Risk Analysis最新文献

Human drivers and machine drivers (i.e., automated vehicles or AVs) will share roads and interact with each other, creating mixed traffic. In this perspective, we develop two mental models about them and their social interactions, aiming to understand the risk implications of AVs and mixed traffic. Based on Mental Model I (i.e., machine drivers are superior drivers without human weaknesses), many simulation-based safety assessments, which often overlook or oversimplify human-AV social interactions, have predicted significant safety benefits when machine drivers interact with or replace human drivers. In contrast, Mental Model II considers human and machine drivers as heterogeneous and incompatible, suggesting that their interactions may lead to unexpected and occasionally negative outcomes, particularly in imminent mixed traffic. This perspective gains support from recent comparative empirical studies that employ various methods such as survey experiments, driving simulators, test-tracks, on-road observations, and AV accident analysis. These studies provide initial evidence of emerging traffic risks arising from human-AV social interactions, including human drivers' aggression and road rage toward AVs, human drivers exploiting AVs, AVs exerting negative peer influences on human drivers, and their incompatibility increasing human drivers' challenges in joining mixed traffic and thus risky behaviors. We propose specific suggestions to mitigate problematic human-AV social interactions and the associated emerging risks.

Working with data about homicide victims and perpetrators from 50 of America's largest cities, we investigate the explanatory power of some familiar explanations for why murder in those cities rose sharply in 2020. The analysis reveals that the distribution of risk by race was essentially the same in 2020 as in 2019. That empirical finding challenges some theories of how racial tensions after the death of George Floyd may have driven homicide increases. Similarly, homicide growth was not concentrated in those cities with the greatest availability in 2020 of new and older guns, or among the cities that suffered the most from the COVID-19 pandemic. At a minimum, the cross-city outcomes should reduce confidence that some combination of "race, guns, and COVID-19" explains all of the most important aspects of what happened in 2020.

The Horn of Africa Drylands are increasingly experiencing severe droughts, which impose a threat on traditional livelihood strategies. Understanding adaptation behavior in rural communities is key to helping reduce the impact of these droughts. We investigate adaptation behavior by assessing four established economic and social psychological theories on decision making under risk: expected utility theory (EUT), rank dependent utility theory (RDU), protection motivation theory (PMT), and theory of planned behavior (PMT). To measure adaptation behavior and the theory constructs, we conducted a household survey in Kenya (N = 502). Regression analysis shows that the economic theories (EUT and RDU) have the best fit for our data. Risk and time preferences are found to play an important role in adaptation decisions. An analysis of differences in decision making for distinct types of adaptation measures shows that risk averse (agro-)pastoralists are more likely to implement adaptation measures that are adjustments to their current livelihood practices, and less willing to invest in adaptation measures that require a shift to other livelihood activities. Moreover, we find significant effects for elements of the social psychological theories (PMT and TPB). A person's belief in their own ability to implement an adaptation measure (perceived self-efficacy) and adaptation by family and friends are important factors in explaining adaptation decisions. Finally, we find that the type of adaptation measures that people implement is influenced by, among others, gender, education level, access to financial resources, and access to government support or aid. Our analysis gives insights into the drivers of individual adaptation decisions, which can enhance policies promoting adaptation of dryland communities.

This article presents a risk analysis of large language models (LLMs), a type of "generative" artificial intelligence (AI) system that produces text, commonly in response to textual inputs from human users. The article is specifically focused on the risk of LLMs causing an extreme catastrophe in which they do something akin to taking over the world and killing everyone. The possibility of LLM takeover catastrophe has been a major point of public discussion since the recent release of remarkably capable LLMs such as ChatGPT and GPT-4. This arguably marks the first time when actual AI systems (and not hypothetical future systems) have sparked concern about takeover catastrophe. The article's analysis compares (A) characteristics of AI systems that may be needed for takeover, as identified in prior theoretical literature on AI takeover risk, with (B) characteristics observed in current LLMs. This comparison reveals that the capabilities of current LLMs appear to fall well short of what may be needed for takeover catastrophe. Future LLMs may be similarly incapable due to fundamental limitations of deep learning algorithms. However, divided expert opinion on deep learning and surprise capabilities found in current LLMs suggests some risk of takeover catastrophe from future LLMs. LLM governance should monitor for changes in takeover characteristics and be prepared to proceed more aggressively if warning signs emerge. Unless and until such signs emerge, more aggressive governance measures may be unwarranted.

As countries and communities grapple with climate change, they seek to rapidly decarbonize their economies and cultures. A low-carbon future will likely depend on more distributed solar energy, the electrification of mobility, and more efficient homes and buildings. But what emergent risks are evident within this low-carbon society? This exploratory study first reviews the existing literature to identify 75 risk-risk tradeoffs by their category, medium of distribution, and type. It builds on these 75 examples to apply a typology of Risk Offsets, Risk Substitution, Risk Transfer, and Risk Transformation. Based on extensive document analysis, it applies that typology to three low-carbon innovations: solar energy, battery electric vehicles, and building energy efficiency retrofits, identifying 36 distinct risk-risk tradeoffs in total. As such, the paper moves to discuss complexities and challenges in risk management. In doing so, it calls for a more refined risk assessment that better accounts for decision-making considerations such as the magnitude or probability of risk, size of population exposed, certainty in risk estimation, severity of adverse outcome, distributional considerations, and the timing of risk impacts. It also summarizes emergent research gaps. Risk management in the context of climate action becomes a three-dimensional chess game of weighing risk transmission, risk mediums, and risk categories.

Distributed clean, reliable energy resources like solar plus battery storage (solar + storage) can reduce harmful emissions while supporting resilience. Solar + storage-powered resilience hubs provide energy for critical services during disasters while increasing human adaptive capacity year round. We studied where utility rates, local climate, and historical injustice make solar + storage resilience hubs more valuable and more challenging. We modeled the economic and climate impacts of outfitting candidate hub sites across California with solar + storage for everyday operations and identified designs and costs required to withstand a range of outages considering weather impacts on energy needs and availability. We integrated sociodemographic data to prioritize the siting of resilience hubs, to focus potential policy and funding priorities on regions where solar + storage for resilience hubs is hard or expensive, and where populations are most in need. We identified almost 20,000 candidate buildings with more than 8 GW of total rooftop solar potential capable of reducing CO₂ emissions by 5 million tons per year while providing energy for community resilience. Hub capacity for one of the most challenging missions-providing emergency shelter during a power outage and smoke event-could have a statewide average lifetime cost of less than $2000 per seat. We identified regional challenges including insufficient rooftop solar capacity in cities, low sunlight in northern coastal California, and high costs driven by utility rate structures in Sacramento and the Imperial Valley. Results show that rates and net metering rules that incentivize solar + storage during everyday operations decrease resilience costs.

Due to the pervasive uncertainty in human society, super large and megacities are increasingly prone to becoming high-risk areas. However, the construction of urban resilience in this new era lacks sufficient research on the core conditions and complex interactive mechanisms governing it. Hence, this study proposes a specialized event-oriented framework for governing urban resilience in China based on the pressure-state-response (PSR) theory. We examined COVID-19 cases in 30 cities across China and analyzed the distribution of prevention and control achievements between high-level and non-high-level conditions. Our findings reveal the following key points: (1) High-level achievements in COVID-19 prevention and control rely on three condition configurations: non-pressure-responsive type, pressure-state type, and pressure-responsive type. (2) High economic resilience may indicate a robust state of urban systems amid demographic pressures. In cities experiencing fewer event pressure factors, the application of digital technology plays a crucial role in daily urban management. (3) The implementation of flexible policies proves beneficial in mitigating the impact of objective pressure conditions, such as environmental factors, on urban resilience.

Global pandemics restrict long-haul mobility and international trade. To restore air traffic, a policy named "travel bubble" was implemented during the recent COVID-19 pandemic, which seeks to re-establish air connections among specific countries by permitting unrestricted passenger travel without mandatory quarantine upon arrival. However, travel bubbles are prone to bursting for safety reasons, and how to develop an effective restoration plan through travel bubbles is under-explored. Thus, it is vital to learn from COVID-19 and develop a formal framework for implementing travel bubble therapy for future public health emergencies. This article conducts an analytical investigation of the air travel bubble problem from a network design standpoint. First, a link-based network design problem is established with the goal of minimizing the total infection risk during air travel. Then, based on the relationship between origin-destination pairs and international candidate links, the model is reformulated into a path-based one. A Lagrangian relaxation-based solution framework is proposed to determine the optimal restored international air routes and assign the traffic flow. Finally, computational experiments on both hypothetical data and real-world cases are conducted to examine the algorithm's performance. The results demonstrate the effectiveness and efficiency of the proposed model and algorithm. In addition, compared to a benchmark strategy, it is found that the infection risk under the proposed travel bubble strategy can be reduced by up to 45.2%. More importantly, this work provides practical insights into developing pandemic-induced air transport recovery schemes for both policymakers and aviation operations regulators.

Benzophenone (BP) and BP derivatives (BPDs) are widely used as ultraviolet (UV) stabilizers in food packaging materials and as photoinitiators in UV-curable inks for printing on food-contact materials. However, our knowledge regarding the sources and risks of dietary exposure to BP and BPDs in cereals remains limited, which prompted us to conduct this study. We measured the levels of BP and nine BPDs-BP-1, BP-2, BP-3, BP-8, 2-hydroxybenzophenone, 4-hydroxybenzophenone, 4-methylbenzophenone (4-MBP), methyl-2-benzoylbenzoate, and 4-benzoylbiphenyl-in three types of cereals (rice flour, oatmeal, and cornflakes; 180 samples in total). A Bayesian Markov-chain Monte Carlo (MC) simulation approach was used for deriving the posterior distributions of BP and BPD residues. This approach helped in addressing the uncertainty in probabilistic distribution for the sampled data under the detection limit. Through an MC simulation, we calculated the daily exposure levels of dietary BP and BPDs and corresponding health risks. The results revealed the ubiquitous presence of BP, BP-3, and 4-MBP in cereals. Older adults (aged >65 years) had the highest (97.5 percentile) lifetime carcinogenic risk for BP exposure through cereals (9.41 × 10^-7), whereas children aged 0-3 years had the highest (97.5 percentile) hazard indices for BPD exposure through cereals (2.5 × 10^-2). Nevertheless, across age groups, the lifetime carcinogenic risks of BP exposure through cereals were acceptable, and the hazard indices for BPD exposure through cereals were <1. Therefore, BPD exposure through cereals may not be a health concern for individuals in Taiwan.

The enhancing risk from human action and multi-hazard interaction has substantially complicated the hazard-society relationship. The underlying vulnerabilities are crucial in predicting the probable impact to be caused by multi-hazards. Thus, the evaluation of social vulnerability is decisive in inferring the driving factor and preparing for mitigation strategies. The Himalayan landscape is prone to multiple hazards as well as possesses a multitude of vulnerabilities owing to changing human landscape. Thus, an attempt has been made to inquire into the underlying socioeconomic factors enhancing the susceptibility of the region to multi-hazards. The social vulnerability index (SVI_ent) has been introduced, consisting of 13 indicators and 33 variables. The variables have been standardized using the maximum and minimum normalization method and the relative importance for each indicator has been determined using Shannon entropy methods to compute SVI_ent. The findings revealed that female population, population above 60 years old, net irrigated area, migrant population, dilapidated house, nonworkers, bank, and nonworkers seeking jobs were found to be relatively significant contributors to the vulnerability. The western part of the study area was classified as the highly vulnerable category (SVI > 0.40628), attributed to high dependence, and higher share of unemployed workers and high poverty. The SVI_ent was shown to have positive correlation between unemployment, socioeconomic status, migration, dependency, and household structure significant at two-tailed test. The study's impact can be found in influencing the decision of policymakers and stakeholders in framing the mitigation strategies and policy documents.