Applied Sciences doi: 10.3390/app16167996

Authors: Muhammad Shahbaz Jun Guo Jiajun Liao Tianhao Ye

Wicking geotextiles are specialized geosynthetics designed to regulate soil moisture through a combination of capillary-barrier and lateral-drainage mechanisms. Laboratory and field studies demonstrate their effectiveness in reducing volumetric water content, restricting capillary rise, and maintaining subgrade and base stability under rainfall, dry–wet cycles, freeze–thaw cycles, and traffic loading. Performance is influenced by soil type, fines content, installation depth, edge exposure, and environmental conditions. Field applications show benefits in pavements, expansive soils, pumping-prone sections, cold-region roadbeds, and permeable urban infrastructure. Multi-layer and composite geotextiles further enhance hydraulic and mechanical performance. Despite these advantages, gaps remain in standardized testing, long-term monitoring, design methods, and durability assessments under aggressive conditions. This review synthesizes recent experimental and field evidence, highlighting mechanisms, performance factors, and research needs to guide the optimal design and application of wicking geotextiles in geotechnical engineering.

Applied Sciences doi: 10.3390/app16167995

Authors: Jorge Manuel Araújo Teixeira Filipe Alexandre de Sousa Pereira

The modernization of legacy industrial machines is a major challenge in intelligent automation, particularly when critical assets must be upgraded without replacing high-value mechanical and hydraulic infrastructure. This paper presents an industrial case study on the intelligent PLC-based retrofit of an obsolete Neyrpic Digipid speed governor installed in a Kaplan turbine. The proposed solution replaces a closed, vendor-dependent controller with an open Siemens ET 200SP architecture programmed in TIA Portal, integrating existing sensors, hydraulic actuators, redundant speed acquisition, sequential state-machine control, and a digital distributor–runner blade Cam Curve. A key technical contribution is the diagnosis and mitigation of hydraulic hunting in the distributor position loop. The instability was traced to the interaction between integral control action and the intrinsic integrating behavior of the hydraulic actuator, leading to the adoption of a proportional-only position tracking strategy. The system was validated through Factory Acceptance Tests (FATs) and Site Acceptance Tests (SATs), including signal verification, startup, synchronization, load acceptance and emergency load rejection. Quantitative results demonstrate that during initial commissioning of the new PLC-based PI position loop, the LVDT position error reached 41.59% peak-to-peak, with 351.2 servo-valve reversals per minute. Disabling the integral action reduced the peak-to-peak position error to below 1.5% and eliminated steady-state valve reversals under the tested operating conditions. Separately, the historical 0.966 V oscillation detected in the legacy analog-input chain was resolved during the retrofit. During no-load startup, the unit reached 97% of nominal speed in 49.4 s, with a maximum overshoot of 2.3% and a speed tracking standard deviation of 1.1%. The complete operational cycle was successfully validated under real industrial conditions, including a near-nominal load-rejection test (approximately 2.25 MW), during which the measured speed peaked at 126.4% of nominal speed and the shutdown sequence was completed without protection-system malfunction. The results show that open PLC-based retrofits can improve maintainability, diagnostics and operational reliability in safety-critical industrial machines, while establishing a foundation for future SCADA integration and condition-based maintenance.

Applied Sciences doi: 10.3390/app16167993

Authors: Muhammad Hamza Nain Tara Afzal Akram El Barbary Hassan

Per- and polyfluoroalkyl substances (PFAS) are synthetic fluorinated compounds widely recognized for their stability and persistence in the environment. These characteristics make PFAS valuable in many applications but also pose serious health risks because they do not break down easily. These chemicals can accumulate in living organisms and persist in water systems. PFAS typically transfer from water to other media rather than being completely degraded by conventional treatment technologies such as adsorption and membrane filtration. Chemical degradation techniques, i.e., advanced oxidation processes or electrochemical conversions, require harsh reaction conditions, which make them unsustainable. Microbial degradation offers a green, sustainable alternative for PFAS remediation in water. This review critically analyzes advances in the use of bacterial, fungal, and microbial consortia for PFAS transformation via reductive and oxidative defluorination and/or metabolic reactions. Later, the influence of PFAS structural attributes, i.e., chain length and functional head groups, on microbial activities has been discussed. In the following section, the current progress toward the complete mineralization of PFAS is evaluated. Considering the critical evaluation of microbial degradation processes, several research gaps have been identified, including the lack of detailed mechanistic studies of enzymatic degradation pathways, the need to optimize microbial systems for the sustainable degradation of PFAS, and the integration of biological approaches with other technologies to achieve complete PFAS mineralization. The most important is scaling up microbial techniques by cost evaluation of the process to compete with other physicochemical techniques.

Applied Sciences doi: 10.3390/app16167994

Authors: Bartosz Sojka Anna Sojka Agnieszka Chwałczyńska

The aim of the study is to compare changes in physical fitness in children aged 7–10 training in two martial arts: Korean International Taekwon-Do Federation (ITF) Taekwon-do and Japanese Kyokushin Karate. A group of 163 children (51.5% boys) aged 7–10 years (median-7.7 years (Q1-7.4, Q3-8.1)) was examined. The study participants were divided into: Training group (T = 55), group TT (n = 19, Taekwon-do), group TK (n = 36, Kyokushin Karate) and the control group (C), consisting of 108 children. The study was multi-stage: Stage I—an initial screening for group selection, height, weight, and segmental body composition, as well as physical fitness using selected EUROFIT tests. In Stage II, children from groups TT and TK completed a 12-week training programme, while group C did not change their physical activity routines. After the program was completed, all groups (TT, TK, C) repeated the tests from Stage I. In the group of children training Taekwon-do, positive changes were observed in the range of fitness tests, including flexibility, speed of hand movements and abdominal strength; the differences are statistically significant. In the group of children training Karate, positive, statistically insignificant changes were observed only in functional strength. Introducing martial arts elements into physical education classes can have a positive impact on improving the physical condition of children aged 7–10.

Applied Sciences doi: 10.3390/app16167991

Authors: Zhongwen Chen Qingbing Zeng Zihua Chen Yixiao Zhang Heng Yang Qihao Wang

Slender obstacles, such as ropes, cables, and rubber hoses, may interfere with the operation of autonomous street-sweeping vehicles because of their narrow shapes and weak visual features. This study presents RFP-YOLO26 as an applied detector-design and systems-integration approach that combines established SPDConv, C3k2_Faster_EMA, and SimAM modules with a fixed single-cycle feedback feature pyramid consisting of an initial top-down pass, one bottom-up feedback pass, and a second top-down refinement pass. Experiments were conducted on the proprietary USLO dataset using a random image-level split. On the current internal test set, RFP-YOLO26 achieved 97.9% mAP@0.5, 65.0% mAP@0.5:0.95, 98.1% precision, and 95.6% recall. Compared with YOLOv26n, these values represent increases of 2.9, 2.2, 1.7, and 4.3 percentage points, respectively. RFP-YOLO26 contains 4.21 M parameters and requires 9.9 GFLOPs, compared with 2.38 M parameters and 5.2 GFLOPs for YOLOv26n. Deployment on the Jetson Orin Nano indicates embedded execution feasibility under the reported configuration. The primary benchmark remains a seed-0 descriptive comparison. The supplementary five-seed analysis and the seed-0 principal-model comparison under route-disjoint Split A retained the same model ordering; the additional route-disjoint runs provided descriptive route-level summaries. However, the available evidence does not establish universal statistical superiority, external generalization, or improved operational safety.

Applied Sciences doi: 10.3390/app16167990

Authors: Jinhao Shen Hua Zhang Xueneng Su Yiwen Gao Kun Zheng Cheng Long Xinbo Liu

Parallel-inverter microgrids are prone to PCC voltage distortion during islanded operation with nonlinear and unbalanced loads. Virtual-impedance methods can reshape inverter output impedance, but they also add control complexity and may introduce extra voltage drops. This paper proposes a droop-based PI-QPR control strategy to improve PCC voltage quality in islanded parallel-inverter microgrids. The droop scheme generates the fundamental voltage and frequency references, and the PI-QPR voltage outer loop regulates the fundamental, negative-sequence, and dominant low-order harmonic voltage components in the dq synchronous reference frame. The PI regulator is used for the fundamental component, while QPR branches at 2ω0 and 6ω0 compensate the negative-sequence component and the dominant fifth- and seventh-order harmonics. No additional virtual-impedance loop is introduced. Two-inverter hardware-in-the-loop (HIL) tests were conducted under nonlinear and unbalanced load conditions. Compared with the traditional voltage controller, the proposed controller reduces the measured three-phase PCC-voltage THD from 5.79–6.02% to 2.41–2.79%, confirming improved PCC voltage quality in the tested islanded condition.

Applied Sciences doi: 10.3390/app16167992

Authors: Md Rejaul Karim Md Nasim Reza Md Ashikur Rahman Dae-Hyun Lee Sun-Ok Chung

Conventional methods of plant distance and volume measurements are limited by low efficiency, limited spatial coverage, and high measurement error. LiDAR and RGB-D imaging offer cost-effective, precise, and non-destructive techniques for plant distance and volume measurements. This study aimed to measure cabbage height, volume, and distance using LiDAR and RGB-D imaging. The sensors were mounted on a 1.6 kW electric field scouting platform (EFSP) for data collection. Point cloud (PCD) data were collected using LiDAR, whereas data processing, visualization, and measurements were done using commercial software and open-source programming scripts. A total of 20 cabbage plants were analyzed. LiDAR data processing included data frame screening, outlier removal, denoising, voxelization, and generation of 3D PCD density maps. Depth image processing included importing raw data and metadata shaping using intrinsic camera parameters, visualization, extraction of depth points, and pixel-level measurements of distances and volume. RGB image processing involved image conversion, segmentation, normalization, binary masking, mask cleaning, region extraction of cabbages, separation of ROI and preparation of contours, Delaunay triangulation and convex hull preparation, ROI overlay, bounding box preparation, sharing boundary between two boxes, conversion to pixel distances, and for visualization, plant height, volume measurements, and center to center distance measurement for measuring the plant distance. LiDAR demonstrated higher measurement accuracy for cabbage plant height, circumferential volume (geometric canopy volume), and plant distance, followed by RGB-D imaging, while RGB imagery showed comparatively lower performance under the study field conditions. Overall, LiDAR and RGB-D imaging provided reliable and non-destructive approaches for cabbage geometric characterization under field conditions, although accurately capturing complex plant geometry remains challenging. Positive and negative values of bias represent the over- and under-estimated results, respectively. Future studies should include larger and more diverse plant datasets exhibiting diversified size, shape, and geometric structure to further improve the robustness and general applicability of the proposed sensing approaches.

Applied Sciences doi: 10.3390/app16167989

Authors: Caroline H. Claudino Mateus Zanotto Paula Homem-de-Mello José Miranda de Carvalho Júnior Juliana S. Souza

Developing efficient photoelectrocatalysts is crucial for advancing sustainable energy solutions to generate green hydrogen, particularly in photoelectrochemical water splitting. This study addresses these limitations by synthesizing and characterizing WO3/(W)BiVO4 heterojunctions incorporated with graphitic carbon nitride (C3N4). Our approach combines a microwave-assisted synthesis for heterojunction formation with the direct polymerization of C3N4, aiming to enhance the charge separation and light absorption. Structural and morphological analyses confirmed the presence of well-defined heterojunctions with homogeneous element distributions, while spectroscopic studies demonstrated enhanced visible light absorption. Our results show the potential of WO3/(W)BiVO4/C3N4 systems as durable and efficient photoanodes. The further optimization of polymerization conditions and band alignment strategies may unlock greater efficiency, paving the way for more effective solar-driven hydrogen production.

Applied Sciences doi: 10.3390/app16167988

Authors: Andrés Fernando Reyes Rodríguez Silvia A. Quijano Sandra S. Arango-Varela Sandra Patricia Castro Narváez

Agriculture plays a key role in global food production, whereas organic waste management remains an important environmental challenge. Organic residue-derived compost represents a sustainable strategy for nutrient recycling, soil fertility improvement, and reduced reliance on synthetic fertilizers in horticultural systems. This study systematically reviewed the scientific literature on the application of compost in cherry tomato (Solanum lycopersicum var. cerasiforme) cultivation to evaluate its effects on agronomic performance and production systems. A database search from 2004 to July 2025 identified 86 studies that met the predefined eligibility criteria. The analysis showed an increasing trend in research over the last two decades, with publications peaking in 2021. Most studies were conducted in Latin America, particularly in Colombia, Mexico, and Brazil, highlighting the relevance of the cherry tomato crop. The application of compost was associated with improved plant growth, nutrient availability, and crop productivity. The production cycles ranged from 37 to 220 days, with shorter cycles commonly observed in tropical environments. Rice husk compost is among the most frequently used organic amendments. Overall, compost represents an effective strategy to enhance agronomic performance and promote sustainable cherry tomato production worldwide.

Applied Sciences doi: 10.3390/app16167986

Authors: Tomasz Borzyszkowski Janusz Telega Sławomir Telega Małgorzata A. Śmiałek Stanisław Grzywiński Ryszard Szwaba

Self-start of lift-type Darrieus vertical-axis wind turbines remains one of the main obstacles limiting their wider application, especially at low wind speeds and low Reynolds numbers. This review examines the physical basis of the self-start problem and critically discusses the main aerodynamic and design strategies proposed to improve start-up behaviour. Particular attention is paid to the distinction between the mere initiation of rotation and true self-start, understood as the autonomous acceleration of the rotor from rest to its operating regime. The reviewed approaches include changes in rotor solidity, blade-shape modification, auxiliary blades, slot-based flow control, and active or passive pitch-control systems. The available studies show that increasing solidity generally improves low-TSR torque and facilitates start-up, but usually at the expense of lower peak efficiency. Blade-shape optimization, auxiliary blades, and flow-control concepts may improve start-up-related aerodynamic characteristics; yet, the reported evidence often concerns only motion initiation or low-TSR performance rather than full passive self-start. Pitch-control mechanisms appear to be the most direct way to reduce negative torque regions, but they introduce additional mechanical complexity. Overall, no reviewed solution can yet be regarded as a universally reliable passive self-start method for Darrieus turbines operating under low-wind-speed conditions.

Applied Sciences doi: 10.3390/app16167987

Authors: Ward L. Johnson Jared Tarr Anne-Françoise Obaton

Complex geometries of additively manufactured (AM) metal parts present challenges for rapid post-build part qualification with most conventional nondestructive measurement techniques. This study explores the potential of nonlinear reverberation spectroscopy (NRS) for detecting defects in parts with complex geometries. It focuses on measurements of two CoCrMo hollow pentagonal star-shaped specimens with and without designed-in mesoscale cavities and trapped unmelted powder. Nonlinearity and loss of the specimen with cavities are found to be greater than those of the specimen without cavities, and these differences are found to increase after six years of aging at room temperature. Different behavior of the two specimens with respect to aging and extended acoustic excitation lead to the conclusion that the dominant sources of nonlinearity and loss in the two specimens are different: contacting unbonded internal interfaces in the specimen with cavities, and dislocations in the other specimen. This conclusion is partly supported by calculations showing that the time scale of an aging-induced decrease in nonlinearity of the specimen without cavities is consistent with expected rates of migration of vacancies and nitrogen interstitials to dislocations. The results support the idea that NRS measurements would be useful for rapid nondestructive qualification of AM parts with complex geometries.

Applied Sciences doi: 10.3390/app16167985

Authors: Giuseppe Ferrara Pier Paolo G. Bruno

Pre-stack depth migration can improve imaging shallow structures in fault zones, but its use is limited by the difficulty of defining a suitable starting velocity model. This problem is relevant in small intramontane basins, where strong lateral heterogeneity, limited aperture, complex wavefields, and the absence of direct velocity measurements make first-arrival tomography and conventional semblance analysis difficult to use as standalone constraints. More demanding approaches, such as full-waveform inversion or machine-learning-based velocity modelling, are difficult to apply because these surveys commonly lack low frequencies, long offsets, and calibration data. We examine this problem on a high-resolution seismic profile from the Pantano di San Gregorio Magno basin, Southern Apennines. We compare three starting models: a first-arrival tomographic model, a semblance-derived interval-velocity model, and a horizon-guided model that combines robust elements of both. Tomography constrains the main refractors but includes short-wavelength and locally high-velocity features that are unsuitable for Kirchhoff depth migration. The semblance-derived model better matches the reflected wavefield in the basin fill but is poorly constrained where coherent reflections are weak or absent. The horizon-guided macromodel provides the most appropriate starting point for residual-moveout refinement and offers a practical strategy for shallow depth imaging where direct velocity control is unavailable.

Applied Sciences doi: 10.3390/app16167984

Authors: Yunkai Hao Jian Yang Zhigang Ji

Short-term building-electricity-load forecasting is important for energy management and smart-building operation. However, existing evaluations mainly focus on prediction accuracy and provide limited assessment of reliability under cross-building shifts and data quality variations. This study develops an empirical evaluation framework based on the HEEW dataset. A five-fold cross-building validation protocol was designed, including 55 main buildings and 27 additional buildings for extended sensitivity analysis. The prediction performance, sensor missingness robustness, prediction interval calibration, and model interpretation stability were evaluated. The results show that the LightGBM residual model outperforms the Lag-1 persistence baseline in cross-building forecasting. It reduces Normalized Mean Absolute Error (NMAE) by 11.3% in the unseen-building future-year scenario. Missingness experiments indicate that historical load features have a stronger influence on one-hour-ahead forecasting than weather features. A 50% missing rate in auxiliary load history increases NMAE by 96.71%, while weather feature missingness causes limited degradation. Conformal prediction maintains reasonable coverage for seen buildings but shows reduced coverage after transferring to unseen buildings. SHAP analysis further shows stable feature importance patterns across different scenarios. Overall, this study shows that the reliability of cross-building load forecasting should be evaluated from multiple angles, beyond point prediction accuracy alone.

Applied Sciences doi: 10.3390/app16167982

Authors: Luis Sánchez-Fernández Alessia Nizzoli María Barrera-Báez Orly Enrique Apolo-Apolo Manuel Pérez-Ruiz

Weed management in woody perennial crops relies mainly on broadcast herbicide application, with well-documented costs to soil health, biodiversity, and crop physiology. Robotic platforms offer a path toward selective, site-specific control, but orchard environments present challenges such as irregular geometries, trunks, and strong illumination variability under the canopy that have limited fully integrated solutions. This work presents an autonomous robotic platform for selective under-canopy weed control in woody crops, combining multi-sensor perception, a six-degree-of-freedom robotic arm with a mechanical trunk-avoidance mechanism, and a precision spraying module with independently controlled nozzles. A weed image dataset tailored to Mediterranean orchard conditions was built from controlled-cultivation and commercial-orchard imagery under a two-phase training strategy, and the platform was evaluated in a commercial almond orchard in southern Spain. Field trials confirmed the platform’s ability to avoid tree trunks (presenting an average of 1.28% coverage near the tree trunks) and spray only selected targets under typical orchard operation but weed detection accuracy dropped substantially between winter conditions (mAP@0.5 = 93.5%) and summer conditions (mAP@0.5 = 47.8%), with uneven canopy lighting identified as the main cause. These results confirm the technical feasibility of integrating perception, navigation, and actuation into a single autonomous platform, while highlighting robust perception under canopy-induced illumination heterogeneity and tighter perception–navigation integration as the main remaining challenges.

Applied Sciences doi: 10.3390/app16167983

Authors: Xiuli Zhang

Under the impact of earthquakes, substation systems may experience equipment failures, prolonged functional recovery periods, and expanded power outage consequences. This paper proposes a collaborative optimization framework for enhancing the seismic performance of substation systems and deploying repair resources, aimed at pre-earthquake planning. The system function is characterized by the outage availability of feeders weighted by load and user importance, and the user outage cost is explicitly defined as an economic consequence indicator with monetary units, rather than a dimensionless resilience index. A directed graph model is constructed to simulate the post-earthquake functional recovery process, and under given seismic hazard and vulnerability parameters, Monte Carlo sampling is used to capture the randomness of equipment condition failures and repair durations. Sensitivity analysis is employed to identify critical equipment, and the elitism-preservation and adaptive evolution non-dominated sorting genetic algorithm II (ERA-NSGA-II algorithm), which integrates heuristic initialization, adaptive evolution, and diversity maintenance mechanisms, is proposed to achieve joint optimization of repair teams and equipment reinforcement plans. A typical 220 kV substation case study demonstrates that the framework is feasible under the analyzed scenarios and exhibits better empirical search performance compared to the selected benchmark algorithm. Due to the limitations of a single topology and certain fixed input parameters, the obtained results are scenario-dependent and cannot be used to infer general applicability or global convergence.

Applied Sciences doi: 10.3390/app16167981

Authors: Lijiao Zhang Haojie Hu Bo Xie Zejun Zhang Qin Zhang Haitao Wang Qi Liu

Borehole effects are a critical source of uncertainty in uranium fission prompt neutron (PFN) logging, particularly in sandstone-type uranium deposits with complex borehole conditions. This study develops a quantitative correction method for borehole diameter effects based on Monte Carlo numerical simulations and the epithermal-to-thermal neutron ratio (E/T ratio). A 1:1 Monte Carlo model of the PFN logging system and a representative sandstone formation model were established to investigate neutron transport behavior under varying borehole diameters and tool positions (centered and eccentered). The simulation results show that both epithermal and thermal neutron count rates decrease with increasing borehole diameter, while the E/T ratio exhibits a nonlinear decreasing trend. This behavior indicates that borehole geometry significantly influences neutron moderation and detection efficiency. Based on the relationship between borehole diameter and the E/T ratio, a correction function for borehole effects was derived using curve fitting methods. The proposed model was validated against Monte Carlo simulation results and experimental measurements, showing relative errors within 10%. The results demonstrate that the proposed method can effectively quantify and correct borehole diameter effects in PFN logging. This provides a physics-based and geometry-dependent correction framework for improving the accuracy of uranium quantification in complex borehole environments.

Applied Sciences doi: 10.3390/app16167980

Authors: Zongjie Mu Jingna Yan Zhaowei Sun Haizhu Wang Wenhao He Zhuang Yan Zhehua Yang Panpan Zhang

Subglacial drilling through polar ice–rock transitional zones is frequently hindered by severe load fluctuations and instability induced by strong formation heterogeneity. This study investigates the coupled drilling mechanics across pure ice, an ice–rock mixture, and bedrock using a fracture-mechanics-based finite–discrete element method (F-DEM) with a strain-softening constitutive model for brittle ice. The simulations quantify the evolution of drilling force components and torque under varying gravel content, weight on bit (WOB), and rotational speed. Results show that drilling responses are strongly governed by formation-dependent mechanical behavior, transitioning from stable periodic fluctuations in pure ice to impact-dominated irregular variations in mixtures, and finally to high-level continuous resistance in bedrock. Rotational speed significantly influences drilling stability through its interaction with formation properties; lower speeds effectively suppress transient load fluctuations in pure ice and mixtures, whereas moderate speeds mitigate stick–slip effects in rock. Conversely, WOB primarily controls axial penetration while also influencing load fluctuation characteristics. The optimal WOB is identified as 10 kN for stable shearing in pure ice, reduced to 7.5 kN in mixtures to buffer lateral impacts, and increased to 10 kN in rock for efficient penetration and stable cutting. Furthermore, increasing gravel content amplifies force and torque sensitivity, transitioning the system from stable periodic behavior to impact-driven instability. These findings emphasize the necessity of a layer-dependent adaptive control strategy, providing a quantitative framework to optimize parameters and drill-bit design for safer, more efficient subglacial operations.

Applied Sciences doi: 10.3390/app16167978

Authors: Ying Liu Zeguan Qiao

Author verification is a fundamental task in natural language processing with potential applications in digital forensics, copyright dispute resolution, and social media account linkage. While existing deep pre-trained models achieve high accuracy, they incur substantial computational costs, high inference latency, and poor interpretability, limiting their deployment in resource-constrained environments. To address this gap, a lightweight and interpretable author verification method is proposed that relies solely on hand-crafted stylistic features and classical machine learning classifiers, requiring no GPU acceleration. Specifically, three complementary feature sets are extracted per text: 37 punctuation features, 22 text-style features, and an 800-dimensional character-level TF-IDF vector. For each text pair, we compute the absolute difference and element-wise product for each feature group separately, and we also concatenate all resulting vectors to form the final classification representation. On a public review dataset under a strict author-level split, the proposed method achieves competitive verification performance with significantly improved training and inference efficiency. Notably, while the lightweight neural architectures evaluated in this study yield marginal accuracy gains, they incur substantially longer inference and training times, validating the efficiency-performance trade-off of our approach. Ablation studies confirm the contribution of each feature group, with text statistics being the most influential. The proposed method offers an efficient, transparent, and easily deployable solution for author verification in resource-limited settings and provides empirical evidence for integrating traditional stylistic features with lightweight models.

Applied Sciences doi: 10.3390/app16167979

Authors: Assamae Chabni Celia Bañares Carlos F. Torres

Olive oil production is an economically and culturally important agro-industrial activity, particularly in Mediterranean regions, but conventional extraction processes generate significant quantities of by-products and may lead to the loss of valuable bioactive compounds. In this context, improving extraction efficiency while preserving oil quality and promoting sustainable processing has become a major research focus. This review examines the composition of virgin olive oil with particular attention to minor bioactive compounds and discusses how conventional processing technologies influence their distribution and retention. The environmental implications of traditional extraction systems, especially the generation of olive mill wastewater and solid residues, are also addressed. Furthermore, emerging and alternative extraction technologies are analyzed, including screw press (expeller) and supercritical CO2 extraction, highlighting their potential to enhance the recovery of bioactive compounds and reduce environmental impacts. Particular emphasis is placed on water-free or reduced-water extraction approaches and on the valorization of olive mill by-products as part of a circular economy strategy. Overall, innovative extraction technologies can reconcile efficiency, sustainability and quality by minimizing polyphenol losses and wastewater generation, requiring further industrial optimization.

Applied Sciences doi: 10.3390/app16167977

Authors: Maoqian Zhang Elena Pasternak Arcady Dyskin

Understanding the mechanical behaviour of blocky materials and structures is critical in engineering fields dealing with rock masses, masonry, ice crust, and fault gouge. The diverse block kinematics make mechanical responses difficult to predict. An important mechanism controlling deformation of block assemblies (e.g., blocky rock mass) is so-called block elbowing, a process in which rotating blocks push neighbouring blocks apart. Previously, this mechanism has been studied using 1D chain models; however, the effect of higher dimensionality has not been fully understood. In this paper, a two-dimensional blocky structure (block panel) is analysed. It is found that adding a kinematic degree of freedom fundamentally alters structural behaviour. The rotation angle of blocks remains the same such that the assembly breaks into layers sliding over each other; however, the mutual sliding is complex and dependent on frictional conditions. First, the structure becomes skewed, with the sliding directions both subhorizontal and/or subvertical. Second, the sliding modes are not necessarily purely subhorizontal or purely subvertical; combined sliding patterns are more prevalent. These observations highlight the fundamental role of elbowing in governing the block kinematics. In frictionless conditions, coordinated block rotations and complex block sliding are observed. The only mechanism capable of producing this behaviour is elbowing. In frictional conditions, sliding is considerably restrained and the assembly tends to rotate as a whole. These findings contribute to the understanding that neglecting block elbowing may overlook critical deformation mechanisms. The results provide a mechanical basis for analysing blocky structures in rock engineering and related geophysical systems.

Applied Sciences doi: 10.3390/app16167975

Authors: Youquan Yuan Jiandong He

Redistributing a nearly fixed hydrogen–air inventory can alter early flame response under confinement, while terminal-state comparisons may confound mixture placement with unequal flame development. Three-dimensional WALE–PaSR calculations in a 0.240×0.020×0.020m semi-open channel compare eight redistribution archetypes whose hydrogen and oxygen inventory mismatches do not exceed 0.385% and 0.011%, respectively. Cases are aligned over χ=0.05–0.15 using flame-front speed, whole-domain gauge pressure, positive heat-release, hydrogen-consumption, and wake/backflow diagnostics. The common-window front speed spans 26.5–62.3ms−1, while the monitored responses do not collapse onto one trajectory. A one-dimensional check using the same 10-species/27-reaction mechanism differs from the Dayma–Halter–Dagaut experimental values by at most 6.85%. Selected grid, maximum-Courant-number, and PaSR perturbations, together with representative positive-Sutherland AX-I/U10 calculations, bound numerical and transport-model sensitivity. For that pair, the response direction is retained for front speed, instantaneous whole-domain maximum gauge pressure, and the domain-integrated positive heat-release-rate indicator, whereas the domain-mean pressure contrast remains within the selected 6% interpretive envelope. Spatial redistribution is therefore supported as a model-bounded screening variable, not as facility-level validation, a universal severity ranking, or safety certification.

Applied Sciences doi: 10.3390/app16167976

Authors: Yuan Ren Zhanfang Chen Zeming Du Xiaoming Jiang

This study evaluates whether an offline system can conservatively rerank future learning activity profiles under temporal distribution shift; it does not test whether an intervention improves actual student learning. An activity profile comprises temporally binned activity type proportions, click intensity, and active bin indicators. A multilayer perceptron (MLP) predicts a base profile from the first 30% of a course, seven local residual candidates are transferred from similar training students, unsupported candidates are excluded, and a ridge outcome model fitted with inverse probability weighting (IPW) estimates simulator-defined utility. The proposed support-constrained bootstrap lower quantile rule (SC-LQ) switches only when the empirical 10th percentile of a candidate’s bootstrap gain distribution is positive; this quantity is a ranking statistic, not a calibrated confidence bound. Experiments used five Open University Learning Analytics Dataset (OULAD) courses and frozen semi-synthetic potential utilities. Under natural decision rules, SC-LQ switched for 47.5% of development students and 52.2% of holdout students, whereas IPW argmax switched for 81.8% and 86.9%, respectively. At exactly matched switching coverage, SC-LQ did not significantly improve the mean simulator value over IPW (paired difference 0.00065, 95% confidence interval (CI) [−0.00030, 0.00161], Holm-adjusted p = 0.135) but reduced overall harm by 0.01533 and utility loss above 0.01 by 0.02494. Candidate-level empirical coverage of the q10 (10th-percentile) statistic was only 0.763 with 200 bootstrap models, confirming that SC-LQ provides empirical risk ranking rather than a safety guarantee. The value–risk pattern persisted across temporal resolutions, candidate set sizes, validation-selected anchors, and several outcome and utility models but failed or weakened in deliberately discontinuous and misspecified settings. These findings support conservative fallback as a simulator-tested risk control principle for active OULAD learners, not as evidence of causal learning improvement.

Applied Sciences doi: 10.3390/app16167973

Authors: Yuchu Lu Wei Dong Chengxuan Tang Yuewei Tian Xun Huang Yueheng Meng Yongxiang Cai Youzhuo Zheng Haonan Cui Niancheng Zhou

Existing EMTR methods typically adopt passive location schemes that rely on transient signals generated by faults. When faults occur at a low inception angle, the resulting traveling-wave signals are typically weak and suffer from poor detectability. Active pulse injection provides controllable excitation and improves signal identification. The transfer function similarity method achieves high accuracy; its practical application is constrained by the requirement for transient voltage at the fault point. To address these limitations, this paper proposes a fault location method based on active pulse injection and systematically investigates the characteristics of fault voltage in both frequency and time domains. First, the frequency-domain formulation of fault voltage in the reversed-time process is derived, and the applicable scope of the energy metric is evaluated. The waveform features of the reversed-time fault voltage are then analyzed to assess the similarity between the fault voltage and the injected pulse voltage, as well as the applicability of the MCCC metric. An improved IMCCC criterion is developed to quantify the similarity between the fault voltage and the forward-time voltage. Furthermore, a symmetry similarity coefficient (SSC) is defined by leveraging the inherent symmetry property of fault voltage waveforms. The four metrics were validated using reduced-scale experiments and simulation studies. All metrics achieved accurate fault location in simple lines. In complex networks, the energy metric showed significant deviation from the real fault location. The IMCCC and SSC metrics provided higher accuracy than the MCCC metric and maintained reliable fault location for grounding faults up to 300 Ω.

Applied Sciences doi: 10.3390/app16167974

Authors: Wonbae Kim Hee-Kyong Yoo Nammee Moon

The deployment of Large Language Model (LLM)-generated SQL in Artificial Intelligence of Things (AIoT) systems introduces critical security risks, as prompt injection attacks can manipulate LLMs into producing unauthorized queries that expose sensitive data or execute destructive operations. Existing Natural Language to SQL (NL2SQL) research targets query accuracy, while current Model Context Protocol (MCP) servers offer only SQL-level protection without fine-grained, role-based access control. This paper proposes SecureMCP, a policy-enforced framework that integrates Role-Based Access Control (RBAC) with an MCP server to establish multi-layer defense for LLM-generated SQL execution. Grounded in an explicit threat model, the framework chains five defense modules in a sequential fail-closed pipeline addressing six prompt injection types spanning four adversary goals. We evaluate SecureMCP on the IoT-SQL dataset using Qwen3-8B, reporting filter performance—false positive rate (FPR) and false negative rate (FNR)—separately from LLM generation quality. On benign queries, the framework maintains a low false positive rate (0.3–2.2%) across four RBAC roles while keeping execution accuracy among allowed queries within 65.1–76.4%, matching the unprotected baseline of 63.8% and confirming that the defenses act as a transparent pre-execution filter. On 2400 adversarial queries, SecureMCP limits the effective false negative rate—computed over realized threats in which the injection payload was actually incorporated—to 3.98%, and an ablation confirms that RBAC and MCP-level defenses are complementary, as neither blocks the full range of injection vectors alone. The 72.5% injection incorporation rate confirms high LLM susceptibility, establishing the necessity of external policy enforcement.

Applied Sciences doi: 10.3390/app16167972

Authors: Konstantina-Sevasti Komnou Athanasia K. Tolkou

Triclosan (TCS) is an antimicrobial agent belonging to the bisphenol class, and is widely used in healthcare applications and in personal care products (PCPs). As an emerging pollutant frequently detected in aquatic environments, its toxicity to aquatic organisms and the male reproductive system highlights the need for effective removal methods such as adsorption. In this study, bio-based activated carbon was prepared from chestnut shell as waste biomass (CnSAC), modified by polyethylenimine (PEI) (CnSAC/PEI), manganese oxide MnO2 (CnSAC/Mn) or a combination of these (CnSAC/Mn-PEI). The prepared adsorbents were evaluated for their efficiency in TCS removal. The composite material CnSAC500PEI exhibited improved adsorption efficiency for TCS (99.5%) at pH 3, whereas MnO2 modification alone did not provide a significant improvement over the pristine activated carbon. The pHpzc analysis determined the surface charge of the adsorbent, while SEM, FTIR, EDS, and XRD characterization confirmed that the modification formed a thin, homogeneous PEI layer. This layer introduced N–H and C–N functional groups without affecting the porous structure of the activated carbon. The adsorption kinetics showed excellent agreement between both the PSO and PFO models, a combination of physical adsorption mechanisms and surface interactions. Isotherm analysis revealed a Freundlich behavior, indicating adsorption on a heterogeneous surface and a maximum capacity of 250.33 mg/g at 303 K for the CnSAC500PEI material.

Applied Sciences doi: 10.3390/app16167971

Authors: Fei Jiang Chengshuai Liu Xiaofeng Liu Yongsheng Sun Chenglin Han Luhan Wang Shaolong Jiang Xiaocai Liu Guoqing Yao

The spatial distribution of ground control points (GCPs) is a critical factor affecting the accuracy of UAV photogrammetry in hilly terrain. In existing studies on GCP distribution, researchers have largely focused on planar uniformity metrics in flat terrain or on the effects of flight parameters in mountainous areas, with limited attention to the distinct roles of horizontal and vertical placement. In this study, we utilized a consumer-grade RTK-equipped UAV to acquire aerial imagery in a typical hilly area, with 27 high-precision GCPs deployed as a reference dataset. Four comparative experiments combining random/uniform distributions in both horizontal and vertical dimensions were designed to quantitatively analyze the impact of different distribution patterns on aerial triangulation and mapping accuracy. Our results demonstrate that the dual-uniform distribution strategy (i.e., uniform in both planimetric layout and elevation stratification) achieves the highest accuracy among the four tested configurations, with horizontal RMSE of 0.045 m and vertical RMSE of 0.039 m. Furthermore, we propose the Spatial Distribution Balance Index (SDBI), which integrates the Planar Uniformity Index (PUI) and Vertical Uniformity Index (VUI) with a terrain-adaptive weighting mechanism. The VUI weight, exemplified as β = 0.714 for this study area via a Sigmoid nonlinear amplification function (k = 15, x0 = 0.15), enables the SDBI to adaptively reflect terrain sensitivity to vertical control. The enhanced SDBI exhibits a correlation coefficient of r = −0.93 with final accuracy, validating its effectiveness as a GCP layout evaluation tool. In this study, we establish the SDBI as a diagnostic metric that quantitatively links GCP distribution characteristics to photogrammetric accuracy outcomes, providing both theoretical insights into anisotropic error propagation and practical guidance for deployment design in hilly regions.

Applied Sciences doi: 10.3390/app16167970

Authors: Noelia García Rosalía Rodríguez Gemma Vicente Juan J. Espada Luis Fernando Bautista

Endocrine-disrupting compounds (EDCs) are frequently detected in wastewater and aquatic environments because of their incomplete removal in conventional wastewater treatment plants. Evaluating alternative treatment technologies, including microalgae-based systems, requires analytical methods suitable for low-volume samples. In this work, a solid-phase extraction (SPE) procedure coupled with high-performance liquid chromatography with diode array detection (HPLC-DAD) was developed for the simultaneous determination of methylparaben (MeP), propylparaben (PrP), butylparaben (BuP), benzophenone (BP), bisphenol A (BPA) and estrone (E). The method combines relatively low sample volume and limited eluent consumption with the analysis of six chemically diverse EDCs in biomass-containing samples using accessible HPLC-DAD instrumentation. Sorbent mass, sample volume, elution volume, and pH were evaluated experimentally, with Random Forest used for exploratory multivariable interpretation. Selected conditions were a 500 mg cartridge, 30 mL of sample volume at pH 7 and 2 mL methanol elution volume. The method showed satisfactory linearity (R2 ≥ 0.98), limits of quantification between 1.67 and 6.67 ppb, and stable recoveries across the evaluated concentration range. Applicability was demonstrated in a Scenedesmus sp.-based system. AGREEprep assessment yielded a score of 0.34. The developed SPE method provides a low-volume approach for monitoring EDCs in microalgae-based wastewater treatment systems.

Applied Sciences doi: 10.3390/app16167968

Authors: Jiangbing Sun Xinyi Zhu Yan Zhang Wei Qian Hongbing Zhang Yihang Ge Zhenyi Song

Rock image recognition based on deep learning serves as a critical task in intelligent geological analysis and core documentation. Although recent advances in deep learning have established the technical foundation for rock image identification, its practical deployment remains constrained by two main bottlenecks: intricate feature representations and ubiquitous label noise. Complexity of rock image features limits the effectiveness of intelligent model applications. Additionally, manual mislabeling and coarse marking issues during lithology annotation are often overlooked, leading to label noise in training datasets. To resolve these challenges, this study proposes a robust learning framework that integrates a novelty-driven feature extraction module with noise-resistant loss functions. Specifically, a Rock Hierarchical Heterogeneous Attention (RHHA) module is designed for enhancing rock feature extraction capabilities by applying distinct spatial and channel attention biases at shallow and deep layers of the network. Furthermore, Symmetric Cross Entropy (SCE) loss is used for loss calculation to enhance the robustness of models under noisy conditions. Comparative experiments were carried out on a real-world dataset of metamorphic rock images from northern Jiangsu Province, China. The results show that the proposed training framework achieves the best performance in rock image identification tasks, outperforming baseline and other comparative methods. The RHHA module outperforms other existing attention modules in capturing the texture and semantic features of rocks. In particular, the SCE loss effectively mitigates overfitting and maintains excellent generalization capability in the presence of noisy labels. The proposed framework holds promise for providing new insights into rock image recognition tasks.

Applied Sciences doi: 10.3390/app16167969

Authors: Sudha Kamaraj Muthumeenakshi Kailasam Dhanasekaran Subramanian

This study focuses on reducing harmonic distortion in residential electrical systems caused by the use of nonlinear household appliances. A combined prediction and control framework is proposed to estimate and reduce total harmonic distortion (THD) under different operating conditions. Measurements were collected from common domestic appliances, along with environmental factors such as temperature and humidity. An auto-optimized neighborhood fuzzy rough set (AO-NFRS) method is used to identify important input features. These features are then used in a physics-informed machine learning model to predict THD. Based on the predicted values, a Bayesian-optimized ANFIS controller is applied to decide the suitable filtering mode in real time. The results show that the proposed method improves prediction accuracy and reduces harmonic distortion compared to existing methods. It also provides stable filter switching under changing load conditions. The study demonstrates that combining measurement data, physical relationships, and adaptive control can improve power quality in residential systems.

Applied Sciences doi: 10.3390/app16167967

Authors: Myat Noe Kabyar Qian Huang Zekun Xu Jun Chen

The generation of realistic wind speed time histories is essential for wind engineering analysis but remains challenging due to the scarcity of high-quality measured data and the non-stationary, stochastic nature of atmospheric turbulence. While existing artificial intelligence-based data-driven methods in wind engineering mainly focus on conditional forecasting tasks, the unconditional generation of independent wind speed time histories has received limited attention. To address this gap, a novel spectrogram-based generative framework, called DiffWind, is proposed for wind speed time history generation based on denoising diffusion probabilistic models (DDPM). Wind speed time histories are transformed into magnitude spectrograms using the short-time Fourier transform (STFT), modeled in the spectral domain using a U-Net-based diffusion model, and reconstructed through the Griffin–Lim algorithm (GLA). Field-measured wind speed records were employed to tune the STFT-GLA hyperparameters and train the DDPM. The effectiveness and accuracy of the STFT-GLA combination were validated through numerical experiments, while the diffusion-based spectrogram generation was evaluated using quantitative metrics. The results indicate that the proposed framework can generate high-fidelity wind speed time histories that reproduce key statistical, temporal, and spectral characteristics of measured wind data while demonstrating the capability to generate longer-duration records, highlighting its potential for wind engineering applications.

Applied Sciences doi: 10.3390/app16167964

Authors: Emanuel-Crăciun Trînc Beatrice Arvinti Emil-Radu Iacob Cristina Stolojescu-Crișan

Chest X-ray pneumonia assessment is commonly formulated as a binary classification problem that does not explicitly represent radiographic burden. In this work, we reinterpret the RSNA 2018 Pneumonia Detection Challenge dataset by using expert bounding-box annotations to derive three burden-oriented groups: low/mild (Severity 1), moderate (Severity 2), and severe (Severity 3). We evaluate a classical binary baseline, a unified four-class classifier, and three severity-specific binary specialist branches. The selected binary Vision Transformer achieved 95.07±0.05% maximum validation accuracy, while the unified four-class model reached 79.16±0.63%. To complement exact-match accuracy, we report three task-specific ordinal measures: Severity Consistency Score (SCS), probability-aware Severity Consistency Score (pSCS), and Adjacent-Class Accuracy (ACA). These reached 91.80%, 89.18%, and 96.10%, respectively, indicating that many non-exact predictions remained close to the reference burden category. The specialist branches achieved maximum validation accuracies of 94.59±0.33%, 97.91±0.05%, and 99.03±0.08% for Severity 1, Severity 2, and Severity 3, respectively. At the joint-system level, maximum-probability fusion of the three specialist outputs achieved a Severity Binary Accuracy (SBA) of 98.17±0.17% on the common evaluation cohort. Overall, the results show that RSNA bounding-box extent can support a reproducible, annotation-derived radiographic burden analysis, while ordinal-aware measures provide complementary information to strict four-class accuracy.

Applied Sciences doi: 10.3390/app16167966

Authors: Wenxin Dong Zhuohang Zhang Shizhou Zhu Yalina Qi Fei Gao Minke Duan

The prevention of coal mine fires and the sequestration of CO2 represent two grand challenges that have traditionally been addressed separately. Here we report a CO2-modified bentonite-based sealing material that concurrently achieves fire resistance, gas sealing and mineral-carbonation CO2 uptake through rational materials engineering. In this work, “carbon-negative” is used as a comparative property: the material’s cradle-to-gate embodied emissions combined with its measured 28-day mineral uptake are lower than the cradle-to-gate footprint of a conventional cement-based benchmark under the stated system boundary. High-pressure CO2 intercalation expanded the montmorillonite d-spacing from 12.48 to 14.79 Å and introduced carbonate functional groups (1435 cm−1), as confirmed by FTIR and XRD. Systematic optimization of a bicomponent formulation incorporating municipal solid waste incineration slag and CO2-saturated zeolite yielded a material with 28-day compressive strength of 37.9 MPa, a fire resistance limit of 186 s and O2 reduction from 13.7% to 4.9%. Notably, carbon sequestration reached 24–40 kg CO2 per ton through mineral carbonation, validated by carbonate peaks in FTIR and calcite detection in XRD. Carbon accounting based on a cradle-to-gate inventory (342–408 kg CO2-eq/t) combined with the measured 28-day mineral uptake (24–40 kg CO2/t) yields a comparative net balance of −46 to −188 kg CO2-eq/t relative to a conventional cement-based benchmark under the stated system boundary. The synergistic mechanism involves CO2-modified bentonite-regulating layer spacing, alkali-activated slag releasing Ca2+/Mg2+ for carbonate precipitation, and zeolite providing endogenous carbon slow-release. This work establishes a materials-chemistry paradigm for transforming industrial waste streams into functional carbon sinks while addressing critical mining safety needs.

Applied Sciences doi: 10.3390/app16167965

Authors: Yılmaz Seryar Arıkuşu Alexandra Catalina Lazaroiu

The problem of load frequency control (LFC) becomes more severe with the extensive integration of photovoltaic (PV) generation owing to the intermittent nature of the source. In this study, a machine learning approach is developed to design the proportional–integral–derivative (PID) controller of a two-area PV–thermal LFC system, extending a prior proportional–integral (PI) benchmark to full PID action. A Random Forest model is trained to predict the relationship between the six PID gains and the closed-loop integral of time-multiplied absolute error (ITAE), yielding an accurate performance model (test R2 = 0.933) that is subsequently searched by a metaheuristic optimizer to determine the controller gains; the resulting controller is termed ML-PID. The novelty of the approach lies in employing the learned model not as a controller or a physical-quantity predictor, as in existing ML-based LFC studies, but as a reusable performance model that maps the controller gains directly to the closed-loop index and guides the PID design. To isolate and quantify the contribution of the learned model, the same three optimizers, namely the Cheetah Optimizer (CO), the Grey Wolf Optimizer (GWO), and Particle Swarm Optimization (PSO), are also applied directly to the plant, yielding purely metaheuristic controllers (CO-PID, GWO-PID, and PSO-PID) that are compared against the machine learning-assisted designs under identical algorithms and computational budget, with CO selected on the basis of the Friedman and Wilcoxon tests. The proposed ML-PID-CO controller attains the minimum ITAE under a step-load disturbance, approximately 70% lower than that of the reference SCHO-PI controller and comparable to the directly optimized controllers, with reduced control effort. Under a simultaneous variation in the plant time constants, it is the most robust of all controllers, exhibiting the smallest Δf1 undershoot and a performance that degrades about 4.2 times less than that of the reference. The results show that a learned performance model provides a good and reusable basis for PID design. It can be searched over repeatedly once built and reduces the per-design simulation burden relative to direct metaheuristic tuning, while the design is largely independent of the optimizer used.

Applied Sciences doi: 10.3390/app16167963

Authors: Danuta Kołożyn-Krajewska Marta Pokora-Carzyńska Agnieszka Rudzka Arkadiusz Żarski

One of the reasons for food waste is the inefficient use of by-products. This paper presents a current overview of technological solutions for processing waste, aimed at giving it a proverbial “second life.” To this end, the most important extraction, conversion, and valorization methods are summarized and compared. Extraction using green solvents (ionic liquids, deep eutectic solvents, supercritical fluids) and green methods such as micro-wave-assisted extraction (MAE) and ultrasound-assisted extraction (UAE) are discussed. Advances in biorecycling and biotechnological methods, taking into account the role of genetically modified microorganisms (GMOs), are also described. Attention was paid to innovative approaches and current trends, such as 3D food waste printing, hybrid technologies, and the concept of smart factories involving artificial intelligence (AI), machine learning (ML), and the Internet of Things (IoT). Although all of the mentioned food waste upcycling technologies are already in use, several hurdles limit their use to their full potential. A major problem is the smooth flow of food waste to upcycling points to ensure the required quality and processing optimization. Furthermore, societal and regulatory concerns related to the use of food waste for producing value-added dietary products need to be considered.

Applied Sciences doi: 10.3390/app16167962

Authors: Wenzheng Lv Juncheng Dai Zongyang Li Bing Luo Yuanyuan Yan Peidong Huang Yuchen Peng Jun Lu Siyao Li

Azimuthal traveltime analysis provides comparatively stable fracture-orientation estimates but limited information on fracture-related elastic changes, whereas amplitude variation with offset (AVO) inversion is sensitive to residual azimuthal moveout. We propose a sequential offset vector tile (OVT)-domain azimuthal traveltime-constrained AVO inversion workflow for reservoir-scale fracture characterization. Traveltime responses are first used to estimate the locally dominant fracture orientation and reduce azimuth-dependent event misalignment. Prestack AVO inversion is then applied separately to representative fracture-parallel and fracture-perpendicular gathers to obtain two sets of apparent elastic parameters, from which a relative tangential-weakness attribute (Δe) and Poisson’s ratio ratio (Rν) are derived. Synthetic tests show that the mean fracture-orientation error is no greater than 1.1° for prescribed noise levels of 0–50%, and that varying the maximum incidence angle from 20° to 40° causes no systematic deterioration in the recovered attributes. In the field application, the predicted dominant orientation of N 45° E–N 60° E agrees with the approximately N 50° E fracture trend identified from an independent structure-tensor-based seismic interpretation near ZT3. Compared with the uncorrected results, the corrected Δe and Rν maps exhibit improved continuity and better spatial correspondence with the independently interpreted fracture pattern. These results demonstrate that the proposed workflow effectively integrates the stable directional constraint provided by azimuthal traveltime analysis with the fracture-related elastic information obtained from directional AVO inversion, enabling reservoir-scale characterization of the dominant fracture orientation, relative fracture-related weakness variation, and possible fluid sensitivity. Within the stated assumptions and applicability conditions, the method provides a practical seismic framework for fractured-reservoir characterization and evaluation.

Applied Sciences doi: 10.3390/app16167961

Authors: Yuhui Lin Zhiwei Shen Chaopeng Li Weiwei Yu

Underwater images are frequently degraded by wavelength-dependent absorption and scattering, resulting in color casts, low contrast, blurred textures, and loss of structural details. Existing enhancement networks may struggle to balance global context modeling, local detail recovery, and computational efficiency. To address this problem, we propose MambaUNet, an efficient U-shaped state-space network for underwater image enhancement. Its core VMEC pipeline integrates visual state-space scanning to capture long-range spatial dependencies, multi-scale alignment and adaptive aggregation to improve skip-feature coherence, efficient channel attention to recalibrate feature responses, and cross-channel state-space modeling to represent channel-dependent degradation. These components are assigned stage-specific roles within the U-shaped network, forming a spatial–scale–response–channel restoration pipeline. By coordinating these components within an encoder–decoder architecture, MambaUNet improves global tone consistency and structural recovery without relying on computationally expensive self-attention. Experiments on the full-reference LSUI and UIEB benchmarks and the no-reference C60 and S16 test sets show that the proposed network achieves competitive or superior restoration quality compared with representative conventional, CNN- or GAN-based, Transformer-based, and recent Mamba-based methods. Ablation and complexity analyses further demonstrate the complementary roles of the VMEC components and the favorable balance between enhancement quality, model size, and inference speed. MambaUNet can therefore serve as a lightweight preprocessing component for underwater imaging and vision-based applications.

Applied Sciences doi: 10.3390/app16167959

Authors: Renata Žvirdauskienė Renata Baranauskienė Dalia Čižeikienė Daiva Žadeikė Simona Paulikienė

Fusarium head blight of cereal, caused by fungi of the genus Fusarium spp., reduces yield and accumulates mycotoxins. The increasing resistance of pathogens to synthetic fungicides encourages the search for natural alternatives. This work aims to determine the chemical composition of essential oils (EOs) of caraway, sage, and pine and to evaluate their antifungal activity against Fusarium graminearum and Fusarium culmorum. The composition of EOs was determined by gas chromatography and gas chromatography–mass spectrometry, and activity was assessed by agar dilution to determine mycelial growth inhibition and EC50, with the effect evaluated by two-way ANOVA. Caraway EO, belonging to the carvone chemotype (carvone 57.57%, limonene 35.86%), had the highest antifungal activity (EC50 = 727 ppm), sage EO (cis-thujone 29.59%, camphor 21.74%) had an average activity (EC50 = 1017 ppm), and pine EO, rich in α-pinene and δ-3-carene, had the lowest activity (EC50 > 3000 ppm). The antifungal activity was determined by the compounds’ chemical nature, not by their amount. Low EO concentrations (125–250 ppm) elicited a biphasic response, promoting mycelial growth (zero effect point ≈ 269–295 ppm); therefore, insufficient dosage could increase the risk of mycotoxin contamination. Carvone-rich caraway EO exhibited the highest antifungal activity among the oils tested and is worthy of further investigation as a potential tool for sustainable control of Fusarium head blight in cereals.

Applied Sciences doi: 10.3390/app16167960

Authors: Haojie Liu Jincheng Wei Zhengchao Zhang Fangchuan Wang Fan Ye Wenjian Wang

Extending the service life of asphalt pavement and cutting maintenance costs are two reasons self-healing technologies have drawn growing attention. Based on 66 studies published between 2011 and 2026, this review covers four engineered self-healing technologies: induction heating, microwave heating, microcapsule-based healing, and microbial-induced calcium carbonate precipitation (MICP). Quantitative comparisons are made for the first three; microbial healing is discussed separately because only four studies were found. Induction heating has been tested in the field and can reach healing efficiencies of up to 96.5% (maximum reported value). Microcapsule-based healing has a reported peak efficiency that approached 100% under optimal conditions, but can only be used once. Microwave heating heats more evenly across the pavement depth but has not been tested in the field. Microbial healing is at an early stage, with modest strength recovery (UCS, unconfined compressive strength, up to 47%). This review argues that evaluation fragmentation—the use of different indicators, test conditions, and material formulations across studies—makes it difficult to compare technologies and hinders field use. A standardized dual-index evaluation framework is proposed, along with three research priorities: adopting standardized protocols, monitoring existing field trials over the long term, and conducting full life-cycle cost analysis.

Applied Sciences doi: 10.3390/app16167958

Authors: He Tian Hanguang Zhao Xinchao Xu Pengfei Xin Wentao Song Youqing Ma

To address the scarcity of high-precision control points on planetary surfaces and the accumulated drift of conventional relative-localization methods in deep-space exploration missions, this paper proposes a visual absolute-localization method based on salient-landmark contour matching and centroid-consistency constraints. Absolute localization is defined as estimating the rover position in the landing-site North-East-Down (NED) coordinate system or a map-projection coordinate system, rather than in image-pixel coordinates. Stable natural objects, including dunes and impact craters, are treated as generalized feature points. Local terrain is reconstructed from binocular navigation imagery; LiDAR is additionally used in the ground physical-equivalent experiment for multi-source terrain fusion. Multi-class cross-scale contour matching provides homologous landmark associations, after which centroid consistency aligns the local terrain with the global DOM/DEM reference frame. For ten Tianwen-1/Zhurong camera stations, the mean planar error was 0.458 m and the RMSE was 0.491 m. For five ground-test conditions, the mean planar error was 0.494 m and the RMSE was 0.526 m; all tested errors were below 1 m. Because the in-orbit reference DOM has a ground sampling distance of 1 m/pixel, the in-orbit sub-meter values indicate agreement with the adopted reference products and should not be interpreted as absolute accuracy independent of reference-map uncertainty. The results support the feasibility of natural-landmark-based map localization for future Chang’e and Tianwen missions.

Applied Sciences doi: 10.3390/app16167957

Authors: Rui Ye Jialin Wang Zhihao Kong Mingxiong Ou

Binocular stereo vision is a low-cost and scalable 3D perception technology that shows strong potential in agricultural phenotyping and smart agriculture. By estimating depth from multi-view RGB images, it enables non-contact, high-precision sensing of crop structure, canopy morphology, growth dynamics, and livestock traits, providing essential support for digital and intelligent agricultural production. With recent advances in deep learning-based stereo matching, multimodal sensor fusion, and 3D reconstruction, its robustness and accuracy in complex field environments have been significantly improved. This paper systematically reviews recent progress in agricultural applications of binocular stereo vision, covering system architectures, traditional and deep learning-based stereo matching methods, point cloud reconstruction techniques, and emerging supervision strategies such as 3D Gaussian splatting. It further summarizes key applications, including high-throughput phenotyping, fruit localization and robotic harvesting, weed detection and precision spraying, autonomous navigation, and livestock body condition assessment, highlighting its role in multi-task agricultural perception systems. Finally, the paper discusses major challenges, including low-texture matching difficulty, occlusions in complex environments, cross-domain generalization, real-time lightweight deployment, and limited dataset availability. Future directions are outlined in foundation model-based visual perception, self- and weakly supervised learning, multimodal fusion, and edge-efficient model design, aiming to support large-scale deployment in smart agriculture.

Applied Sciences doi: 10.3390/app16167956

Authors: Feng Yan Dequan Zheng Feng Yu Jing Kang

Biomedical named entity recognition (NER) and relation extraction (RE) remain challenging because biomedical texts contain ambiguous abbreviations, complex entity boundaries, domain-specific terminology, and implicit relations. This study proposes a prompt-enhanced and QLoRA-adapted large language model framework for biomedical information extraction. For NER, abbreviation-aware prompting supports candidate detection, contextual interpretation, boundary-aware generation, and schema-constrained outputs. For RE, entity markers identify a predefined target pair, while filtered UMLS and MeSH concepts provide concise evidence. DeepSeek-R1-Distill-Qwen-7B is adapted using LoRA over a 4-bit quantized frozen backbone. Experiments cover three NER and three RE datasets. Across three training seeds, the dataset-level macro-average F1 values are 0.909 ± 0.001 for NER and 0.787 ± 0.001 for RE. Seed-balanced paired bootstrap resampling with 10,000 aligned instance-level resamples confirms significant improvements over a matched deterministic simple-prompt baseline on all six datasets after Holm–Bonferroni correction (adjusted p < 0.001), with absolute F1 gains from +0.091 to +0.131. Repeated-run ablations show low variability and complementary contributions from task-structured prompting, knowledge filtering, deterministic validation, and parameter-efficient adaptation.

Applied Sciences doi: 10.3390/app16167955

Authors: Francesco Buonocore Sara Marchio Simone Giusepponi Randa Assadi Muhammad Y. Bashouti Massimo Celino

First-principles calculations are used to investigate how terminal functionalization of conjugated alkenyl C8 chains, an eight-carbon backbone with alternating single and double bonds, grafted onto hydrogen-terminated silicon nanowires can be used to engineer band alignment and low-energy optical absorption. We focus on functionalization with carboxyl (–COOH), amino (–NH2), or phenyl (–C6H5) groups. All terminations preserve good passivation and comparable charge transfer from the nanowire while shifting the HOMO from slightly below the valence-band maximum to either deeper below (–COOH), within the gap (–NH2), or nearly resonant with the valence band maximum (–C6H5). Calculated longitudinal optical conductivities show that these shifts translate into distinct enhancements of optical absorption over complementary windows in the visible range, driven by transitions below energy gaps from the HOMO energy level to hybridized conduction states in the nanowire that are absent in pristine silicon nanowires and in the isolated molecules. Our results establish terminally functionalized alkenyl chains as an effective method to tailor silicon-based hybrid nanostructures for optoelectronic and photovoltaic applications.

Applied Sciences doi: 10.3390/app16167954

Authors: Sławomir Kudzia Mateusz Szramowiat Adam Kot Marcin Noga

The rapid development of electric vehicles has increased the need for a better understanding of the relationships between vehicle dynamics, energy consumption and control strategies. This study presents an experimental investigation of the acceleration and coasting characteristics of a 2024 Kia Niro EV. Road tests were combined with laboratory measurements of the vehicle mass properties, including the centre of gravity. Vehicle motion parameters were recorded using a GNSS/INS measurement system, while electric powertrain data were acquired from the vehicle CAN bus using proprietary software developed by the authors. The influence of driving mode, accelerator pedal position and regenerative braking intensity was analysed. The results showed that the selected driving mode significantly affects the acceleration characteristics only at intermediate accelerator pedal positions, whereas identical maximum performance is obtained with the accelerator pedal fully depressed. The energy required to accelerate the vehicle to 90 km/h remained nearly constant under most operating conditions, indicating high electric powertrain efficiency. During coasting, regenerative braking recovered up to 50% of the energy previously required for acceleration. The obtained results provide valuable experimental data for the validation of vehicle dynamics and energy consumption models and support the development of more efficient electric vehicle control strategies.

Applied Sciences doi: 10.3390/app16167953

Authors: Shengxue Liu Haixin Fan Xiaofeng Li

With the rapid development of the global shipping industry, the efficiency and accuracy of ship lifesaving systems remain significant challenges. To address the key issues of slow response and low intelligence in traditional maritime lifesaving systems, this study designs and implements an intelligent lifesaving system that integrates motion object detection with remote positioning and communication. A GPS positioning module is integrated for precise location acquisition, and GPRS technology is utilized to remotely transmit alarm information and location data to a rescue center. This constitutes a comprehensive technical solution comprising data acquisition, intelligent decision-making, wireless communication, and auxiliary rescue modules. Experimental results in a controlled wave pool environment demonstrate the system’s high efficacy, achieving a detection time of 15 s with a 96% accuracy and a 23.35% improvement in the rescue success rate under extreme conditions. The novelty of the proposed system lies in the integrated architectural approach and adaptive workflow. It intelligently synthesizes data from multiple sensors through a rule-based and model-driven decision pipeline. These findings suggest that the proposed integration enhances maritime rescue efficiency and reliability, highlighting its potential practical value for safe operations.

Applied Sciences doi: 10.3390/app16167952

Authors: L. Lanuwabang S. Suprakash

Low-and-slow data exfiltration by malicious insiders remains among the most operationally difficult threat classes to detect: the behavior is unremarkable in any individual session and becomes recognizable only across weeks of otherwise routine activity. This paper presents a UEBA-based machine learning framework that constructs per-user behavioral profiles from enterprise proxy and access log data, scoring sessions against a 30-feature behavioral representation spanning temporal patterns, data-transfer anomalies, domain interactions, HTTP characteristics, and session-device signals. The contribution is an operationally integrated and empirically audited UEBA pipeline that combines a pre-specified behavioral representation, isolated evaluation regimes, and session-level analyst explanations. Training used a hybrid corpus of approximately 8.96 million sessions drawn from 160 GB of real and behaviorally parameterized simulated logs. On the full hybrid held-out partition, the LightGBM classifier achieved 96.84% overall accuracy, 95.38% balanced accuracy, 91.80% malicious-class precision, 92.90% recall, 92.35% F1-score, and 98.2% ROC-AUC. Because the hybrid test set uses an enriched 20.5% malicious-session evaluation prevalence, balanced accuracy and malicious-class F1 are emphasized alongside overall accuracy. In the strict real-only evaluation, the independently trained and calibrated LightGBM model achieved 95.99% overall accuracy, 93.80% balanced accuracy, 90.30% malicious-class precision, 90.10% recall, 90.20% F1-score, and 95.8% ROC-AUC, while hybrid-to-real transfer achieved 90.1% F1-score, 89.8% recall, and 96.1% ROC-AUC. These settings are reported separately to distinguish full-corpus benchmark performance from real-log generalization. Each scored session is accompanied by a SHAP-based decomposition that identifies which behavioral signals drove the alert, supporting analyst triage in operational SOC environments. The framework was further validated on a confirmed 17-day insider exfiltration incident that existing organizational controls had not detected. BiLSTM and a Tabular Transformer evaluated under the same regime-specific partitions produced closely matched performance, with malicious-class F1 differences of at most 0.4 percentage points; LightGBM retained the strongest measured performance–deployment trade-off for the engineered tabular pipeline.

Applied Sciences doi: 10.3390/app16167951

Authors: Shajan Jacob Muthayyan Kamalam Jeyakumar

Automatic License Plate Recognition (LPR) is an essential technology in modern intelligent transportation systems, facilitating the identification of vehicles without manual intervention. It supports a wide range of applications, including traffic monitoring, electronic toll payment, parking automation, secure access management, and law enforcement operations, thereby improving transportation efficiency, security, and operational effectiveness. However, real-world license plate images are frequently affected by motion blur, noise, illumination variations, adverse weather conditions, and low resolution, which significantly degrade recognition performance. To address these challenges, this study proposes an Adaptive Hybrid CNN–Mamba–Multi-Head Attention Network with Northern Goshawk Optimization (AHCMM-Net-NGO) for robust LPR. The proposed framework combines license plate detection, progressive image restoration and enhancement, hierarchical multi-scale feature learning, efficient contextual modeling using Vision Mamba, adaptive attention refinement, and automatic hyperparameter optimization within a unified end-to-end architecture. The framework was evaluated on the UFPR-ALPR dataset containing 4500 fully annotated vehicle images captured under real-world driving conditions. Experimental outcomes demonstrate superior recognition performance, achieving 98.96% accuracy, 98.89% precision, 98.81% recall, and a 98.85% F1-score. Comprehensive experimental evaluations, including ablation studies, hyperparameter sensitivity analysis, cross-validation, and comparative performance analysis, further demonstrate the efficacy, robustness, and generalization capability of the proposed framework. Overall, the proposed AHCMM-Net-NGO framework provides an accurate, reliable, and robust solution for license plate recognition under challenging imaging conditions and demonstrates strong potential for intelligent transportation systems, although practical deployment should consider the computational requirements of the integrated deep learning framework.

Applied Sciences doi: 10.3390/app16167950

Authors: Ju Liu Hui Liu Dengfeng Ren Longcang Huang Xin Qiao Cheng Huang Kun Li Yaoyao Sun Xiaoguang Wu Zhongwei Huang

Ultra-deep lattice-like fractured carbonate reservoirs, formed by multi-period tectonic movements, feature strong heterogeneity, multi-scale fracture nesting, and anisotropic in situ stress. However, hydraulic fracture (HF) propagation behaviors within these complex formations remain poorly understood. In this study, using an unstructured fracture network approach, we simulated HF propagation in two typical fault-controlled lattice-like structures: compressive-torsion and pull-apart overlap zones. The performance of commingled, staged, and temporary plugging fracturing was evaluated, alongside sensitivity analyses of wellbore orientation, plugging timing, pump rate, and fluid viscosity. Results indicate that HFs in compressive-torsion zones exhibit long, straight geometries with local tensile activation points. Conversely, pull-apart overlap zones promote step-shaped, multi-branched fractures with superior lateral connectivity. The optimal timing for temporary plugging exhibits a delayed trend with increasing natural fracture density, ranging from 50% to 70% of the fracturing process in compressive-torsion zones, whereas an earlier implementation is preferred in pull-apart overlap zones, occurring at 33–65% of the fracturing process. Furthermore, HFs in compressive-torsion zones are less sensitive to viscosity and pump rate. To optimize stimulated volume, a moderate viscosity of 50–60 mPa·s is universally recommended. Regarding pump rates, 8–10 m3/min is ideal for balanced connectivity in pull-apart overlap zones, whereas >12 m3/min is required for compressive-torsion zones. These findings provide critical theoretical and engineering guidelines for differentiated fracturing strategies in ultra-deep reservoirs.

Applied Sciences doi: 10.3390/app16167949

Authors: Toni Cortez Jonathan Willis Himanshu Raje Ramaraj Boopathy

Constructed wetland has evolved into an effective treatment method of effluent, an important consideration in mitigating the escalating problem of antibiotic resistance in environmental and human health contexts. This study evaluated the performance of a 21-acre horizontal flow constructed wetland (HFCW) located at the Nicholls State University Farm for reducing nutrients, fecal coliform bacteria, and antibiotics, in water originating from Bayou Folse receiving treated sewage wastewater. Water samples were collected monthly in triplicate from input and output sites from January 2025 to February 2026. Water quality parameters, such as nitrate, phosphate, ammonia, sulfate, COD and antibiotic concentrations and fecal coliform abundance were analyzed to evaluate treatment efficiency. Significant reductions in sulfate (99.2%) and phosphate (40%) concentrations were observed, along with reduction in other nutrients, between input and output sites over the 14-month study period, indicating nutrient attenuation within the wetland. Antibiotic analysis revealed significant decreases in amoxicillin (81.81%), erythromycin (62.96%), sulfamethoxazole (62.02%), trimethoprim (68.72%), bacitracin (69.77%), tetracycline (69.94%), and penicillin (88.14%) concentrations. Fecal coliform bacterial numbers decreased from input to output site with 32.77% reduction. Overall, the results demonstrated that the constructed wetland effectively reduced nutrients, antibiotics concentration, and fecal coliforms. This study highlights the potential of HFCWs as sustainable mitigation tools for improving water quality.

Applied Sciences doi: 10.3390/app16167948

Authors: Orestis Stavrousis Andreas Kampitsis

This study maps the dynamic response of the IEA 15 MW reference wind turbine mounted on the UMaine VolturnUS-S semi-submersible platform across seven wave headings, spanning β = 0–180°. A fully coupled aero–hydro–servo–elastic OpenFAST model with lumped-mass catenary mooring is simulated over three environmental groups: below-rated operation, a severe sea state at rated wind, and a parked extreme, combined with seven wave headings and supplemented by a 77-case operational envelope sweep across hub wind speeds of 4–24 m/s. Responses are analyzed through time-domain and frequency-domain statistics, drift kinematics, and exceedance curves. At operating states, the maximum side-to-side moment rises from 80.44 to 273.60 MNm between following and beam seas, while the maximum fore–aft moment reduces from 566.90 MNm to 485.23 MNm, respectively. The parked extreme LC-C maximum pitch response (2.77° at following seas, 2.59° at beam seas) is roughly half that of the severe operating LC-B (6.29° and 5.38°, respectively). This directional coupling is robust across environmental severities. In the parked group, the lateral motions (sway, roll) are amplified, as feathering reduces the rotor’s aerodynamic contribution to lateral damping. The developed wind speed misalignment response atlas condenses the aforementioned data, providing a compact basis for rapid early screening of ultra-large floating offshore wind turbines.

Applied Sciences doi: 10.3390/app16167947

Authors: Shankui Ding Kun Tan Ying He

Accurate identification of mango varieties holds substantial significance for the elevation of product added value and the facilitation of market differentiation through quality-based pricing. Near-infrared (NIR) spectral analysis offers a rapid, non-destructive solution for mango variety identification. To address the challenges in fine-grained classification of NIR spectra, namely, high spectral similarity and severe overlap of absorption peaks, which make it difficult to extract nonlinear features using chemometrics, as well as the excessive complexity of existing deep learning models, a lightweight multi-scale spatial global context network is proposed. One-dimensional NIR spectra are converted into two-dimensional images through the Gramian angular difference field. Multi-scale partial convolution, coordinate-aware global context, efficient multi-scale attention, and structural re-parameterization are integrated to capture local spectral features and long-range band correlations effectively. Evaluated on two mango spectral datasets with different distributions, the proposed model achieves variety identification accuracies of 99.46% and 97.83%, with only 19.08 M parameters. Computational complexity, throughput, and latency reach 120.29 M FLOPs, 2848.5 FPS, and 0.351 ms, respectively, realizing a balance between classification accuracy and computational speed. Ablation and robustness experiments demonstrate that the accuracy of the model is improved by 5.91% and 2.15% compared with one-dimensional convolutional neural network and FasterNet, respectively. Important wavelengths obtained by threshold screening of activation maps exhibit consistency with the majority of conclusions from analysis of variance and VIP methods, while the remainder represent newly identified important bands. Validation across different temperature and batch scenarios reveals strong generalization capability. Future refinement will be pursued through increased sample diversity. Overall, high-precision identification is attained by the model at comparatively low computational overhead, indicating potential for advancing the practical application of NIR spectroscopy in agricultural quality inspection.

Applied Sciences doi: 10.3390/app16167946

Authors: Fangcao Zhang Kunpeng Chen Fei Guo Hao Yan

To improve the accuracy of fatigue assessment in patients undergoing upper limb rehabilitation, this study proposes a lightweight fatigue detection algorithm based on ensemble learning and single-channel surface electromyography (sEMG) signals. Thirty healthy subjects without upper limb injuries or severe chronic diseases were recruited, and dynamic sEMG signals of the biceps brachii were collected during dumbbell bicep curls at a sampling frequency of 2048 Hz, yielding 6650 valid experimental samples. To address the class imbalance in the sEMG dataset, the SMOTETomek hybrid sampling algorithm was employed for data balancing. Three ensemble strategies—voting, stacking, and mean fusion—were integrated and combined with four different sets of base classifiers to construct a total of 12 fusion models, and the optimal model was identified through comparative screening. The experimental results demonstrated that, after SMOTETomek sample balancing, the LightGBM-LR-MLP soft voting ensemble model achieved the best overall performance, with the accuracy, recall, precision, and F1-score for both fatigue categories all exceeding 0.93. Multiple statistical analyses, including paired t-tests, effect sizes, and 95% confidence intervals, verified the reliability of the model selection and confirmed that the soft voting algorithm significantly outperformed the other comparison schemes. Machine learning can efficiently interpret dynamic biceps brachii sEMG signals; the proposed method improves the accuracy of fatigue recognition and provides an objective, quantitative fatigue reference index for upper limb rehabilitation training. It holds promise for assisting the dynamic adjustment of rehabilitation training intensity, thereby potentially reducing the risk of overtraining injuries and enhancing the safety of rehabilitation training.

Applied Sciences doi: 10.3390/app16167945

Authors: Mohamed Dafalla Wael S. El-Sayed Ani Memuduaghan Hanaa Omar Wael Ismail Rania Hamza

Per- and polyfluoroalkyl substances (PFAS) are persistent environmental pollutants. Conventional destructive technologies, such as advanced oxidation and electrochemical processes, can achieve partial or complete defluorination. However, their high energy demand, chemical inputs, and operational complexity limit widespread implementation. Increasing evidence indicates that biological systems provide complementary mechanisms for PFAS management through partial biotransformation, defluorination of selected structurally susceptible compounds, and biomass-driven partitioning. This review evaluates the evidence for partial, largely precursor-directed PFAS biotransformation and biocatalytic defluorination through reductive, oxidative, and hydrolytic pathways. It also examines enzymatic carbon–fluorine bond cleavage by fluoroacetate dehalogenases, haloacid dehalogenases, and reductive systems, while recognizing that their demonstrated activity is generally limited to monofluorinated, activated, or polyfluorinated substrates rather than conventional fully perfluorinated PFAS. Laboratory and field observations demonstrate substantial PFAS enrichment within aquatic biomass, including intracellular compartments and extracellular polymeric substances (EPS), indicating that living systems can function as dynamic concentrators that partition PFAS from the aqueous phase. Building on these findings, this review advances a separation-first framework in which PFAS are initially captured and concentrated within biological matrices before the application of targeted destruction, regeneration, or residual-management technologies. By decoupling concentration from transformation, this approach enables independent optimization of each step, potentially reducing treatment volumes and improving overall process sustainability.

Applied Sciences doi: 10.3390/app16167942

Authors: Xiaotian Kuang Mingshi Gao Liang Xue Xin Yu Jinyu Sun Shifan Zhao Maoxing Ran

Deep longwall mining under large dip and advance inclinations can induce strongly asymmetric roof deformation and nonuniform load transfer, complicating the assessment of hydraulic support loading. Taking the 2902 working face of Chensilou Coal Mine as an engineering case, this study combines field monitoring, theoretical analysis, and FLAC3D simulation to investigate a recurrent M-shaped pressure distribution under a fixed geometry with dip and advance-direction inclinations of approximately 35°. Representative field profiles at cumulative advance distances of 40, 60, 80, and 100 m consistently exhibited an M-shaped hydraulic support pressure distribution. To interpret this spatial pattern, the main roof was represented as a bidirectionally inclined equivalent elastic plate with three clamped edges and one free edge, while the hydraulic support–immediate roof system was represented by a local Winkler foundation. The analytical solution identified two spatially separated roof-deflection concentration zones, and the FLAC3D results exhibited a broadly similar bimodal pressure tendency. The complete 28-day monitoring record was subsequently analyzed to examine whether the observed pattern persisted beyond the four representative advance stages, revealing two recurrent high-pressure sectors approximately within supports 45–75 and 105–135. The field, analytical, and numerical results showed approximate sector-level correspondence. This study therefore provides a case-specific mechanical framework relating recurrent zonal hydraulic support loading to spatially variable overburden loading, asymmetric equivalent boundary constraints, nonuniform main-roof flexure, and local roof–support load transfer.

Applied Sciences doi: 10.3390/app16167944

Authors: Runhan Liu Zilong Liu Zhudan Chen Xinglin Tong Jinglin Zhou Dazi Li

Complex industrial processes are characterized by strong coupling, nonlinear interactions, and dynamic causal dependencies between variables, posing significant challenges for accurate fault diagnosis. Conventional data-driven methods often fail to effectively exploit structural prior knowledge, limiting their ability to model complex spatiotemporal relationships. To address this issue, this paper proposes a graph iTransformer-based fault diagnosis method enhanced by reinforcement learning (RL) optimization. First, a signed directed graph (SDG) is constructed to represent the causal topological relationships between process variables, and a graph embedding algorithm is employed to generate positional encodings, enabling the incorporation of structural prior knowledge into the time series modeling framework. Subsequently, a graph iTransformer model is developed by treating individual variable sequences as tokens, thereby enhancing the learning of multivariate dependencies and complex spatiotemporal features. Furthermore, a multi-dimensional discrete Q-network (MDDQN) is introduced to jointly optimize the iTransformer modules through adaptive collaborative parameter tuning. Experiments conducted on a three-phase flow process dataset demonstrate that the proposed method achieves superior fault diagnosis performance compared with existing approaches. The results validate the effectiveness of the proposed framework in improving fault diagnosis accuracy and generalization performance for complex industrial processes.

Applied Sciences doi: 10.3390/app16167941

Authors: Reynold E. Silber Elizabeth A. Silber Kyle Staggs Carman Melendrez

Radon-222, a decay product of the uranium-238 series, is the principal source of natural ionizing-radiation exposure in most indoor environments and an established cause of lung cancer. In New Mexico, uranium-bearing geology, a legacy of uranium mining and milling, and arid, variably constructed housing create elevated but poorly characterized geogenic radon potential. This evidence-informed narrative review examines radon protection in New Mexico by synthesizing the radiological basis of the hazard (uranium-series sources, radium-226 emanation, and soil–gas transport into buildings) with measurement, monitoring, and mapping evidence relevant to under-resourced regions. We emphasize that representative, high-resolution indoor radon measurements for the state are still lacking. We show that the central challenge is not only geologic potential but the uneven distribution of measurement, testing, and mitigation capacity across housing type, tenure, geography, and jurisdiction, including Tribal lands governed by consent-based data agreements. We evaluate monitoring and outreach interventions by evidence strength and outline a phased, standards-based program (statewide measurement and data systems, high-resolution mapping, school and rental testing, workforce development, and mitigation support) to convert radiological knowledge into measurable, equitable exposure reduction. New Mexico serves as a well-documented representative example; the synthesis is intended to inform other rural, Tribal, and under-resourced jurisdictions.

Applied Sciences doi: 10.3390/app16167943

Authors: Moisés Moreira Vitor Carvalho Duarte Duque

Immersive virtual reality (VR) may support mindfulness and related therapeutic skills, but the contribution of wearable sensors and external hardware to adherence remains un-certain. This systematic review narratively synthesized randomized trials published from 1 January 2020 to 10 November 2025 that evaluated VR-based mindfulness or meditation in therapeutic populations. Four bibliographic or publisher sources were searched. Technology configuration, adherence, engagement, safety, and clinical outcomes were extracted, and methodological quality was appraised with the Joanna Briggs Institute checklist. Fifteen reports were included in the current synthesis. Hardware and software reporting was inconsistent, and no trial directly compared an otherwise-equivalent VR intervention with versus without an added wearable or peripheral. Retention was frequently described as high or attrition as low, but adherence definitions and denominators varied; discomfort, cybersickness, technical barriers, clinical events, and time demands were also reported. Physiological peripherals were used mainly for measurement or biofeedback. Because populations, interventions, comparators, and outcome definitions were hetero-generous, no meta-analysis was undertaken. The evidence supports feasibility in several therapeutic settings but does not establish a causal effect of added hardware on adherence or clinical outcomes. Future trials should preregister hardware-specific hypotheses, report complete technology and safety specifications, and directly compare otherwise-equivalent VR configurations.

Applied Sciences doi: 10.3390/app16167938

Authors: Xiaotong Guo Guang Yang Yuebao Wang Wangze Lu Rongxiang Liu

Accurate road extraction from high-resolution remote sensing imagery plays a vital role in numerous geospatial applications, including urban planning, disaster emergency response, intelligent transportation, and map updating. However, significant variations in road width, geometry, and orientation, together with complex backgrounds such as shadows, vegetation, and occlusions, often lead to incomplete extraction and poor structural continuity. To address these challenges, this paper proposes an adaptive scale-aware road extraction network, termed ASAR-Net, which jointly improves multi-scale feature representation and structural continuity. Specifically, an Adaptive Bidirectional Enhancement Module (ABEM) is introduced in the encoder to improve the representation of roads with diverse spatial scales through adaptive scale-aware convolution and bidirectional attention. Furthermore, a Directional Fusion Module (DFM) is incorporated into the decoder to guide feature reconstruction along road orientations using dynamic snake convolution, thereby facilitating the recovery of continuous and complete road structures. Extensive experiments on two public benchmark datasets, Massachusetts Roads and DeepGlobe, demonstrate that ASAR-Net consistently outperforms several representative state-of-the-art road extraction methods in terms of mIoU and F1-score. The proposed network effectively improves both the semantic completeness and structural continuity of extracted road networks, demonstrating its robustness and effectiveness for road extraction in complex high-resolution remote sensing scenarios.

Applied Sciences doi: 10.3390/app16167939

Authors: Di Wang Yixiang Zhang Yifei Zhu Jiaxin Wu Yan Di Jiawei Huang Bo Zhang Linjun Wang

Steep rock slopes are widely distributed in the karst mountainous regions of southwestern China, where structurally controlled rockfalls frequently threaten transportation infrastructure and human safety. Accurate identification of unstable rock blocks (URs) and quantitative assessment of their post-failure hazards remain major challenges because of complex discontinuity networks and fragmentation during rockfall motion. Taking the Zuojiaying steep rock slope in Guizhou Province as a representative case, this study integrates high-resolution UAV photogrammetry, automatic discontinuity identification, unstable rock block detection, and three-dimensional rockfall simulation to investigate the formation mechanisms and hazard characteristics of discontinuity-controlled rockfalls. A high-resolution three-dimensional terrain model was reconstructed from UAV imagery, and six dominant discontinuity sets were automatically identified using the I-MinPts-constrained DBSCAN algorithm. Combined with the Rock Occurrence Kinematic Analysis (ROKA) algorithm and Block Theory, 54 unstable rock blocks were identified, with wedge failure and toppling failure representing the dominant instability modes. The results indicate that discontinuity combinations govern both rock mass segmentation and unstable rock block geometry. Specifically, discontinuity sets J1, J3, and J5 mainly control wedge-shaped blocks, and J2 and J4 dominate columnar toppling blocks, whereas J6 further promotes the formation of isolated unstable rock blocks. Three-dimensional RockGIS simulations considering fragmentation reproduced the complete rockfall process from detachment to final deposition. The maximum travel distance, kinetic energy, and bounce height reached 395 m, 748.5 kJ, and 40.1 m, respectively. Fragmentation increased the number of rock blocks from 54 to 1013, substantially enlarging the potential impact area. A raster-based Rockfall Hazard Index (RHI) further revealed that the middle–lower slope and slope toe constitute the principal high-hazard zones, and under extreme scenarios, high-energy fragments may reach the G246 National Highway and adjacent infrastructure. This study revealed the formation mechanisms and hazard characteristics of unstable rock blocks controlled by discontinuity combinations in the study area, providing a case reference for rockfall hazard identification and mitigation on similar high-steep rock slopes in karst mountainous regions.

Applied Sciences doi: 10.3390/app16167940

Authors: Hristo Ivanov Iglika Sotkova-Ivanova Veselina Goranova-Marinova

Clonal hematopoiesis of indeterminate potential (CHIP) is an age-associated premalignant state defined by somatic mutations in hematopoietic cells at a variant allele frequency (VAF) ≥2% in the absence of overt hematologic malignancy. Among CHIP-associated mutations, JAK2 V617F is of particular interest because it occupies a dual biological and clinical role: it is both the principal driver of BCR::ABL1-negative myeloproliferative neoplasms (MPNs) and a clonal hematopoiesis variant conferring approximately 12-fold cardiovascular risk in selected cohorts, exceeding that reported for common DTA CHIP variants. Quantitative assessment of JAK2 V617F allele burden is therefore clinically relevant across the full disease continuum—from subclinical clonal expansion to MPN diagnosis, prognostic stratification, and therapeutic monitoring—as VAF thresholds correlate with disease phenotype, thrombotic risk, molecular response, and fibrotic progression. Digital PCR platforms, including droplet digital PCR (ddPCR) and chip-based digital PCR, have emerged as highly sensitive and reproducible methods for absolute JAK2 V617F quantification without the need for standard curves, with reported limits of detection as low as 0.01%. In this review, we synthesize current evidence on the molecular biology of JAK2-driven clonal hematopoiesis, the clinical significance of allele burden quantification, and the analytical performance of digital PCR compared with quantitative PCR and next-generation sequencing. We interpret these findings through a systems biology lens that draws together JAK-STAT signaling networks and thrombo-inflammatory pathways including inflammasome-dependent IL-1 signaling, clonal architecture, and bone marrow microenvironmental remodeling. We also discuss published quantitative models in which JAK2 V617F allele burden is treated as a dynamic state variable, while emphasizing that the present review offers a conceptual synthesis rather than a new computational model. We provide a structured comparative synthesis of published digital PCR analytical performance data, a stage-adapted proposal for clinical monitoring, and schematic models to guide future implementation. Overall, the evidence supports digital PCR as a precision tool for monitoring JAK2 V617F clonal dynamics across the CHIP–MPN spectrum, and points to several priorities: assay standardization, harmonized reporting, external quality assessment, and prospective clinical validation.

Applied Sciences doi: 10.3390/app16167937

Authors: Sergey Pogorelskiy Angelina Rybakova Imre Kocsis

Modular construction requires a comprehensive assessment of the quality of design solutions, taking into account interconnected architectural, production, logistical, and digital factors at all stages of the lifecycle. Existing approaches, including Design Quality Indicator (DQI), Life Cycle Assessment (LCA), and Design for Manufacturing and Assembly (DFMA), focus on individual aspects and do not provide an integrated assessment, necessitating the development of a comprehensive methodology. The aim of this study is to develop a BIM-oriented design quality assessment system for modular facilities (DQIMC), providing a quantitative and partially automated assessment based on information model data. The methodology is based on the formation of a hierarchical system of criteria and indicators, their classification by automation level (machine-readable, partially automated, expert), and integration with BIM model parameters in industry foundation classes (IFC). Criteria weighting factors are determined using the analytic hierarchy process (AHP). The integrated quality indicator is calculated based on a weighted sum of standardized values. The test was conducted on a sample of five BIM models. The results showed that DQIMC values ranged from 0.73 to 0.91, with the greatest impact coming from the information quality of the BIM model and manufacturing and design characteristics, as confirmed by correlation analysis. The proposed system provides a reproducible quality assessment and creates a foundation for digitalizing the design of modular facilities, improving the validity of decisions and the effectiveness of lifecycle management.

Applied Sciences doi: 10.3390/app16167936

Authors: Robin Lee Yoon Sang Kim

Mitigating in-vehicle VR sickness is a critical challenge for the use of immersive content in autonomous vehicles. In this study, to solve in-vehicle VR sickness, the cause of motion sickness in each field is identified by conducting a scoping review in three fields (medicine, control engineering, and computer engineering). As a result of the confirmation, motion sickness was defined in all three academic fields based on the theory of sensory conflict and posture instability, so we defined and identified seven major SFs based on these universal theories. Based on the defined SF, the analysis of the current research trend confirmed that there is an important research gap that shows that motion-induced motion sickness (MIMS) research focuses on the dynamic control of the vehicle while at the same time overlooks cognitive-sensory factors such as Mismatched Environmental Perception and feedback prediction error. At the same time, visually induced motion sickness (VIMS) research is focused on rendering control elements, which often reduces user immersion. This review suggests the need to focus not only on approaching the form of motion sickness in the vehicle, but also on the possible sensory-cognitive conflict of the passenger in the vehicle. By connecting the three academic fields related to motion sickness, it provides a strategic roadmap for the final in-vehicle content control system. Ultimately, this framework is expected to serve as a fundamental technology resource for in-vehicle content viewing, laying the groundwork for in-vehicle VR content consumption.

Applied Sciences doi: 10.3390/app16167935

Authors: Sebastian Popescu Dan-Gheorghe Dimitriu

One of the consequences of driving a plasma discharge away from the thermodynamic equilibrium is the appearance of one or more fireballs in front of the anode. The formation of this self-organized space charge structure marks the transformation of part of the thermal energy of the electrical charges into the electric potential energy of the structure. The self-assembly of the fireball as well as its existence in a stable stationary state prove to be good candidates for testing the validity/selection of the proposed extremal (maximum or minimum) entropy production principles in the case of systems driven far from the thermodynamical equilibrium. The use of experimental results and of a basic mathematical model shows that the emergence of a fireball in a plasma diode maximizes the entropy production rate, while its subsequent existence minimizes the entropy production rate.

Applied Sciences doi: 10.3390/app16167934

Authors: Aleksandras Chlebnikovas Dainius Paliulis Mantas Pranskevičius

Soil organic carbon (SOC) is a key indicator of soil quality, ecosystem functioning, and climate regulation. However, the combined influence of land use and geochemical landscape position on the vertical distribution of SOC remains insufficiently understood, particularly in protected landscapes. This study evaluated SOC concentration in agricultural land, grasslands, and forests located within autonomous and superaqual geochemical landscapes of the Neris Regional Park (Lithuania). Composite soil samples were collected from three depth intervals (0–10, 10–20, and 20–30 cm) following the European Soil Sampling Protocol, and SOC concentration was determined using a Shimadzu SSM-5000A analyzer. SOC concentration generally decreased with increasing soil depth regardless of land use. Agricultural soils generally exhibited lower SOC concentrations (0.84–4.41%) than forest and grassland soils, although a comparatively high value of 4.41% was recorded at one agricultural plot. In forest ecosystems, SOC concentration ranged from 1.12–1.47% in autonomous landscapes and from 4.17–4.98% in superaqual landscapes, indicating approximately a threefold difference between geochemical positions. Grassland soils contained 1.33–2.54% SOC in autonomous landscapes and 2.16–3.28% in superaqual landscapes. These findings characterize observed patterns of SOC concentration variability associated with land use, geochemical landscape position, and soil depth within the investigated protected landscape. The reported findings contribute to understanding the observed vertical and spatial distribution of SOC concentration within protected landscapes and provide a basis for future investigations incorporating complementary soil physicochemical measurements and broader spatial sampling designs.

Applied Sciences doi: 10.3390/app16167933

Authors: Mingsheng Liu Qi Xia Yaopu Xu Chengming Zhao Zhenhu Lyu Haizhu Wang Guoxin Zhang Bin Wang Zongjie Mu

Reservoir stimulation through fracturing is essential for the commercial development of unconventional oil and gas resources. Supercritical CO2 (scCO2) combines liquid-like density with gas-like viscosity and compressibility, enabling efficient conversion of stored energy into shock waves and jet impacts. This study introduces an innovative scCO2 shock fracturing technique, in which a downhole pressure-control valve rapidly releases compressed scCO2 to generate transient shock pressures that induce rock fracture initiation and propagation. A series of scCO2 shock fracturing experiments were conducted on sandstone and shale to evaluate the influence of different impact loads on both macroscopic and microscopic damage. Rock damage evolution was characterized using computed tomography (CT), nuclear magnetic resonance (NMR), mercury intrusion porosimetry (MIP), and quantitative analysis of fracture surface morphology. The results showed that increasing shock pressure enhanced fracture surface roughness, shear slip, and particle spalling in sandstone, producing rough tensile–shear fracture surfaces with a potential self-supporting tendency. NMR results indicated that sandstone mainly exhibited a single-peak T2 response, and scCO2 shock loading primarily affected pores and pore-fracture spaces larger than 0.08 µm. In contrast, shale showed a broader and more heterogeneous pore-fracture response, with preferential enlargement and connection of large pore-fracture spaces. The NMR-MIP-calibrated equivalent pore-fracture diameter distribution showed that scCO2 shock fracturing mainly promoted pore-fracture spaces larger than 0.2 μm in shale; at 40 MPa, the volume of this pore-fracture range increased by approximately 6.75 times. However, the characteristic equivalent pore-fracture diameter decreased at 45 MPa, which is attributed to severe specimen fragmentation, fragment displacement, scCO2 escape, and energy dissipation. These findings suggest that scCO2 shock fracturing is a promising stimulation approach for enhancing macroscopic fracturing and microscopic pore-fracture reconstruction in unconventional reservoirs.

Applied Sciences doi: 10.3390/app16167931

Authors: Md Rejaul Karim Md Nasim Reza Arnab Majumder Dae-Hyun Lee Sun-Ok Chung

LiDAR (Light Detection and Ranging) and RGB-D camera imaging have emerged as essential tools in agricultural applications, particularly for plant size and distance measurements, enabling non-destructive, cost-effective, and precise estimation. The objective of this study was to measure the plant canopy dimensions and distance between apples using commercial LiDAR, and an RGB-D camera with a speed sprayer platform was used to determine whether LiDAR provides a higher measurement accuracy under field conditions. Data were collected in an apple orchard in Muju, Republic of Korea. Commercial 3D LiDAR, a terminal box, an RGB-D camera, a microcontroller, a power supply, and individual display monitors were integrated into a customized data acquisition (DAQ) box for LiDAR point cloud (PCD), RGB, and depth imagery data collection. Commercial software was used for data acquisition, data conversion (pcap to PCD), segmentation of regions of interest (ROI), and pre-processing of data. PCD processing and measurement consisted of data frame selection, data conversion, outlier removal, downsampling, denoising, ground point removal by filtering, voxelization, and density map generation using an open access programming language script. Depth image processing included importing raw data, shaping metadata using intrinsic camera parameters, visualizing depth images, extracting depth points, and measuring the plant canopy at the pixel level. RGB image analysis involved grayscale conversion, thresholding, segmentation of ROI, contour preparation, noise removal, and binary masking for eliminating the background. Estimated results were compared to measured results. LiDAR measurements showed the closest agreement with the measured results for plant height, canopy volume, plant spacing, and row distance, outperforming both RGB and depth imaging. Under field conditions, plant spacing and row distance were estimated with accuracies of 97.5% and 94.7%, respectively, exhibiting higher measurement accuracies than RGB and depth imagery data results. Despite some discrepancies due to complex plant geometry and dynamic data collection, the results support data collection strategies critical for precision horticulture.

Applied Sciences doi: 10.3390/app16167930

Authors: Jiaxuan Lyu Jiang Guo Fang Yuan Yingbing Ran Haipeng Gong Tao Wu Tong Zhang

Hydropower-generator rotors contain numerous closely spaced bolted joints, making full-coverage contact instrumentation impractical, while single-view vision methods are vulnerable to missing or ambiguous evidence during rotation. This study proposes a detector-guided multi-view visual monitoring framework that separates region-of-interest (ROI) localization from explicit geometric interpretation. On a simulated hydropower-generator rotor platform operating at 15 rpm, YOLO-family detectors localize candidate bolttop, boltside, and starmarker regions. Quality-retained top-view ROIs yield the image-space angular indicator θimg from the relative orientation of nut-side and disk-side anti-loosening lines; side-view ROIs yield the pixel-space thread-exposure indicator Lpx from exposed-thread endpoints and, when marker geometry is sufficiently visible, an auxiliary angular cue. A star-shaped reference marker organizes accepted frame-level observations into approximate rotation intervals, while hierarchical checks of ROI completeness, endpoint availability, image quality, geometric plausibility, and temporal membership retain both usable evidence and explicit rejection reasons. YOLO11n achieved precision 0.9987, recall 1.0000, mAP50 0.9950, and mAP50–95 0.7798 for laboratory ROI localization. After geometric screening, evidence availability was 16.7% for the top-view branch and 76.3% for the side-view branch. In a supplementary 169-image operational field subset, the principal ROI model achieved precision 0.9782, recall 0.9942, mAP50 0.9946, and mAP50–95 0.7948. The field results support appearance-level localization under complex rotor-bolt conditions, and the framework provides a traceable, reliability-aware basis for organizing, screening, and interpreting multi-view evidence in hydropower-generator rotors and similar rotating structures.

Applied Sciences doi: 10.3390/app16167932

Authors: Mladen Lujić Ekrem Bektašević Luka Crnogorac Kemal Gutić

This study presents an integrated multidisciplinary framework for evaluating the reuse of abandoned underground mining infrastructure in Bosnia and Herzegovina as sustainable underground energy storage systems that support the energy transition and decarbonization. The research focuses on the Central Bosnia and Tuzla coal basins, using case studies from the Zenica and Tuzla mining regions to assess Underground Pumped Hydroelectric Energy Storage (UPHES), Compressed Air Energy Storage (CAES), and gravity-based energy storage technologies. The methodology integrates geological and geotechnical characterization, thermo-hydro-mechanical (THM) analysis, thermodynamic calculations, and Multi-Criteria Decision Analysis (MCDA) to evaluate technical, operational, and safety performance. Methane mitigation, smart ventilation, thermal stability, and geomechanical behavior under cyclic loading were also considered. The results indicate that sedimentary coal basins are well suited for UPHES and gravity-based storage systems, with UPHES capacities reaching 1.75 GWh per cycle under optimized conditions, while the separately evaluated solid-mass gravity storage system provides a capacity of 6.15 MWh. Evaporite formations in the Tuzla Basin offer favorable conditions for CAES because of the low permeability and plasticity of halite, enabling storage capacities exceeding several GWh. THM analysis confirmed acceptable geomechanical stability during cyclic operation, while the economic assessment based on the Levelized Cost of Storage (LCOS) demonstrated the long-term competitiveness of Abandoned Mine Energy Storage (AMES) compared with battery technologies. Overall, the findings highlight abandoned mines as strategic low-carbon assets for renewable energy integration and regional post-mining transition.

Applied Sciences doi: 10.3390/app16167929

Authors: Antonio Di Marino Vincenzo Bevilacqua Angelo Ciaramella Ivanoe De Falco Giovanna Sannino

Deep learning classifiers have achieved high accuracy on plant disease recognition tasks, but their decision-making processes remain opaque. Counterfactual explanations (CFs), minimally modified inputs flipping a classifier’s prediction, reveal which input changes are sufficient to alter the decision. While diffusion-based counterfactual generation has been studied primarily on controlled face datasets (CelebA), its systematic evaluation for fine-grained plant disease classification, particularly under in-the-wild conditions, remains limited. In this work, we apply for the first time DiME (Diffusion Models for CFs) to plant disease classification, evaluating its behavior on both controlled (PlantVillage) and in-the-wild (PlantWild) data. We adapt the pipeline using Stable Diffusion with LoRA fine-tuning as the generative backbone, and we propose Plant Verification Accuracy (PVA), a domain-adapted variant of the Face Verification Accuracy (FVA) metric, to measure species identity preservation in the generated counterfactuals. We compare DiME against three established baselines: Wachter (pixel-space gradient), xGEM+, and DiVE (both based on Variational Autoencoders). On PlantVillage, DiME achieves a 7.1 percentage-point PVA drop versus 60–65 pp for Variational Autoencoders baselines while preserving the target flip rate. Extension to PlantWild yields larger PVA drops (16.6–18.6 pp) and qualitative degradation on in-the-wild imagery, identifying current limitations of the approach when applied outside controlled settings.

Applied Sciences doi: 10.3390/app16167928

Authors: Teresa Inchingolo Elena Sibilano Antonio Brunetti Giuseppe Lucarelli Michele Battaglia Vitoantonio Bevilacqua

Robot-assisted surgery has significantly expanded the possibilities of minimally invasive procedures by providing enhanced dexterity and visualization. However, the lack of direct haptic feedback still limits the surgeon’s ability to accurately assess instrument–tissue interactions, motivating the need for automatic intraoperative assistance systems. During vesicourethral anastomosis (VUA) in robot-assisted radical prostatectomy (RARP), accurate engagement of the bladder and urethral mucosa is essential to ensure proper tissue approximation and watertight closure. Nevertheless, automatic identification of fine-grained needle–tissue interactions during this phase remains largely unexplored. In this work, we propose a proof-of-concept framework for vision-based needle–tissue interaction analysis in RARP endoscopic videos, combining semantic segmentation, geometric proximity analysis, and motion coherence estimation to identify biomechanically plausible interaction events. Two independent transformer-based models were fine-tuned for semantic segmentation of the mucosal tissue and the surgical needle using a patient-level split of six real-world RARP procedures, comprising four procedures for training, one for validation, and one for independent testing. The models achieved Dice scores of 0.837 and 0.774, respectively. The segmentation outputs were subsequently used to drive a motion-aware interaction analysis pipeline, combining geometric proximity estimation between the needle endpoint and the mucosal tissue with optical-flow motion coherence analysis. The proposed interaction framework was evaluated on an independent test set, achieving a specificity of 0.933 and a recall of 0.667. An ablation study further demonstrated the complementary contribution of geometric proximity and motion coherence cues for needle–tissue interaction detection. Although limited by the retrospective nature and size of the dataset, this study introduces a low-latency, end-to-end framework for interaction-aware surgical scene understanding during RARP. The proposed approach represents an initial step toward the development of context-aware intraoperative guidance systems for robotic urologic surgery.

Applied Sciences doi: 10.3390/app16167927

Authors: Evangelia D. Karvela Athena Stergiou Velisaria-Eleni Gerogianni Evgenia N. Nikolaou Eirini K. Nikolidaki Eftychios Apostolidis Vaios T. Karathanos Antonia Chiou

Olive oil production generates large amounts of agricultural and agro-industrial residues, with olive leaves representing one of the most abundant by-products. This study investigated the valorization of olive leaves from two Greek cultivars, Koroneiki and Megaritiki, collected from the Aeghion, Heraklion, and Megara regions, by means of an optimized green extraction process for the effective recovery of phenolic compounds. Response surface methodology (RSM) was applied to optimize solvent composition and extraction time using ethanol/water mixtures (0/100, 30/70, 50/50, and 70/30, v/v) and extraction times ranging from 15 to 90 min. Total phenolic content (TPC), antioxidant activity (DPPH and FRAP), and color parameters were determined spectrophotometrically, while individual phenolic compounds were analyzed by HPLC-DAD. Both solvent composition and extraction time significantly affected phenolic recovery and antioxidant activity. The optimal extraction conditions were 37.5/62.5 (v/v) ethanol/water and 52.5 min. Extracts from Megara showed the highest TPC (32 ± 1 g GAE kg−1 LL), antioxidant capacity (2556 ± 162 mmol Trolox kg−1 LL), and phenolic compound levels. Significant differences were observed between the two cultivars, whereas Koroneiki samples from Heraklion and Aeghion exhibited similar phenolic profiles, indicating that cultivars had a greater influence than geographical origin.

Applied Sciences doi: 10.3390/app16167926

Authors: Constantin Bîrțan Monica Roman Dan Selișteanu Dragoș Laurențiu Popa

As proved by the studies carried out but also from practical situations, the acoustic behaviour of an enclosed performance hall is strongly dependent on occupancy level. This happens because human bodies and upholstered seats contribute significantly to sound absorption and in addition modify the modal structure of the acoustic field. This work presents a complex finite element method (FEM) simulation, conducted in ANSYS Workbench 2023, which compares the acoustic performance of a concert hall under three discrete occupancy scenarios: 57, 43, and 28 attendees. Seven humanoid solid models were used to represent audience members and extract the air volume accurately from the hall geometry. Key acoustic metrics were evaluated, including acoustic pressure distribution, Sound Pressure Level (SPL), A-weighted SPL (dBA), far-field SPL, frequency band SPL, and frequency response phase angle, at representative frequencies of 400 Hz, 10,800 Hz, and 20,000 Hz. The obtained results demonstrate that occupancy reduction from 57 to 28 persons increases the effective modal density, elevates low-frequency standing wave contributions, reduces broadband absorption, and shifts the acoustic pressure distribution toward higher spatial non-uniformity. The results obtained provide quantitative assistance for the design of adaptive acoustic scenarios in spaces with variable occupancy.

Applied Sciences doi: 10.3390/app16167925

Authors: Antonia Terpou Christos Bontsidis Ioanna Mantzourani Stavros Plessas

Rising demand for functional dairy products has stimulated interest in incorporating traditional fermented ingredients into modern food systems. This study investigated the use of traditional Greek trahanas as a functional ingredient in yogurt and evaluated its effects on physicochemical, microbiological, antioxidant-related, and sensory characteristics. Five formulations were produced: a control yogurt inoculated with a commercial yogurt starter culture, two starter-inoculated yogurts supplemented with 2% and 4% (w/w) trahanas, and two starter-free fermented dairy formulations supplemented with 2% and 4% (w/w) trahanas. All products were evaluated over 30 days of refrigerated storage (4 °C) in terms of pH evolution, lactose consumption, lactic acid production, protein content, total solids, syneresis, antioxidant activity, total phenolic content, microbial stability, and sensory properties. The trahanas-enriched yogurts inoculated with a starter culture exhibited enhanced acidification, higher protein and total solid contents, reduced syneresis, and increased antioxidant activity and phenolic content compared with the control and starter-free formulations. In particular, the 4% trahanas formulation inoculated with a commercial starter culture exhibited the highest DPPH and ABTS radical scavenging activities, suggesting a synergistic effect between trahanas-derived bioactive compounds and fermentation-associated lactic acid bacteria. Microbiological analyses confirmed the maintenance of high lactic acid bacterial populations throughout storage, particularly in the 4% trahanas formulation, which exhibited the highest Streptococcus thermophilus counts during refrigerated storage. The detected lactic acid bacterial populations originated both from the commercial yogurt starter culture and from the traditional preparation of trahanas itself, which involves fermentation with sour sheep’s milk and traditional yogurt. In addition, coliforms, Enterobacteriaceae, and Staphylococci remained undetectable in all formulations, while only limited yeast and mold growth was observed during the later stages of storage, mainly in starter-free samples. Sensory evaluation indicated good overall acceptability of trahanas-enriched formulations, particularly at 4% (w/w). Overall, the incorporation of traditional Greek trahanas into yogurt systems represents a promising approach for the development of fermented dairy products with improved technological, antioxidant-related, and functional properties.

Applied Sciences doi: 10.3390/app16167924

Authors: Jian Wang Xingyang Li Wei Shen

Pointer gauges remain widely deployed in industrial environments because they are inexpensive, resistant to electromagnetic interference, and readable from a distance. However, automatic reading remains difficult in practice because reliable prediction requires jointly interpreting pointer geometry, scale layout, and unit-type consistency under challenging conditions such as glare, scratches, blur, and oblique viewpoints. Although multimodal large language models (MLLMs) offer a promising unified interface for visual understanding and structured output, their direct application to gauge reading is limited by weak geometric grounding, unit confusion, and unstable numeric generation. Rather than claiming a new model architecture or a new reading algorithm, this work frames the contribution as a practical adaptation and evaluation framework for applying existing MLLM and PEFT components to structured industrial gauge reading. Our framework combines three components: (i) a dedicated dataset and VQA-style annotation protocol covering multiple noise types and intensity levels; (ii) a unified screening pipeline for selecting a suitable MLLM backbone under zero-shot settings; and (iii) parameter-efficient adaptation of the selected model with Projector-LoRA, together with training and decoding mechanisms designed to improve reading robustness and output consistency. On our test set, the fine-tuned Granite-Vision 3.2 model achieves 99.9% type accuracy, 43.77% reading accuracy, and 43.60% joint accuracy. It obtains an MAE of 3.98 over valid numerical predictions, a parsing coverage of 98.65%, and an all-sample penalized normalized MAE of 0.052. These results substantially outperform the evaluated zero-shot MLLM baselines in structured prediction accuracy, although the lightweight CNN baseline remains slightly better in all-sample normalized numerical error.These results should be interpreted as evidence of promise and measurable improvement over untuned MLLMs, not as evidence that the system is already sufficient for safety-critical or fully autonomous industrial deployment. More broadly, the proposed framework offers a practical, traceable path for adapting large multimodal models to visual measurement tasks that require structured numerical outputs.

Applied Sciences doi: 10.3390/app16167923

Authors: Mehdi Rahbar Masoud Ziabasharhagh Rambod Rayegan

This study numerically investigates the capability of cylindrical phase change material (PCM) encapsulations to attenuate inlet-temperature fluctuations in water as the heat transfer fluid (HTF). A sinusoidal inlet profile with a 20 K amplitude is imposed, and melting and solidification are modeled using the enthalpy–porosity method. The analysis begins with a single encapsulation, which reduces the outlet temperature amplitude by 45.06%, and extends to three, nine, and 15 cylinders in aligned and staggered arrangements. Increasing the cylinder count enhances fluctuation reduction but with diminishing returns, as the HTF thermal energy reaching downstream cylinders decreases. Spatial arrangement is equally important: a staggered array of nine encapsulations achieves a 65.91% reduction, surpassing a 15-cylinder aligned configuration (65.80%), while the highest reduction, 73.01%, is obtained with a 15-cylinder staggered configuration. Across all configurations, the outlet fluctuation reduction follows a single near-linear relationship with the total melted PCM mass (R2 = 0.96), across cylinder count and arrangement, identifying melted mass as a controlling parameter for fluctuation mitigation rather than the cylinder-averaged liquid fraction. These results indicate that the total melted PCM mass is a practical criterion for comparing PCM encapsulation configurations, while the arrangement remains a distinct factor.

Applied Sciences doi: 10.3390/app16167922

Authors: Shinhao Yang Po-Chen Hung Hsiao-Chien Huang Ying-Fang Hsu

Airborne viral transmission necessitates effective indoor air purification strategies. Conventional methods often face operational challenges, including potential secondary aerosolization and performance degradation under high-humidity conditions. This study evaluates a hybrid control system integrating a multi-walled carbon nanotube (MWCNT) field-emission plasma with a UV-LED/TiO2 photocatalyst to continuously inactivate airborne bacteriophages. The system’s performance was assessed under varying applied voltages and relative humidity (RH) levels. The kinetic results demonstrated that the hybrid configuration yields a synergistic inactivation effect compared to the isolated plasma or photocatalytic treatments. Based on the kinetic enhancement, it is hypothesized that trace ozone generated by the plasma discharge serves as an electron acceptor on the UV-illuminated TiO2 surface, thereby mitigating electron–hole recombination and enhancing the generation of hydroxyl radicals (·OH). Furthermore, the hybrid system exhibited operational resilience under high-moisture conditions, maintaining a robust active inactivation constant (ka = 0.190 min−1) at 70% RH without statistical degradation. This stability indicates that the continuous field emission effectively utilizes ambient moisture for secondary radical generation rather than being quenched by water condensation. Ultimately, this hybrid technology presents a continuous and adaptable engineering control measure for mitigating airborne pathogens in enclosed occupational environments, including those in high-humidity climates.

Applied Sciences doi: 10.3390/app16167921

Authors: Shanfeng Ge Nijia Qian Jingxiang Gao Xin Liu Wenyuan Zhang Yong Feng Dehu Yang

Open-pit mines contain rapidly changing terrain, discontinuous bench structures, and mixed artificial–natural objects, which complicate automated three-dimensional mapping. This study presents a dual-module workflow for UAV LiDAR point clouds. Module A characterizes local geometry using normal and curvature descriptors, constructs local plane support through RANSAC fitting, and detects candidate bench-line points using an angular-gap criterion, followed by regional grouping and Kalman-filter refinement. Qualitative overlay with the orthophoto showed coherent correspondence with principal platform–slope transitions. Module B segments buildings, roads, and vegetation using a PointNet++ network enhanced by local Transformer self-attention and inverted residual feature transformation. Under a fixed spatial hold-out setting, the network achieved an overall accuracy of 97.6% and a mean intersection over union of 96.4%. It obtained the highest overall accuracy, mean intersection over union, and class-wise intersection over union among the selected baselines, whereas Point Transformer achieved a slightly higher mean class accuracy. The two independently operated modules provide complementary structural and semantic information for open-pit mine mapping. Broader applicability requires reference-based bench-line assessment and evaluation across additional mines and survey periods.

Applied Sciences doi: 10.3390/app16167920

Authors: Monika Sujka Daniil Samchenko Sofiya Sauchuk

The aim of this study was to develop functional crackers enriched with young barley powder and to evaluate the young barley powder’s impact on the product quality. The experiment was designed as a single-factor study, in which the effect of young barley powder addition (0, 2, 4, 6, 8, 10, and 15% wheat flour substitution) on the physicochemical, nutritional, and sensory properties of crackers was evaluated. The results showed that young barley powder incorporation affected color and selected texture parameters of both the dough and the baked crackers. The enriched crackers exhibited higher antioxidant capacity and improved nutritional value compared with the control sample, including increased total polyphenol (TP) content and higher levels of chlorophylls and carotenoids. The product containing 15% young barley powder had over three times higher TP content (2.074 mg GAE/g dm vs. 0.464 mg GAE/g dm in the control), approximately eightfold and sixfold higher levels of chlorophyll a and chlorophyll b, respectively, and over 16-fold higher carotenoid content than the control. These crackers also contained over threefold more Ca and K, nearly twice as much Mg, and fourfold more Fe. Dialyzable iron constituted 8.4–12.3% of total iron present in the crackers. Consumer evaluation showed that crackers containing 2% and 4% young barley powder were rated the highest in terms of taste and had optimal hardness and crispness. These findings suggest that young barley powder can be successfully incorporated into cracker formulations not only to modify their technological properties but, more importantly, to enhance their nutritional quality, particularly their mineral content.

Applied Sciences doi: 10.3390/app16167919

Authors: Mesut Uğurlu İbrahim Alper Doğru

Industrial Internet of Things (IIoT) environments are increasingly exposed to sophisticated cyberattacks, creating a growing need for intrusion detection systems (IDSs) that balance high accuracy with computational efficiency. Although existing studies primarily focus on classification performance, they often overlook deployment costs and model interpretability. To address these challenges, this study presents BOSE (Bayesian-Optimized Stacked Ensemble), an intrusion detection system that combines hybrid feature selection, Bayesian hyperparameter optimization, Out-of-Fold (OOF) stacking with leakage-free meta-feature generation, and SHAP-based explainability. Hybrid feature selection reduces the original feature space from 84 to 46 informative features, while Bayesian Optimization is used to determine the hyperparameter configuration of the ensemble models and leakage-free OOF stacking enables reliable meta-learning. Experimental results on the 50-class DataSense benchmark show that BOSE achieves a Macro-F1 score of 88.43% over ten independent runs, outperforming all directly comparable baseline models evaluated under the same end-to-end 50-class classification setting while remaining competitive with recent hierarchical multi-stage frameworks. In addition, the proposed model achieves an inference latency of 0.1421 ms, a throughput of 7035 packets/s, a runtime memory footprint of 19.11 MB, and a model size of 309.12 MB. These results demonstrate the CPU-based inference feasibility of the suggested framework under the evaluated workstation configuration and indicate its potential applicability to industrial edge gateways and industrial PCs. The proposed dual-level SHAP solution improves model transparency by explaining both feature-level contributions and ensemble-level decisions, making BOSE an accurate, computationally efficient, and interpretable solution for fine-grained IIoT intrusion detection.

Applied Sciences doi: 10.3390/app16167917

Authors: Višnja Križanović Krešimir Grgić Ana Pejković Drago Žagar

This study examines energy savings and efficient use through the application of smart adaptive lighting within the framework of “smart energy”, “smart city”, and “smart village”. Adaptive street lighting systems have emerged as an effective solution for reducing energy consumption while maintaining traffic safety. However, existing studies typically evaluate energy performance under predefined traffic conditions or focus primarily on AI-based control strategies without systematically investigating the influence of object speed on energy savings. This study proposes a comprehensive methodology that combines CupCarbon traffic simulation, Shape-Preserving Cubic Hermite Interpolation (PCHIP), Monte Carlo simulation, sensitivity analysis, and machine learning to evaluate and predict the energy-saving performance of adaptive street lighting. Unlike previous approaches, the proposed framework establishes a continuous relationship between object speed and energy consumption, enabling the estimation of energy savings across the entire operating speed range while quantifying the effects of object speed, pole spacing, and pre-activation time. The machine learning models are employed to predict energy-saving results generated by the simulation framework. Six regression models were trained and validated using simulated datasets, with Gradient Boosting achieving the highest predictive accuracy. Moreover, the analysis demonstrated that adaptive street lighting scenarios operating at 50% power (50 W) under no-object conditions and 100% power (100 W) during object detection achieved energy savings of 19–37% per luminaire compared with conventional street lighting, depending on object speed. Furthermore, adaptive lighting operating at a constant 50% power (50 W) during object detection yielded substantially higher energy savings of 59–77% per luminaire, highlighting the significant influence of the lighting control strategy on overall energy efficiency. The results demonstrate that lower object speeds yield the greatest savings. The proposed methodology provides a robust and scalable framework for the design, optimization, and intelligent control of adaptive street lighting systems and offers a benchmark for evaluating the maximum theoretical energy-saving potential under controlled traffic conditions.

Applied Sciences doi: 10.3390/app16167918

Authors: Weiwei Li Xin Yang Kang Niu Shengze Lu Jiying Liu Yuxuan Zhao

Heating load is jointly affected by meteorological conditions, system operating states, and historical evolution, exhibiting nonlinear, time-varying, and locally fluctuating characteristics. To improve short-term forecasting accuracy, this study proposes a particle swarm optimization (PSO)-based Trend–convolutional neural network (CNN)–bidirectional long short-term memory (BiLSTM)–Attention model. The model constructs trend-enhanced features from meteorological variables, operating parameters, temporal periodicity, historical lags, rolling statistics, differenced features, and exponentially weighted moving averages. CNN is used to extract local temporal features, BiLSTM captures bidirectional temporal dependencies, and Attention identifies key time steps. PSO further optimizes key hyperparameters. Two datasets are constructed from hourly heating-season operating data, and the proposed model is compared with BiLSTM, CNN-BiLSTM, CNN-BiLSTM-Attention, and Trend-CNN-BiLSTM-Attention models. The proposed model achieves the best performance on both datasets. For Dataset 1, the mean absolute error (MAE), root mean square error (RMSE), mean absolute percentage error (MAPE), and coefficient of determination (R2) are 0.628 GJ, 0.928 GJ, 9.308%, and 0.851, respectively; for Dataset 2, they are 0.3987 GJ, 0.5954 GJ, 10.32%, and 0.8709. These results indicate that trend-enhanced features and PSO improve forecasting performance and can support heating system operation scheduling.

Applied Sciences doi: 10.3390/app16167916

Authors: Anita Konieczna Roman Padyuka Anatoliy Tryhuba Pavlo Lub Vadym Ptashnyk Kinga Borek Anna Rygało-Galewska Barbara Dybek Dorota Anders Kamila Klimek Adam Koniuszy Grzegorz Wałowski

Improving the performance of computer networks using seamless Wi-Fi can be achieved by implementing a number of strategies and technologies. Strategies include, first of all, the optimal location of routers and access points, the use of a multi-band network or routers supporting different bands. Routers with support for beamforming technology, which directs the Wi-Fi signal directly to connected devices, allow you to improve the signal quality and data transfer speed. Increasing the performance of Wi-Fi computer networks is also provided by the use of network monitoring and management software, which allows you to monitor its performance and respond to possible problems in the network infrastructure. This is an important task, because it determines the quality and convenience of access to network resources. First of all, it allows you to achieve a high data transfer rate, which is especially important in conditions of high traffic necessary for demanding applications. Seamless Wi-Fi technologies also promote increased mobility and flexibility of users, allowing them to connect to the network in any place with a good signal without having to use wired connections. Network management becomes more efficient with automatic switching between access points and increased fault tolerance in the face of changing traffic usage scales. Quantitative results: Implementation of the Wi-Fi roaming mechanism using the IEEE 802.11 specification; Wi-Fi performance measurements obtained for various IEEE 802.11n HT20 and IEEE 802.11a client ratios; the original test environment included 50 laptops and netbooks from various manufacturers, equipped with various operating systems and wireless network adapters; seamless Wi-Fi technologies based on IEEE 802.11k, IEEE 802.11v, and IEEE 802.11r improve communication continuity during device mobility and support real-time AI-based decision making; Wi-Fi based on local communication standards (WLAN-Wireless Local Area Network). It allows data transmission speeds from 1 Mb∙s−1 to 6.75 Gb∙s−1. Indoors, the Wi-Fi range is 20 m, and outdoors 100 m; WiMax (Worldwide Interoperability for Microwave Access) is a built-in set of wireless broadband standards that provide a constant data rate of 1 Gb∙s−1 and 100 Mb∙s−1 in a cellular network; LR-WPANs (Low-Rate Wireless Personal Area Networks) are standards that are the basis for higher communication protocols, ZigBee. They offer data rates ranging from 40 kb to 250 kb∙s−1. In devices with limited resources, these standards operate at 2.4 GHz at higher transmission speeds and 868/915 MHz at lower. The novelty in the article is the implementation of the Wi-Fi roaming mechanism, presentation of Wi-Fi scenarios, discussion of module generations, indication of integrated agriculture in terms of modern digitalization technologies, and characteristics of smart farming.

Applied Sciences doi: 10.3390/app16167912

Authors: Chang Peng Bao Yang Meiqi Li Ge Zhang Hui Sun Zhenyu Jiang

Ground-penetrating radar (GPR) is a rapid and non-destructive technique for road sub-surface distress (RSD) detection. However, reliable interpretation of GPR images remains challenging because subsurface anomalies often present weak boundaries, ambiguous textures, and high similarity to non-distress targets. This study proposes a cross-verification intelligent algorithm that exploits complementary information from different views of 3D GPR data. Three YOLO-based detectors are trained on view-specific GPR images to identify RSD-related targets, including voids, loose structures, and manholes. By sequentially verifying detection results across different views, the proposed method improves recognition reliability under vague subsurface imaging conditions. The models are trained and evaluated on an expert-annotated field 3D GPR dataset containing 2134 location-level multi-view samples. At the selected operational thresholds, the complete cross-verification procedure achieved 95.9% precision and 98.6% recall for RSD detection in the testing subset. In a field evaluation on 15 roads, all 69 RSD locations in the expert-identified reference set were matched by automatic indications. When integrated into an automatic detection system, the method reduced manual inspection workloads by approximately 90% while maintaining high field reliability. These results demonstrate the potential of multi-view cross-verification for post-survey RSD screening and expert-assisted review.

Applied Sciences doi: 10.3390/app16167913

Authors: Aslıhan Şair Zeynep Bala Duranay

Accurate one-hour-ahead wind speed prediction is critical for the efficient use of wind energy, energy generation planning, and the reliable integration of wind energy systems. However, the time-dependent variability and nonlinear behavior of wind speed make this prediction problem challenging. In addition, abrupt wind-speed peaks and temporal dependence in hourly data require models that can effectively capture short-term wind-speed dynamics without temporal data leakage. In this study, an artificial neural network (ANN)-based model for one-hour-ahead wind speed prediction was developed using hourly meteorological data. The model inputs included temperature, humidity, radiation, pressure, the sine and cosine components of wind direction, lagged wind speed variables, and the most recent wind speed observation. The output variable was defined as the wind speed one hour ahead. The modeling procedure consisted of constructing lag-based input variables, aligning the input–output dataset for the one-hour-ahead prediction task, applying a chronological training–validation–test split, and selecting the final ANN model through validation-based tuning. The model was trained and tested using a total of 8757 hourly data points from 2024 obtained from a meteorological station in Elazig, Türkiye. A chronological training, validation, and test split was used to prevent temporal data leakage, and the proposed model was compared with persistence-based, linear, tree-based, and meteorological-only ANN benchmark models. On the independent test dataset, the proposed tuned lag-based ANN achieved an RMSE of 0.8904 m/s, an MAE of 0.6394 m/s, and an R of 0.8897. The proposed model obtained the lowest RMSE among the evaluated models, although the persistence model achieved slightly better MAE and R values. The results show that using lagged variables allows the model to learn the temporal dependencies of wind speed more effectively and produce competitive RMSE-based prediction performance, while the persistence model remains competitive according to MAE and R. Therefore, the proposed lag-based ANN approach provides a competitive RMSE-based framework for one-hour-ahead station-level wind speed prediction under similar onshore meteorological station conditions.

Applied Sciences doi: 10.3390/app16167915

Authors: Rodrigo Tavares de Pina Simões Xavier Larriva-Novo Carmen Sánchez-Zas Victor A. Villagrá Andrés I. Marín López

Security Operations Centers (SOCs) rely on Level 1 analysts to triage increasing alert volumes amid alert fatigue and tool fragmentation. LLM-based multi-agent systems using the Model Context Protocol (MCP) are being adopted to automate these tasks, but their autonomy and tool access expose them to attacks such as tool poisoning, indirect prompt injection, and confused deputy exploitation. To address this gap, this work proposes a security framework for MCP-based multi-agent SOC pipelines, implemented as a middleware layer comprising a tool registration validator and five execution layers: access control, rate limiting, input validation, output validation, and audit logging. The framework is applied to a triage-enrichment-response pipeline connected to a Wazuh SIEM through a custom MCP server. Of the 35 attack vectors considered in a threat model derived from different threat taxonomies, including OWASP, MITRE ATLAS, and ATFAA, 29 are addressable at the middleware level and are covered by the framework’s controls. These controls are then validated experimentally using a purpose-built malicious MCP server and targeted test-harness injections, organized into six test suites that together exercise the covered vectors across 600 executions. Every attack instance in the evaluated threat model was blocked, none bypassed the framework, and no legitimate call in the evaluated set was incorrectly rejected; obfuscated variants, however, evade the lexical content-inspection controls, delimiting the scope of this result. A full-pipeline demonstration confirms that the framework preserves benign operational outputs. These results indicate that systematic middleware controls can secure MCP-based agentic SOC deployments without modifying the underlying agents or MCP servers.

Applied Sciences doi: 10.3390/app16167910

Authors: Andrey Kryukov Iliya Iliev Aleksandr Kryukov Hristo Beloev Alexey Kolotygin Ivan Beloev Konstantin Suslov

The objective of the research presented in this paper was to develop methods for simulating the operating conditions of traction power supply systems (TPSSs) equipped with asynchronous generators (ASGs), which may be driven by wind or hydraulic turbines as prime movers, thereby significantly reducing train traction energy costs and lowering carbon monoxide emissions. Using phase-coordinate methods and the Fazonord AC-DC industrial software package, simulations were performed for a TPSS configuration comprising three traction substations (TSs) with ASGs connected to the 6 kV busbars. The results demonstrate that connecting the ASG reduces the maximum active power flow from the utility grid by 27%, decreases peak losses in the 220 kV primary supply line by 44%, and lowers voltage unbalance levels at the 220 kV busbars by 71–76%. Additionally, electromagnetic safety conditions along the 220 kV overhead lines feeding the substations are improved, and the temperature at the hottest points of the traction transformers is reduced. The developed ASG models, implemented using three controlled current sources, are universal and can be applied to TPSSs of various configurations and design layouts.

Applied Sciences doi: 10.3390/app16167914

Authors: Tingting Wu Xin Li Hongyan Tian Mingwei Liu

In high-density urban areas, pedestrian route choice behavior is influenced differently by the attributes of the objective environment, depending on the purpose of the trip. To explore this issue, this paper first constructs the traditional Path Size Logit (PSL) model to perform baseline estimation of the effects of objective attributes such as path distance, intersections, number of lanes, greenery, and commercial facilities. It then introduces a hierarchical Bayesian framework to build the Hierarchical Bayesian Path Size Logit (HB-PSL) model, using the No-U-Turn Sampler (NUTS) for posterior sampling to quantify parameter uncertainty and capture inter-group heterogeneity. The model achieves inter-group information sharing through a hierarchical prior structure and uses Automatic Differentiation Variational Inference (ADVI) to provide rapid approximate estimation. An empirical analysis shows that different importance is given to objective environmental attributes for different travel purposes. Furthermore, this paper proposes the “Equivalent Distance (ED)” index, which transforms the preference of different groups for different environmental attributes into an actionable spatial length. For shoppers, the greening level increases by one level for every 100 m, which is equivalent to shortening the path by about 31.31 m; commercial facilities increase by one for every 100 m, which is equivalent to shortening the path by about 76.63 m. The results provide behavioral support for the formulation of differentiated walking strategies in high-density urban areas: priority should be given to strengthening the green coverage and commercial facility layout in commercial blocks to synergistically improve walking efficiency, safety, and comfort.

Applied Sciences doi: 10.3390/app16167911

Authors: Darlan Quinta Brito Daphne Heloisa de Freitas Muniz Flávia Nogueira Sá Carlos Henke-Oliveira Eduardo Cyrino Oliveira-Filho

This study evaluates the short-term effects of nine commercially available fire retardants (FRs) on water quality under controlled laboratory conditions. FRs were diluted to a 1:10 FR:water stock solution to approximate readily water-extractable fractions, and ion solubility was quantified together with key physicochemical parameters, including pH, electrical conductivity (EC), and dissolved oxygen (DO). Results revealed substantial heterogeneity among formulations in the concentrations of solubilized ions. Notably, FR4 (NP+) exhibited exceptionally high concentrations of NH4+, NO3−, NO2−, PO43−, and Br−, as well as elevated Ca2+ and Na+. Other formulations also released considerable loads of common nutrients, particularly NO3− and PO43−, although at lower magnitudes. Despite the pronounced ionic enrichment, pH and DO remained relatively stable across most treatments. However, DO decreased in several cases at higher FR concentrations, suggesting increased oxygen demand associated with dissolved constituents. EC displayed strong, concentration-dependent increases across treatments, reflecting the rapid dissolution of ionic components in water. Cluster analysis identified three distinct groups of FRs based on their water-quality responses. The PCA biplot further revealed two main gradients structuring the formulations: mineralization and compositional diversity (PC1), and physicochemical solution conditions (PC2), which together distinguish the chemical and physicochemical signatures of the tested products. Overall, the results highlight substantial formulation-specific differences in nutrient and metal release, with potential short-term implications for aquatic chemistry and ecosystem functioning. Although derived from controlled laboratory conditions, these findings provide a first-order assessment of FR impacts on water quality and underscore the importance of considering formulation composition in watershed risk assessments and post-fire water quality management strategies.

Applied Sciences doi: 10.3390/app16167909

Authors: Aleksandra Krzemień Jakub Bartyzel Łukasz Chmura Anna Czaplicka

Artificial light at night (ALAN) alters natural nocturnal environments and affects ecosystems, human health, and night sky visibility, while cloud cover strongly modulates night sky brightness (NSB) by enhancing or attenuating skyglow. This study investigated the relationship between cloudiness and NSB at three locations in southern Poland representing contrasting light environments: urban Kraków, high-altitude Kasprowy Wierch, and rural Chochołów. Cloudiness was quantified using a custom binary image-processing method applied to monochrome all-sky camera images acquired under automatic exposure conditions. The procedure included exposure-time normalisation, dual-threshold segmentation with hysteresis, and site-specific histogram-based classification. Cloud-cover time series were synchronised with continuous SQM-LU night sky brightness measurements collected during moonless astronomical nights. An independent comparison was performed using a Hukseflux NR01 net radiometer and a radiative cloudiness proxy derived from surface–sky temperature differences. Increasing cloudiness was associated with brighter night skies at all sites, with the strongest effect observed in Kraków (r=−0.97; slope ≈−2.8 mag/arcsec2), a weaker response at Kasprowy Wierch (r=−0.77; slope ≈−0.4 mag/arcsec2), and a moderate relationship in Chochołów (r=−0.48; slope ≈−0.7 mag/arcsec2). Camera-derived cloudiness showed moderate agreement with the radiometric cloudiness proxy (robust R2=0.56, rs=−0.79). The proposed method proved practically useful across diverse lighting conditions, demonstrating its usefulness for long-term light pollution studies based on monochrome all-sky imagery.

Applied Sciences doi: 10.3390/app16167908

Authors: Zhijun Zhu Xinyi Fang Linjun Lu

Autonomous vehicles (AVs) will operate alongside human-driven vehicles for an extended transition period, during which aggressive human driving may become a major source of risk. This study proposes an integrated safety-control framework that combines real-world-data-driven behavior modeling with deep reinforcement learning to design longitudinal AV control strategies for mixed traffic. Aggressive, general, and defensive driving patterns are calibrated from the CitySim dataset, and dynamic aggressiveness is incorporated into an improved car-following model. A proximal policy optimization algorithm with a Kullback–Leibler penalty is then used to learn multi-objective strategies balancing safety, efficiency, comfort, and fuel economy in freeway and signalized-intersection scenarios. The results show that the behavior-aware strategies exhibit different strengths across traffic environments. On the freeway, the defensive-threshold strategy maintains a larger time headway, reduces positive acceleration, and lowers system-level fuel consumption, whereas the default, aggressive, and general strategies preserve higher traffic efficiency. At the intersection, signal control narrows the differences among strategies and limits the influence of longitudinal threshold settings on most evaluated indicators. These findings provide a quantitative basis for selecting behavior-aware control thresholds and designing robust AV strategies for mixed-autonomy traffic containing aggressive human drivers.

Applied Sciences doi: 10.3390/app16167907

Authors: Haodong Shi Chunjian Hua

Bearing, as a critical component in industrial production, directly dictates equipment performance. Bearing surface defects are typically minute in scale, indistinct at their boundaries, and prone to interference from surface textures and light reflections, thereby posing substantial challenges to detection. This paper proposes a novel deep segmentation network, Bearing-TFUNet, which incorporates the skip connection structure of U-Net to preserve low-level detail information, introduces an enhanced feature pyramid network (UFPN) to bolster multi-scale feature representation capabilities, and integrates a lightweight LE-Transformer for efficient modeling of global contextual information. This triple-fusion mechanism enables the decoder to concurrently integrate skip connection features from the encoder, upsampled features, and multi-scale information furnished by UFPN, facilitating comprehensive feature interaction and fusion. The DWConv-based LE-Transformer attention module effectively augments the model’s capacity for collaborative modeling of local texture details and global dependencies. Experimental results demonstrate that, compared with the baseline U-Net model, Bearing-TFUNet achieves a 12.19% improvement in the Dice coefficient, a 21.94% increase in the IoU metric, and a reduction in Hausdorff Distance from 5.6 to 4.3.

Applied Sciences doi: 10.3390/app16167906

Authors: Zhi-Cheng Wang Tao Zhang Yi-Meng Zhu Qi-Ang Liu

Accurate traffic flow prediction is a core task in intelligent transportation systems because urban traffic observations are spatially distributed, temporally dynamic, and commonly held by different regional management entities. Existing centralized and local models remain limited when traffic data are non-independent and identically distributed across regions, and when raw data cannot be directly exchanged because of privacy, ownership, and communication constraints. To address these challenges, this study proposes a personalized similarity-aware federated spatiotemporal learning framework for multi-regional traffic flow prediction. The framework integrates three mechanisms: client-specific adaptation for regional distributional heterogeneity, adaptive delayed graph learning for dynamic congestion propagation, and similarity-aware federated aggregation for information-quality-based cross-client collaboration. Spatial dependency, temporal evolution, traffic-flow-theory-informed variables, road attributes, and temporal contextual features are jointly modeled without sharing raw client data. Experiments on controlled synthetic data and the Q-Traffic real-world dataset demonstrate that the proposed method consistently outperforms independent training, FedAvg, FedProx, FedSTN-inspired, and FedAGCN-inspired baselines. On the Q-Traffic grid-level setting, the proposed adaptive graph version reduces MSE by 35.3% compared with FedAvg, while the CNN version reduces MSE by 27.5%. Under the cluster-level setting, the adaptive graph version reduces MSE by 26.3% compared with FedAvg. Ablation, sensitivity, communication-cost, differential-privacy, and client-dropout analyses further show that the proposed framework improves predictive accuracy, cross-client stability, and robustness under heterogeneous federated traffic scenarios.

Applied Sciences doi: 10.3390/app16167905

Authors: Viktor V. Baklanov Yerbolat T. Koyanbayev Kuanysh Samarkhanov Yuliya Yu. Baklanova Olga S. Bukina Vadim Bochkov Radmila Sabitova Amina Nokanova

Legacy liquid radioactive waste (LRW) from the BN-350 sodium-cooled fast reactor is chemically heterogeneous and requires updated characterization for further management. This study investigated the physicochemical properties and gamma-emitting radionuclide composition of 13 LRW samples collected from four tanks (B-02/1, B-02/2, B-02/5, and B-02/6) at surface (“mirror”), middle, and lower levels. Five samples were analyzed as LRW and eight as evaporated LRW residues. Physicochemical analysis included pH, density, dry residue, alkalinity, and major inorganic components. Portable gamma spectrometry was applied to all samples, and laboratory HPGe measurements were performed for selected aqueous LRW samples. The saline LRW samples were strongly alkaline and highly mineralized, with dry residue values of 139.30–295.10 g/dm3. Tank B-02/6 contained distinct oil-containing, emulsion, and carbonate-rich alkaline aqueous phases. 137Cs was the dominant identified gamma-emitting radionuclide, with specific activities ranging from (1.4 ± 0.3) × 105 to (1.1 ± 0.2) × 108 Bq/kg in liquid samples and up to (6.5 ± 1.3) × 108 Bq/kg in evaporated residues. The results show that BN-350 LRW is not a homogeneous waste stream; therefore, future monitoring, retrieval, treatment, and conditioning should be planned according to matrix type, sampling depth, vertical phase heterogeneity, and 137Cs-dominated radiological characteristics.

Applied Sciences doi: 10.3390/app16167904

Authors: Tihomir Orehovački

Large language models (LLMs) are expanding the capabilities of socially assistive robots (SARs) through natural dialogue, personalisation, multimodal reasoning, retained interaction context, and adaptive behaviour in healthcare. Integrating generative language models into robots, however, complicates evaluation because fluent output may exaggerate perceived competence and increase the risks of hallucination, overtrust, privacy exposure, relationship dependency, and unsafe reliance on advice or actions. This PRISMA-informed review synthesises healthcare robotics, human–robot interaction, LLM-enabled systems, ethics, implementation, and care delivery. Database searches returned 128 records, of which 110 were unique after deduplication. Supplementary retrieval and assessment yielded 85 substantive sources spanning background mapping, primary analysis, and governance. Studies focused mainly on feasibility, usability, acceptability, dialogue quality, and short-term engagement, whereas longitudinal safety, governance of retained interaction context, comparative effectiveness, workflow integration, and sustained healthcare value received limited attention. These gaps indicate that evaluation of LLM-enabled SARs must account for physical presence, social role, interaction memory, and potential actions rather than focus on conversational performance alone. The review therefore proposes HEART, a healthcare-specific evaluative architecture comprising Human-Centred Communication, Ethical and Trustworthy Deployment, Adaptive and Embodied Intelligence, Relationship Continuity, and Translational Healthcare Value. HEART uses boundary rules, operational indicators, qualitative labels, and non-additive deployment gates to separate evaluative domains, define assessable outcomes, summarise reported support, and prevent strengths in one area from masking critical safety or governance failures. Future research should validate HEART through longitudinal and comparative assessment of hallucination severity, language-to-action safety, long-term effects, equity, and post-deployment monitoring.

Applied Sciences doi: 10.3390/app16167903

Authors: Xiaoning Wang Zhutang Li Changzhen Hu Shengjun Wei

Long-context modeling is important for long-horizon generation tasks, yet larger context windows do not necessarily ensure stable organizational continuity. We propose Dimension-Constrained Organizational Relay (DCOR), a workflow-level framework that reformulates long-horizon generation as the ordered propagation of finite organizational states rather than repeated replay of complete token histories. DCOR introduces Order and Dimension to represent sequential organizational evolution and task-oriented constraints, and comprises Organizational Set Extraction, Dimension-Constrained Generation, Organizational Relay, and Organizational Convergence. We evaluate DCOR on cumulative multi-instance advertisement generation and long-form article generation. Across ten advertisement runs, DCOR generated 259–315 structured creative blocks per run, averaging 279.3 ± 17.9, while maintaining the required format and producing no exact duplicate blocks. At the matched 30-block scale, its main advantage was cumulative structured production rather than the lowest character-level repetition. In long-form generation, DCOR achieved the lowest mean character-level 4-gram and 6-gram repetition compared with Direct Generation, Rolling-Summary Generation, Hierarchical-Outline Generation, and Neural RAG-Memory, while maintaining high Distinct-2 and Distinct-3 values. Ablation results further showed that removing Dimension, Organizational Relay, or Order increased repetition and reduced sustained multi-section expansion. These findings support organizational-state propagation as a complementary mechanism for structured long-horizon generation under restricted local-context conditions.

Applied Sciences doi: 10.3390/app16167902

Authors: Zundong Yang Bengao Yang Jing Xie Gan Feng Fei Li Junjun Liu Yunlong Wang Xiyuan Zhao Longhua Xu Hongfei Duan Mingzhong Gao

Intact deep sandstone cores are scarce, heterogeneous with poorly repeatable, limiting systematic laboratory studies of deep-rock mechanical behavior. This study selected dense sandstone recovered from 1050 m in the Pingdingshan mining area as the prototype system for the development of a targeted-sandstone-constrained screening strategy for simulated deep sandstone. The strategy integrates mineral-composition matching, orthogonal mixture design, mechanical testing, PCA-based comprehensive similarity evaluation, GMM classification, stress–strain curve comparison and fracture-morphology verification. Candidate materials were prepared using a cement–silica-fume matrix with quartz sand, K-feldspar, Na-feldspar, nanoclay and superplasticizer. Results show that the water–binder ratio dominated uniaxial compressive strength, tensile strength and elastic modulus, whereas superplasticizer and nanoclay had secondary effects. The PCA-based index assigned weights of 52.9%, 29.2% and 17.9% to uniaxial compressive strength, elastic modulus and brittleness index, respectively. Among the 25 mixtures sampled, S5 showed the highest mechanical similarity, with a simulation index of 71.65% and a stress–strain curve similarity of 0.958. GMM clustering identified S5 and S10 as the closest high-strength, high-stiffness and high-brittleness group, while S10 better reproduced natural crack geometry. These results indicate that the optimal simulated sandstone depends on the target response and provide a task-oriented route for reproducible simulated deep sandstone.

Applied Sciences doi: 10.3390/app16167900

Authors: Dolly Bhati Maria Hayes

An increase in flexitarians, vegetarians, and vegan consumers coupled with soaring egg prices post covid has increased demand in the food industry for alternative ingredients, including emulsifier products. Plant-derived emulsifiers, including chickpea aquafaba and oat proteins, can imitate egg protein techno-functional attributes like foam formation and emulsification properties. Oat milk proteins provide less protein than soy or dairy and, as they are carbohydrate rich, can cause rapid blood sugar spikes. More alternative emulsifiers are required. This work describes the development of a red seaweed Chondrus crispus Stakehouse emulsifier ingredient applied in a vegan fairy cake recipe. The techno-functional attributes of the developed C. crispus hydrolysate were assessed using in vitro assays, and a prototype vegan fairy cake was made using the generated hydrolysate as an emulsifier. Product proximate qualities including protein, ash, and lipid content were compared to those of a traditionally made fairy cake. Results suggest that hydrolysis of C. crispus enables the extraction of 17.37% of available protein and makes the application of C. crispus in food products more attractive by improving emulsification properties. Moreover, novel bioactive peptides with sequences GIPDEWMGL, VLPSLPFM, GGPAGGGPAGGDAGLA, and AQVPPLPNMPRMPA were characterized and were shown to have in silico anti-inflammatory and anti-diabetic potential activities.

Applied Sciences doi: 10.3390/app16167901

Authors: Kai Song Yiran Feng Xu Ji Qiaosheng Han

Kelp is a high-moisture, flexible, sheet-like marine biomass whose drying behavior is strongly affected by the coupled effects of radiative heating, convective vapor removal, internal moisture migration, tissue shrinkage, curling, and material overlap. Traditional sun drying and hot-air drying remain widely used but are limited by long processing cycles, environmental dependence, high energy consumption, and inconsistent product quality. With the development of infrared heating, heat-pump dehumidification, Internet of Things (IoT)-enabled sensing, fifth-generation (5G) mobile communication, multiphysics simulation, and digital control, kelp drying is progressively shifting toward monitored, model-assisted, and intelligent processing. This review critically summarizes recent advances in kelp and related seaweed drying, with particular emphasis on infrared-assisted heat and mass transfer, drying kinetics, coupled computational fluid dynamics–finite element method (CFD–FEM) simulation, quality evaluation, and intelligent control. Representative published studies demonstrate the engineering potential of these approaches. In a suspended infrared-array kelp drying system, an infrared power density of 1.2 kW m−2 combined with an air velocity of 3 m s−1 maintained the drying temperature at approximately 55–62 °C, while relative humidity decreased from about 80% to 20–30%. Under these conditions, the Page model achieved R2 = 0.987 and RMSE = 0.019, the rehydration ratio exceeded 94%, and the total color difference remained below ΔE = 6.5. A recent CFD–FEM–MATLAB workflow further reported a composite operating-condition index of J = 0.4535, with mapped mean and maximum kelp surface temperatures of 62.23 and 63.57 °C, respectively. These quantitative results indicate that the key challenge in infrared kelp drying is not simply to increase heat input, but to coordinate radiation distribution, airflow organization, internal moisture transport, structural response, and quality preservation. Future research should therefore focus on experimentally validated heat–mass-transfer models, adaptive sensing and control, multi-objective optimization, and pilot-scale verification under realistic production conditions.

Applied Sciences doi: 10.3390/app16167899

Authors: Mi-Ae Jeong Sang-Dol Kim

The growth of single-person older adult households has heightened concerns about social isolation, chronic disease management, and diminished quality of life (QoL). A systematic review with meta-analysis was conducted to examine the efficacy of human-centered AI healthcare interventions on QoL among older adults living alone. Five electronic databases (PubMed, Web of Science, CINAHL, Embase, and Cochrane CENTRAL) were searched (January 2020–March 2026) following PRISMA 2020 guidelines. Only randomized controlled trials (RCTs) were eligible. Methodological quality was appraised with the Cochrane RoB 2 tool; pooled effect estimates were derived via random-effects modeling. Fourteen RCTs (N = 2840) were included. Intervention types comprised conversational agents, socially assistive robots, remote monitoring systems, integrated platforms, and AI-driven mHealth applications. Meta-analysis demonstrated significant improvements in overall QoL (SMD = 0.40, 95% CI: 0.27–0.52, I2 = 59%), depression (SMD = −0.35, 95% CI: −0.43 to −0.26), and social connectedness (SMD = 0.38, 95% CI: 0.26–0.51). Subgroup analyses showed stronger effects for interventions with personalized feedback and human interaction. GRADE certainty was moderate for all three primary outcomes. GRADE certainty for secondary outcomes was moderate for physical health but low for self-management and cognitive function. Human-centered AI healthcare interventions significantly improve QoL and psychosocial outcomes among older adults living alone. Future research should prioritize long-term effectiveness, ethical implementation, digital inclusion, and culturally adaptive models.

Applied Sciences doi: 10.3390/app16167897

Authors: Thang Ngoc Pham Hoa Thi Phuong Nguyen Hong-Son Vu Khanh Tuan Do Huy-Du Dao

This paper investigates the problem of model-order reduction for linear systems arising from minimum-phase circuits and filters, where stability and frequency characteristics must be preserved. The MPPBT framework uses a Riccati–Lyapunov Gramian pair and constructs a balancing transformation with the sequence of MPPBT singular values, from which a relative error bound in the H∞ norm and a scoring function, Sβ, are derived to select the model order. We apply the algorithm to a fourth-order Butterworth low-pass filter with a full-order state dimension, n=4, and reduced-order models with r=1, 2, 3 are examined. The results show that the model with r=3 yields an H∞ error of approximately 6.3×10−3 and an H2 error of approximately 2.2×10−3. The model with r=1 gives an H∞ error of approximately 1.46 and an H2 error of approximately 4.8×10−1. The model with r=2 attains a composite score of Sβ≈0.16, preserves stability and the minimum-phase property, and is regarded as a balanced choice between accuracy and complexity. A further comparison on an RLC ladder circuit of order n=15 shows that at r=3, MPPBT achieves the lowest H∞ error among BT, PRBT, and MPPBT, while retaining an H2 error close to that of BT.

Applied Sciences doi: 10.3390/app16167898

Authors: Sinem Tola Joaquim Tinoco

Bridge scour threatens bridge safety and serviceability, yet the growing application of Machine Learning (ML) has not produced a unified understanding of how sensing sources, learning tasks and validation practices influence monitoring capability. This article addresses this gap through a taxonomy of 53 ML-based bridge-scour studies and a simulation-based railway case study examining the relatively underexplored use of displacement-related drive-by measurements for scour detection and support-level localisation. The dataset comprises 2924 simulated crossings involving four bridge models and three train types, with scour represented by reductions in vertical support stiffness. Full-bridge and support-centred Discrete Wavelet Transform and Autoregressive with exogenous inputs-derived Markov features are used to train Artificial Neural Network, Random Forest (RF) and Support Vector Machine classifiers under nested cross-validation. Batch inputs incorporate information from consecutive crossings. Among the highest-performing detection configurations, rear-mounted RF using 31-crossing batches achieved 0.990 accuracy and F1 scores of 0.994 and 0.977 for the scoured and healthy classes. The best localisation configuration, front-mounted RF using seven-crossing batches, achieved 0.968 accuracy, with both F1 scores approximately 0.97. These simulation-based results do not demonstrate transfer to unseen bridges or field-confirmed scoured conditions but indicate potential for bridge screening and targeted inspection.