, which are fundamental to understanding the processes of an accident in complex systems. This study aims to apply FRAM to the analysis of a rail accident involving human errors, which occurred recently in South Korea. Through the application of FRAM, we found that it can be a useful alternative to traditional methods in the analysis and assessment of accidents in complex systems. In addition, it was also found that FRAM can help analysts understand the interactions between functional elements of a system in a systematic manner.
본 연구의 목적은 실제 선박 충돌사고 사례를 조사하여 선박 충돌상황에서 충돌 원인별 선장 및 당직 해기사의 인적과실 유발 요인을 통계적으로 분석하여 규명함으로써 해기사의 선박 충돌예방에 기여하고자 하는 것이다. 연구대상은 2010년부터 2016년까지 7년 동 안의 상선과 상선, 상선과 어선 간의 선박 충돌사고 중 분석기준에 적합한 총 109건 218척(피항선109척, 유지선 109척)을 대상으로 하였으 며 선종, 피항선과 유지선, 인적과실 측면의 충돌원인 등의 항목으로 구분하여 데이터를 수집하였고 상선에서 해기사의 충돌사고 유발 요 인 규명에 중점을 두고 통계분석 도구인 SPSS를 이용하여 빈도분석과 교차분석을 실시하여 해기사의 인적과실 유발요인을 도출하였다. 분석결과 피항선에서는 레이다 감시를 포함한 경계소홀(74.3%) 및 상대선 지속관찰 소홀(17.4%) 순 이었으며, 유지선 에서는 적절한 피항 협력동작 미 이행(63.3%)이 주요 요인이었다. 특히 상선의 경계소홀 유형 대부분이 상대선 초인 후 지속관찰을 소홀히 한 점이며 미 경계 원인과 당직근무 태만의 공통요인은 항해당직 시간에 다른 업무에 치중하였기 때문인 것으로 나타났다.
To verify the effect of driver's personal characteristics of driver on the accident frequency through railway accidents caused by human errors and the relationship with aptitude test. To prove the relevance between the driver's personal characteristics and human error accidents. Accident data from 2010 to 2011 was analyzed which collected from a train crew department in K national corporation, and 31 drivers gave an personal interview from Sep. 2011 to Nov. 2011 who had controlled a train alone and caused an accident. Compared between driver's personal characteristics and accident rate, and accident induction possibility surveyed from normal person and disqualified in aptitude tests. Accidents was occurred with the age 40s (27%) and 50s (25%), and with the experience between 15 years and 20 years (38%) and over 20 years (30%). Because more aged, more experienced, it can be seen in the correlation between driver's age and accidents induction caused by human errors like illusion. First of all it must be checked whether working conditions and environmental factors are human error-prone. Most accidents occur when received civil complaints or manager at the riding. Therefore accidents can be prevented when investigated through subsequent surveys how often human error happens, even though no accident, and safety device installed based on the error frequency.
Humans are well-known for being adept at using intuition and expertise in many situations. However, human experts are still susceptible to errors in judgment or execution, and failure to recognize the limits of knowledge. This would happen especially in semi-structured situations, in multi-disciplinary settings, under time or other stress, under uncertainty, or when knowledge is outdated Human errors are caused by cognitive biases, attentional slips/memory lapses, cultural motivations, and missing knowledge. The purpose of this research is to study errors of human experts committed in judgment and the general idea of critiquing systems as corresponding plan. Compared to expert systems, critiquing systems are narrowly focused programs useful in limited situations for collaborating with and supporting experts in their task activities. It supports an expert by detecting the human's errors by deploying various strategies that stimulate humans to improve their performance. A variety of types of critiquing systems has spread through numerous application areas.
There is no universally agreed classification of human error, nor is there one in prospect. Thus, a taxonomy is usually made for a specific purpose. To seek the types of human errors in the environment of man-machine interface under the railway industry, we develop a cognitive information processing model incorporating the human's mental states. Using the model, this study investigates the types of human errors about the railway workers. Thus, a survey is conducted for railway safety personnel-locomotive engineers, station employees, and train commanders- in Korean railway company. Through the survey that is designed to investigate four types of human errors from the Questionnaires composed of thirty Questions, we analyze the types of human errors related to railway safety according to affiliated offices, operation shifts, age, and working years. Finally, from the insights of the results some guidelines for the railway safety management are presented.
Formal Safety Assessment (FSA) has been mostly implemented on the hardware aspects of vessels. Although there are guidelines regarding human error FSAs, there have not been many assessments in such areas. To this end, this study seeks to use precedent studies for the safe operation of DP vessels, conducting an FSA regarding human error of DP LOP (Loss of Position) incidents. For this, the study referred to precedent studies for the frequency of DP LOP incidents caused by human errors, adding the severity of LOP incidents, and then applying them to the Bayesian network. As a result, the study was able to confirm that among DP LOP incidents caused by human errors, the drive-off from skill-based errors was 74.3% and the drive-off from unsafe supervision was 50.5%. Based on such results, RCOs (Risk Control Options) were devised through a brainstorming session with experts coming up with proposals including providing mandatory DPO training, installing DP simulator on the vessels, drawing up measures to understanding the procedures for safe operation of DP vessels. Moreover, it was found that mandatory DPO training is reasonable in terms of cost benefits and that while installing a DP simulator is not suitable in terms of cost benefits, it can significantly reduce risks when operating DP vessels.
지금까지(1988∼2000년) 국내 해양안전심판원에서 재결한 선박충돌사고의 원인분류 통계데이터를 살펴보면 전체 2,290건의 인적과실(Human Error) 중에서 상대선박에 대한 경계의무소홀이 929건으로 약 40.6%나 되는 가장 많은 비중을 차지하고 있는 것으로 파악되고 있다. 선박충돌사고는 좌초사고와 더불어 인적과실로 기인한 수많은 인명피해와 재산피해 및 해양환경오염을 유발하는 심각한 해양사고로서 이에 대한 사고원인을 철저히 분석하고 그 예방대책 마련이 무엇보다 중요한 과제가 되고 있다. 따라서, 본 연구에서는 선박충돌사고의 인적과실을 분석하기 위한 목적으로 목포해양안전심판원에서 재결한 선박충돌사고(1990∼2002, 65건)에 대하여 상대선 경계의무소홀이나 동정감시 불충분으로 기인하여 발생한 충돌사고를 조사항목별로 분석하고, GEMS 동적모델을 이용하여 선박충돌사고의 인적과실 유형을 체계적으로 분류하였다.