Probability Of Accident Occurrence Research Articles

A Bayesian network-based model for risk modeling and scenario deduction of collision accidents of inland intelligent ships

Safety is an important premise and foundation for the operation of intelligent ships. This paper introduces a novel scenario analysis framework that employs disaster system theory to produce more comprehensive results for identifying scenario elements and calculating collision risks for inland intelligent ships. The framework is utilized to investigate the collision accident risk evolution mechanism. This process is incorporated into Bayesian Network (BN) modeling for ship collisions on inland rivers. By comparing the change in occurrence probability and consequence severity of risk factors for inland ship collision accidents with and without selected intelligent technologies, the collision risk of intelligent ships is quantified. The results indicate that the application of intelligent technologies, such as ship speed optimization and situational awareness, can reduce the occurrence probability of collision accidents and mitigate the severity of their consequences. Moreover, it has been discovered that such intelligent technologies have a greater impact on accidents with severe consequences than those with minor consequences. This research provides a framework for the preliminary safety evaluation of inland intelligent ships. It is of great significance to accelerate the improvement of navigational risk prevention and response-ability of inland intelligent ships in the future.

Significance ranking and correlation identification of accident causes in process industry based on system thinking and statistical analysis

AbstractAccidents in process industry occur frequently with serious casualties and property losses. This paper builds an accident causation model of process industry based on system thinking by dividing the accident causation system into 4 subsystems and 22 factors. A combination of grey relational analysis and correspondence analysis is conducted to carry out a structured analysis of the collected data. The research contains three main parts: (1) Grey relational analysis is used to obtain the significance ranking of 22 cause factors in process industry, and three critical cause factors are identified as “Security inspection,” “Risk identification,” and “Security awareness.” (2) Through correspondence analysis, the correlations between three sets of variables are analyzed and the cause factors requiring focused attention are identified as “Electric spark,” “Temperature,” “Raw material control,” “Punching phenomenon,” “Equipment clogging,” and “Combustible gases.” (3) An intelligent monitoring scheme is developed for the critical factors of each subsystem, which aims to achieve real‐time monitoring and early warning by means of video surveillance and sensor placement for the human, equipment, and environment subsystems. The conclusions obtained from this study can be used to enhance the efficiency of safety management and reduce the probability of accident occurrence in the process industry.

A Route Choice Model for the Investigation of Drivers’ Willingness to Choose a Flyover Motorway in Greece

The constant evolution of many urban areas ultimately reaches a point where the current infrastructure cannot further serve the needs of citizens. In the case of transport networks, congested roads, increased delay, and low level of service are among the indicators of a need for road infrastructure upgrade. Thessaloniki is the second-largest city in Greece with a population of over 1 million inhabitants in its metropolitan area. Currently, a significant share of the city’s traffic demand is served via its ring road, whose capacity is set to be enhanced through the construction of a flyover highway with the simultaneous upgrade of the existing ring road. The current study aims at investigating the key factors determining the final route choice of drivers between the two road axes. To that end, data from a combined revealed and stated preference survey targeting car drivers were collected, which were later exploited as the basis for the development of binary route choice regression and machine learning models. The results reveal that drivers’ choice is affected by criteria such as total travel time, the probability of accident occurrence, and closure time due to accident. The results of this paper could prove beneficial to transport researchers in forecasting drivers’ behavior in terms of route choice and to practitioners during the planning phase of similar infrastructure projects.

Modelling and Simulating Automated Vehicular Functions in Critical Situations—Application of a Novel Accident Reconstruction Concept

Our paper introduces new reconstruction techniques of real-life critical road traffic accidents focusing on highly automated functions. The investigation method presented here focuses on the effect of relevant control parameters and environmental factors following the concept of sensitivity analysis. Two reconstruction tools are applied, the choice depending on the relevant causal factor of the accidents. Our measurement proves that the technical parameters of the control process, like time to collision or braking pressure that affects user satisfaction directly, can significantly influence the probability of accident occurrence. Thus, it is reasonable to consider safety with an increased weight compared to the user experience when identifying these parameters’ values. On the other hand, the effects of the investigated environmental factors were also found to be significant. Accordingly, future ADAS applications need to consider the change of environmental factors in the case of increased risk level, and driver-mode should be adapted to the new situation.

Research on forecasting approach of key parameters of PWR pressurizer water level

Nuclear power has special safety requirements and must be generated on the premise of nuclear safety. It is necessary to forecast the abnormal situation of system and equipment as soon as possible, so we can control the accident beforehand and prevent the nuclear safety risk caused by the failure and accident. At present, the pre-accident control method in nuclear power plant mainly relies on the subjective experience of personnel. In contrast, intelligent forecasting and timely reasoning of the change trend and possible impact of the key safety-related parameters in nuclear power plant can bring data reference to the staff of the power plant, help to control the accident before it happens and effectively reduce the probability of accident occurrence. In PWR(Pressurized Water Reactor), safety is closely related to the key parameters of primary loop such as pressurizer water level. Adequate water level of pressurizer is the key to ensure the normal discharge of reactor heat, and abnormal changes of pressurizer water level may lead to nuclear accidents. Therefore, our paper analyzes and models the water level control function of the pressurizer in PWR, the operation history of nuclear power plant related to this function is collected as well. Then a network model is built based on transformer to do information mining, which can successfully forecast the trend of key parameters related to pressurized water level and provides reference information for staff.

Human Intrusion Detection in Static Hazardous Areas at Construction Sites: Deep Learning–Based Method

Construction sites have complex environments and high accident rates. Human intrusion into static hazardous areas is a significant cause of accidents. Traditional engineering safety management mainly relies on manual methods, such as patrol inspection by safety supervisors, which is time-consuming and labor-intensive, and it is difficult to achieve a complete safety supervision. In recent years, the emergence of artificial intelligence technology and computer vision has provided a new scheme for intrusion detection. However, existing studies have used a single method for human intrusion judgment in static dangerous areas, without in-depth consideration of the influence of human posture, intrusion direction, and other factors. In this study, a computer vision–based intrusion detection method was developed, mainly aimed at static hazardous areas. The object detection was based on the You Only Look Once (YOLO) V5 module to extract the image feature information. Subsequently, the basic rule of intrusion judgment based on the key points of bounding boxes was formulated, in which the workers’ intrusion direction was recognized and postured using two auxiliary detection modules. Finally, the intrusion rule base was constructed as the basis for human intrusion detection, containing rules with different sensitivities for different intrusion states. The case study indicated that the precision and recall rate of the algorithm were 96.05% and 90.05%, respectively. Overall, this method can effectively address the defects of manual supervision in engineering safety management, reducing the probability of accident occurrence and enhancing safety at construction sites.

THE SAFETY AND EFFICIENCY OF WATER TRANSPORT: STATISTICAL ANALYSIS

This paper provides comprehensive research on the linearity between the efficiency and safety of inland waterway transport (IWT). For this purpose, methods of statistical analysis are used. The efficiency of IWT is expressed by the transportation outputs, which today are strongly affected by the COVID-19 pandemic. This paper aims to find whether the demand for IWT has an impact on the accident occurrence probability. The results show the linearity between shipping accidents and the outputs of freight and passenger IWT. The COVID-19 crisis has a higher impact on the efficiency of passenger navigation than on freight navigation. With reduced demand, the level of safety is higher but at the cost of declining efficiency in the transport sector.

Identification of Critical Risks in Hosting Sports Mega-events: a Social Network Perspective

Identifying the potential risks and their influencing factors in hosting sports mega-events can effectively reduce the probability of accident occurrence in the future. However, traditional risk identification primarily focuses on investigating these factors affecting each category of risk individually and ignores the interaction of risks, which leads to poor performance in risk management efficiency. As such, this study first collects a risk list for sports mega-events by means of content analysis using grounded theory, the original data of which comes from 26 research articles and 88 risk management cases. Then a risk network is constructed through the social network analysis (SNA), followed by a quantitative analysis from both the node level and the whole network level. The results indicate that organization and coordination, personnel training, traffic facilities, political conflicts, racial discrimination, and economic crisis are the critical risk factors in hosting sports mega-events. The corresponding suggestions are finally put forward to help organizers control risks at the source, prevent the spread of risk events, and improve efficiency of risk management.

건축공사 주요 공종의 안전사고 발생확률 예측에 관한 연구

In a construction site with high uncertainty and a lot of risks, a management supervisor can reduce safety accidents by recognizing and removing hazards in advance. Therefore, the prevention possibility of safety accidents may vary depending on the judgment of the management supervisor as follows. 1) If the management supervisor judges that there are no accident possibility because the management supervisor is not aware of the risk. 2) If the management supervisor judges that there are no hazards and there are no accident possibility. 3) If the management supervisor recognizes the hazards and judges that there are accident possibility. This study presented a prediction model of occurrence probability of safety accident using the probability that the management supervisor will make an appropriate judgment about the presence or absence of a risk and the probability that the management supervisor will make an improper judgment as variables. This study defined the causal relationship for the causes and results of safety accidents at construction sites using a fault tree analysis and structured using a Bayesian network. In addition, a basic model for predicting the occurrence probability of safety accident at construction sites was presented using Bayes' theorem, an inductive approach and applied to virtual formwork to examine the consistency of the study model through case studies. As a result of the case study, it was confirmed that the occurrence probability of safety accident in the formwork changed according to the change in the probability of determining that there are hazards but the management supervisor does not recognize them and judges that there are no accident possibility and the change of the probability that the management supervisor recognizes hazards and judges that there are accident possibility. In addition, when the probability of appropriate judgment that there are accident possibility by recognizing hazards decreased by 10%, the occurrence probability of safety accidents decreased by more than 10%. This phenomenon was the same in all types of safety accidents. On the other hand, when there were hazards but the probability of inappropriate judgment that there were no accident possibility increased by 10%, the increase rate of the occurrence probability of safety accident was more than 10%. Through this study, it was confirmed that the occurrence probability of safety accident was updated depending on whether the management supervisor was aware of the safety accident in advance. When the management supervisor did not recognize the hazards, the occurrence probability of safety accident increased. And, the increase range in the occurrence probability of a safety accident was larger than the increase range in the probability of unawareness of the hazards. Therefore, in order to prevent safety accidents at construction sites, it is judged that it is necessary to strengthen the pre-management activities to recognize and remove the hazards inherent in the work through the safety review of the design.

Research on Fault Tree Reconstruction Based on Contingency

The fault tree analysis (FTA) method is an important analysis method for safety system engineering. Traditional accident analysis theory agrees that basic events lead to top events, but it does not fully consider that the accident process is accidental, and the calculation results exaggerate the probability of accident occurrence. This paper selects typical collision accidents, analyzes the shortcomings of the existing fault tree, indicates that there is a contingency in the accident process, constructs a probability fault tree based on the traditional fault tree, and puts forward concepts of “probability AND gate” and “probability OR gate”. In addition, based on the traditional quantitative analysis method of fault trees, calculations of the occurrence probability, probability importance coefficient, and critical importance coefficient of top events are modified, and the modified quantitative calculation is applied to accident cases.

Multi-Factor Rear-End Collision Avoidance in Connected Autonomous Vehicles

According to World Health Organization (WHO), the leading cause of fatalities and injuries is rear-ending collision in vehicles. The critical challenge of the technologically rich transportation system is to reduce the chances of accidents between vehicles. For this purpose, it is especially important to analyze the factors that are the cause of accidents. Based on these factors’ results, this paper presents a driver assistance system for collision avoidance. There are many factors involved in collisions in the existing literature from which we identified some factors which can affect the accident occurrence probability. However, with advancements in the technologies of autonomous vehicles, these factors can be controlled using an onboard driver assistance system. We used MATLAB’s Fuzzy Inference System Tool to analyze the categories of accident contributing factors. Fuzzy results are validated using the VOMAS agent in the NetLogo simulation model. The proposed system can inform the vehicle’s automated system when chances of an accident are higher so that the vehicle may take control from the driver. The proposed research is extremely helpful in handling various kinds of factors involved in accidents. The results of the experiments demonstrated that multi-factor-enabled vehicles could better avoid collision as compared to other vehicles.

Two-dimensional phase lag index image representation of electroencephalography for automated recognition of driver fatigue using convolutional neural network

Driving in fatigue state will increase the occurrence probability of related traffic accidents and cause severe economic and societal problems. To tackle the issue, a deep learning approach is proposed for the automated recognition of driver fatigue using electroencephalography (EEG) signals obtained from real driving. The methodology here proposed consists of converting the multi-channel EEG recording into functional brain network (FBN) adjacency matrices based on phase lag index (PLI) and feeding them into various convolutional neural networks (CNN) as input. These CNN models with convolutional layer, rectifier linear activation unit (ReLU), pooling layer and fully connected layer are designed to extract hidden features from images representing FBN adjacency matrices and then to achieve the two-ways classification task. The experimental results indicate that the highest classification accuracy of 95.4 ± 2.0%, highest sensitivity of 93.9 ± 3.1%, highest precision of 95.5 ± 2.4%, highest F1 score of 94.7 ± 2.0% and highest value of area under the receiver operating curve (AUC-ROC = 0.9953) are achieved using Model 4 based on PLI adjacency matrices as input with the 10-fold cross validation strategy. Indeed, all the CNN models considered in this research achieved accuracy higher than 94.40%. It is hence concluded that the proposed CNN models have the ability to self-learn and pick up more distinguishable features from the input data without a separate feature extraction or feature selection procedure. The experimental results also confirmed the effectiveness of the combination of FBN and CNN for the recognition of driver fatigue.

A Bayesian network based study on determining the relationship between job stress and safety climate factors in occurrence of accidents

BackgroundJob stress and safety climate have been recognized as two crucial factors that can increase the risk of occupational accidents. This study was performed to determine the relationship between job stress and safety climate factors in the occurrence of accidents using the Bayesian network model.MethodsThis cross-sectional study was performed on 1530 male workers of Asaluyeh petrochemical company in Iran. The participants were asked to complete the questionnaires, including demographical information and accident history questionnaire, NIOSH generic job stress questionnaire, and Nordic safety climate questionnaire. Also, work experience and the accident history data were inquired from the petrochemical health unit. Finally, the relationships between the variables were investigated using the Bayesian network model.ResultsA high job stress condition could decrease the high safety climate from 53 to 37% and increase the accident occurrence from 72 to 94%. Moreover, a low safety climate condition could increase the accident occurrence from 72 to 93%. Also, the concurrent high job stress and low safety climate could raise the accident occurrence from 72 to 93%. Among the associations between the job stress factor and safety climate dimensions, the job stress and worker’s safety priority and risk non-acceptance (0.19) had the highest mean influence value.ConclusionThe adverse effect of high job stress conditions on accident occurrence is twofold. It can directly increase the accident occurrence probability and in another way, it can indirectly increase the accident occurrence probability by causing the safety climate to go to a lower level.

Risk-informed based comprehensive path-planning method for radioactive materials road transportation

The optimization of transportation routes of radioactive materials (RADMAT) is recognized as one of the most important measures to reduce the radioactive risks and costs during transportation. However, the route optimization indicators have not been fully considered and quantified in current researches. Here, we propose a Risk-Informed based Comprehensive Path-Planning method (RICPP) for RADMAT Road Transportation (RMRT) route optimization by considering radiological risk cost, time cost, and economical cost. Among which, the ET/FT based Probabilistic Safety Assessment (PSA) model is established to quantitatively assess the accident occurrence probability, and then the PEHR (personnel, environment, hazard target, and rescue force) comprehensive severity indexes considering vulnerability and resilience are established and quantified to characterize the accident radiological consequence. Finally, the gray relation analysis (GRA) model is employed to select an optimal path considering the multi-objective route planning indicators. The case study results show that the obtained path could reasonably reduce risks and costs during the transportation process, which demonstrates the effectiveness of the proposed method. The proposed RICPP methodology is beneficial for selecting a safer and more economical route for radioactive materials road transportation.

A Novel Cap-LSTM Model for Remaining Useful Life Prediction

In recent years, the Remaining Useful Life (RUL) prediction has become a hot spot in Prognostics and Health Management (PHM) research. High-accuracy RUL prediction can reduce the probability of accident occurrence and improve the reliability of the mechanical system. This paper proposes a novel two-channel hybrid model for RUL prediction based on a Capsule Neural Network and a Long Short-Term Memory Network (Cap-LSTM). Conventional Convolutional Neural Networks (CNN) are widely used in RUL prediction. However, the pooling layer of the conventional CNN only extracts the most active part of the multivariate time-series sensor data. Moreover, conventional CNN is not sensitive to the direction and spatial position of features and thus the feature information may not be fully used. To overcome these shortcomings, this paper uses the capsule neural network to directly extract the highly correlated spatial feature information, from the multivariate time-series sensor data, thus avoiding the loss of the spatial position relationship between local features and reducing the complexity of the model. In order to obtain the training and test samples, this paper uses the sliding time window method to preprocess the data. Meanwhile, the piece-wise linear function is used to represent the actual performance degradation of the engine. The NASA C-MAPSS dataset is used to verify the prediction effectiveness of the Cap-LSTM model. The state-of-the-art RUL prediction models are also introduced and compared to the proposed approach, providing results that show the superiority of the presented method.

3D Sight Distance Calculation and Estimation of its Effect on Road Accidents in GIS Environment

Nowadays, one of the main causes of human death is driving accidents. Across the world 1.2 million people die and several million people get injured annually as a result of road accidents. One of the main solutions that is important in resolving the problem of accidents is identifying the factors and their role and impact and the contribution of each in the accident. The stopping sight distance in the route is one of the initial factors in the driver’s performance error as well as the occurrence of an accident, financial losses, and deaths. The geometric design of roads is generally designed according to two-dimensional rules and regulations. Hence, today, given the remarkable advances in computer science and programming, there are many possibilities for 3D modeling of the route. Therefore, the calculation of the stopping sight distance should be based on existing facts and new scientific achievements. On the contrary, the stopping sight distance is one of the factors affecting driving crashes; therefore, this paper tries to investigate the accident occurrence probability in other spots of the existing route by calculating 3D stopping sight distance and using drivers’ free flow speed, as well as using the GIS software. In this way, the results of this study can pave the way for improving the geometric design of existing roads as well as prioritizing the correction of the accidental points of existing routes.

Using N-K Model to quantitatively calculate the variability in Functional Resonance Analysis Method

In this paper, we formulate the formation mechanism of variability of Functional Resonance Analysis Method (FRAM) based on the risk pulse theory for the first time. The coupling and interactions among human errors, mechanical failures, terrible environment and organization factors might cause the system state change, and cause the variability during the operation processes of the system. Based on the collected historical occurrence frequency of each risk factor and the determined coupling and interaction relationships among the risk factors, the N-K model is used to calculate the coupling risk intensity, which provides a quantitative method to evaluate the variability of the functional module. The more frequent the coupling times, the greater the probability of the coupling. The greater the coupling intensity, the greater the risk and the greater the occurrence probability of accidents. Compared with previous improvement works, the historical statistical frequency of each risk factor is need during the calculation based on N-K model, the results are not influenced by experts. A case study is conducted by taking a railway dangerous goods transportation accident in 2001 as the background. The results show that N-K model is able to calculate the variability of the functional module, quantitatively and effectively.

BRIDGING ACADEMICIAN AND COMMUNITY THROUGH KNOWLEDGE TRANSFER PROGRAM (KTP) ON WORKPLACE RISK ASSESSMENT TECHNIQUE- A CASE STUDY

Since technical education involves equipment and workshop/laboratory activities, safety and health risks due to hazardous situations have become the primary issue and require immediate mitigation measures. The main issue that needs immediate attention here is that most ADTEC (Advanced Technology Training Center) staff lack expertise in workplace risk assessment techniques. Some of them are also unaware of these life-threatening safety and health risks. They are unable to anticipate the long-term impact of their everyday activities on their safety and health. Thus, knowledge of hazard identification and prevention strategies is essential for them to improve their workplace environment. This program has been executed in four stages, which are off-the-job training, walkabout and hands-on activity, oral presentation, and followed by an individual assessment. As for an individual assessment, all participants need to sit the written test and an oral interview. The purpose of an individual assessment is to gauge participant understanding and the effectiveness of the program. The program has been conducted successfully whereby all the 32 participants have undergone the training and passed the assessment (written test and interview). All participants gave very encouraging feedback and believed this program would equip them with knowledge on conducting the risk assessment for all activities in their organization. The authors strongly believe that this initiative and program will ultimately assist ADTEC in reducing the probability of accident occurrence in their organization.

Optimal Education Plan of Employees Using Maintenance Model

The employee education is indispensable for companies to improve productive efficiency and product quality. In general, the employee education is divided into two types, i.e., On-job trainings and Off-job ones, and Off-job trainings are divided into two types, i.e., compulsory educations and non-compulsory ones. Compulsory educations such as safety program and compliance education, are necessary to maintain daily production without any accidents. Although all employees of a department and a division gather in a classroom annually in conventional compulsory educations, daily e-learning education complements and strengthens conventional compulsory ones today because employees forget what they learn by annual learning. In past studies, the logit model is used for modeling the influence from education receipt to its memory condition and quantitative relationship between the effect of traffic safety education and the accident-related human-errors was clarified. The effectiveness of the safety driving educational program was indicated by Structural Equation Modelling method. In this paper, an annual compulsory which is complemented and strengthened by e-learning, is discussed. The expected cost rate of education is expressed using imperfect maintenance models and optimal policies which minimizes it is considered. Although we use the exponential function which denotes the occurrence probability of accidents which are caused by forgetting lessons, the actual occurrence probability which is approximated from the actual data would be a complicated one. For solving such complex optimization problems, metaheuristic methods can be applied.

Oil Spill Response Planning For Amazon Region–A Crude Carrier Study Case

It is well-known that an adequate planning process is paramount for a response success in oil spill events. This is especially true for tankers leakages due to its dynamic nature and potential to cause severe environmental harms. Moreover, one of the most challenging planning dimensions for an adequate response is, certainly, the human and material mobilization logistics, which can be strongly limited in remote regions. In South America, one of the most worldwide preserved rainforest areas, the Amazon, which consists of more than 5 million square kilometers and contains a thousand rivers, is also one of the most remote area, with important logistics challenges for almost any human activity. Despite all these difficulties, however, since 1986 Brazil has been producing an average of 35.000 bpd of oil at Urucu Field, located close to the city of Coari, around 400 miles from the province capital, Manaus. Once produced, this oil is pumped through a pipeline until a terminal at Solimões river, where is loaded in tankers destined to Manaus. Although contingency facilities exist in these two cities, there is no support installation in any point over the 342 nautical miles of the journey. Therefore, this study aimed at identifying the major operational risks of this naval transport and at analyzing some specific contingency planning aspects. Initially, a total of 12 safety critical areas were identified during a field inspection trip along Solimões and Amazon rivers. For each point, a qualitative risk analysis was developed, using tropical environmental sensitivities indexes and an empirical oil slick forecast modelling, as well as an oil weathering modelling using the ADIOS2 computational package. Finally, a verification checklist with a scorecard system was developed and applied, in order to assess the oil spill response feasibility and rank the critical spots over the rivers courses. For all the points evaluated, around 30% of all accidental hypothesis were associated with a catastrophic risk, with most being identified as potentially causing a severe environmental impact. Besides, all the estimated response times calculated were insufficient to avoid oiling of the riverbanks. Finally, due also to strong currents and other limiting environmental conditions, all critical point displayed low scores for the response feasibility evaluation, with poor logistics being a key parameter. From these results, navigational safety recommendations and oil spill response planning guidelines were developed in order to minimize an accident occurrence probability and, hence, the risk associated with it.