安全科学与韧性(英文)最新文献

Continuous Renal Replacement Therapy (CRRT) serves as an intervention strategy for the management of acute kidney injury (AKI) in critically ill patients. However, owing to its complex nature and the potential for complications, the implementation of CRRT demands continuous monitoring to prevent patient safety risks. This study aims to identify and validate prevalent risks linked to CRRT within a real-world clinical setting, intending to propose preventive measures grounded in expert insights. To systematically categorize and visually depict the risks, their consequences, preventive measures, and recovery controls, our study employed the Bowtie method in conjunction with the Systems Engineering Initiative for Patient Safety (SEIPS) model. In addition to considering patient-related factors that exhibit variability among critically ill individuals, our key findings showed that the most influential risks impacting the effective delivery of CRRT are incidents of clotted filters, bleeding risks arising from the necessity of anticoagulation for filter efficacy, vascular catheter-related bloodstream infections, variations in proficiency levels among healthcare professionals regarding CRRT modalities, especially in operating the CRRT machines, high nursing workload, frequent nursing turnover, occurrences of hypophosphatemia, variability in CRRT prescribing patterns, and issues related to communication among stakeholders. This research sheds light on the primary risks associated with CRRT and provides practical and viable strategies for effective management. Furthermore, the Bowtie diagram developed as part of this study serves as a valuable tool for visually representing the healthcare system and facilitating the identification of system-related risks within healthcare settings.

During emergency evacuation, it is crucial to accurately detect and classify different groups of evacuees based on their behaviours using computer vision. Traditional object detection models trained on standard image databases often fail to recognise individuals in specific groups such as the elderly, disabled individuals and pregnant women, who require additional assistance during emergencies. To address this limitation, this study proposes a novel image dataset called the Human Behaviour Detection Dataset (HBDset), specifically collected and annotated for public safety and emergency response purposes. This dataset contains eight types of human behaviour categories, i.e. the normal adult, child, holding a crutch, holding a baby, using a wheelchair, pregnant woman, lugging luggage and using a mobile phone. The dataset comprises more than 1,500 images collected from various public scenarios, with more than 2,900 bounding box annotations. The images were carefully selected, cleaned and subsequently manually annotated using the LabelImg tool. To demonstrate the effectiveness of the dataset, classical object detection algorithms were trained and tested based on the HBDset, and the average detection accuracy exceeds 90 %, highlighting the robustness and universality of the dataset. The developed open HBDset has the potential to enhance public safety, provide early disaster warnings and prioritise the needs of vulnerable individuals during emergency evacuation.

Guided by the intelligent construction of coal mines and integrating the concept of the Cyber-Physical System (CPS), a novel approach to a coal mine information-physical fusion system is proposed. This concept emerges from the advancement of intelligent mines and the growing need for enhanced safety in coal mine operations. From a system safety perspective, this approach aims to effectively evaluate security issues by better integrating the informational and physical layers of intelligent mines. Currently, research in areas such as power grids and water networks has widely adopted concepts of cascade failures and robustness. It is suggested that the CPS framework be adapted to the mining sector. Considering the critical role of coal mines in China's energy security and their vulnerability to cyber attacks, a mine-specific CPS system is designed. This system incorporates a model of mine information-physical fusion, including networks for mine operations, management, and control, tailored to the unique conditions of mine shafts and underground settings. Utilizing Markov process theory, the paper delves into the theoretical study of the cascade failure process under deliberate attack strategies. This analysis is essential to effectively assess the safety issues within the information-physical system, providing vital support for fault prediction and safety analysis. The model of cascading failures in the mine information-physical system is developed, and based on an analysis of resilience and robustness, the study identifies key factors influencing the security of mine information-physical systems. Critical factors affecting the reliability of these systems include the number of system nodes, node coupling, edge count, and the interconnection methods of the dual-layer dependency network. The paper concludes by discussing the challenges in analyzing the security of information-physical systems and suggesting directions for future research.

Staircase choice is one of the most critical factors leading to the difference in pedestrian flow and evacuation routes in buildings with multiple staircases. Neither the shortest path to the building exit nor the locally quickest path to the nearest staircase can represent the natural mode of evacuation path choices for an authentic evacuation simulation. Thus, a prediction-based approach is established to predict and simulate evacuation choices, which helps to address three key issues: (1) extracting evacuation data through a controlled experiment; (2) establishing a Logit model for staircase choice prediction based on experimental data; (3) developing a prediction-based cellular automaton model. The proposed approach has achieved the coupling between choice prediction and evacuation simulation. A comparison with Pathfinder software is conducted to reveal the superiority of the prediction-based CA model for simulating staircase choice.

Safety training is the exercise normally conducted for all the current and future employees of a company to identify and recognize occupational hazards and diseases as well as determine the appropriate controlling methods. Moreover, virtual reality (VR) is a technology developed to virtually simulate the surrounding environment to ensure immersive experience and interaction through artificial three-dimensional (3D) platforms. VR devices have been developed to be more compact, easy to use, and affordable to enable people to enjoy immersive virtual experiences and provide interactive and realistic content. This has made technology one of the most popular forms of media for different kinds of training, such as safety-related ones. Therefore, this study aimed to review the use of VR in safety training through the systematic literature review (SLR) method. The process focused on developing 4 primary questions (PQs) classified into 11 systematic research questions (SRQs) for discussion points concerning current developments in VR technology applications. These were further combined with the preferred reporting items for systematic reviews and meta-analyses (PRISMA) flow diagrams in selecting the relevant literature. The questions were also used to investigate the scenarios, methods, objectives, and outcomes of previous studies. The results showed the need for further studies on the application of VR technology in safety training in other fields such as firefighting, chemical industry, maritime, etc. Furthermore, several scenarios such as construction design, disaster response, rescue procedures, and others need to be included. This study also provides information on the gaps for future study, including the exploration of a broader range of industries and VR scenarios.

In recent years, research has focused heavily on the investigation of functionalized ammonium polyphosphate (APP) flame retardants to improve the fire safety of epoxy resins (EP). The reason for this is the dual nature of APP's performance in fire protection of EP. This article provides a comprehensive overview of the advances in the use of functionalized APP flame retardants to improve the fire resistance of EP materials. It then presents the improvement of the modification of the functionalized APP flame retardants in terms of the hydrophobicity, compatibility and catalytic ability of the flame retardants, as well as the effects on the fire resistance, heat resistance, smoke reduction and mechanical properties of the EP composites. After the summary and comparison of the relevant studies, it is clear that the functionalized APP flame retardants can effectively improve the fire safety of EP composites and offset the adverse effects of APP in EP flame retardant applications. In addition, APP flame retardants can obtain various excellent functions through the use of materials with different properties, and the interaction between APP and materials can also lead to more efficient fire protection. However, the current problem is to find ways to streamline the process and minimise the costs associated with functionalized APP flame retardants, as well as to use them effectively in industrial production. We hope that this review can provide valuable hints and insights for the practical application of functionalized APP in EP and perspectives for future research.

Evaluating the resilience of the innovation ecosystem to maintain its performance, in the sense of resistance to disruption and recovery after it, has recently received more attention. Several studies have been conducted to model different ecosystems and evaluate their resilience. However, modeling the innovation ecosystem from a holistic perspective and performing a quantitative assessment of its resilience have received less attention. This paper models the innovation ecosystem holistically and evaluates its resilience index using a quantitative approach through five main steps. In the first step, a case study related to the innovation ecosystem of Iran's Ministry of Energy, called the Power Innovation Ecosystem, is modeled by combining system dynamics and agent-based modeling. Upon validating the model in the second step, the disruption of the loss of experts is investigated in the third step, and all possible actions to recover each actor are analyzed. In the fourth step, the performance of the ecosystem is simulated before and after the disruption using the data gathered in the previous steps. Finally, resilience is calculated in two different ways in the fifth step. Several improvement solutions are also suggested when considering that the resilience index of the innovation ecosystem is at a medium level. This research may assist policymakers in observing the resilience level of the innovation ecosystem based on the proposed model. By applying strategic changes to this model, they can determine the effects of their policies and make the most appropriate decisions to increase the resilience of the innovation ecosystem.

The implementation of early and accurate detection of aircraft cargo compartment fire is of great significance to ensure flight safety. The current airborne fire detection technology mostly relies on single-parameter smoke detection using infrared light. This often results in a high false alarm rate in complex air transportation environments. The traditional deep learning model struggles to effectively address the issue of long-term dependency in multivariate fire information. This paper proposes a multi-technology collaborative fire detection method based on an improved transformers model. Dual-wavelength optical sensors, flue gas analyzers, and other equipment are used to carry out multi-technology collaborative detection methods and characterize various feature dimensions of fire to improve detection accuracy. The improved Transformer model which integrates the self-attention mechanism and position encoding mechanism is applied to the problem of long-time series modeling of fire information from a global perspective, which effectively solves the problem of gradient disappearance and gradient explosion in traditional RNN (recurrent neural network) and CNN (convolutional neural network). Two different multi-head self-attention mechanisms are used to classify and model multivariate fire information, respectively, which solves the problem of confusing time series modeling and classification modeling in dealing with multivariate classification tasks by a single attention mechanism. Finally, the output results of the two models are fused through the gate mechanism. The research results show that, compared with the traditional single-feature detection technology, the multi-technology collaborative fire detection method can better capture fire information. Compared with the traditional deep learning model, the multivariate fire prediction model constructed by the improved Transformer can better detect fires, and the accuracy rate is 0.995.

Complex disaster systems involve various components and mechanisms that could interact in complex ways and change over time, leading to significant deep uncertainty. Due to deep uncertainty, decision-makers have severe inadequacy of knowledge and often encounter unpredictable surprises that may emerge in the future, thus making it difficult to specify appropriate models and parameters to describe the system of interest. In this paper, we propose a dynamic exploratory hybrid modeling framework that fits data, models, and computational experiments together to simulate complex systems with deep uncertainty. In the framework, one needs to develop multiple plausible models from a hybrid modeling perspective and perform enormous computational experiments to explore the diversity of future scenarios. Real-time data is then incorporated into diverse forecasts to dynamically adjust the simulation system. This ultimately enables an ongoing modeling and analysis process in which deep uncertainty would be gradually mitigated. Our approach has been applied to a human-involved car-following system simulation under complex traffic conditions. The results show that the proposed approach can improve the prediction accuracy while enhancing the sensitivity of the simulation system to uncertain changes in the system of interest.

The COVID-19 outbreak had a significant negative impact on the world, and the fifth wave of COVID-19 in Hong Kong brought a considerable shock to Chinese society. There is a growing call for more resilient cities. However, empirical evidence and validation of modeling studies of resilience indicators for urban community responses to the COVID-19 pandemic still need to be provided. In this study, a resilience assessment indicator model comprising 4 subsystems, 7 indicators, and 12 variables was developed to assess the resilience of Hong Kong communities in response to COVID-19 (i.e., Resilience Index). Furthermore, this study utilized regression models such as geographically weighted regression (GWR) and multiscale GWR (MGWR) to validate the resilience model proposed in this study at the model and variable levels. In the regression model, the Resilience Index and the individual variables in the resilience model are explanatory variables, and the outcomes of the COVID-19 pandemic (confirmed cases, confirmation rate, discharged cases, discharge rate) are dependent variables. The results showed that: (i) the resilience of Hong Kong communities to the COVID-19 pandemic was not strong in general and showed some clustered spatial distribution characteristics; (ii) the validation results at the model level showed that the Resilience Index did not explain the consequences of the COVID-19 pandemic to a high degree; (iii) the validation results at the variable level showed that the MGWR model was the best at identifying the relationships between explanatory variables and the dependent variable; and (iv) compared with the model-level assessment results, the variable-level assessment explained the consequences of the COVID-19 pandemic better than the model level assessment results. The above analysis and the spatial distribution maps of the resilience variables can provide empirically based and targeted insights for policymakers.