安全科学与韧性(英文)最新文献

Volunteer teams provide valuable support after large-scale disasters. However, excessive volunteer participation poses challenges for formal operations. Therefore, an appropriate decision-making method is required to quickly determine the number of volunteers required after a disaster. This study proposes a data-driven decision-making (D³M) method for typhoon disaster volunteerism that can effectively predict the number of volunteers required. Disaster data from actual cases were gathered, analyzed, and preprocessed to prepare the model. Feature selection, D³M model training and optimization, and model validation were performed to fine-tune the volunteer participant predictions. Using data from an actual typhoon in the Philippines, the rationality and efficacy of the method were verified through a comparative analysis of the experimental results. The proposed method learns from disaster-event data to quickly predict the number of volunteers needed, such that it not only reasonably allocates volunteers to assist professional teams in rescue but also avoids secondary problems caused by an overwhelming response.

Investigative identification is a routine criminal investigative procedure, the results of which can be used as evidence in litigation. However, some suspects often deny their involvement in the case, and some witnesses may withhold information or misrepresent it, all of which may lead to a miscarriage of justice. This study created a stressful environment and conducted a simulated crime experiment to explore whether eye movement data can be an effective feature for distinguishing perpetrators, innocents, and insiders. The eye movement features—such as the total fixation duration, number of fixations, and first fixation duration—within an area of interest were collected from 83 participants sorted into informed, involved, and innocent groups. The results revealed the following: (1) compared with the object and scene stimuli, subjects with different identities were more likely to exhibit significant differences in eye movement data for the involved and irrelevant portraits. The total fixation duration and the number of fixations can provide a reference for judging whether someone is involved in a case, and the first fixation duration effect was not obvious. (2) Using machine learning algorithms to predict subjects’ identities through eye movement features, it was demonstrated that the involved portrait-object-scene model had the best predictive effect. (3) Multiple algorithmic models were used to distinguish subjects’ identities, and the highest accuracy of 92.7% was achieved for the informed × innocent group, 88% for the innocent × suspect group (including the informed and involved groups), and 84.5% for the involved group. The eye movement analysis method can provide a reference for criminal investigators to distinguish between the perpetrator, insider, and innocent, and offer a novel approach to determining the direction of further investigation and uncovering and verifying case clues.

China's natural disaster situation presents a complex and severe scenario, resulting in substantial human and material losses as a result of large-scale emergencies. Recognizing the significance of aviation emergency rescue, the state provides strong support for its development. However, China's current aviation emergency rescue system is still under construction and encounters various challenges; one such challenge is to match the dynamically changing multi-point rescue demands with the limited availability of aircraft dispatch. We propose a dynamic task assignment model and a trainable model framework for aviation emergency rescue based on multi-agent reinforcement learning. Combined with a targeted design, the scheduling matching problem is transformed into a stochastic game process from the rescue location perspective. Subsequently, an optimized strategy model with high robustness can be obtained by solving the training framework. Comparative experiments demonstrate that the proposed model is able to achieve higher assignment benefits by considering the dynamic nature of rescue demands and the limited availability of rescue helicopter crews. Additionally, the model is able to achieve higher task assignment rates and average time satisfaction by assigning tasks in a more efficient and timely manner. The results suggest that the proposed dynamic task assignment model is a promising approach for improving the efficiency of aviation emergency rescue.

Financial resilience refers to a start-up's capacity to anticipate, plan for, respond to, and adapt to gradual change and abrupt unforeseen shocks to survive and thrive by enacting appropriate economic policies to decrease budget deficits. Economic history tells us that more companies fail to emerge from a downturn than go into or during it. Many studies have been done on financial resilience in many dimensions, but no one has studied start-ups’ organizational readiness for financial resilience. This gap inspires the current research, which uses the Total Interpretive Structural Modelling (TISM) approach to identify financial resilience factors and analyze hierarchical interrelationships start-ups’ organizational readiness factors for financial resilience. This article aims to identify, assess, and categorize start-up organizational preparation elements for financial resilience. The result shows that the first importance should be given to digital financial innovation, liquidity planning, going concern consideration, financial strategy of CFOs, and cyberthreats. Managers of start-ups can utilize the findings of this study to prepare for financial resilience professionally. In a fast-paced environment, start-ups may use financial resilience to gain a competitive edge.

Fire-detection technology plays a critical role in ensuring public safety and facilitating the development of smart cities. Early fire detection is imperative to mitigate potential hazards and minimize associated losses. However, existing vision-based fire-detection methods exhibit limited generalizability and fail to adequately consider the effect of fire object size on detection accuracy. To address this issue, in this study a decoder-free fully transformer-based (DFFT) detector is used to achieve early smoke and flame detection, improving the detection performance for fires of different sizes. This method effectively captures multi-level and multi-scale fire features with rich semantic information while using two powerful encoders to maintain the accuracy of the single-feature map prediction. First, data augmentation is performed to enhance the generalizability of the model. Second, the detection-oriented transformer (DOT) backbone network is treated as a single-layer fire-feature extractor to obtain fire-related features on four scales, which are then fed into an encoder-only single-layer dense prediction module. Finally, the prediction module aggregates the multi-scale fire features into a single feature map using a scale-aggregated encoder (SAE). The prediction module then aligns the classification and regression features using a task-aligned encoder (TAE) to ensure the semantic interaction of the classification and regression predictions. Experimental results on one private dataset and one public dataset demonstrate that the adopted DFFT possesses high detection accuracy and a strong generalizability for fires of different sizes, particularly early small fires. The DFFT achieved mean average precision (mAP) values of 87.40% and 81.12% for the two datasets, outperforming other baseline models. It exhibits a better detection performance on flame objects than on smoke objects because of the prominence of flame features.

This study assessed the influence of occupational stress, individual resilience, and organizational resilience on the safety performance of healthcare providers during the COVID-19 pandemic. Demographic variables including age, work experience, and gender were explored. Data were collected from 344 healthcare providers employed at a teaching hospital. The entropy method and the multi-criteria decision-making (MCDM) method were used to examine the influence of occupational stress, individual resilience, and organizational resilience on the safe performance of healthcare providers. The results of the entropy method showed that organizational resilience was the most influential factor in the safe performance of older healthcare providers. In contrast, individual resilience was the most significant factor in enhancing the safety performance of younger healthcare providers. Analyses of work experience indicated that individual resilience was the most influential factor in the safe performance of less experienced healthcare providers. Gender-based analysis revealed that individual resilience had a major effect on the safety performance of both women and men. The findings of this study could assist managers in improving the performance of the healthcare sector during pandemics by using and implementing resilience concepts at both the individual and organizational levels.

Crisis information dissemination plays a key role in the development of emergency responses to epidemic-level public health events. Therefore, clarifying the causes of crisis information dissemination and making accurate predictions to effectively control such situations have attracted extensive attention. Based on media richness theory and persuasion theory, this study constructs an index system of crisis information dissemination impact factors from two aspects: the crisis information publisher and the published crisis information content. A multi-layer perceptron is used to analyze the weight of the index system, and the prediction is transformed into a pattern classification problem to test crisis information dissemination. In this study, COVID-19 is considered a representative event. An experiment is conducted to predict the crisis information dissemination of COVID-19 in two megacities. Data related to COVID-19 from these two megacities are acquired from the well-known Chinese social media platform Weibo. The experimental results show that not only the identity but also the social influence of the information publisher has a significant impact on crisis information dissemination in epidemic-level public health events. Furthermore, the proposed model achieves more than 95% test accuracy, precision rate, recall value and f1-score in the prediction task. The study provides decision-making support for government departments and a guide for correctly disseminating crisis information and public opinion during future epidemic-level public health events.

An outbreak of Ebola virus disease (EVD) (Sudan virus) was first reported in the Republic of Uganda on September 20, 2022. As of November 17, 2022, 151 confirmed cases have been reported, including 55 deaths (with a case fatality rate among confirmed cases of 39%).

During the EVD outbreak in 2013–2016, international travel played a significant role in the spread of the disease across national borders. During that time, several tasks requiring improvement were identified in the World Health Organization (WHO) Southeast Asia Region (SEAR), including inadequate risk communication and risk assessment, data management gaps for surveillance purposes, inadequate capacity in molecular diagnostic techniques, lack of adequate planning for a surge of cases, and inadequate isolation rooms.

It is therefore recommended that all countries of the WHO SEAR revisit their level of Ebola preparedness and address existing gaps. The emergence and rapid global spread of coronavirus disease 2019 (COVID-19) have reiterated that the world has become a global village. This was confirmed by the spread of the monkeypox virus with cases reported in the WHO SEAR. Therefore, given the weak health infrastructure in the region, complacency could wreak havoc on the healthcare system if another epidemic emerges without an adequate level of preparedness.

Artificial intelligence generated content (AIGC) is a production method based on artificial intelligence (AI) technology that finds rules through data and automatically generates content. In contrast to computational intelligence, generative AI, as exemplified by ChatGPT, exhibits characteristics that increasingly resemble human-level comprehension and creation processes. This paper provides a detailed technical framework and history of ChatGPT, followed by an examination of the challenges posed to political security, military security, economic security, cultural security, social security, ethical security, legal security, machine escape problems, and information leakage. Finally, this paper discusses the potential opportunities that AIGC presents in the realms of politics, military, cybersecurity, society, and public safety education.