Search Results (27,739)

The expansion of the Internet of Things (IoT) has intensified cyber risk in Edge deployments with limited computational resources, where conventional intrusion detection systems are unfeasible. Aiming to reduce this gap, a lightweight anomaly detection system for IoT networks has been designed and implemented, structured through the CRISP-ML(Q) (Cross-Industry Standard Process for Machine Learning with Quality Assurance) lifecycle. Unsupervised models (Isolation Forest, OCSVM, K-Means, and Autoencoder) and supervised models (Decision Tree, Extra Trees, and Random Forest) were trained and compared on the BoT-IoT dataset, validating their operational behavior through real-time traffic captures during “TCP RST Flood”, “TCP SYN Flood”, “UDP Flood”, and “ICMP Flood” attacks on a Raspberry Pi 4B platform. Isolation Forest provided the most consistent balance between detection capability and computational cost, while the Autoencoder achieved detection rates of up to 98.8%. Among the supervised models, Random Forest achieved the highest ROC–AUC, reaching 98.5%, whereas Extra Trees showed the lowest inference time under attack conditions (0.50 s). The system was deployed on a Raspberry Pi 4B with a Streamlit interface integrating traffic capture, preprocessing, inference, visualization, resource monitoring, and automated alerts. The results demonstrate the feasibility of combining complementary detection strategies in an operational Edge-based IDS, while highlighting the trade-off between detection performance and computational cost. Full article

The digital economy has become an important force driving regional transformation and green development, yet existing evidence on its net carbon-balance effects on land-use carbon emission efficiency (LUCEE) remains limited. Using panel data for 41 cities in the Yangtze River Delta from 2002 to 2022, this study first calculates city-level net carbon-emission pressure by jointly considering anthropogenic carbon emissions and land-based carbon sequestration, and then evaluates LUCEE using a super-efficiency SBM model with undesirable outputs, in which urban construction land is incorporated as an independent input rather than used as a common scaling denominator. A two-way fixed-effects model, mediation model, geographical detector, and spatial econometric model are further employed to examine the temporal dynamics, transmission mechanisms, spatial heterogeneity, and spatial associations of the digital economy. This study finds that, in the short term, digital-economy development is significantly negatively associated with LUCEE, while this negative association gradually weakens over time, with no clear evidence of a positive reversal within the observed lag structure. Industrial tertiarisation shows partial and limited evidence as a transmission channel, although it does not automatically translate into greener land-use performance. The digital economy also helps explain LUCEE heterogeneity and shows enhanced explanatory power when combined with government intervention, land-use intensity, industrial tertiarisation, and economic development. Moreover, the spatial analysis provides supplementary evidence of a negative association between digital-economy development in spatially connected cities and local LUCEE, although this finding should be interpreted cautiously given the limited evidence of endogenous spatial dependence. These findings deepen the understanding of the low-carbon consequences of digitalisation and provide policy implications for green digital development, industrial transformation, and regionally coordinated carbon mitigation. Full article

Water inequality arises when regional water use is disproportionate to local socioeconomic and resource conditions. Yet, little is known about whether artificial intelligence (AI) technological innovation is associated with this mismatch. Using panel data for 31 provincial-level regions in mainland China from 2015 to 2024, this study examines the relationship between AI technological innovation and regional water inequality, measured from a water footprint perspective. The annual number of patent applications related to AI is used to measure regional AI technological innovation, and an instrumental variable generalized method of moments (IV-GMM) approach is employed to address potential endogeneity. The estimates show that AI technological innovation is significantly associated with lower regional water inequality: a 1% increase in AI patenting is associated with an approximately 0.249% decrease in the water inequality index. This result remains robust after adding further controls and applying alternative instrumental variable estimators. The quantile estimates remain negative across the conditional distribution of water inequality, although formal tests do not indicate statistically significant differences. AI technological innovation is also positively associated with financial development and industrial structure upgrading, providing evidence consistent with their potential roles as channels. The association is statistically significant in provinces with lower tax burdens and higher levels of industrialization, although formal differences between groups are statistically significant only for tax burden. These findings suggest that AI innovation policies may be more effective when combined with practical applications in water governance and adapted to local fiscal and industrial conditions. Full article

Urban ecological resilience is a crucial foundation for sustainable urban development. Based on balanced panel data from 285 prefecture-level and higher cities in China from 2007 to 2023, this paper treats the implementation of the 2018 “Environmental Protection Tax Law of the People’s Republic of China” as a quasi-natural experiment and employs a continuous-intensity difference-in-differences model to examine the impact of the environmental protection tax reform on urban ecological resilience. The study finds that the environmental protection tax reform significantly enhances urban ecological resilience, with industrial structure upgrading and green technology innovation serving as key transmission channels. Fiscal pressures weaken the reform’s effects, while the level of marketization reinforces its positive impact. Heterogeneity analysis indicates that the reform’s effects are more pronounced in non-resource-based cities, coastal cities, and cities in the eastern region. This study expands the research on the ecological effects of environmental protection tax reform and provides empirical evidence for optimizing differentiated environmental tax policies, strengthening incentives for green innovation, and improving regional environmental governance. Full article

Reliable prediction of provincial energy consumption is essential for translating national energy-saving and carbon-neutrality objectives into differentiated provincial action. National estimates can obscure regional heterogeneity, whereas prediction studies that focus only on accuracy provide limited guidance on which industrial, electricity, and emissions conditions require intervention. Using 1003 observations from 31 provincial-level regions for 1990–2025, this study develops an interpretable artificial-intelligence-assisted framework combining six tree-ensemble models, Optuna optimisation, Shapley additive explanations (SHAP), and partial dependence analysis. On the original random test split, the gradient boosting decision tree (GBDT) achieved a coefficient of determination (R²) = 0.9890 and root mean square error (RMSE) = 1011.59. Across 30 repeated random seeds using Optuna-optimised hyperparameters, categorical boosting (CatBoost) and GBDT obtained mean R² values of 0.9871 and 0.9865, respectively. In a blocked temporal holdout using 1990–2017 for training and 2018–2025 for testing, adaptive boosting (AdaBoost) and GBDT retained R² values of 0.9202 and 0.9131. SHAP and partial dependence results identify industrial value added, electricity generation, carbon emissions, industrial structure, population, and regional context as the principal sources of differentiated energy consumption pressure; the highest predictions occur when industrial output, electricity supply, and carbon emissions are jointly elevated. These results support a result-linked policy framework: provinces dominated by industrial activity should strengthen sector-specific efficiency benchmarks and technological retrofits; provinces with high industrial and electricity contributions should coordinate industrial load management with green electricity use and power-sector decarbonisation; and provinces with strong carbon emission contributions should implement joint energy–carbon monitoring and fossil fuel structure reviews. Regional and population signals further support zoned energy budgets and infrastructure planning. The framework therefore provides interpretable quantitative evidence for differentiated energy-saving and carbon-reduction policy under China’s low-carbon transition. Full article

Traditional craft practices face growing marginalization under industrialization yet remain central to cultural heritage and community identity. The Horizon Europe project CRAEFT addresses this through a Valorization Pilot, applying digital technologies across four dimensions: cultural experiences, games, retail support, and maker-oriented activities. The pilot adopted the Europeana Impact Framework, built on the Balanced Value Impact Model, to structure and assess impact using strategic perspectives and value lenses. Using surveys, interviews, observations, and utilization-based evaluations by independent teams, we examine how the framework can be operationalized in a complex, multi-site craft context, what short- and medium-term impacts can be evidenced, and what methodological trade-offs arise in real-world deployment. The findings indicate that digital augmentations, when usable and context-aware, strengthen cultural connection, support learning, and improve practitioner visibility, with utilization-based replications providing initial evidence of transferability. The study documents measurable social and selected economic value aligned with sustainability goals concerning education, decent work, sustainable communities, and responsible consumption, while environmental and long-term legacy impacts remain only partially evidenced. We conclude that learning and socio-cultural outcomes are the most readily evidenced within funded pilot timeframes, and offer recommendations for heritage organizations adopting impact-oriented evaluation. Full article

In a global landscape characterized by interconnected risks and the deepening of digital transformation, understanding the impact and mechanisms of artificial intelligence on social–ecological resilience is crucial for advancing resilience governance and the green transition. This paper utilizes panel data from 30 provinces in China (2013–2023) to construct a comprehensive index of social–ecological resilience and artificial intelligence. It employs two-stage fixed-effects models, mediation effect models, and robustness and endogeneity tests to systematically examine the impact of AI and regional heterogeneity. The results demonstrate that: first, AI significantly enhances social–ecological resilience, and this conclusion remains robust after controlling for measurement errors, trimming, and instrumental variable estimation; second, employment structure and innovation levels play a partial mediating role in the influence of AI on social–ecological resilience; and third, the promoting effect of artificial intelligence on social–ecological system resilience exhibits a regional heterogeneity pattern: strongest in the central region, followed by the western region, and weakest in the eastern region. The study highlights that AI not only directly enhances resilience by improving resource allocation and governance responsiveness, but also indirectly optimizes the social–ecological coupling through employment restructuring and innovation diffusion. Therefore, efforts should be directed towards promoting AI-enabled resilience governance through digital capacity building, industrial-employment synergy, institutional innovation, and infrastructure improvements. Full article

This study develops and tests an integrated model explaining why travellers substitute conventional watches with smartwatches, incorporating functional, emotional, and health-related factors: travel convenience, perceived multifunctionality, digital travel attachment, health monitoring orientation, and smartwatch satisfaction. The majority of studies have explored technology adoption and continuation or health-oriented usage of a product but not its substitution by travellers, making it necessary to identify the main determinants of this behaviour. Because the phenomenon of interest is substitution rather than first-time adoption, the target population is travellers who already use a smartwatch; switching intention is accordingly operationalised as the intention to complete the substitution—that is, to make the smartwatch the primary timepiece during travel and to relinquish the conventional watch—rather than as an initial adoption decision. Data were gathered using an online questionnaire of 406 respondents in Indonesia who were users of smartwatches and travelled outside of their city in the last six months. The measurement model was first purified through exploratory factor analysis and then validated through confirmatory factor analysis, with convergent and discriminant validity established before the structural model was estimated. The results obtained suggest that travel convenience is positively associated with perceived multifunctionality, digital travel attachment, and health monitoring orientation. Moreover, perceived multifunctionality, digital travel attachment and health monitoring orientation are positively associated with smartwatch satisfaction, which in turn shows the strongest association with travellers’ intention to switch to smartwatches. Smartwatch satisfaction shows the strongest association with switching intention (β = 0.861, 95% CI [0.744, 0.978]) and the model accounts for 74.2% of its variance, while all ten indirect effects tested are significant, indicating that satisfaction is the channel through which the functional, emotional and health-related evaluations reach the substitution decision. These results mean that travellers are more likely to use smartwatches because of their multifunctionality and the possibility of using them to facilitate mobility, connectivity, health care, and improve the travel experience. With regard to sustainability, the findings are interpreted in relation to SDG 3 (Good Health and Well-being), through the proactive use of wearable technology for health monitoring during travel; SDG 9 (Industry, Innovation and Infrastructure), through the reconfiguration of travel consumption by digital innovation; and SDG 12 (Responsible Consumption and Production), through the device convergence that reduces the number of single-function products travellers carry. These linkages are developed conceptually in ^{Section 5.3} and ^{Section 6.3} and were not measured directly in the present study. Full article

Oleogels have attracted considerable interest as structured lipid systems due to their ability to immobilize liquid oils and tailor the physicochemical properties of formulations. Although much of the fundamental knowledge on oleogel development derives from food-related research, the unique structural, rheological, and textural properties of these systems have increasingly supported their exploration also for cosmetic and pharmaceutical purposes. This review aims to explore the relevance of oleogels in these industries by compiling and critically evaluating the current knowledge on cosmetic and pharmaceutical formulations, with particular emphasis on their composition, preparation methods, and the rationale underlying their selection and use. Studies on oleogels for cosmetic uses are limited, while in pharmaceuticals, oleogels have been extensively studied for topical drug delivery, being their use in oral, transdermal, and intraocular delivery less explored. Oleogels are versatile and biocompatible delivery systems, ideal for transporting lipophilic active ingredients. Their inner structure modulates drug delivery, making them promising vehicles for innovative cosmetic and pharmaceutical applications. Full article

In the global efforts to combat climate warming and promote low-carbon development, the transition risks stemming from continuous adjustments and changes in policies, technologies, and markets have become another important driver exacerbating inter-city economic divergence with disparate industrial foundations, resource endowments, and transition prerequisites. Based on a mathematical economic modeling framework, this article constructs a Recentered Influence Function (RIF) regression model to empirically examine the direct impacts and heterogeneous characteristics of climate transition risks (proxied by climate policy uncertainty) on inter-city economic disparities across 285 Chinese cities from 2005 to 2024 and further analyzes the underlying mechanisms and feedback effects through mediation and moderation effect models. The empirical results reveal that climate transition risks widen inter-city economic disparities and amplify the polarization effect in inter-city economic development. Moreover, the positive impact of transition risks on inter-city economic disparities is more pronounced in the types of high-growth, non-resource-based, and Eastern Chinese cities. The mediating transmission channels through which climate transition risks affect inter-city economic disparities include the lock-in effect of industrial structural imbalances and disparities in green technology; improvements in energy efficiency and green credit availability can mitigate the amplifying influence of transition risks on inter-city economic inequality. This study enriches the theoretical foundation in the field of climate finance regarding transition risks and balanced regional development, which provides practical guidance and policy evidence for addressing inter-city economic disparities arising from climate change and energy transition, as well as advancing coordinated and balanced urban economic development via green transformation. Full article

Corporate governance guidelines often emphasize board independence to minimize agency costs. Yet, in highly concentrated emerging markets, strict compliance with outside-monitoring mandates can be associated with structural bottlenecks and decision friction during economic shocks. We evaluate how board independence and corporate social responsibility (CSR) relate to bank valuations during systemic disruptions, using data from the 10 systemically important commercial banks listed on the Nigerian Exchange (NGX) from 2013 to 2024. This panel captures 120 bank-year observations, representing 87.4% of total commercial banking industry assets. Long-run parameters are estimated using a Pooled Mean Group (PMG) Panel ARDL framework with robust Driscoll–Kraay standard errors, supplemented by a non-parametric Random Forest machine learning feature importance diagnostic. The machine learning model identifies CSR expenditure (LTCSR) as the most important predictor of bank value restoration, outranking traditional balance-sheet controls like equity book value and asset scale. The parametric estimations reveal a significant long-run independence discount (β = −0.342, p < 0.05), where a 10 percentage point increase in outside directors is associated with an absolute 0.034 unit market valuation penalty, a trend theoretically consistent with crisis-driven decision friction. However, the underlying banking network shows high recovery elasticity, absorbing 69% of external valuation shocks within a single annual cycle (φ = −0.690, p < 0.01). Finally, a quadratic inflection point (translating to an actual annual monetary expenditure threshold of approximately ₦1.67 billion Naira) paired with asymmetric quantile distributions indicates that CSR serves as a plausible emergency reputational shield for lower-quantile institutions (q25) but acts as a strategic asset for market leaders (q90). These findings suggest that macroprudential supervisors should consider shifting from rigid, headcount-based compliance toward functional capability thresholds. Full article

Large-scale battery energy storage systems (BESSs) continuously generate cell-level monitoring data. These data increase storage demand and communication load in cloud-edge systems. Exact measurement values are required for archival and diagnostic tasks, which motivates lossless compression. This paper releases an industrial cell-level dataset collected from a 215 kWh BESS and analyzes its compression-relevant properties. The data exhibit finite decimal precision, strong temporal dependence, spatial similarity, and different value ranges across measurement variables. Based on these properties, we propose a structure-aware lossless compression method. The method groups the data by measurement variable and reversibly scales the values to integers. It applies temporal and spatial differencing, followed by signed-value remapping, bit-width packing, and final lossless coding. Experiments on AMD and Raspberry Pi platforms evaluate global block-based compression using all 59,452 records and real-time compression. The proposed method achieves compression ratio of 0.779% in the global block-based setting, while its real-time ratio is 1.801%. The real-time configuration requires 0.198 ms per time record on AMD and 1.252 ms on Raspberry Pi. Full article

In laser vision-based robotic welding, the seam feature point must be located, under strong arc-light and spatter interference, in every frame acquired while the weld is being made, so that the torch can be steered in real time; the detector must therefore keep pace with the camera on the modest processor of an industrial controller. Conventional deep learning models encounter structural redundancy and high computational latency under this constraint. To address these limitations, this paper proposes a lightweight object detection model, whose box-center output is mapped to the weld seam feature point, optimized via Bayesian variational neural architecture search (BayeNAS), in which architecture parameters are modeled as Gaussian distributions and updated through natural-gradient variational inference. A novel dual-selection optimization strategy is introduced to jointly search for optimal candidate operators and network topology branches, achieving an explicit balance between model complexity and hardware latency. Furthermore, an in-search channel pruning strategy is incorporated to compress the search space and prevent the combinatorial growth of the channel configuration space. Comparative analysis and real-world welding experiments demonstrate that the proposed model achieves a high processing speed of 78 frames per second (FPS) on a legacy central processing unit (CPU)—above the 60 fps acquisition rate of the sensor—while maintaining a mean positioning error within 0.15 mm on the measurement plane. These results demonstrate that the proposed model achieves detection accuracy comparable to heavyweight detectors while offering an order-of-magnitude advantage in inference speed and model size, validating its suitability for low-cost industrial deployment. Full article

Anthropogenic emissions of carbon oxides (CO_x) and nitrogen oxides (NO_x) arising from industrial activities contribute to environmental and health problems, motivating the development of effective strategies for their mitigation and valorization. Among the available approaches, electrochemical reduction offers the additional possibility of converting CO_x and NO_x species into value-added products that may serve as chemical feedstocks or energy carriers, thereby contributing to carbon and nitrogen resource recovery. However, the practical development of these processes requires catalysts with high activity, selectivity, and stability, supported by synthesis strategies capable of controlling their active structures. Several metals, including Ir, Pt, Rh, Ru, Pd, Ni, Ag, Au, and Cu, have been investigated for the electrochemical reduction of CO_x and NO_x in aqueous media. Among them, Cu is particularly attractive because of its relatively high abundance, tunable oxidation states, and ability to catalyze both reaction families. This review examines Cu-based nanomaterials prepared through different synthesis strategies and analyzes how composition, oxidation state, morphology, exposed facets, defects, alloying, and interfacial structure influence CO_x and NO_x electroreduction. By comparing both reaction families, the review identifies shared structure–active-site–performance relationships together with reaction-specific requirements for intermediate stabilization and product formation. Particular attention is given to in situ and operando evidence showing that Cu-based catalysts may reconstruct under electrochemical conditions, so that the catalytically relevant material can differ from the as-synthesized structure. These insights support catalyst optimization with mechanistically informed design of dynamic Cu-based interfaces for selective CO_x and NO_x electroreduction. Full article

Industrial heritage is increasingly adapted for cultural, tourism, and consumption-oriented uses. Where industrial production was historically embedded within difficult institutional settings such as prisons, regeneration also involves the ordering of historical attributes, the governance of difficult memories, and the reorganisation of public narratives. This study examines the transformation of the former Hefei Prison into Hechai 1972 Cultural and Creative Park through a qualitative interpretive single-case design combining historical and policy analysis, qualitative content analysis, structured spatial observation, and semi-structured interviews. The findings show that industrial production was institutionally embedded within the prison and reform-through-labour system, but these historical attributes have been reordered in contemporary public interpretation. Carceral memory is mainly represented through symbolic retention, low-intensity interpretation, targeted access, and educational reframing. Industrial buildings, products, and achievements remain highly visible, while their links to prison institutions, reform-through-labour production, labour organisation, and historical actors are weakened or decontextualised. The study identifies processes through which historically embedded institutional relationships shape heritage-attribute ordering, while differentiated memory governance, narrative practices, and spatial practices reshape their public interpretation, with implications for interpretive integrity and the ethical reuse of carceral–industrial composite heritage. Full article