Search Results (13,891)

Multi-temporal orbital satellite imagery is an alternative for measuring behavioral patterns or trends in different physiognomies through vegetation indices (VIs) and Spectral Linear Mixture Models (SLMMs). In this study, time series of Landsat 7/8/9 and Sentinel-2 have been used to classify a considerable quantity of areas spread over the São Paulo state from 2021 to 2024. Because the large amount of samples considered in our analysis, self-organizing maps (SOMs) have been applied as a convenient method to group similar satellite image time series samples with respect to a certain vegetation index or green vegetation fraction (VEG). Since every dataset area belongs to different types of physiognomies, each cluster has been labeled according to the plurality technique. Additionally, we obtained the mean spectral behavior of the VIs and VEG in the 2021–2024 seasonal cycle of all samples. The results showed similar variations from the rainy to the dry season for most of the physiognomies. On the other hand, this research indicates that the proposed method for classification the Brazilian areas spread over the São Paulo state is consistently good, obtaining the best performance (quantization error) associated with Normalized Difference Vegetation Index (NDVI) time series samples. Full article

Recent studies have shown that common household activities produce characteristic patterns in indoor air pollutants, enabling activity inference using environmental measurements alone. However, pollutant-based approaches are usually formulated as flat multi-class classification problems, even though indoor environments are dominated by long baseline periods with no emission-generating activity, leading to false alarms and unstable predictions. This work proposes a gated hierarchical inference framework for recognizing activities from indoor air quality data. A first-stage gate detects whether a time window contains activity-induced pollutant dynamics, while a second-stage classifier conditionally identifies the specific activity only when activity relevance is detected. Multivariate time-series measurements of particulate matter, volatile organic compounds, nitrogen oxides, carbon dioxide, temperature and relative humidity were collected using a portable monitoring system during controlled household cooking and cleaning experiments. Temporal windows were processed using recurrent neural network models in both stages. By separating activity detection from activity identification, the proposed method aligns inference with the physical generation of indoor pollutant signals and improves robustness in baseline-dominated monitoring scenarios while maintaining reliable discrimination among activities. The framework supports unobtrusive activity recognition and enables applications in exposure-aware monitoring and intelligent indoor environmental management. Full article

Under strict spatial constraints and environmental interference, autonomous navigation of vessels in inland bridge-restricted waterways demands precise coordination between collision avoidance and trajectory tracking. To meet these operational demands, an integrated framework that directly combines spatiotemporal risk assessment with dynamic control execution is developed. Based on a digital traffic model integrating bridge piers and channel boundaries, collision risks are evaluated by combining trajectory-predicted time to safe distance with the velocity obstacle interval. Such a formulation quantifies the actual spatial difficulty of evasion rather than relying solely on temporal urgency. Driven by this continuous assessment, a time-series rolling strategy calculates feasible maneuvering intervals, generating trajectories that comply strictly with inland navigation rules and physical vessel limits. Subsequently, an adaptive model predictive control algorithm executes these commands, implicitly compensating for the localized hydrodynamic disturbances typical of bridge areas. The effectiveness of the architecture is validated through comprehensive simulations covering rule-based encounters and complex multi-vessel scenarios. Quantitative results indicate that under wind and current disturbances, the maximum route tracking deviation is constrained below 53 m, while the minimum encounter distance with target ships is consistently maintained above 51 m. These performance metrics confirm the capacity to execute safe, rule-compliant maneuvers while preserving high navigational precision in confined inland environments. Full article

Wildfire smoke contributes substantially to episodic PM_2.5 exposure, yet outdoor measurements may not represent indoor conditions. We analyzed indoor PurpleAir sensors and nearby outdoor monitors from U.S. residences (2019–2024) to estimate smoke-day changes in the outdoor–indoor PM_2.5 difference and characterize heterogeneity across regions. After data quality control and the application of completeness criteria, 509 monitor pairs contributed 250,873 monitor-days. Smoke days were assigned using the NOAA Hazard Mapping System smoke-plume polygons. Pair-specific time-series models estimated smoke-day changes in the outdoor–indoor PM_2.5 difference, which were pooled using random-effects meta-analysis; heterogeneity was summarized by clustering indoor and outdoor smoke–non-smoke contrasts. In the unadjusted summary, the mean outdoor PM_2.5 was 8.61 vs. 5.63 µg/m³ on smoke vs. non-smoke days and the mean indoor PM_2.5 was 6.33 vs. 5.09 µg/m³, reflecting an increase in the mean outdoor–indoor difference from 0.54 to 2.27 µg/m³ (p < 0.001). The pooled smoke-day effect on the outdoor–indoor difference was 0.88 µg/m³ (95% CI: 0.80, 0.96). Clustering identified four distinct response patterns, most commonly outdoor increases exceeding indoor increases, with smaller subsets showing extreme outdoor amplification or net indoor reductions under modest outdoor increases. These results indicate that indoor protection during smoke episodes is common but variable and support exposure characterization beyond outdoor concentrations alone. Full article

Neural ordinary differential equations (Neural ODEs) describe the feature evolution of deep networks by continuous-time dynamical systems and enable end-to-end learning through differentiable numerical solvers. Nevertheless, in closed-loop rolling prediction for small-sample time series, conventional Neural ODEs remain vulnerable to error accumulation and numerical instability. To improve the controllability of long-term evolution, this study proposes a neural ordinary differential equation framework based on fractional-order operators. Rather than directly introducing full-history convolution kernels into the governing dynamics, the proposed approach constructs a fractional effective step size from the closed-form expression of the Riemann–Liouville fractional integral of a constant function and consistently embeds it into all sub-steps of a fourth-order Runge–Kutta solver. In this way, the scale of continuous-depth propagation is regulated by a single tunable parameter. Combined with a residual output structure, the method preserves the interpretability of continuous dynamics while effectively suppressing trajectory drift in closed-loop prediction and improving training stability. To investigate the impact of the fractional-order parameter on fitting and extrapolation, particle swarm optimization is employed to search automatically for the optimal order. Experimental evaluations on the linear spiral system and Lorenz continuous dynamical systems and on a small-sample provincial annual electricity-consumption dataset show that the proposed model achieves lower prediction errors across multiple tasks and exhibits superior trajectory preservation and robustness under long-horizon forecasting. Full article

This study addresses the critical need for accurate sea wind speed predictions to support ocean wind farm operations, equipment maintenance, and maritime navigation safety. To enhance prediction accuracy for any location within target sea areas, we propose a short-to-medium-term wind speed prediction method that effectively explores spatiotemporal correlations in ocean reanalysis grid data. The method involves collecting and reanalyzing data, as well as spatial processing, to reconstruct the historical wind speed sequence at the target point. Finally, a future wind speed time series is generated using an LSTM network and a Transformer encoder. Test results validated against NOAA buoy data demonstrate the effectiveness of our spatiotemporal prediction model, achieving RMSE values of 1.161 m/s, 1.500 m/s, and 1.854 m/s for 1 h, 6 h, and 12 h predictions, respectively, outperforming comparative methods. The conclusions are threefold: (1) The proposed hybrid model effectively captures spatiotemporal dependencies and achieves more accurate spatiotemporal predictions compared to the benchmark model; (2) taking into account seasonal factors and forecasting time periods, the method proposed in this paper maintains good stability; (3) this framework provides a reliable technical approach for generating operational references in maritime navigation and wind power maintenance, with potential applications in wind farm siting and resource assessment. Full article

This research paper examines the spatial and time variation of demographic dependency in Europe in a 30-year horizon of the evolution of the demographic dividend regarding the economic dependency ratio (ADR1). We used the Curve Fit Forecast tool to estimate the trends of ADR1 in each of the EU Member States using data on Eurostat projections and a sophisticated geostatistical analysis tool developed in ArcGIS Pro 3.2.2. The findings indicate that the dependency in all countries has increased significantly in a statistically significant manner as the Gompertz function has appeared as the best curve in a third of the cases. It is an S-shaped asymptotic behaviour of this function that effectively describes the nonlinear patterns of acceleration and saturation of demographic ageing. As indicated in the analysis, the European regions are increasingly moving apart, with the southern and eastern nations such as Romania demonstrating the most alarming decline in ADR1. These trends highlight the need to reform labour market policies and social protection mechanisms to an ageing population. The paper combines the curve-fitting, descriptive statistics (median, skewness, interquartile range (IQR)) with time clustering (value, correlation, and Fourier) to provide an effective, replicable approach to early warning and policy prioritisation. Overall, the results highlight the importance of integrating predictive spatial modelling and demographic economics to support anticipatory and evidence-based policy decisions. The proposed approach proves to be a robust and transferable framework, applicable to a wide range of socio-economic phenomena characterised by inertia and structural change. Future research should extend the analysis to subnational levels, incorporate additional explanatory variables, and develop scenario-based simulations, including multivariate Gompertz-type models, to further enhance both predictive accuracy and policy relevance in the context of emerging structural labour scarcity. Full article

Human Activity Recognition (HAR) from inertial sensors must run continuously on battery-powered wearables under tight latency, memory, and energy budgets. While tiny Transformers can be effective on inertial time series, end-to-end co-design across quantized inference and heterogeneous low-power platforms remains underexplored. We present STORM (Small Transformer for On-node Recognition of Motion), a deployment-oriented [round-mode=places, round-precision=1]19.7k-parameter 1D Transformer co-designed with X-HEEP, an open-source low-power single-core RISC-V SoC, and a tightly coupled streaming CGRA for nonlinear primitives (e.g., softmax). We build a cross-source 8-class benchmark by harmonizing 3 public datasets under a stringent, deployment-aligned protocol that exposes both cross-subject and cross-source shift. Using 1.280 s windows with 0.640 s stride, the protocol models continuous on-node HAR under cross-dataset generalization. After quantization-aware training and INT8 C inference export, STORM achieves [round-mode=places, round-precision=3]0.799/[round-mode=places, round-precision=3]0.801 accuracy/macro-F1 on this benchmark. Deployed on an FPGA prototype of X-HEEP with the streaming CGRA backend, STORM requires round(6739790/ (100* 1000000)* 1000, 1) ms per inference at 100 MHz, while activity-based power analysis estimates a total inference energy of 632.4 $\mathsf{\mu }$J, satisfying the stride-driven real-time constraint. These results support the practical viability of compact attention-based HAR on low-power wearable-class embedded platforms. Full article

Publicly available electricity datasets from operational industrial facilities remain limited due to instrumentation cost, retrofit complexity, and data governance constraints. This paper presents an openly accessible dataset of asset-level electrical energy measurements collected from a medium-scale industrial manufacturing facility over an approximately one-year observation window, with staged commissioning resulting in heterogeneous temporal coverage. The dataset includes time-series measurements from production machinery, auxiliary systems, and distribution-level assets instrumented using a heterogeneous fleet of Ethernet and RS-485 energy meters integrated via industrial gateways and programmable logic controllers. Measurements were acquired via a SCADA-based logging infrastructure and exported from an operational SQL historian. The publicly released dataset comprises fixed 15 min aggregated energy and power metrics derived from high-frequency SCADA telemetry. In its released ALL-phase representation, the dataset comprises measurements from 43 monitored assets and 1,039,873 15 min windows, corresponding to 2.96 GWh of measured electrical energy. Mean window-level data coverage is 99.99%, and 97.72% of ALL-phase windows satisfy the dataset’s reliability criterion. Interval records include energy consumption, demand, data coverage metrics, and reliability indicators. The dataset reflects real-world industrial monitoring conditions, including mixed communication pathways and irregular sampling behaviour, and is intended to support research in industrial energy analytics, data quality assessment, load profiling, and operational energy modelling. Full article

This study aims to evaluate the potential of Global Navigation Satellite System Interferometric Reflectometry (GNSS-IR) based on signal-to-noise ratio (SNR) analysis for monitoring crop structure and moisture. Data were collected using a GNSS antenna placed within an experimental meadow located in NW Italy. GNSS-IR exploits the interference between direct and ground-reflected signals to derive physical parameters such as the vegetation phase center height and soil moisture. In this work, by analyzing and modeling the oscillations in SNR time series, the sensitivity to crop growth dynamics was assessed. Vegetation height and dielectric parameters were compared against corresponding ground-surveyed values collected using a ruler and buried soil moisture sensors. Results suggest that GNSS-IR can detect canopy height with a high degree of consistency (Pearson’s r = 0.89, MAPE = 18%). Results also show that changes in the amplitude and phase of the interference pattern are sensitive to biomass density and dielectric properties of the reflecting surface (r = −0.81 and r = 0.86 respectively). GNSS-IR observables were analyzed across four representative measurement campaigns capturing distinct seasonal stages of meadow development. Despite the limited temporal sampling (n = 4), the selected observations correspond to contrasting vegetation and soil moisture conditions, allowing the identification of systematic variations in crop biophysical properties. These findings open promising perspectives for the development of innovative monitoring strategies in precision agriculture, leveraging existing GNSS infrastructure to obtain key biophysical parameters with minimal additional equipment and operational complexity. Full article

With the continuous advancement of intelligence in power systems, traditional unimodal fault diagnosis methods can no longer satisfy the demand for precise monitoring of complex power equipment. To address the challenges of multimodal data fusion and fault diagnosis in intelligent sensing scenarios, this paper proposes a multimodal intelligent diagnosis model for power transformers based on causal dynamic fusion reasoning. By introducing a causal reasoning mechanism, the proposed model overcomes the limitations of conventional multimodal fusion approaches that rely solely on statistical correlations. A four-layer architecture is constructed, consisting of a Causal Dynamic Fusion layer, a Graph Reasoning layer, a State Prediction layer, and a Meta-Reinforcement Learning Optimizer, thereby forming a complete closed-loop framework from multimodal feature extraction to intelligent diagnostic decision-making. This study focuses on key issues including causal discovery and dynamic fusion in multimodal data, cross-sample contextual enhancement, equipment state prediction, and early warning. Performance evaluation experiments are conducted on a large-scale synchronized dataset containing image, audio, and time-series modalities. Experimental results demonstrate that the proposed CDFMD model outperforms conventional methods in diagnostic accuracy and real-time performance, providing a novel technical pathway for intelligent operation and maintenance of power equipment. Full article

With the in-depth advancement of the “Digital China” initiative, policies and research discourses related to the digital economy have continued evolved, making it necessary to systematically examine their stage-specific characteristics and underlying logic from a long-term perspective. Accordingly, this study adopts information society theory as the analytical framework and selects the annual series of reports on China’s digital economy development published by the China Academy of Information and Communications Technology (CAICT) from 2015 to 2024 as the research corpus. Using text mining techniques and Latent Dirichlet Allocation (LDA) topic modeling, this paper conducts a longitudinal examination of the stage-wise systemic evolution of key topics in China’s digital economy development. The findings indicate that over the past decade, the topic structure of China’s digital economy has followed a clear evolutionary trajectory, progressing from “informatization-driven development” to “platform expansion,” and subsequently to “data factors and institutional governance.” In the early stage, the focus was on information infrastructure development and industrial integration; the middle stage shifted toward the platform economy and enterprise growth; more recently, the emphasis has increasingly been placed on the construction of data factor markets and the improvement of governance frameworks. This process of topic evolution not only reflects changes in the practical forms of the digital economy but also reveals the ongoing adjustment of the state’s cognitive framework and governance logic regarding digital economy development. These findings provide empirical evidence for understanding the systemic evolution of China’s digital economy over time. By identifying the stage-specific pathways of China’s digital economy, this study extends the application of information society theory within this context and provides new empirical evidence for understanding the evolutionary logic underlying high-quality digital economy development. Full article

This study empirically examines the joint effect of financial anomaly risk and ESG performance on accounting conservatism using accrual models, market models, and earnings time-series models. Financial anomaly scores are obtained using unsupervised machine learning to identify reporting anomalies for firms. Our findings suggest that higher financial anomaly risk is negatively related to accounting conservatism through delayed or reduced loss recognition. ESG engagement serves as a moderating variable to mitigate conditional conservatism losses partially for both accrual- and earnings-based models, conditional on financial anomaly risk; otherwise, ESG engagement has a weak or insignificant effect on market-based models. ESG practice is therefore a state-dependent conditional governor to complement traditional governance structures, depending on both levels of anomaly risk as well as accounting models used to derive conservatism measures. Our findings have practical implications for investors and government regulators, as well as managers, which emphasize that ESG practice is not universally beneficial to conservatism but can further improve reporting quality, conditional on certain risk levels. Full article

Recruitment is a time-consuming process, and AI systems are increasingly being used to support the decision-making process. However, machine learning models used in such systems can inherit bias if the underlying training data reflects biased human preferences. It is essential to analyze and quantify these biases in order to develop fairer AI systems. To address this issue, we collected human judgments of colleague preference for 2200 face images. The face image set includes images of different ethnicities and genders, as well as both real and synthetically generated faces. The images were annotated by humans from diverse backgrounds in terms of age, gender, and ethnicity. Annotators were shown series of pairs of face images and asked to select which individual they would prefer as a colleague. We gathered responses from 451 annotators and aggregated the annotations to compute a preference score for each image. This dataset provides a basis for understanding human bias in colleague preference and can support the development of fair and unbiased AI models for use in recruitment settings. Full article

The high-precision, continuous monitoring of the surface water level is of great importance for water resource management and the conservation of ecological systems. This study proposes a GNSS-IR-based water-level estimation method using NMEA observations collected from low-cost GNSS receivers. First, the NMEA-recorded satellite elevation angle, azimuth angle, and signal-to-noise ratio (SNR) are processed using time-series characteristics for improving the resolution and applicability of these GNSS observations. Then, the multi-frequency GNSS signal-based reflector height inversion models are developed by making use of the Lomb–Scargle periodogram method. Finally, the Velocity Pausing Particle Swarm Optimization (VPPSO) algorithm is employed to calculate the reflector height estimation and thus the water level. Two experimental data sets collected in two different environments were used to test the proposed method. The experimental results show that the root mean square error (RMSE) of the water-level estimation error is less than 6 cm for the proposed method when the in situ ones are in the range of 196.4 cm to 296.1 cm. This study provides a theoretical and technical foundation for the development of the low-cost GNSS-IR water-level measuring instrument. Full article