Search Results (44,308)

This study presents an integrated design methodology for earthquake demolition rescue robots by combining UXMs, Kano, and QFD to improve design rationality and performance in extreme rescue scenarios. It addresses key gaps in existing approaches, particularly the lack of systematic experiential data acquisition, quantitative requirement analysis, and effective design translation. UXMs are applied to reconstruct critical task scenarios and identify high-load nodes and user experience variations. The Kano model is used to prioritise and classify user requirements, which are then translated into engineering characteristics through QFD. Based on this framework, a conceptual robot design is developed using the FBS model and evaluated through process-level simulation and usability assessment. The results demonstrate that the proposed method enables structured requirement transformation and supports traceable design decisions. Simulation indicates the consistency of task workflows and coordination among functional modules at the process level. A System Usability Scale score of 80.22 indicates a relatively high level of perceived usability at the conceptual evaluation stage. The proposed methodology provides a structured and traceable conceptual design framework for earthquake rescue robots. While the current validation is based on conceptual-level evaluation, the methodology offers a traceable design pathway that may be extended to other high-risk emergency equipment with further empirical testing. Full article

Indigenous peoples in Amazonia face major mental health inequities, including high rates of suicidal behaviour among adolescents and young adults in some settings. We conducted a scoping review of the peer-reviewed literature on community health workers (CHWs) and equivalent cadres involved in Indigenous and remote contexts, with a focus on their roles in relation to mental health, psychosocial support, and suicide prevention among Indigenous populations in Amazonia and the Guiana Shield. We reported this review in line with PRISMA-ScR. Searches (September–November 2025) were conducted in PubMed/MEDLINE, Scopus, Web of Science and SciELO, complemented by targeted searches in major publisher platforms and JSTOR. We included English, French, Spanish and Portuguese publications that (i) described CHWs or functionally equivalent cadres in Indigenous/remote contexts and/or (ii) reported CHW-related roles, models, or experiences relevant to mental health, psychosocial support or suicide prevention in Amazonian settings. Global documentation of CHW designations used in Indigenous/remote contexts was compiled; we compiled evidence from Amazonia and the Guiana Shield on CHW roles, programme models, implementation conditions and reported outcomes. Data were charted into a structured template (cadre designation, setting, population, study type, functions, programme features and reported mental health/suicide-related outcomes) and synthesised descriptively and thematically. CHWs commonly function as cultural and linguistic brokers between Indigenous communities and biomedical systems, supporting early detection of distress, psychosocial accompaniment, referral navigation and dialogue with local healing practices. Reported programme models differ markedly: Brazil’s institutionalised Indigenous Health Agents (AIS) offer stability and formal recognition, whereas French Guiana relies more heavily on project-based mediation with innovative practices but greater funding fragility. The available literature remains heterogeneous and uneven across countries, with limited evaluative designs and substantial reliance on descriptive reports. Future work should prioritise stronger implementation and impact evaluation, alongside Indigenous-led governance and sustainable support for CHW cadres. Full article

Recent advances in edge computing and Tiny Machine Learning (TinyML) have enabled the deployment of artificial intelligence models directly on microcontrollers with extremely limited computational and memory resources. In this context, this work presents the design, implementation, and validation of a real-time cardiac arrhythmia classification system based on a quantized one-dimensional convolutional neural network (1D-CNN), deployed on an 8-bit Arduino UNO microcontroller. The proposed system integrates end-to-end processing, including ECG signal acquisition using a low-cost AD8232 analog front-end, signal preprocessing, heartbeat segmentation, classification, and real-time visualization on an OLED display. The model was trained and evaluated using the MIT-BIH Arrhythmia Database, considering a reduced three-class problem (Normal, Ventricular, and Supraventricular) to meet the constraints of ultra-low-cost hardware deployment. Under benchmark conditions, the quantized model achieved an accuracy of 97.6%, with a memory footprint below 24 KB and an average inference time of 200 ms per heartbeat, enabling real-time operation on a resource-constrained microcontroller. Real-time experiments were conducted using signals acquired from healthy volunteers to validate system functionality, although no annotated ground truth was available for these recordings, and therefore no diagnostic performance was derived from them. The results demonstrate the feasibility of deploying lightweight deep learning models on ultra-constrained embedded systems using the TinyML paradigm, implemented using TensorFlow 2.15 and TensorFlow Lite. This work should be interpreted as a proof-of-concept platform that highlights the trade-off between classification performance and hardware limitations, providing a foundation for future development of low-cost cardiac monitoring technologies in resource-limited environments. Full article

The densely distributed, scale-varying objects in unmanned aerial vehicle (UAV) images, together with their dynamic, diverse, and unconstrained backgrounds, make conventional detection methods prone to missed detections, false alarms, and localization biases. To improve UAV vision tasks, we propose AD-YOLO, a unified method tailored for small object detection in traffic-dense settings. First, a module combining an adaptive rotation convolution unit and grouped directional attention with mixed-kernel features is introduced to enhance the model’s orientation invariance and multi-scale discrimination. Then, a dual-path collaborative feature pyramid network is proposed to jointly refine the model’s semantic and spatial details via a multi-directional context aggregation path and a hierarchical semantic progressive fusion path. Last, a hierarchically dense reparameterized large-kernel module is designed to produce broader receptive fields with reduced computational complexity. Extensive experiments on the VisDrone2019 and UAVDT datasets demonstrate that AD-YOLO outperforms state-of-the-art methods in detection accuracy while maintaining favorable computational efficiency. Full article

As multi-view datasets expand across diverse practical fields, feature selection (FS) has become an indispensable preparatory stage for machine learning models. Nevertheless, real-world multi-view data is often unlabeled and distributed among isolated clients, posing significant challenges to traditional centralized methods due to privacy concerns and communication constraints. Furthermore, existing centralized and federated approaches frequently suffer from entrapment in local optima and lack robust convergence guarantees. To address these issues, we propose Fed-MUFSHT, a federated framework for multi-view unsupervised FS (MUFS) that integrates tensor learning with a novel metaheuristic optimizer, Hierarchical-Cognitive Tianji’s Horse Racing Optimization (HC-THRO). Within the federated learning paradigm, Fed-MUFSHT follows a dual-stage local optimization process. Stage 1 applies HC-THRO, which integrates Hierarchical Competitive Learning and Adaptive Cognitive Mapping to simulate multi-level strategic competition and cognitive adaptation among individuals. This design enhances global exploration, adaptive learning, and fine-grained feature selection in high-dimensional spaces. Stage 2 employs a TL module based on canonical polyadic (CP) decomposition to perform missing-view imputation and refine latent representation learning. At the global level, a privacy-preserving aggregation strategy based on Normalized Mutual Information (NMI) and feature weights enables efficient model coordination without exposing raw data. Comparative experiments on several public benchmark datasets reveal that Fed-MUFSHT maintains clear advantages over strong competing methods, showing better optimization results together with more dependable convergence characteristics. The overall evidence suggests that the proposed approach is both robust and effective for distributed optimization tasks involving privacy protection. Full article

To address the issue of motion artifact interference faced by wearable ECG monitoring devices in dynamic environments, this paper proposes an adaptive motion artifact removal framework based on improved Variational Mode Decomposition (VMD). By designing a parameter self-adjustment mechanism and a multi-feature fusion mode selection strategy, the algorithm’s adaptability to non-stationary ECG signals and noise separation accuracy are enhanced. Experiments on the MIT-BIH Arrhythmia Database demonstrate that the improved VMD algorithm outperforms traditional wavelet transform, Recursive Least Squares (RLS), and conventional VMD methods in multiple performance metrics. Specifically, the signal-to-noise ratio (SNR) is improved by 5.17 dB, the Percentage Root Mean Squared Difference (PRD) is reduced to 49.13%, the correlation coefficient is increased to 0.88, and high real-time processing capability (Real-Time Processing Ratio, RTR = 22.5) is maintained, meeting the low-latency requirements of wearable devices. Moreover, case studies on pathological recordings (e.g., Wolff–Parkinson–White syndrome and third-degree atrioventricular block) reveal that the improved VMD better preserves clinically significant features such as delta waves and dissociated P waves. Furthermore, a downstream arrhythmia classification task using a CWT-CNN classifier achieves 91.67% accuracy on denoised heartbeats, which is 2.67 percentage points higher than that on raw noisy signals (89.00%), confirming the practical benefit of the proposed preprocessing for AI-based diagnosis. This study provides an effective processing solution for improving the signal quality of wearable ECG monitoring. Full article

Flavanones retrieved in the leaves of Glycyrrhiza glabra (licorice), specifically glabranin (GLA), pinocembrin (PIN) and licoflavanone (LIC), represent a valuable source of bioactive natural products, although their isolation and handling are often complicated by their structural similarity and unfavorable physical properties. In this work, crystal engineering strategies were explored both to facilitate the selective separation of licorice flavanones and to improve their solid-state characteristics. Co-crystallization was investigated as a tool for the selective recognition of PIN from a GLA-rich chromatographic fraction. Guided by structural considerations and predictive analyses performed using the Co-Crystal Design and Hydrogen Bond Propensity (HBP) tools in CCDC Mercury (within CCDC-Materials), co-crystallization experiments were performed with pyridinic co-formers. 4,4′-Bipyridine (BPY) selectively formed a new co-crystal with PIN, enabling the capture of traces of this flavanone directly from the GLA-rich fraction. In contrast, nicotinic acid (NIC) did not form a co-crystal with PIN, consistently with the predicted preference for NIC self-association. In addition, a co-amorphous system between LIC and BPY was obtained by quench cooling, yielding a fully amorphous solid with improved handling properties compared to the waxy precursor. These results highlight the potential of crystal engineering approaches for the selective separation and solid-state modification of natural flavanones. Full article

Additive manufacturing (AM) enables the fabrication of complex three-dimensional components with embedded internal flow channels, but the as-built inner surfaces often exhibit high roughness and poor surface-quality uniformity, particularly at non-coplanar corner regions such as sharp bends and junctions. Conventional abrasive flow machining (AFM) can improve the overall surface finish of such channels; however, corner regions commonly remain weak-removal zones because of local flow stagnation and insufficient abrasive action. To address this limitation, this study proposes a six-degree-of-freedom (6-DOF) robotic-arm-assisted liquid metal-driven abrasive flow (LM-AF) polishing strategy in which robotic pose regulation is used to guide the liquid metal droplet to designated corner regions while preserving its responsiveness to the electric field. Numerical simulations and conventional AFM experiments on S-shaped and M-shaped spatial channels were first conducted to identify the corner regions as the primary sources of polishing non-uniformity. A robotic posture-control framework was then established through manipulator kinematics, point-cloud-based flow-direction identification, and Rodrigues-matrix-based pose transformation. On this basis, localized secondary polishing was experimentally performed on an S-shaped channel using an AC electric-field-driven liquid-metal abrasive system. The results show that corner-region roughness was significantly reduced and approached the straight-channel benchmark after secondary polishing, demonstrating a marked improvement in inner-surface uniformity. This study provides a practical route for targeted compensation polishing in complex three-dimensional internal channels and offers a new framework for robotic-assisted post-processing of AM-fabricated flow paths. Full article

Black TiO₂-impregnated electrodes were prepared via a modified dip-coating method, using six deposition layers to investigate the influence of the TiO₂ precursor phase (anatase, rutile, and P25) on their structural and optical properties, as well as their photoelectrooxidation performance toward acetaminophen degradation. A reductive thermal treatment under a H₂/Ar atmosphere successfully modified the band gap energy and promoted the formation of oxygen vacancies (Vo) and Ti³⁺ species, as evidenced by UV–Vis diffuse reflectance spectroscopy and photoluminescence analysis. Among the precursor phases, anatase exhibited the most significant band gap reduction, whereas rutile and P25 showed greater structural stability after the reduction process. Photoelectrochemical experiments revealed that the supporting electrolyte plays a dominant role in the degradation process, with significantly higher removal efficiencies observed in chloride medium (0.1 M NaCl) compared with sulfate medium (0.1 M Na₂SO₄) due to the formation of active chlorine species. Among the tested materials, rutile- and P25-derived electrodes showed the highest degradation efficiencies, reaching concentrations (C/C₀) of 0.631 and 0.650, respectively. The results highlight the combined influence of precursor phase, defect structure, and electrolyte composition on the photoelectrooxidation behavior of black TiO₂ electrodes and provide insights for the design of electrochemical systems for pharmaceutical contaminants removal. Full article

In this paper, we propose a secure protocol to compute the quadratic optimisation problem under a three-party outsourcing architecture in the scenario of cyber-physical systems. To enable real-world implementation, we propose an encoding framework that uses a fixed-point expression and a truncated-mapping scheme to map real numbers into multiple data blocks, improving the protocol’s efficiency. Based on this, we define the recovery operations for decryption, addition, and multiplication. Considering computations involving three parties to solve the quadratic optimisation problem, we thoroughly analyse privacy issues during the interaction process. Then, a secure protocol is developed by designing privacy-preserving addition, multiplication, and comparison protocols based on the additive homomorphic encryption scheme. The data blowup and “0”-privacy leakage problems are addressed specifically for the gradient descent process by designing a secure addition protocol for block data and a secure comparison protocol. The efficiency and security of the proposed protocol are formally analysed in depth. Finally, through intensive experiments, we demonstrate the efficiency and security of our protocol. Full article

Achieving robust environmental responsiveness in early stage architectural spatial layout design remains a critical challenge under the imperatives of global carbon neutrality and climate-adaptive building practice. Conventional parametric design and multi-objective optimization approaches suffer from computational inefficiency, inadequate constraint satisfaction, and opaque generative logic when operating in high-dimensional design spaces. This paper presents a mathematically rigorous, climate-responsive spatial layout generation framework that unifies category theory with conditional diffusion modeling. The proposed method formalizes site-specific environmental parameter systems and architectural spatial topologies as two small categories, and establishes structure-preserving environment-to-space mappings via covariant functors; natural transformations are further introduced to characterize morphological transitions across distinct design strategies. A conditional diffusion model (CDM) serves as the generative engine, producing candidate spatial topological configurations subject to environmental parameter conditioning. A three-stage categorical constraint screening mechanism—constructed from groupoid structures and pullback limits—enforces simultaneous compliance with functional adjacency requirements, topological coherence, and multi-criteria environmental performance targets. Extensive experiments across three climatically contrasting sites (Hangzhou, Qingdao, and Lijiang) demonstrate that the framework substantially enhances environmental response performance while preserving spatial topological rationality, achieving competitive generation efficiency and constraint satisfaction relative to conventional parametric optimization baselines. These findings establish that categorical structures can serve as interpretable, mathematically consistent constraint engines within AI-driven generative design pipelines, offering a principled computational paradigm for climate-responsive architectural layout synthesis. Full article

Madagascar’s Central Highlands have experienced extensive deforestation and soil degradation, limiting the success of reforestation efforts. Poor soil fertility, particularly nitrogen limitation, constrains early seedling growth in degraded landscapes. This study evaluated the field performance of composted Black Soldier Fly frass (CBSFF) as a soil amendment for the native pioneer tree Dodonaea madagascariensis within the Ambohitantely Special Reserve. Four treatments were compared across four sites using a randomized complete block design: unfertilized control, cattle manure (4 g N), CBSFF one-fold (4 g N), and CBSFF two-fold (8 g N). The experiment was conducted on seedlings aged 16 months at the start of the study, and their growth was monitored over a six-month period. Growth responses were analyzed using generalized linear mixed-effects models with site included as a random factor. Seedling survival remained near 100% across all treatments, indicating no phytotoxic effects of composted frass under field conditions. Fertilization significantly enhanced both basal stem diameter and height growth. When standardized by nitrogen input, cattle manure and CBSFF produced comparable growth responses, indicating that nitrogen availability, rather than fertilizer identity, primarily drove early seedling performance. Height growth exhibited a clear dose-dependent response, with the double-dose CBSFF treatment producing the greatest increase. Planting method had a modest effect on height but did not alter the relative performance of fertilizer treatments. These findings demonstrate that composted BSF frass functions as an effective nitrogen source for early tree establishment in degraded tropical soils and performs comparably to traditional manure under field conditions. By validating insect-derived fertilizer within a restoration context, this study supports the integration of circular nutrient systems into sustainable reforestation strategies in biodiversity-rich yet resource-limited landscapes. Full article

This study explores the sustainability of pedagogical innovation in higher education by examining a faculty-collaborative, ADDIE-based selective flipped learning framework in an engineering education context. It addresses a persistent challenge in engineering classrooms, namely the mismatch between traditional teaching approaches and the diverse learning needs of students, while also considering whether such innovations can be realistically sustained over time. A mixed-methods sequential explanatory design was implemented in a foundational Circuits I course at Princess Sumaya University for Technology (PSUT), involving 110 undergraduate students and eight faculty members. The ADDIE model guided the design and implementation of selectively flipped instructional materials. Quantitative data were analyzed using independent samples t-tests, while qualitative data from surveys, interviews, and focus groups were examined through thematic analysis to better understand faculty experiences and implementation processes. The findings show statistically significant improvements in student performance. Independent samples t-tests indicated significant differences in overall scores and final exam performance (p < 0.001), while additional analyses of formative assessment components also demonstrated statistically significant improvements. At the same time, the results reveal important implementation challenges. Although the course was collaboratively designed, implementation varied across instructors, and faculty were divided in their willingness to continue using the approach. This highlights a clear gap between instructional design and classroom practice, shaped by both human and institutional factors. Overall, the study suggests that well-structured instructional design models such as ADDIE can support improved learning outcomes. However, the findings do not provide conclusive evidence of long-term sustainability; rather, they highlight the conditions under which pedagogical innovations may be sustained, including institutional support, faculty engagement, and alignment with teaching realities. Full article

With the rapid development of new energy vehicles, fast and reliable prediction of power battery collision damage has become increasingly important. Traditional finite-element analysis is computationally expensive and difficult to deploy for rapid prediction under varying conditions. Although learning-based methods are faster, they usually rely on single-fidelity data: high-fidelity data is accurate but scarce and costly, while low-fidelity data is abundant but less reliable. Existing multi-fidelity methods alleviate this issue, yet often suffer from imbalanced sample allocation and weak cross-fidelity modeling. Moreover, current adaptive sampling strategies cannot dynamically determine the appropriate fidelity for different regions of the design space. To address these challenges, we propose HNGP-LCA, a multi-fidelity active learning framework for battery pack collision damage prediction. Our method consists of two components: (1) an Ensemble Nested Gaussian Process module that integrates single-layer and double-layer nested Gaussian process regression to better capture high–low fidelity correlations; and (2) a Location Information Cost-aware Active Learning strategy that leverages positional information to reconstruct expected improvement under different fidelities, enabling dynamic fidelity selection during sampling. Experiments on multiple synthetic benchmarks and a real battery pack engineering case demonstrate that HNGP-LCA achieves a better trade-off among accuracy, efficiency, and cost than strong baselines such as NARCO and MFBO. In the engineering case, it improves prediction accuracy by 0.6% over NARCO and 1.29% over MFBO, while reducing dependence on expensive high-fidelity data. These results show that HNGP-LCA provides an effective and practical solution for battery collision damage prediction. Full article

Non-alcoholic fatty liver disease affects approximately thirty percent of the global population, yet public awareness remains dangerously low among young adults facing occupational risk factors. This study introduces the Fatty Liver Awareness Game (FLAG), an educational serious game designed to improve liver health literacy among software engineering students at the University of Guayaquil. While evaluated with this specific sample, FLAG is intended for the broader target population of young adults in developing nations who face occupational sedentary risk and limited access to preventive health education. Through a controlled experiment with fifty participants randomly assigned to game-based or traditional lecture instruction, the game demonstrated superior effectiveness, with a twenty-percentage-point advantage in post-test scores and a seventy-two percent reduction in incorrect responses compared to fifty percent in the lecture group. The large effect size (Cohen’s d = 1.43) and reduced performance variability among game participants indicate that interactive, feedback-rich learning environments can outperform passive instruction for this population and content domain. While the present design does not isolate the contribution of individual game elements—such as narrative framing, explanatory feedback, or mini-game interleaving—the results establish FLAG as a replicable model for digital health interventions targeting underserved populations at critical developmental junctures. Future component analyses are needed to determine which specific design features drive the observed advantages. Full article