Search Results (1,475)

Falls are a prominent issue in society and the second leading cause of unintentional death globally. Traditional gait analysis is a process that can aid in identifying factors that increase a person’s risk of falling through determining their gait parameters in a controlled environment. Advances in wearable sensor technology and analytical methods such as deep learning can enable remote gait analysis, increasing the quality of the collected data, standardizing the process between centers, and automating aspects of the analysis. Real-world gait analysis requires two problems to be solved: high-accuracy Human Activity Recognition (HAR) and high-accuracy terrain classification. High accuracy HAR has been achieved through the application of powerful novel classification techniques to various HAR datasets; however, terrain classification cannot be approached in this way due to a lack of suitable datasets. In this study, we present the Context-Aware Human Activity Recognition (CAHAR) dataset: the first activity- and terrain-labeled dataset that targets a full range of indoor and outdoor terrains, along with the common gait activities associated with them. Data were captured using Inertial Measurement Units (IMUs), Force-Sensing Resistor (FSR) insoles, color sensors, and LiDARs from 20 healthy participants. With this dataset, researchers can develop new classification models that are capable of both HAR and terrain identification to progress the capabilities of wearable sensors towards remote gait analysis. Full article

Firefighting requires high and multidimensional fitness to ensure operational readiness and public safety. In Portugal, there is limited data regarding firefighters’ fitness and exercise programs to improve readiness are lacking. This study evaluated the effects of a 6-month minimal-equipment exercise program on the physical fitness of firefighter recruits. Thirty-five male subjects (23.0 ± 2.72 years) were assessed at baseline,3 and 6 months for body composition, handgrip strength, running speed, cardiovascular endurance, anaerobic power, and upper- and lower-body strength. The intervention entailed daily sessions with 15 min of continuous running (50–65% HRmax) and active stretching, followed by alternating routines, including endurance running, free weights, interval sprints, calisthenics, and drills. A repeated-measures ANOVA and Bonferroni adjusted post hoc comparisons identified time-based changes. Significant improvements occurred across all fitness variables. Body fat fell by 8.4% and VO₂max increased (p < 0.001), surpassing occupational thresholds required for extended suppression tasks. Bench press and sit-up performance improved by 88% and 81%, respectively, while countermovement jump showed double-digit gains (13%), all of which can translate directly to hose advancement, victim rescue, and forcible entry. These results highlight that resource-constrained departments can implement effective, low-cost exercise programs for enhancing pivotal fitness components, supporting policy initiatives to include structured training throughout firefighters’ careers. Full article

In this study, polystyrene (PS) and PS/TiO₂ nanofibers were fabricated through electrospinning and quantitatively characterized to analyze and predict fiber diameters. To advance predictive methodologies for materials design, artificial neural network (ANN) models based on multilayer perceptron (MLP) and radial basis function (RBF) architectures were developed using system- and process-level parameters as inputs and the fiber diameter as the output. Two data classes were constructed: Class 1, consisting of PS/TiO₂ nanofibers, and Class 2, containing both PS and PS/TiO₂ nanofibers. The architectural optimization of the ANN models, particularly the number of neurons in hidden layers, had a critical influence on the correlation between predicted and experimentally measured fiber diameters. The optimal MLP configuration employed 40 and 20 neurons in the hidden layers, achieving mean square errors (MSEs) of 4.03 × 10⁻³ (Class 1) and 7.01 × 10⁻³ (Class 2). The RBF model reached its highest accuracy with 30 and 250 neurons, yielding substantially lower MSE values of 1.42 × 10⁻³² and 2.75 × 10⁻³² for Class 1 and Class 2, respectively. These findings underline the importance of methodological rigor in data-driven modeling and demonstrate that carefully optimized ANN frameworks can serve as powerful tools for predicting structural features in nanostructured materials, thereby supporting rational materials design and synthesis. Full article

Conversational agents are transforming digital interactions across various domains, including healthcare, education, and customer service, thanks to advances in large language models (LLMs). As these systems become more autonomous and ubiquitous, understanding what constitutes high-quality interaction from a user perspective is increasingly critical. Despite growing empirical research, the field lacks a unified framework for defining, measuring, and designing user-perceived interaction quality in human–artificial intelligence (AI) dialogue. Here, we present an integrative review of 125 empirical studies published between 2017 and 2025, spanning text-, voice-, and LLM-powered systems. Our synthesis identifies three consistent layers of user judgment: a pragmatic core (usability, task effectiveness, and conversational competence), a social–affective layer (social presence, warmth, and synchronicity), and an accountability and inclusion layer (transparency, accessibility, and fairness). These insights are formalised into a four-layer interpretive framework—Capacity, Alignment, Levers, and Outcomes—operationalised via a Capacity × Alignment matrix that maps distinct success and failure regimes. It also identifies design levers such as anthropomorphism, role framing, and onboarding strategies. The framework consolidates constructs, positions inclusion and accountability as central to quality, and offers actionable guidance for evaluation and design. This research redefines interaction quality as a dialogic construct, shifting the focus from system performance to co-orchestrated, user-centred dialogue quality. Full article

Non-Intrusive Load Monitoring is an effective method for disaggregating the power consumption of individual appliances from the aggregate load data of a building. The advent of smart meters, Internet of Things devices, and artificial intelligence technologies has significantly advanced the capabilities of non-intrusive load monitoring. However, challenges such as varying sampling frequencies and measurement sensitivities remain. This paper introduces an innovative model incorporating an Adaptive Multi-Scale Attention Integration Module (AMSAIM) to address these issues. The model leverages deep learning and attention mechanisms to improve the accuracy and real-time performance of non-intrusive load monitoring. Validated on the standard UK-DALE dataset, the model consistently demonstrated superior performance. In seen scenarios, our model achieved average F1-scores approximating 0.94 and notably reduced Mean Absolute Error (MAE) values. For washing machines, it achieved an F1-score of 0.99 and MAE of 41.64, outperforming the next best method’s F1-score by 1 percentage point. In challenging unseen scenarios, the model showcased strong generalization, achieving an F1-score of 0.91 for washing machines and reducing MAE to 7.66. Furthermore, an ablation study rigorously confirmed the necessity of the AMSAIM module, showing that the synergistic integration of the efficient multi-scale attention (EMA) and the selective kernel (SK) adaptive receptive field unit is crucial for enhancing model robustness and generalization. Our results highlight the model’s potential for enhancing energy efficiency and providing actionable insights for energy management across various conditions. Full article

With the rapid advancement of smart grids driven by renewable energy integration and the extensive deployment of supervisory control and data acquisition (SCADA) and phasor measurement units (PMUs), addressing the escalating alarm flooding via intelligent analysis of large-scale alarm data is pivotal to safeguarding the safe and stable operation of power grids. To tackle these challenges, this study introduces a pioneering alarm optimization framework based on symmetry-driven crowdsourced active learning and interpretable deep reinforcement learning (DRL). Firstly, an anomaly alarm annotation method integrating differentiated crowdsourcing and active learning is proposed to mitigate the inherent asymmetry in data distribution. Secondly, a symmetrically structured DRL-based hierarchical attention deep Q-network is designed with a dual-path encoder to balance the processing of multi-scale alarm features. Finally, a SHAP-driven interpretability framework is established, providing global and local attribution to enhance decision transparency. Experimental results on a real-world power alarm dataset demonstrate that the proposed method achieves a Fleiss’ Kappa of 0.82 in annotation consistency and an F1-Score of 0.95 in detection performance, significantly outperforming state-of-the-art baselines. Additionally, the false positive rate is reduced to 0.04, verifying the framework’s effectiveness in suppressing alarm flooding while maintaining high recall. Full article

Diagnosing faults in Permanent Magnet Synchronous Motors (PMSMs) is critical for ensuring their reliable operation, particularly in detecting internal short-circuit faults in the stator windings. These faults, such as inter-turn and inter-coil short circuits, can significantly affect motor performance and lead to costly downtime if not detected early. However, detecting these faults accurately, especially in the presence of operational noise and varying load conditions, remains a challenging task. To address this, a novel methodology is proposed for diagnosing and classifying fault severity in PMSMs using vibration and current data. The key innovation of the method is the combination of signal processing for both vibration and current data, to enhance fault detection by applying advanced feature extraction techniques such as root mean square (RMS), peak-to peak values, and spectral entropy in both time and frequency domains. Furthermore, a cooperative gain transformation is applied to amplify weak correlations between vibration and current signals, improving detection sensitivity, especially during early fault progression. In this study, the publicly available dataset on Mendeley, which consists of vibration and current measurements from three PMSMs with different power ratings of 1.0 kW, 1.5 kW, and 3.0 kW, was used. The dataset includes eight different levels of stator fault severity, ranging from 0% up to 37.66%, and covers normal operation, inter-coil short circuit, and inter-turn short circuit. The results demonstrate the effectiveness of the proposed methodology, achieving an accuracy of 96.6% in fault classification. The performance values for vibration and current measurements, along with the corresponding fault severities, validate the method’s ability to accurately detect faults across various operating conditions. Full article

This paper presents the design and implementation of the Power Stage GAIA (PSG), a high-current digital bias board developed by the Italian National Institute for Astrophysics (INAF) to extend the capabilities of the GAIA bias system. The PSG was developed within the Advanced European THz Receiver Array (AETHRA) project to support next-generation cryogenic receivers for millimeter-wave astronomy. Specifically, the AETHRA Work Package 1 (WP1) W-band downconverter integrates Monolithic Microwave Integrated Circuits (MMICs) requiring currents significantly exceeding the 50 mA limit of standard bias boards. To address these requirements, the PSG introduces a modular extension providing ten independent channels, each capable of delivering up to 500 mA with a programmable output range of 0–5 V. A key feature of the design is the adoption of a fully linear architecture based on LT1970 power amplifiers and INA225 precision sensors managed via an I²C digital interface. This approach ensures the high current capability required by modern power amplifiers while strictly avoiding the spectral noise and Radio Frequency Interference (RFI) typical of switching power supplies. Experimental validation confirms the system’s robustness and precision: the board demonstrated linear operation up to 460 mA and exceptional long-term stability, with a measured RMS voltage deviation below 50 µV. These results establish the PSG as a scalable, low-noise solution suitable for biasing high-power MMICs in future cryogenic receiver arrays. Full article

Battery management systems (BMSs) are integral components of electric vehicles (EVs), as they ensure the safe and efficient operation of lithium-ion batteries. State of health (SoH) estimation is one of the core functions of BMSs, providing an assessment of the current condition of a battery, while prognostics aim to predict remaining useful life (RUL) as a function of the battery’s condition. An accurate SoH estimation allows proactive maintenance to prolong battery lifespan. Traditional SoH estimation methods can be broadly divided into experiment-based and model-based approaches. Experiment-based approaches rely on direct physical measurements, while model-driven approaches use physics-based or data-driven models. Although experiment-based methods can offer high accuracy, they are often impractical and costly for real-time applications. With recent advances in artificial intelligence (AI), deep learning models have emerged as powerful alternatives for SoH prediction. This paper offers an in-depth examination of AI-driven SoH prediction technologies, including their historical development, recent advancements, and practical applications, with particular emphasis on the implementation of widely used AI algorithms for SoH prediction. Key technical challenges associated with SoH prediction, such as computational complexity, data availability constraints, interpretability issues, and real-world deployment constraints, are discussed, along with possible solution strategies. Full article

Accurate modeling of Battery Energy Storage Systems (BESS) is essential for optimizing system performance, ensuring operational safety, and extending service life in applications ranging from electric vehicles (EV) to large-scale grid storage. However, the simplifications inherent in conventional battery models often hinder optimal system design and operation, leading to conservative performance limits, inaccurate State-of-Energy (SOE) estimation, and reduced overall efficiency. This paper presents a framework for advanced battery modeling, developed to achieve higher fidelity in SOE estimation and improved power-capability prediction. The proposed model introduces a dynamic energy-based representation of the charging and discharging processes, incorporating a functional dependence of instantaneous power on stored energy. Experimental validation confirms the superiority of this modeling framework over existing state-of-the-art models. The proposed approach reduces SOE estimation error to 0.1% and cycle-time duration error to 0.82% compared to the measurements. Consequently, the model provides more accurate predictions of the maximum charge and discharge power limits than state-of-the-art solutions. The enhanced predictive accuracy improves energy utilization, mitigates premature degradation, and strengthens safety assurance in advanced battery management systems. Full article

This paper presents a multichannel electronics measurement system that uses microwave sensors to perform real-time microplastics assessment in aqueous environments. The system is capable of simultaneously reading up to four microwave sensors, enabling the use of multiple sensors that target microplastic particles with different sizes and properties. The multichannel capability allows the measurement of multiple MW sensors integrated with different microfluidic channel designs while targeting different MPs’ dimension ranges, although experimental validation in this work was limited to a single sensor. Each readout channel is implemented combining radio-technology-integrated circuits with a microprocessor that has advanced analog peripherals used for signal conditioning and acquisition. An ADF4351 wideband frequency synthesizer is used for excitation signal generation while an ADL5902 power detector converts the sensor output to a DC voltage. Baseline removal and amplification of the power detector output is realized with a MSP430FR2355 microprocessor which is also responsible for its acquisition at 40 kHz and digital decimation. Characterization results show the system’s capability to generate excitation signals between 700 MHz and 3.5 GHz with power levels around 0 dBm. Sensor output can be detected with a power between −50 dBm and −5 dBm and a 230 Hz bandwidth. A compact form factor of 15 cm × 10 cm × 3 cm was realized together with a low power consumption of 6.6 W. Validation was realized with a previously developed microwave sensor, demonstrating the detection of polyethylene spheres with 400 μm diameters animated in 10 mL/min flux within the microfluidics device. Full article

The aviation industry, responsible for approximately 2.5–3.5% of global greenhouse gas emissions, faces increasing pressure to adopt sustainable energy solutions. Hydrogen, with its high gravimetric energy density and zero carbon emissions during use, has emerged as a promising alternative fuel to support aviation decarbonization. However, its large-scale implementation remains hindered by cryogenic storage requirements, safety risks, infrastructure adaptation, and economic constraints. This study aims to identify and evaluate the primary technical and operational risks associated with hydrogen utilization in aviation through a comprehensive Monte Carlo Simulation-based risk assessment. The analysis specifically focuses on four key domains—hydrogen leakage, cryogenic storage, explosion hazards, and infrastructure challenges—while excluding economic and lifecycle aspects to maintain a technical scope only. A 10,000-iteration simulation was conducted to quantify the probability and impact of each risk factor. Results indicate that hydrogen leakage and explosion hazards represent the most critical risks, with mean risk scores exceeding 20 on a 25-point scale, whereas investment costs and technical expertise were ranked as comparatively low-level risks. Based on these findings, strategic mitigation measures—including real-time leak detection systems, composite cryotank technologies, and standardized safety protocols—are proposed to enhance system reliability and support the safe integration of hydrogen-powered aviation. This study contributes to a data-driven understanding of hydrogen-related risks and provides a technological roadmap for advancing carbon-neutral air transport. Full article

Rapid advancements in the field of deep learning have had a profound impact on a wide range of scientific studies. This paper incorporates the power of deep neural networks to learn complex relationships in longitudinal data. The novel generative approach, Variational Deep Alliance (VaDA), is established, where an “alliance” is formed across repeated measurements via the strength of Variational Auto-Encoder. VaDA models the generating process of longitudinal data with a unified and well-structured latent space, allowing outcomes prediction, subjects clustering and representation learning simultaneously. The integrated model can be inferred efficiently within a stochastic Auto-Encoding Variational Bayes framework, which is scalable to large datasets and can accommodate variables of mixed type. Quantitative comparisons to those baseline methods are considered. VaDA shows high robustness and generalization capability across various synthetic scenarios. Moreover, a longitudinal study based on the well-known CelebFaces Attributes dataset is carried out, where we show its usefulness in detecting meaningful latent clusters and generating high-quality face images. Full article

Obstructive sleep apnea (OSA) is characterized by recurrent partial or complete upper airway collapse during sleep. Accurate assessment of airway anatomy is crucial for risk stratification, diagnosis, and treatment planning. While polysomnography (PSG) is considered the gold standard for OSA diagnosis, it provides limited anatomical insights. Cone-beam computed tomography (CBCT) has emerged as a valuable tool with lower radiation dose for three-dimensional (3D) assessment of the upper airway space and craniofacial structures. CBCT enables precise measurement of critical airway parameters including total airway volume and length, minimum cross-sectional area, linear dimensions of anteroposterior and lateral diameters, as well as soft tissue structures such as tongue, tonsils, and adenoids. This review aims to explore and comprehensively review the role of CBCT, primarily in upper airway assessment for OSA, with an emphasis on airway measurement parameters, anatomical reference landmarks, and the variabilities, in addition to its clinical applications in treatment planning and simulation and post-treatment efficacy evaluation. This review also highlights the technical considerations such image acquisition protocols, machine specifications and software algorithm, and patient positioning, which may affect measurement reliability and diagnostic accuracy. CBCT serves as a powerful adjunct in OSA diagnosis and management, enabling comprehensive assessment of the airway space and hard and soft tissue structures. It complements PSG by guiding personalized interventions such as maxillomandibular advancement or CPAP optimization. Standardized imaging protocols and consideration of patient positioning can further improve its clinical utility. Full article

Compared to traditional marine productivity, marine new-quality productivity (MNQP) is composed of advanced productive forces driven by the deepening application of new technologies, is characterized by the rapid emergence of new industries, new business models, and new modes of operation, and is marked by a substantial increase in total factor productivity in the marine economy. It has, therefore, become a new engine and pathway for China’s development into a maritime power. The main research approaches and conclusions of this paper are as follows: ① Using a combined order relation analysis method–Entropy Weight Method (G1-EWM) weighting method that integrates subjective and objective factors, we measured the development level of China’s MNQP from 2006 to 2021 across two dimensions: “factor structure” and “quality and efficiency”. The findings indicate that China’s MNQP is developing robustly and still holds considerable potential for improvement. ② Utilizing Gaussian Kernel Density Estimation and Spatial Markov Chain analysis to examine the dynamic evolution of China’s MNQP, the study identifies breaking the low-end lock-in of MNQP as crucial for accelerating balanced development. Spatial imbalances in China’s MNQP may exist both at the national level and within the three major marine economic zones. ③ To further examine potential spatial imbalances, Dagum Gini decomposition was employed to assess regional disparities in China’s MNQP. The DER polarization index and EGR polarization index were used to analyze spatial polarization levels, revealing an intensifying spatial imbalance in China’s MNQP. ④ Finally, geographic detectors were employed to identify the factors influencing spatial imbalances in China’s MNQP. Results indicate that these imbalances result from the combined effects of multiple factors, with marine economic development emerging as the core determinant exerting a dominant influence. The core conclusions of this study provide theoretical support and practical evidence for advancing the enhancement of China’s MNQP, thereby contributing to the realization of the goal of building a maritime power. Full article