• 한국재난정보학회 논문집
• /
• 제19권3호
• /
• pp.498-509
• /
• 2023

연구목적: 본 논문은 CCTV 영상을 활용한 딥러닝 객체 인식 기술을 적용해 지하공동구 내 쓰러진 관리인력의 검출 방법을 제시하고, 제안 방법의 관리인력 모니터링 적용성을 평가한다. 연구방법: 사람 검출 목적으로 사전 훈련된 YOLOv5와 OpenPose 모델의 추론 결과로부터 쓰러짐을 판별할 수 있는 규칙을 제안하고, 각 모델의 결과를 통합해 지하공동구 내 작업자 쓰러짐 검출에 적용하였다. 연구결과: 제안된 모델로 작업인력의 감지 및 쓰러짐을 판단할 수 있었으나, CCTV와 작업자 간격 및 작업자가 쓰러진 방향에 의존해 검출성능이 영향을 받았다. 또한 지하공동구 작업자에 대해 YOLOv5 기반 쓰러짐 판별 규칙 적용 모델이 거리 및 쓰러짐 방향 의존성이 낮아 OpenPose 기반 모델에 비해 우수한 성능을 보였다. 그 결과 통합된 이중 딥러닝 모델의 쓰러짐 검출 결과는 YOLOv5 결과에 종속되었다. 결론: 제안 모델을 통해 지하공동구 작업자의 이상상황 검출이 가능함을 보였으나, 개별 딥러닝 모델별 사람 감지 성능 차이로 인해 YOLOv5 기반 모델 대비 통합 모델의 쓰러짐 검출 성능 개선은 미미하였다.

• 한국안전학회지
• /
• 제11권1호
• /
• pp.108-120
• /
• 1996

It is a problems in industrial accidents that the knowledge for industrial accidents is obtained by experience, not by experiment. This experiential knowledge is obtained by Investigating accident cases and utilizing those for safety education. Therefore, in this paper, the situation about the serious accident of construction is analyzed by occupation, a kind of construction, time group, season, type of accident, and accidental cause. And the mutual · relations of these factors are studied. The most frequent type of the serious accidents of construction Is the falling accident. It happenes most frequently at apartment construction among kinds of construction and to structural worker, finishing worker, normal worker in order among occupations. And it is found that the most critical causes of the falling accident are the imperfection of safety facilities and unwearing of protection equipments, so a number of accidents can be reduced by the expansion of safety facilities and wearing of protection equipments absolutely. The counterplan of prohibition of accidents and the direction of government policy are presented by a series of nalyses for accident cases.

• 시스템엔지니어링학술지
• /
• 제16권1호
• /
• pp.68-77
• /
• 2020

The shipbuilding and offshore industry is a large and complex assembly industry, which causes many safety accidents. The major accidents in the shipbuilding and offshore industry workplaces are stenosis, falling objects, dust, fire, explosions, and gas poisoning. The accident by worker in this industry mainly has three factors: frequent movement, narrow work space, and increased use of subcontractors. To control these factors, it is necessary to monitor the worker's location and work status. In this paper, a worker location monitoring system using inaudible sound wave was designd that can be used in environments with many metal barriers. The process included deriving stakeholder requirements, transforming to system requirements, designing system architecture, and developing prototype. The prototype was validated by third-party testing agency. As a result, it satisfied the designed performance and verified its feasibility.

• 대한안전경영과학회지
• /
• 제16권3호
• /
• pp.81-88
• /
• 2014

A number of screening tools have been developed to evaluate the human balance capability. Many of them were designed to identify the elderly with falling risk. Three different screening tools, which have been well used many clinical fields, were used in this study. The purpose of this study was to evaluate balance capability in facilities maintenance workers in korea. There were no statistical significance between male and female when evaluated with three different screening tools. However, significant differences were found among the age groups irrespective of gender when evaluated with three different screening tools. The results of three different screening tools in korea showed poor values compared with previous results. These results revealed that facilities maintenance workers faced more critical falling risk in korea.

• 국제학술발표논문집
• /
• The 9th International Conference on Construction Engineering and Project Management
• /
• pp.353-360
• /
• 2022

Due to the dense and complicated working environment, the construction industry is susceptible to many accidents. Worker's fall is a severe problem at the construction site, including falling into holes or openings because of the inadequate coverings as per the safety rules. During the construction or demolition of a building, openings and holes are formed in the floors and roofs. Many workers neglect to cover openings for ease of work while being aware of the risks of holes, openings, and gaps at heights. However, there are safety rules for worker safety; the holes and openings must be covered to prevent falls. The safety inspector typically examines it by visiting the construction site, which is time-consuming and requires safety manager efforts. Therefore, this study presented a worker-driven approach (the worker is involved in the reporting process) to facilitate safety managers by developing integrated computer vision and inertia sensors-based mobile applications to identify openings. The TensorFlow framework is used to design Convolutional Neural Network (CNN); the designed CNN is trained on a custom dataset for binary class openings and covered and deployed on an android smartphone. When an application captures an image, the device also extracts the accelerometer values to determine the inclination in parallel with the classification task of the device to predict the final output as floor (openings/ covered), wall (openings/covered), and roof (openings / covered). The proposed worker-driven approach will be extended with other case scenarios at the construction site.

• 국제학술발표논문집
• /
• The 2th International Conference on Construction Engineering and Project Management
• /
• pp.229-238
• /
• 2007

This paper presents a study on construction safety in the People Democratic Republic of Laos (PDRL). Fifty workers experienced certain injuries in their construction sites and 15 top managers were interviewed in twenty-six construction projects in Vientiane, the capital of PDRL. Research results show that stepping on and/or striking against objects (48%), struck by falling objects (24%), falling of persons (12%) are major types of construction injuries. The paper stresses that the ignorance of top managers about their crucial role in safety improvement, using safety incentive to raise safety performance, lack of thorough understanding about benefit from labor safety performance, and the willingness to cut off safety performance expenditures is considered as obstructions of safety improvement programs. The survey indicates that physical working conditions, relationship among workers, foremen behaviors and the monthly wage were influencing factors to worker's job satisfaction. The study also highlights afternoon as dominant time that led to a large number of injuries.

• 대한안전경영과학회지
• /
• 제16권4호
• /
• pp.203-213
• /
• 2014

Sickness absence is one of the most important indicators for worker's health and occupational safety and health performance. Sickness absence is primarily depended upon sickness but psycho-social factors in workplace may moderate sickness absence. Even though worker is falling into illness, sickness absence can be prevented by job satisfaction. In Korea it is very difficult to find research output about the association of sickness absence with job satisfaction. This study is planned to investigate the effect of job satisfaction on sickness absence. The third Korean Working Conditions Survey done by Occupational Safety and Health Research Institute in 2011 was used to analyze by logistic regression analysis. The result has shown that job satisfaction has statistically significant effect on sickness absence and simultaneously diminish the effect of symptoms experience on sickness absence. The effect of job satisfaction is greater in short term sickness absence than in long term sickness absence. This study has some limitation because of the cross sectional data of Korean Working Conditions Survey. In future, sophisticated statistical analysis may be done with modelling.

• 대한안전경영과학회지
• /
• 제14권3호
• /
• pp.49-57
• /
• 2012

Proportion of falling from height accident at construction work accounts for more than 40%, and the number of injuries is getting increased. So without considering falling from height, we can say that it's hard to achieve our goal(accident prevention). Another critical point that we have to think about theses days is the fact that construction workers are getting older. To be specific, the number of workers who are above 50 years old accounts for 65.6% among the fatalities(2007~2011, KOSHA inspection). Accordingly, to effectively prevent construction accedents, especially falls from height, we need to focus on motion analysis of older construction workers and then make suitable measures for growing accident rates at construction work.

• 한국건축시공학회지
• /
• 제23권1호
• /
• pp.81-92
• /
• 2023

According to the 2020 industrial accident reports of the Ministry of Employment and Labor, the number of fatal accidents in the construction industry over the past 5 years has been higher than in other industries. Of these more than 50% of fatal accidents are initially caused by fall accidents. The central government is intensively managing falling/jamming protection device and the use of personal protective equipment to eradicate the inappropriate factors disrupting safety at construction sites. In addition, although efforts have been made to prevent safety accidents with the proposal of the Special Act on Construction Safety, fatalities on construction sites are constantly occurring. Therefore, this study developed a model that automatically detects the wearing state of the worker's safety helmet and belt using computer vision technology. In considerations of conditions occurring at construction sites, we suggest an optimization method, which has been verified in terms of the accuracy and operation speed of the proposed model. As a result, it is possible to improve the efficiency of inspection and patrol by construction site managers, which is expected to contribute to reinforcing competency of safety management.

• 한국재난정보학회 논문집
• /
• 제20권2호
• /
• pp.302-314
• /
• 2024

연구목적: 본 연구는 지하공동구 내 다수 작업자의 낙상을 자동으로 판별하기 위한 Top-down 방식의 딥러닝 자세 추정 모델 기반 낙상 검출 모델을 제안하고, 제안 모델의 성능을 평가한다. 연구방법: Top-down 방식의 자세 추정모델 중 하나인 YOLOv8-pose로부터 추론된 결과와 낙상 판별 규칙을 결합한 모델을 제시하고, 지하공동구 내 2인 이하 작업자가 출현한 기립 및 낙상 이미지에 대해 모델 성능지표를 평가하였다. 또한 동일한 방법으로 Bottom-up 방식 자세추정모델(OpenPose)을 적용한 결과를 함께 분석하였다. 두 모델의 낙상 검출 결과는 각 딥러닝 모델의 작업자 인식 성능에 의존적이므로, 작업자 쓰러짐과 함께 작업자 존재 여부에 대한 성능지표도 함께 조사하였다. 연구결과: YOLOv8-pose와 OpenPose의 모델의 작업자 인식 성능은 F1-score 기준으로 각각 0.88, 0.71로 두 모델이 유사한 수준이었으나, 낙상 규칙을 적용함에 따라 0.71, 0.23로 저하되었다. 작업자의 신체 일부만 검출되거나 작업자간 구분을 실패하여, OpenPose 기반 낙상 추론 모델의 성능 저하를 야기한 것으로 분석된다. 결론: Top-down 방식의 딥러닝 자세 추정 모델을 사용하는 것이 신체 관절점 인식 및 개별 작업자 구분 측면에서 지하공동구 내 작업자 낙상 검출에 효과적이라 판단된다.