Applied Sciences doi: 10.3390/app16147228

Authors: Junyan Han Shize Zhao Benwei Hou Zhongxian Liu Mohamed Hesham El Naggar Chengshun Xu

Buried pipelines with corrosion defects crossing reverse faults exhibit complex seismic responses under the combined effects of ground motion and fault displacement. Traditional finite-element analysis is computationally inefficient to comprehensively address such problems. This paper proposes a backpropagation neural network (BPNN)-based method for seismic response prediction and fragility assessment. A three-dimensional finite-element model is first employed to analyze the effects of corrosion depth-to-thickness ratio, diameter-to-thickness ratio, internal pressure, and burial depth on the axial compressive strain of the pipeline. Consequently, a BPNN model is constructed with these parameters, along with fault displacement, as inputs with the peak compressive strain as the output. The BPNN model demonstrated excellent predictive performance, with a maximum prediction error below 15%. The incremental dynamic analysis (IDA) method is then applied to map strength and damage indices of the pipeline, enabling quantitative evaluation of its failure probability and functional integrity under various conditions. It is found that higher diameter-to-thickness ratio (D/t) corresponds to a higher likelihood of the pipeline reaching adverse performance levels; this is also accompanied by a reduction in functional integrity. Specifically, as D/t increases from 72 to 144, the probability of pipe wall damage and the risk of transmission function loss rise significantly, highlighting the pronounced fragility of thin-walled pipelines subjected to fault movement. Moreover, corrosion defects exacerbate pipeline fragility: a corrosion depth equivalent to 10% of the wall thickness substantially amplifies strain responses, resulting in an approximately 80% probability of moderate damage, while a corrosion depth of 40% elevates the probability of severe damage beyond 60%.

Applied Sciences doi: 10.3390/app16147227

Authors: Jack Pessaud Eric Moran John Nguyen Joel Palko

With the increasing adoption of iris recognition systems and the expansion of large-scale enrollment databases, there is a growing need to efficiently assess iris image quality at the time of acquisition. Image quality may degrade for several reasons, such as eyelid occlusion or pupil dilation. Although previous studies have shown that occlusion and changes in the pupil-to-iris ratio negatively impact recognition performance, these investigations were typically limited by small sample sizes and did not examine the combined effects of eyelid and pupil variations. In this study, we analyze both dilation and eyelid occlusion using a large dataset of 827 distinct irises and demonstrate a strong correlation between probe image visible iris area and the Hamming distance of iris code pairs. Building on this finding, we develop a probe-adaptive row-weighted Hamming distance model that scales per-row contributions by a spline function of probe covariates, significantly reducing the false nonmatch rate on the held-out evaluation split. These results establish visible iris area as a robust, acquisition-time quality signal and demonstrate that it can be directly operationalized to improve match confidence without modifying the underlying iris recognition pipeline.

Applied Sciences doi: 10.3390/app16147225

Authors: Yisen Zhang Zhenzhou Wang Suizi Jia

Near-fault pulse-like ground motions produce coupled intensity, duration, and period-matching effects, making nonlinear seismic assessment of multi-ribbed composite slab structures (MCSS) computationally expensive and difficult to generalize. To address this problem, a 4000-case OpenSees nonlinear time-history analysis (NLTHA) database is generated from Wenchuan ground motions through Latin hypercube sampling, and a mechanically informed feature-enhanced deep neural network (MIFE-DNN, previously denoted as PE-DNN in the first submission) is trained using equivalent stiffness, equivalent yield strength, mass proxy, demand-capacity ratios, period-matching ratio, normalized duration, and energy-capacity proxy; a validation-weighted stacked surrogate is further constructed from multi-seed MIFE-DNN and residual learners. On the independent test set, the mean R2 increases from 0.9645 for the ordinary deep neural network (DNN) and 0.9686 for the single MIFE-DNN to 0.9782 for the stacked mechanically informed surrogate, while the maximum inter-story drift-ratio R2 reaches 0.9541. Additional checks include 16 active-learning OpenSees enrichment cases, 12 analyses under two external near-fault records, 3 out-of-domain parameter cases, 100 cross-story tests, SHAP-based interpretation, and multi-EDP fragility post-processing. These checks show that the surrogate is reliable for interpolation and screening within the calibrated equivalent-model domain, but direct OpenSees recalculation is required for boundary, out-of-domain, and cross-configuration use. Parameter-importance, SHAP, and fragility analyses identify peak ground acceleration (PGA), pulse index, period matching, rib height, rib spacing, and damping ratio as dominant factors, indicating that mechanically informed feature-enhanced surrogate modeling provides an interpretable and efficient tool for MCSS response prediction and conditional fragility assessment within the sampled structural and ground-motion domain.

Applied Sciences doi: 10.3390/app16147226

Authors: Jiayu Ling Pengyu Sun Ruijie Sun Yiming Sheng Yi Du Echuan Yan

Cut blasting beneath operating tank farms poses a coupled design problem: the charge must create a cavity while keeping vibration contours outside protected structures. Existing checks evaluate peak particle velocity (PPV) at monitoring points or excavation quality separately, leaving the feasible charge range unclear. This study proposes a site-calibrated dual-threshold framework for a construction adit in a water-sealed cavern. A refined cut-hole damage model defines the lower charge threshold from damage connectivity, damage volume ratio and maximum block volume, while a field-scale vibration model defines the upper threshold from the 3.5 cm/s PPV contour relative to surface structures. Field fragmentation recognition and five vibration records support the comparison, with PPV errors of 0.24–3.37% and waveform correlations of r = 0.62–0.69. At the studied site, charges below 67.2 kg failed to satisfy cut-cavity criteria, whereas 79.8 and 84.0 kg caused the exceedance contour to contact or cross the structural boundary. A 75.6 kg grouped cut charge satisfied both thresholds and reduced cut explosive consumption by 10.0%. These charge values are site-specific; transfer to other excavations requires recalibration against local geology, structural layout and vibration limits. The framework converts charge selection into a bounded safety-and-effectiveness window for constrained underground excavations.

Applied Sciences doi: 10.3390/app16147224

Authors: Junwen Yao Wenxing Sun Xionggang Li Ying Zhang Xiao Yue Xutao Chen Yongfei Ma Yupeng Zou

Maintaining mechanical stability and achieving precise motion control on cylindrical surfaces remain critical challenges for underwater jacket cleaning robots. To address these issues, a systematic mechanical model is initially established for a standard column (diameter 2000 mm, inclination angle 90°) to determine stable climbing conditions. By neglecting equivalent gravity, the critical constraint relationships between adhesion and propulsion forces are derived. Subsequent open-loop maneuvering simulations, conducted via a six-degree-of-freedom dynamic model in MATLAB/Simulink, reveal that the robot achieves a straight-line speed of 0.233–1.601 BL/s and a maximum steady-state thrust of ~80 N at rotational speeds between 500 and 3000 rpm, and this thrust level corresponds to a required maximum adhesion force of approximately 106.4 N. To enable autonomous navigation, a closed-loop control system is developed for two-dimensional waypoint tracking, integrating line-of-sight (LOS) guidance, a waypoint switching radius strategy, and a proportional-derivative (PD) controller. Simulation results verify that the robot successfully navigates through five sequential turning waypoints under operational constraints, maintaining well-converged position and heading errors, with the maximum heading deviation controlled within ±5°. All simulation validations in this study are completed through numerical simulations in the MATLAB/Simulink environment. These findings offer a practical reference and theoretical foundation for the autonomous control and reliable deployment of underwater cleaning robots.

Applied Sciences doi: 10.3390/app16147222

Authors: Ekaterina A. Lysikova Iuliia S. Sukhanova Natalia V. Ekimova Mikhail V. Korokin Michail S. Kukharsky Vladimir L. Buchman Natalia Ninkina

The synuclein family, comprising α-, β-, and γ-synucleins, consists of highly conserved, intrinsically disordered proteins sharing substantial structural homology and exhibiting multifunctional roles in membrane interactions, vesicle trafficking, lipid homeostasis, and cellular signalling. Synucleins are enriched at presynaptic terminals, where they regulate synaptic vesicle dynamics and neurotransmission by contributing to vesicle clustering, trafficking, and the availability of vesicles for release. Aberrant expression, misfolding, and aggregation of synucleins are associated with several human diseases, particularly neurodegenerative disorders and certain cancers, highlighting their biological and clinical relevance. The primary aim of this review was to provide a focused overview of current knowledge regarding the physiological presynaptic functions of synucleins. By concentrating on this relatively underexplored aspect of synuclein biology, we sought to highlight their roles in synaptic transmission and presynaptic regulation, thereby complementing the extensive literature devoted to their pathological significance. Here, the recent findings on the contributions of α-, β-, and γ-synucleins to synaptic vesicle organization, trafficking and neurotransmitter release were summarized, and their emerging roles in maintaining presynaptic homeostasis were discussed. In addition, we consider how disruption of these physiological functions may contribute to synaptic dysfunction and the development of disease.

Applied Sciences doi: 10.3390/app16147223

Authors: In-Young Choi Jundong Kim Ji-Hwan Kim Hye-Yoon Lee Won-Hwan Park Chang-Eop Kim Dong-Woo Lim

Background: History-taking is a core clinical competency in Korean medicine diagnostics, but conventional training methods such as peer role-play and standardized patient-based education have limitations in providing repeated, individualized, and scalable practice opportunities. This study aimed to evaluate the feasibility and educational value of a two-session AI chatbot-based history-taking practicum with automated feedback in Korean medicine diagnostics. Methods: This prospective single-arm repeated-measures educational study was conducted with fourth-year students at the College of Korean Medicine, Dongguk University, in May and June 2026. A total of 76 students participated in two chatbot-assisted history-taking sessions using dizziness and shoulder pain scenarios. Students completed surveys on baseline AI familiarity, chatbot experience, usability, and self-efficacy. Self-efficacy was assessed at three time points: before the first session, after the first session, and after the second session. Chatbot-generated feedback scores were compared between session 1 and 2 for each scenario using paired complete-case analyses. Open-ended responses were descriptively categorized. Results: Students rated the chatbot-based practicum positively in terms of active participation, perceived usefulness, accessibility, and convenience. Item-level self-efficacy analysis showed significant time effects in two domains: planning the conversation, and closing the conversation appropriately. The overall mean self-efficacy score gradually increased from 3.739 ± 0.546 before the first session to 3.887 ± 0.591 after the second session; however, the overall time effect did not reach statistical significance. Chatbot-generated feedback scores showed scenario-dependent patterns. Scores for the shoulder pain scenario increased from session 1 to session 2 before adjustment, but this change did not remain significant after Holm correction; scores for the dizziness scenario showed a non-significant decreasing trend. Open-ended responses indicated that students valued repeated practice and immediate feedback, while also noting limitations related to feedback accuracy, realism of patient responses, and the lack of physical examination or multimodal diagnostic information. Conclusions: The chatbot-assisted practicum was feasible and favorably perceived by students, with selected item-level changes and a modest non-significant upward trend in overall self-efficacy in this exploratory educational study. These findings support the potential role of AI chatbot-based simulation as a supplementary, scalable tool for repeated history-taking practice and formative feedback, rather than as a replacement for performance-based clinical skills training. Given the single-arm design and reliance on learner-reported outcomes, these findings should be interpreted as exploratory. Future controlled studies should incorporate objective performance outcomes, expert-validated automated scoring, non-AI comparison groups, and more realistic multimodal clinical scenarios to determine the educational effectiveness of chatbot-assisted history-taking training.

Applied Sciences doi: 10.3390/app16147221

Authors: Eleni Gesthimani Pachni Nikoletta Solomakou Dimitrios Fotiou Athanasia M. Goula

Hibiscus sabdariffa L. is a valuable source of phenolic compounds, including anthocyanins with recognized antioxidant and health-related properties; however, their instability under processing and storage conditions limits their direct application in food and nutraceutical systems. In this context, encapsulation offers a promising strategy for enhancing the stability and delivery of hibiscus bioactives. The present study comparatively evaluates three encapsulation techniques, spray drying, co-crystallization, and ionic gelation, for the stabilization of aqueous hibiscus extract. Each process was optimized through systematic adjustment of key operating parameters. The encapsulated products were characterized for physicochemical properties (moisture content, bulk density, wetting time, hygroscopicity, color, particle morphology), encapsulation efficiency, antioxidant capacity, anthocyanin content release behavior, and structural interactions through Fourier-transform infrared (FTIR) analysis. The encapsulation method strongly affected phenolic retention, antioxidant activity, color characteristics, and release behavior of the final products. Under the optimized conditions, encapsulation efficiency was approximately 95% for spray drying, while co-crystallization showed an apparent encapsulation efficiency of 99.96%, whereas ionic gelation showed measured values of 7.0–34.0% and a model-predicted, validated optimum of 50.83%. Spray drying better preserved antioxidant activity, anthocyanin content, and color characteristics, while co-crystallization produced low-moisture, low-hygroscopicity powders with rapid aqueous release. FTIR analysis of the spray-dried and co-crystallized samples indicated physical entrapment rather than chemical modification, confirming the successful formation of encapsulated systems. Τhe findings highlight method-dependent differences in the technological and functional performance of the encapsulation systems, providing insight into the stabilization and controlled release of hibiscus-derived bioactive ingredients, with spray drying and co-crystallization emerging as the most promising approaches.

Applied Sciences doi: 10.3390/app16147220

Authors: Cem Peskersoy Basak Singun Ezgi Demir Ilgin Ilgenli

The aim of this in vitro study was to evaluate the internal adaptation and adhesion performance of a novel 3D-printed custom post system compared with three commercially available intra-canal post systems with different physical and surface characteristics. Twenty extracted human mandibular premolars were endodontically treated and divided into four groups according to the post system used: RelyX Fiber Post (3M ESPE), Splendor SAP Post (Angelus), everStick Post (GC), and custom 3D-printed posts fabricated from a permanent restoration resin. Following cementation with a self-adhesive resin cement, micro-computed tomography was used to assess cement thickness and gap formation in the apical, middle, and coronal regions, while push-out bond strength was evaluated using a universal testing machine. The 3D-printed custom posts exhibited the lowest mean cement thickness (39.27 ± 14.47 µm in the apical and 66.25 ± 47.98 µm in the coronal region), whereas the Splendor SAP group showed the highest values (reaching up to 219.76 ± 74.64 µm in the coronal region). In contrast, the highest push-out bond strength was observed in the RelyX Fiber Post (46.52 ± 26.00 MPa) and GC everStick (48.17 ± 4.64 MPa) groups in the coronal region, while the 3D-printed custom posts demonstrated the lowest overall bond strength values, dropping to 8.53 ± 3.56 MPa in the apical region. In all groups, bond strength significantly decreased from the coronal to the apical region. Within the limitations of this study, 3D-printed custom posts provided superior intra-canal adaptation but inferior adhesion performance compared with conventional post systems.

Applied Sciences doi: 10.3390/app16147219

Authors: Lucia Grünermelová Radoslav Koňár Miloš Mičian

Current industrial standards for gas pipeline leak testing often assume inert testing gases are universally interchangeable, neglecting specific fluid dynamics. This study quantifies how testing gas properties affect integral pressure decay leak tests. Experimental measurements (5000 Pa initial pressure) were conducted on a PE100RC (polyethylene resistant to crack propagation) pipe assembly with artificial capillary defects (0.13 mm diameter) using five media: nitrogen, air, argon, carbon dioxide, and a propane−butane (PB) mixture. Results demonstrate that in the continuous viscous flow regime, leak rates depend strictly on the gas’s dynamic viscosity. For identical defects, PB (the lowest viscosity gas tested) increased the leak rate by up to 129% compared to standard nitrogen. This relationship is statistically validated by a strong negative Pearson correlation (r = −0.92). To facilitate industrial application, a preliminary mathematical correction procedure is proposed for safely extrapolating these trends to infrastructure intended for pure hydrogen operation. To prevent false-positive tightness certifications during the transition to low-viscosity alternative fuels like pure hydrogen, implementing a dynamic viscosity correction factor is essential.

Applied Sciences doi: 10.3390/app16147218

Authors: Vicente Bayarri-Porcar Maria-Jesus Agost Joaquín L. Sancho-Bru Margarita Vergara

Product selection often requires visual comparison between multiple alternatives, and hesitation during this process remains less understood than the prediction of final choice. This study investigated whether hesitation during product evaluation could be characterized using eye-tracking metrics in a task involving six jar-opening devices evaluated for robustness, comfort, and effort level required. Sixty-seven participants completed a visual selection task in which each trial was classified according to participants’ self-reported hesitation after each question. Analyses examined first-fixation behavior, global trial-level eye-tracking metrics, attention allocation across grouped alternatives within hesitation trials, pupil diameter, and qualitative written reflections. Hesitation trials were associated with longer decision times, more saccades, and greater total saccadic amplitude. Furthermore, early orienting also differed, with first fixations occurring earlier in hesitation trials. Within hesitation trials, attention was not concentrated on the selected option; candidate alternatives were the most intensively inspected options when attention was considered per individual model. By contrast, baseline-corrected pupil diameter did not significantly differentiate hesitation from no-hesitation trials. These findings suggest that hesitation can be characterized through a structured pattern of visual behavior involving early attention shifts, more extensive exploration, and attention redistributed across competing alternatives.

Applied Sciences doi: 10.3390/app16147217

Authors: Anca Chelmuș Mihaela Constantin Nicolae Băran

Reverse-flow cyclone preselection is often performed with nominal semi-empirical closure values, although swirl and turbulence corrections are family-dependent. This study develops a scenario-based robust geometric screening framework for fixed-throughput reverse-flow cyclones using a family-calibrated reduced-order model for cut size and grade efficiency, combined with a fixed-Eu pressure drop. The design variables are the inlet-area ratio, vortex-finder diameter ratio, and effective residence-height ratio. Closure uncertainty is represented by 27 combinations of the swirl-intensity coefficient, turbulence-correction coefficient, and turbulence intensity. Deterministic and robust Pareto fronts are compared in the pressure-drop–corrected-cut-size plane (Δp,d50,t), and worst-case grade-efficiency indicators are evaluated for 2 and 5 μm particles. The results show that closure uncertainty substantially penalizes nominal fine-particle predictions: across the robust Pareto set, the worst-case corrected cut size is 1.79–1.93 times the nominal value. Within the original design window, the reduced-order screen identifies a balanced candidate near ϕin ≈ 0.10, De/D = 0.40, and Heff/D = 6, with lower predicted nominal and worst-case corrected cut sizes than the baseline at unchanged fixed-Eu pressure drop estimate. The selected geometries are not independently validated in the present study and should be treated as candidates for subsequent high-fidelity assessment. Design-grid refinement and extended-bounds, hydraulic, and logistic-slope sensitivity audits show that the numerical screening results are stable within the adopted formulation but remain conditioned by its closure assumptions.

Applied Sciences doi: 10.3390/app16147216

Authors: Mateusz Radwanski Ewa Zmysłowska-Polakowska Tan Fırat Eyüboğlu Mutlu Özcan Monika Lukomska-Szymanska

Artificial intelligence (AI) has emerged as a promising tool in endodontic diagnostics, particularly in radiographic interpretation and treatment outcome prediction. This narrative review provides a comprehensive overview of current applications of machine learning (ML) and deep learning (DL) algorithms in endodontics, focusing on periapical lesion detection, assessment of root canal morphology, prognosis prediction, and navigation-guided procedures. A literature search was conducted in PubMed, Scopus, and Web of Science for studies published between 2021 and 2026. Following qualitative screening, 45 studies were included in the evidence synthesis. Given the narrative design of the review, no formal meta-analysis was performed. The reviewed evidence indicates that convolutional neural networks (CNNs), U-Net architectures, YOLO-based systems, and other AI models have demonstrated promising diagnostic performance across multiple imaging modalities, with the highest performance generally reported for CBCT-based applications. However, the current body of evidence remains limited by retrospective study designs, small datasets, methodological heterogeneity, and insufficient external validation. Overall, AI should be considered a valuable adjunctive decision-support tool in endodontic diagnostics rather than a replacement for clinical expertise. Further prospective multicenter studies, external validation, and standardized methodological frameworks are required to support its reliable integration into routine clinical practice.

Applied Sciences doi: 10.3390/app16147215

Authors: Lei Liu Shun Jiang Han-Kyo Seo

Lower-limb EMG waveform prediction is difficult to interpret because a model may appear accurate by reproducing average task- and phase-dependent activation patterns rather than learning subject-specific neuromuscular information. This secondary analysis used the processed Gait120 dataset, including 120 healthy adult male participants, 6882 paired kinematic–EMG cycles, seven movement tasks, and 12 processed EMG envelope channels. Under five-fold subject-disjoint validation, a zero-shot task–phase template estimated from training subjects achieved RMSE × 103 = 0.0188, MAE × 103 = 0.0088, R2 = 0.454, and NRMSE_SD = 0.727 in held-out subjects. For k-shot personalization, 1, 3, 5, or 10 EMG-labelled calibration cycles from each held-out subject were used to estimate task-specific shrinkage residual corrections, while separate cycles from the same subject were reserved for evaluation. Ten-shot personalization improved performance to RMSE × 103 = 0.0138, MAE × 103 = 0.0062, R2 = 0.708, and NRMSE_SD = 0.535. Task-specific correction outperformed global residual correction, whereas a fixed-parameter linear Ridge residual comparator did not improve zero-shot performance beyond the template. These findings support task–phase templates as demanding reference baselines for processed EMG envelope prediction, but only under offline conditions with known task labels, known normalized phase, and within-subject EMG-labelled calibration.

Applied Sciences doi: 10.3390/app16147214

Authors: Qi Zhang Kechang Qian Hefei Huang Zhonghe Li Huimin Meng Zhifei Li Hongyan Wang Yaoyao Dong

Quantifying the net hydrological impact of large dams amidst climatic and anthropogenic influences remains a major challenge. This study isolates the effect of Three Gorges Dam (TGD) regulation on discharge at Jiujiang Station in the middle Yangtze River (2009–2016) using a novel scenario-based framework. A Long Short-Term Memory (LSTM) network, optimized by the Sparrow Search Algorithm (SSA), simulated daily discharge with high accuracy (Nash–Sutcliffe Efficiency coefficient > 0.97). By comparing a “with-TGD” simulation against a “without-TGD” scenario—generated by replacing the dam’s regulated outflow with its reconstructed natural inflow—we quantified the net impact (ΔQ). Results show that ΔQ is substantially modulated by river–lake interactions. For example, in December, the backwater effect from Poyang Lake amplified the direct flow reduction by an additional −82.5 m3/s. The “peak-shaving” effect was context dependent: TGD regulation increased high flows (>30,870 m3/s) by an average of +372 m3/s while slightly decreasing low flows (<12,711 m3/s) by −31 m3/s. The impact exhibits strong seasonality alongside considerable intra-seasonal variability, reflecting multi-objective operations (flood control, power generation, water supply). This framework provides a transferable approach for attributing hydrological change in large regulated rivers and supports integrated water resources management.

Applied Sciences doi: 10.3390/app16147213

Authors: Chenyu He Yiyang Li Yang Jin Shiqi Zhao

Wearable devices make it easy to collect long-term records of activity, sleep, heart rate, and other behavioral signals. However, translating these data into reliable personal health insights remains difficult. ReAct-style large language model (LLM) agents can compute over wearable records, yet two issues remain: health interpretations may rely on weak evidence, and long sessions accumulate tool traces that increase context pressure. We present PHIA+, an agent framework designed to address these issues. It introduces LLM-guided medical retrieval, which improves evidence reliability by selecting source-aware passages from a medical corpus, and summary-based hierarchical memory, which reduces context pressure by compressing completed reasoning traces while preserving the current one. We evaluate PHIA+ on the PHIA benchmark with objective and open-ended protocols. In the stacked objective question answering (QA) study, hierarchical memory reduces context length by a factor of 11.8 and API token use by 56.8%, while keeping final accuracy not significantly different from full-history reasoning (within 0.22 percentage points). In the open-ended QA study, under five-pass anonymized LLM judging, the RAG-augmented PHIA attains the highest mean medical source reliability score (2.92 out of 5) and the highest mean response quality score (3.84 out of 5) among five retrieval ablations. These results suggest that explicit mechanisms for evidence retrieval and long-session memory can improve source reliability and context efficiency in wearable-health agents, though the reliability of medical retrieval remains bounded by the coverage of the open-access corpus. These findings are useful to developers and researchers building practical wearable-health assistants.

Applied Sciences doi: 10.3390/app16147212

Authors: Andrei Jorza Adrian Pană Florin Molnar-Matei Alexandru Băloi Ilona Bucatariu

One of the negative effects of the rapid increase in the number and capacity of photovoltaic power sources distributed in consumption areas by connecting to existing low-, medium-, and high-voltage distribution networks is the increased risk of amplifying the non-sinusoidal steady-state condition caused by high-power inverters. This amplification occurs if the peak values of harmonic impedance in the sections of the respective grid zone, which depend on the values of the equivalent shunt capacitances present in the grid, correspond to frequency values that coincide with or are close to the frequencies of the harmonic currents injected by the distributed sources. The risk is particularly high in cases of malfunction or failure of the filters within the installations associated with these sources. High shunt capacitance values are not only caused by the capacitive compensators used to increase the power factor of consumers or to improve the voltage level in the grid, but also by long power lines, particularly long underground lines. Risks increase as the proportion of underground power lines in the network area with sources of harmonic currents rises. This article refers to a real high-voltage grid area where the distribution operator plans to install high-power photovoltaic sources, the connection of which requires grid expansion through the construction of new substations and high-voltage power lines (110 kV). Based on the designer’s intention for the new high-voltage power lines to be underground and of relatively long lengths—which implies the presence of high natural capacitances—the authors conduct two case studies to predictively evaluate the harmonic impedance in the network sections resulting from the expansion, with the aim of identifying the frequencies at which parallel resonances may occur. The authors use two software tools for numerical analysis, Mathcad and MATLAB–Simulink, and compare the results obtained for two design variants of the new high-voltage lines: overhead and underground, respectively. Using the mathematical model of long lines in steady state with uniformly distributed parameters, the analysis highlights that the highest values of harmonic impedance correspond to the overhead line design variant. The use of this design variant increases the risk of parallel resonances, not only by increasing the harmonic impedance values but also by widening the frequency ranges for which the impedance has high values. In both the overhead line and underground line variants, increasing the load leads to a reduction in the maximum values. Compared to the design variant using overhead lines, the variant with underground lines increases the risk of parallel resonances by increasing the number of frequency intervals for which the harmonic impedance has high values, but this increase is limited by narrowing these frequency intervals and reducing the harmonic impedance values. An important contribution of the article lies in arguing the need to extend the conventional node-based evaluation of harmonic impedance by treating harmonic impedance as a continuous spatial frequency characteristic of the transmission network. The impedances seen at nodes become the impedances seen in particular sections/sections of the network, identified by the value of the variable that specifies their position in space, more precisely the distance from the end of one of the feeders on which they are located.

Applied Sciences doi: 10.3390/app16147211

Authors: Domenico Finocchiaro Asia Sarti Nicola Maffei Maria Victoria Gutierrez Marco Saguatti Luciano Morini Agnese Chendi Cecilia Cadioli Alessio Bruni Gabriele Guidi

Accurate dose calculation is crucial in radiation therapy, particularly when employing advanced techniques such as VMAT. Monte Carlo (MC) algorithms are widely recognized as the gold standard for dose calculation. This study aimed to validate an MC dose calculation model in RayStation® TPS and to compare its performance against the Collapsed Cone (CC). The MC model was evaluated using measured data from four matched Elekta® Versa HD LINACs. Validation followed AAPM TG-157, TG-244 and IAEA TECDOC guidelines, including homogeneous media, heterogeneous media and advanced techniques (VMAT, IMRT, dMLC). MC showed high agreement with measurements, with most results within a 2% dose difference or passing a gamma criteria ≥95%. Performance was comparable or superior to CC in many cases, especially in small-field stereotactic plans. Impact of grid size and statistical uncertainty was assessed to find the optimal values. This study indicated a 2 mm grid size with 0.5% uncertainty as the optimal balance between accuracy and computation time. In conclusion, this study demonstrated that the MC model can provide accurate dose calculations for treatment plans, with computation times compatible with clinical practice. The use of MC appears to be particularly advantageous for advanced techniques and in the presence of tissue heterogeneities.

Applied Sciences doi: 10.3390/app16147210

Authors: Pascal Mindermann

Coreless filament winding (CFW) enables the material-efficient digital fabrication of lattice fiber composite lightweight structures with high geometric freedom. However, the unconstrained formation of fiber bundle leads to variable cross-sectional shapes, rugged surfaces, and limited predictability of bundle-level geometry. This study evaluates an industrial epoxy spray coating and a custom epoxy immersion coating as post-processing technologies for improving the cross-sectional geometry of CFW fiber bundles at different compaction levels. Three groups of samples were compared: manually squeezed high-compaction elongated samples with immersion coating, wrapped medium-compaction circularized samples with spray coating, and low-compaction elongated samples with spray coating. Loop specimens were fabricated, modified after winding where applicable, cured, coated, sectioned into samples, and analyzed using digital light microscopy. Image-based analysis quantified layer thickness, coverage interruptions, roundness, ruggedness, roughness, second moments of area distribution, volumetric phase fractions, surface gain, mass gain, and simplified area-estimation errors. Wrapping was the most effective measure for improving global roundness and squeezing improved compaction without circularization. Coating reduced ruggedness in all groups, including the highly rugged low-compaction samples, where part of the contour regularization resulted from bridging indentations, causing higher surface than mass gain. Coating improved the mass-specific second moment in every sample group, with an anisotropy reduction in the immersion-coated sample group.

Applied Sciences doi: 10.3390/app16147209

Authors: Irem Kilinc Andres Abea Yvan Llave Mika Fukuoka

High-frequency ohmic heating (HFOH) represents a significant advancement over conventional ohmic heating (OH) by reducing limitations related to electrode–food interactions. Although conventional OH provides rapid and volumetric heating, its performance is often limited by electrode fouling and corrosion. A key finding of the present review is that the existing literature lacks a unified frequency-based classification of OH, making it difficult to compare processing conditions, interpret frequency-dependent mechanisms, and identify appropriate operating regimes for different food applications. To address this gap, this review proposes a novel functional classification based on the dominant electrical behavior, electrochemical phenomena, and processing characteristics observed across different frequency ranges. Accordingly, this review classifies OH systems into four functional frequency ranges: low-frequency range (<1 kHz); medium-frequency range (1–<10 kHz); high-frequency range (10–30 kHz); and ultra-high-frequency range (>30 kHz–<1 MHz). HFOH operates within these upper ranges to minimize electrochemical reactions at the electrode interface, resulting in faster and more homogeneous temperature distributions. These advantages lead to lower cooking losses and improved retention of nutritional, sensory, and structural attributes, particularly in heterogeneous and multiphase food systems where conventional methods often fail. However, these effects are matrix-dependent, and low-frequency OH can sometimes achieve stronger microbial inactivation via electroporation. Recent research highlights the importance of computational modeling and simulation for process optimization, as these tools enable accurate prediction of thermal profiles and help reduce temperature irregularities. This review analyzes published studies on HFOH and compares its performance with other OH systems. Applications of HFOH across various food processes are evaluated, including microbial and enzyme inactivation, cooking, heating extraction, thawing, tempering and drying/dehydration. Furthermore, this study discusses the benefits and challenges of HFOH and provides strategic recommendations to address these technical limitations, offering insights that may support its future practical implementation and scalability within the modern food processing industry.

Applied Sciences doi: 10.3390/app16147207

Authors: Tingting Huang Shengmin Hong Jundong Chen Liangyi Kang

Classical coherent chaos shift keying (CSK) provides certain physical-layer advantages by exploiting chaotic waveforms, but it relies on fragile chaos synchronization, which is difficult to maintain in multipath fading channels. To overcome this limitation, we propose a novel non-coherent CSK detector based on the spectral centroid (SC)—the center of mass of the power spectral density. We demonstrate that the SC is an intrinsic invariant of a chaotic map, enabling a fixed-threshold detection scheme that requires no synchronization, no reference sequence, and no channel estimation. Simulation results over multipath Rayleigh fading channels show that the proposed system significantly outperforms conventional CSK pairs. Unlike coherent CSK, which suffers from an error floor under frequency-selective fading, the proposed SC-CSK system benefits from frequency diversity and achieves robust performance. Moreover, while conventional differential chaos shift keying (DCSK) degrades severely under high-mobility scenarios, the proposed system maintains a stable BER across a wide range of Doppler shifts. The receiver is lightweight, interpretable, and well-suited for resource-constrained applications such as Internet of Things (IoT) devices.

Applied Sciences doi: 10.3390/app16147206

Authors: Sebastian Mesca Emil Pricop Grigore Stamatescu

Intrusion detection systems (IDSs) for industrial control networks are commonly evaluated using random stratified splits, placing rows from every recorded attack in both training and test sets. Although convenient, this practice measures a model’s ability to recognise repetitions of patterns it has already seen rather than its ability to detect novel attacks. We revisit supervised and unsupervised machine-learning IDSs on the Secure Water Treatment (SWaT) dataset’s network-traffic modality, extending a prior conference study, and quantify the effect of more rigorous evaluation protocols. We evaluate five model families (XGBoost, a convolutional–MLP hybrid, a bidirectional LSTM classifier, an unsupervised LSTM-Autoencoder, and a temporal convolutional network) under three protocols: stratified random, attack-held-out, and leave-one-attack-out (LOO). Under LOO on a 30-file subsample, every supervised classifier scores below random on the majority of held-out attacks; the unsupervised LSTM-Autoencoder retains the best solo mean of 0.550 with a strongly bimodal per-attack distribution spanning 0.046 to 0.894. A sign-adjusted oracle-bound ensemble flips members whose per-attack AUROC inverts achieves a mean LOO AUROC of 0.844; adding the TCN as a fourth ensemble member does not improve the result, providing evidence that what is needed is an additional detection mode rather than another supervised classifier. We additionally report recall at a 5% false-positive-rate budget, paired Wilcoxon significance tests, and bootstrap confidence intervals. The full preprocessing, evaluation, and ensemble pipeline is released, and we argue that attack-held-out and LOO should be standard protocols for network-traffic IDS benchmarks on SWaT.

Applied Sciences doi: 10.3390/app16147208

Authors: Zhengzhe Zhang Shichun Bao Changzi Qu Fan He

During precast beam yard construction on soft ground, prestress tensioning-induced camber transforms the initially uniform load distribution into eccentric concentrated loads at the beam ends, thereby posing risks to foundation stability and geometric accuracy. This study investigates the mechanical responses of natural ground, single-pile reinforcement, and double-pile reinforcement under tension loading through field monitoring and numerical simulations based on a Hangzhou–Ningbo Expressway project. The results show that single-pile reinforcement exhibits a typical point-support behavior, characterized by significant stress concentration at the pile head, a pile–soil stress ratio ranging from 3.0 to 4.0, and a coefficient of variation in base pressure reaching 0.81. In contrast, the double-pile scheme enables load redistribution through a line-support mechanism along the beam length, improving stress uniformity and reducing the coefficient of variation to 0.62. Orthogonal test-based sensitivity analysis further identifies pile diameter and cushion thickness as the dominant factors influencing composite ground. performance. This study clarifies the differences in load-transfer mechanisms among various pile arrangements under tension loading and provides a theoretical basis for the optimized design of casting bed foundations for precast beam yards in soft soil regions.

Applied Sciences doi: 10.3390/app16147205

Authors: Inge Huybrechts Marthe De Boevre Inarie Jacobs Carine Biessy Liesel Claeys Michael Korenjak Geneviève Nicolas Ghislaine Scelo Aurora Perez-Cornago Beatrice Fervers Eva Ardanaz Cecilie Kyrø Anne Tjønneland Krasimira Aleksandrova Maria-Jose Sánchez Verena Katzke Tilman Kühn Sandra Colorado-Yohar Sabina Sieri Isabelle P. Oswald Julien Vignard Matthias B. Schulze Guri Skeie Francesca Romana Mancini Ana Jiménez-Zabala Pietro Ferrari Jiri Zavadil Sarah De Saeger Marc J. Gunter

Mycotoxins are widely consumed food contaminants. Mechanistic studies support potential carcinogenic effects of mycotoxins in the human intestine, yet no studies have investigated these associations in population-based settings. This study aimed to investigate associations between dietary mycotoxin exposures and colorectal cancer risk. Dietary questionnaire data from the European Prospective Investigation into Cancer and Nutrition (EPIC) were combined with mycotoxin food occurrence data from the European Food Safety Authority (EFSA). Associations between mycotoxin exposures and colorectal cancer (CRC) risk and its anatomical sub-sites were evaluated in multivariable Cox proportional hazards regression models to compute hazard ratios (HR) and their 95% confidence intervals (CI), accounting for potential confounders. Food groups that explained the greatest variation in mycotoxin intakes were cereal-based products, vegetables, non-alcoholic beverages and fruits. For individual mycotoxins, deoxynivalenol (DON) was positively associated with CRC risk (HR top vs. bottom tertiles [T3vsT1] = 1.14; 1.04–1.24), while patulin (PAT) was positively associated with rectal cancer risk only (HRT3vsT1 = 1.18; 1.05–1.32). Sensitivity analyses indicated differential results for men versus women with significant positive associations for DON and CRC risk found only in men. In conclusion, dietary mycotoxin exposures may be related to higher CRC risk. Further investigation in independent cohorts and through mechanistic studies is warranted.

Applied Sciences doi: 10.3390/app16147202

Authors: Yanan Cui Xiao Lv Chunyu Zhang Xuanye Zhao Xueqian Fu

The output of photovoltaic power generation exhibits significant randomness, volatility, and intermittency, and its uncertainty will have an impact on the planning assessment, dispatch decision-making, and real-time operation of the power system. To systematically understand the development status of modeling methods for the uncertainty of photovoltaic power generation, this paper conducts a review around the generation of annual scenarios and multi-timescale power prediction of photovoltaic power, and analyzes the correlations between photovoltaic output, influencing factors, and system applications from the perspective of statistical relationships and artificial intelligence. For the annual scale, the focus is on the generation methods of meteorological-driven scenarios for long-term sequences, including probability statistical methods, deep generation methods, constraint relations and engineering application evaluation issues; for the day-ahead scale, the historical power, meteorological variables and numerical weather forecasts are used to explore feature extraction, probability prediction and robust modeling methods; for the intraday scale, the signal decomposition, deep learning, regional collaborative modeling and multi-source perception methods for short-term power fluctuations perception are summarized. On this basis, further analysis is conducted on subsequent research directions such as multi-timescale collaborative modeling, multi-source heterogeneous information fusion, controllable generative modeling, extreme scenario characterization, and engineering closed-loop verification. This paper can serve as a reference for scenario generation, power prediction, and power system operation analysis under the condition of a high proportion of photovoltaic power integration.

Applied Sciences doi: 10.3390/app16147204

Authors: Tadaaki Sone Ryosuke Katagiri Takashi Kawamura Shimpei Aihara

In curling, quantitative assessment of ice surface conditions during actual play is challenging, and existing sensor-on-stone approaches cannot simultaneously capture ice-contact vibration and athlete-related measurements. We developed a self-contained wearable slider unit embedding two IMUs (LSM6DSV16X), a wideband accelerometer (IIS3DWB), and plantar pressure sensors in a commercially compatible curling slider. Measurements were conducted with two participants under four ice surface conditions (used pebble, no-pebble, extra-fine, and coarse-fine pebble) at an actual curling hall. The root mean square (RMS) amplitude of the DC-subtracted IIS3DWB Z-axis vibration signal distinguished no-pebble from pebbled ice but did not distinguish among the three pebbled conditions. The spectral centroid of the no-pebble condition was also biased toward higher frequencies, although participant-dependent effects remained among pebbled conditions. The slider acquired plantar-pressure variation during delivery, and an exploratory pooled analysis showed a pattern consistent with greater right/left balance variation in trials with greater MOCAP-derived lateral wobble. Because only two participants were tested and the lateral pressure channels reached the calibration limit, participant-specific validation requires a wider measurement range and additional participants. These results support the feasibility of simultaneous ice-contact vibration and plantar-pressure measurement during curling delivery.

Applied Sciences doi: 10.3390/app16147203

Authors: Md Sabbir Hossen Gobbi Ramasamy Marran Al Qwaid

The rapid expansion of electric vehicle (EV) adoption has increased the demand for efficient charging infrastructure and data-driven approaches for understanding charging behavior. Analyzing charging patterns and identifying abnormal charging sessions are essential for improving charging network reliability, infrastructure utilization, and operational efficiency. This study proposes a comprehensive machine learning framework for EV charging behavior analysis and anomaly detection using real-world charging session data collected from six charging bays. Four charging behavior indicators, namely energy consumption (Usage), charging duration (Duration), average charging output power (Average Output), and Energy Consumption Ratio (ECR), were extracted through a feature engineering process. K-Means clustering was employed to identify distinct user behavior groups, while Principal Component Analysis (PCA) was utilized to visualize cluster separability. Isolation Forest was subsequently applied to detect anomalous charging sessions and investigate abnormal charging behavior patterns. Statistical validation was conducted using Analysis of Variance (ANOVA), and Pearson correlation analysis was performed to examine relationships among charging features and anomaly occurrence. The results identified four distinct charging behavior clusters representing moderate users, regular users, inefficient users, and high-power users. Clustering validation achieved a silhouette score of 0.6086, while PCA retained 89.8% of the total variance using two principal components. An anomaly detection analysis revealed that inefficient charging behavior exhibited the highest anomaly occurrence, whereas regular users demonstrated highly consistent charging patterns. Analysis indicated that average charging output power and ECR were the most influential variables contributing to anomaly identification. ANOVA results confirmed statistically significant differences among all identified clusters (p < 0.001), while correlation analysis demonstrated a strong positive relationship between charging power and charging efficiency (r = 0.95). The anomaly detection framework achieved accuracy, precision, recall, and F1-score of 80.0%. The proposed framework provides a comprehensive approach for EV charging behavior characterization, anomaly detection, and charging infrastructure assessment. The findings can support charging network operators in improving charging efficiency, identifying abnormal charging activities, and enabling data-driven management of EV charging systems.

Applied Sciences doi: 10.3390/app16147200

Authors: Xin Liang Lihua Hu Liyuan Yu Kai Zhang Jiangcheng Feng

Water significantly modifies rock mechanical performance and fracture characteristics through water content and water distribution. Nevertheless, the evolution laws of rock mechanical properties and underlying fracture mechanisms under variable water conditions remain incompletely clarified. In this study, uniaxial compression tests were carried out on sandstone samples with diverse water immersion durations. Experimental observations reveal that the uniaxial compressive strength (UCS) and elastic modulus of sandstone follow negative exponential attenuation with prolonged immersion time, with maximum reductions of 50.1% and 25.6%, respectively. Under equivalent water contents, samples featuring dry exteriors and wet interiors possess lower strength than those with wet exteriors and dry interiors. A self-developed numerical code incorporating humidity diffusion effects was subsequently adopted to interpret water-controlled sandstone fracture behaviors. Numerical outputs verify that water-induced softening and heterogeneous water distribution exacerbate rock material heterogeneity and internal stress non-uniformity, triggering tensile microcracks along dry–wet interfaces. As the immersion duration rises, the rock failure mode transitions from shear-dominated mixed failure to tension-dominated failure, and finally reverts to shear-dominated mixed failure. Macroscopic rupture is predominantly governed by the gradual coalescence of tension-generated microcracks. This study offers a theoretical foundation to advance the understanding of water-triggered mechanical degradation and fracture mechanisms in sandstone.

Applied Sciences doi: 10.3390/app16147201

Authors: Edward Kozłowski Vytenis Surblys Deividas Navikas Bartosz Przysucha Jonas Matijošius

The present study explores the effect of incorrect tire mounting on vehicle braking performance with a specific focus on its interaction with the Anti-Lock Braking System (ABS). Experimental testing was conducted utilizing a passenger vehicle under controlled conditions according to ISO 21994. Two tire configurations were studied: a complying axle-wise configuration and a non-compliant asymmetric arrangement with the same tires mounted longitudinally on each side of the vehicle. ABS was switched on and off throughout the braking testing. The dynamics of the vehicle and its brake parameters were recorded and analyzed using nonlinear statistical modelling. The obtained coefficients of determination of the suggested nonlinear model were in the range of 0.8265 to 0.9938 and showed a satisfactory agreement with the experimental data. Chow-type statistical testing revealed significant differences across all tire designs and ABS operational modes tested (p < 0.001). Ljung–Box test indicated the lateral acceleration signals were significantly temporally dependent. The results indicate that the incorrect asymmetric tire mounting has a negative influence on the braking dynamics and this effect is more significant when the ABS is turned off. While ABS mitigates these effects to some degree, we still observe statistically significant disparities in performance. The results give empirically validated proof of the need for proper tire installation for the preservation of braking performance and vehicle safety and contribute to a better understanding of tire–road interaction and braking dynamics under asymmetric tire configurations.

Applied Sciences doi: 10.3390/app16147197

Authors: Jorge Hochstetter-Diez Marlene Negrier-Seguel Fernanda Gutiérrez-Gutiérrez Juan Lagos-Obando Claudio Espinoza-Navas Yuliana Puerta-Cruz

One of the factors that has steadily eroded the legitimacy of public institutions is the recurrence of corruption and irregularities in state management, especially in procurement processes, budget allocation, and public appointments. These phenomena have intensified public mistrust and highlight the urgent need for tools to strengthen transparency in the digital sphere. In this context, e-government platforms have become key mechanisms for promoting openness and accountability in public administration. However, ensuring transparency in electronic procedures remains challenging due to the lack of standardized auditing mechanisms and system interoperability. The objective of this study is to design, implement, and technically evaluate a software architecture for an electronic platform that operationalizes a maturity-based diagnostic model to assess transparency in electronic procedures within public organizations. The proposed platform enables real-time monitoring, centralized data management, automated reporting, and evidence-based transparency assessment. The architecture follows a structured design process aligned with ISO 25010 quality attributes, incorporating microservices, database replication, and load balancing to enhance system performance and security. Performance tests identified bottlenecks in authentication, query processing, and concurrent user load, leading to optimizations such as query indexing, caching, and message queue implementation. The results indicate that a well-structured software architecture enhances government transparency by ensuring auditable, traceable, and secure public data access. The platform provides a self-diagnostic tool for institutions to assess their transparency maturity based on open government principles.

Applied Sciences doi: 10.3390/app16147198

Authors: Tonglei Wang Zhengguang Chen Shitianyi Tan Yuandi Lin Yong Ma Shaoqi Wang

The presence of bubbles severely threatens the insulation safety of power transformers. Understanding bubble accumulation characteristics is essential for predicting potential insulation weaknesses. In this study, we propose a Euler–Lagrange two-phase flow model, combined with an experimental platform, to investigate the accumulation and growth characteristics of bubbles in typical oil channels. The continuous phase flow field is solved within the Eulerian framework, while discrete bubble trajectories are tracked using the Lagrangian approach, simultaneously accounting for drag, buoyancy, and friction forces. The results indicate that bubbles primarily accumulate on the top insulating paper surfaces within horizontal and corner oil channels. In horizontal channels, the accumulation rate correlates positively with flow velocity and bubble concentration; furthermore, paper overlapping gaps intercept sliding bubbles, promoting their coalescence into larger ones. Conversely, at channel corners, the accumulation length and rate decrease as flow velocity increases, with accumulation weakening significantly when the velocity exceeds 0.6 m/s. Additionally, bubble growth is highly sensitive to flow velocity and temperature: higher velocities inhibit bubble coalescence, whereas elevated temperatures accelerate both coalescence and growth. Ultimately, this research reveals the fundamental accumulation laws of bubbles in transformer oil channels, providing a reliable analytical tool and theoretical reference for assessing bubble-induced insulation risks.

Applied Sciences doi: 10.3390/app16147199

Authors: Lei Deng Longxiang Chen Haichao Feng Xiaotong Yan Ziwei Zhong Lingkai Zhu Huixiang Chen Kan Kan

The transition process from synchronous condenser pump (SCP) mode to pumping mode determines the response rapidity of the startup procedure and operational stability of pump-turbines; however, the complex gas–liquid interaction and transient hydraulic characteristics during this process remain insufficiently understood. To address this, this study develops a numerical framework for the SCP-to-pumping transition process, incorporating the full-passage system, a multiscale mesh strategy for coupling mainstream and clearance flow regions, and a gas–liquid two-phase flow model based on the Volume of Fluid (VOF) method. The reliability of the numerical model is verified through comparison with model experiments, demonstrating good agreement between simulations and experimental data. Based on the validated model, the transient evolution of hydraulic forces, pressure pulsations, and internal flow structures is systematically analyzed. Axial force analysis reveals a significant internal equilibrium; the crown bears a maximum instantaneous fluctuation of approximately 2800 kN. Conversely, the radial force is primarily dominated by blade hydraulic thrust (1294 kN), showing distinct anisotropic behavior. The runner blade channels and the upper draft tube region are identified as critical areas with intense pressure fluctuations, with peak-to-peak pressure amplitudes reaching 45~48 m and 54 m head, respectively. Furthermore, reducing the duration of the exhaust process constitutes the main strategy for accelerating the transition and mitigating prolonged high-amplitude force and pressure fluctuations. The findings provide new insights into the transient hydraulic mechanisms of SCP-to-pumping transitions and offer guidance for optimizing transition control strategies in pumped-storage units.

Applied Sciences doi: 10.3390/app16147196

Authors: Tinashe Crispen Garidzira William Tichaona Vambe

Brain tumors remain a major cause of mortality worldwide, where early and accurate diagnosis is critical for effective treatment planning. This study compares the performance of lightweight neural networks for brain tumor classification on the “Brain Tumor MRI Dataset” from Kaggle. The Cross-Industry Standard Process for Data Mining (CRISP-DM) methodology guided the experimental workflow, including data understanding, preprocessing, modelling, and evaluation. Two lightweight architectures, a Multilayer Perceptron (MLP) and LeNet, were trained and evaluated for comparison. Results show that LeNet outperformed the MLP, achieving test accuracy of 92.67% with a loss of 1.65, while the MLP attained 84.38% accuracy with a loss of 0.87. LeNet required fewer parameters which were 914,384 than the MLP which were 2,329,754. Experiments were conducted on an NVIDIA RTX 3000 Ada Generation Laptop GPU with 8 GB dedicated memory. Under this setting, LeNet completed training in 1 h 16 min 45.26 s, compared with 1 h 30 min 26.79 s for the MLP. While recent transfer-learning and deeper CNN studies on brain tumor MRI classification commonly report accuracies above 98%, they often rely on substantially heavier architectures; in contrast, the proposed LeNet achieved 93.33% accuracy with only 914,384 parameters and shorter training time, highlighting a practical trade-off between accuracy and efficiency. These findings demonstrate that lightweight convolutional architectures can achieve superior classification performance with lower computational cost, highlighting their suitability for resource-constrained medical imaging applications and edge-based healthcare systems.

Applied Sciences doi: 10.3390/app16147195

Authors: Maja Veršić Bratinčević Ivana Generalić Mekinić Marijana Popović Mira Radunić

Pomegranate juice is a rich source of bioactive compounds, while its quality and organoleptic properties are influenced by the balance between sugars and organic acids. This study comparatively evaluated sweet, sour, and sweet–sour blended pomegranate juices by examining their physicochemical properties, color characteristics, phenolic content and volatile profile. Total phenolics, anthocyanins, and tannins were determined spectrophotometrically, and individual phenolics by HPLC-UV/VIS, while volatile compounds were identified using SPME-GC/MS. Significant differences (p < 0.05) were observed among juice types for all measured physicochemical and bioactive parameters. The sweet juice showed the highest total soluble solids to titratable acidity ratio (TSS/TA = 20.15) and the lowest tannin content (6.71 mg KE/L). The sour juice exhibited the highest acidity (2.28%) and tannin concentration (39.64 mg KE/L), along with lower L* and a* values. The sweet–sour juice presented a balanced TSS/TA ratio (10.50) and higher L*, a*, and b* values. It also contained the highest levels of total phenolics (363.4 mg GAE/L, 10 and 12% higher than in sour and sweet juice, respectively) and anthocyanins (342.05 mg C3G/L, 25 and 65% higher than in sweet and sour juice, respectively). The volatile profile was dominated by (Z)-3-hexen-1-ol and 1-hexanol. Principal component analysis explained 91.9% of the total variance and clearly differentiated the samples according to acidity, sugar balance, and bioactive composition. These results highlight the potential of juice blending to enhance chemical composition and functional properties.

Applied Sciences doi: 10.3390/app16147193

Authors: Shin-Young Park Si-A Lee

Individuals with stroke commonly exhibit impaired gait ability, gait asymmetry, and reduced balance confidence, which limit functional mobility and participation in daily activities. This study aimed to investigate the effects of auditory feedback gait training using a smart walker on gait ability, symmetry, and balance confidence in individuals with stroke. Forty individuals with stroke were randomly assigned to either an experimental group (n = 20) or a control group (n = 20). The experimental group performed gait training using a smart walker equipped with an auditory feedback system, whereas the control group used an identical walker with the feedback function disabled. Both groups trained for 40 min per session, 5 times a week for 4 weeks. Outcomes were measured using the 10 m walk test, Timed Up and Go test, Dynamic Gait Index, Activities-specific Balance Confidence scale, and gait symmetry indices. The experimental group showed significant improvements in the Timed Up and Go test and Dynamic Gait Index compared to the control group (p < 0.05). Notably, step length asymmetry and double support asymmetry significantly decreased in the experimental group (p < 0.05), indicating improved gait quality. Auditory feedback-based smart gait training can aid in gait recovery and balance improvement through real-time self-regulation and may provide a differentiated active rehabilitation strategy for individuals with stroke.

Applied Sciences doi: 10.3390/app16147194

Authors: Ferda Tasci Adem Yagci Selda Daler Harlene Hatterman-Valenti Ozkan Kaya

Bud fertility is a major determinant of grapevine productivity and an important criterion for selecting appropriate pruning strategies, yet information on indigenous Turkish Vitis vinifera cultivars remain limited. This study evaluated the bud fertility characteristics of 22 indigenous grapevine cultivars grafted onto 1103 Paulsen rootstock to identify cultivar-specific pruning requirements. Bud fertility, shoot emergence, and cluster formation were assessed using single-node cuttings collected during winter dormancy and forced under controlled conditions. Marked differences in shoot emergence and cluster production were observed among cultivars and bud positions. In most cultivars, the highest fertility occurred in the middle portion of the cane, whereas Oglak Karası and Rumi Uzum exhibited increasing fertility toward distal buds, indicating a preference for long cane pruning. Based on fertility distribution patterns, the majority of cultivars were classified as suitable for short pruning, while several cultivars showed better adaptation to medium pruning. These results demonstrate substantial variation in bud fertility among indigenous Turkish grapevine cultivars and highlight the importance of cultivar-specific pruning strategies for improving yield potential. The findings provide baseline information for 22 indigenous Turkish cultivars and contribute to the development of more efficient pruning recommendations under local viticultural conditions.

Applied Sciences doi: 10.3390/app16147192

Authors: Dongchul Gu Dongmug Kang Yoon-Ji Kim Youn-Hyang Lee Jung-In Moon Sungwoo Park Jong Kyu Choi Moon Ki Jung Youngki Kim

Background and Objectives: Exoskeletons may be useful in preventing musculoskeletal disorders. This study aimed to evaluate the biomechanical and subjective effects of a passive shoulder exoskeleton during overhead assembly work in a real-world automotive manufacturing environment. Materials and Methods: Sixteen workers performing overhead assembly tasks participated in this study. Muscle activity, joint angles, and joint reaction forces were compared between the No Exo and Exo conditions during overhead work. Muscle activity was measured using surface electromyography (EMG), upper-body kinematics were captured using inertial measurement units (IMUs), and shoulder joint torques and reaction forces were estimated using the AnyBody musculoskeletal modeling system. Additionally, user satisfaction was assessed through a questionnaire. Results: Comparison of biomechanical outcomes between the No Exo and Exo conditions showed that the Exo condition significantly reduced anterior and lateral deltoid activities by approximately 15.2% and 11.1%, respectively, and reduced glenohumeral joint reaction force by 115.24 N (all p < 0.05). Participants also reported lower physical workload (p < 0.001). Conclusions: The study results suggest that wearing an exoskeleton during overhead work may reduce shoulder muscle loading and improve biomechanical indicators associated with shoulder loading.

Applied Sciences doi: 10.3390/app16147191

Authors: Agata Leońska-Duniec

Obesity is a chronic relapsing disease associated with excess adiposity, metabolic dysfunction, low-grade inflammation, and persistent epigenetic alterations. Changes in DNA methylation, histone modifications, chromatin remodeling, and non-coding RNAs may contribute to adipose tissue dysfunction, insulin resistance, impaired metabolic flexibility, and weight regain. This narrative review aimed to critically examine current evidence on the epigenetic effects of physical activity and epigenetically active supplements in obesity, with particular attention to their potential shared and complementary mechanisms. A focused literature search covering publications from January 2000 to May 2026 was conducted using scientific databases. The search strategy was constructed around four main thematic areas: obesity and metabolic dysfunction, physical activity and exercise interventions, bioactive supplementation, and epigenetic regulation. Current evidence indicates that physical activity is a strong epigenetic stimulus, particularly in adipose tissue and skeletal muscle. Exercise-induced epigenetic modifications may regulate biological pathways involved in mitochondrial function, inflammation, insulin sensitivity, and metabolic adaptation. In contrast, evidence for omega-3 fatty acids, vitamin D, polyphenols, and multi-ingredient anti-inflammatory supplementation is more heterogeneous and relies largely on blood-derived or systemic biomarkers. Integrated exercise–supplementation strategies are biologically plausible, but their obesity-specific epigenetic effects require confirmation in long-term, well-controlled human studies.

Applied Sciences doi: 10.3390/app16147190

Authors: Junyan Yi Chen Yu

Retinal nerve fiber layer defect (RNFLD) is a critical early indicator of glaucoma. Existing RNFLD detection methods underutilize clinical anatomical priors, struggle with subtle or diffuse lesions, and rely heavily on high-quality annotations. To address these issues, this paper proposes Inexact Position-Wise Sector Contrastive Learning (IPSCL) for RNFLD detection on retinal fundus images. IPSCL combines position guidance, inexact supervised learning based on a novel Radial Cup–Disc Ratio (RCDR), and sector-level contrastive learning to leverage anatomical correlations and enhance fine-grained feature discrimination. Experiments on the original RNFLD dataset and its extended version show that IPSCL outperforms state-of-the-art methods, achieving sector-level F1 score of 83.84% and image-level F1 score of 94.98%. The method effectively detects subtle and diffuse defects while reducing annotation dependence. This work provides a reliable and interpretable solution for clinical RNFLD detection.

Applied Sciences doi: 10.3390/app16147189

Authors: Seda Yıldırım Erkan Ülker Necdet Poyraz

Breast imaging reports are typically stored as unstructured free-text documents, which limits their use in clinical analytics, research, and downstream computational applications. These challenges are particularly pronounced in Turkish because of its agglutinative linguistic structure, orthographic variability, and the limited availability of standardized clinical corpora. This study presents a data curation framework for transforming heterogeneous real-world Turkish breast imaging reports into structured and machine-readable datasets, with a specific focus on the extraction and standardization of BI-RADS labels already recorded in routine clinical reports. The proposed pipeline integrates domain-specific preprocessing, text normalization, hybrid conclusion-section segmentation, multi-phase BI-RADS label extraction, duplicate and near-duplicate report handling, and modality-based separation within a transparent rule-guided workflow. The framework was applied to two real-world breast imaging datasets comprising 35,104 reports in Dataset 1 and 27,215 reports in Dataset 2 after overlap and duplicate control around the May 2023 transition period. The datasets included ultrasonography, mammography, and magnetic resonance imaging records. The framework achieved conclusion-section segmentation coverage rates of 96.71% and 99.94%, respectively, and BI-RADS label extraction coverage rates of 95.04% and 100.00%, respectively. These coverage values indicate the proportion of reports for which the rule-guided pipeline produced extractable outputs and should not be interpreted as precision, recall, F1-score, or accuracy against an independent gold-standard corpus. The curation process reduced report-level textual redundancy, improved structural consistency, and enabled systematic separation of report narratives from diagnostic assessment labels. Manual review of selected subsets was used as an internal plausibility check rather than as a substitute for independent radiologist-annotated gold-standard validation. By addressing the challenges of structuring free-text breast imaging reports in a low-resource language setting, this study provides a transparent and adaptable methodological basis for future clinical NLP, machine learning, and real-world healthcare analytics.

Applied Sciences doi: 10.3390/app16147188

Authors: Ziyang Jin Meiyan Li

Real-time order arrivals and stringent timeliness demands in e-commerce pose significant challenges to production-distribution coordinated scheduling in distributed manufacturing systems. This paper tackles the integrated production and distribution scheduling problem within a dynamic distributed hybrid flow shop. We formulate a mixed-integer linear programming model aimed at minimizing average order tardiness and total operational cost. To address dynamic uncertainties, we propose an improved decomposition-based multi-objective evolutionary algorithm (I-MOEA/D) operating within a rolling horizon framework. The algorithm integrates three critical components: an urgency-based emergency window management strategy that caps computational complexity, a 2-opt local search operator that boosts vehicle routing efficiency, and acceleration techniques—including order dictionary indexing, lightweight object replication, and distance caching—that guarantee real-time responsiveness. The window size is determined as Wmax = 50 through sensitivity analysis. Extensive experiments conducted on 100, 200, and 300-order scenarios with 30 independent random seeds demonstrate that I-MOEA/D markedly outperforms NSGA-II, MOALNS, and a reinforcement learning-driven hyper-heuristic (RL-HH). For 200 orders, I-MOEA/D reduces average tardiness by 18.5%, 41.0%, and 58.7% compared to NSGA-II, MOALNS, and RL-HH, respectively, while maintaining competitive cost performance (a cost gap of 16.99% relative to the static lower bound). The IGD metric achieves 0.0031, confirming good convergence and diversity. Acceleration techniques cut computation time by 56% without sacrificing solution quality. The algorithm scales efficiently with problem size. These results validate that I-MOEA/D delivers effective real-time decision support for dynamic distributed production-distribution systems.

Applied Sciences doi: 10.3390/app16147187

Authors: Giorgia Stornelli Gennaro Musella Giuseppe Balice Daniela Morari Gioele Gioco Carlo Lajolo Monique Stoffels Gary P. Moran Rossella Grande Vito Carlo Alberto Caponio Vittoria Perrotti

This systematic review aimed to evaluate the effectiveness of indirect cold atmospheric plasma (CAP) in disrupting oral and skin microbial biofilms and to investigate whether the reported effects are mainly attributable to acidic pH, plasma-activated liquids, or specific devices and treatment settings. The review followed PRISMA guidelines and was registered on the Open Science Framework. A comprehensive literature search was conducted across major electronic databases. Studies evaluating indirect CAP on in vitro oral and skin microbial biofilms were included. Comparisons were made with untreated controls, non-activated liquids, and conventional antimicrobials. Outcomes were evaluated based on reductions in biofilm viability or biomass using quantitative assays (e.g., CFU, XTT, and Crystal Violet). Thirteen in vitro studies were included. CFU-based outcomes were reported in 11 of 13 studies (84.61%), and all of these studies showed reductions exceeding 5–6 log under specific conditions. Antimicrobial efficacy varied according to exposure time, microbial species, plasma parameters, and treatment settings, while physicochemical analyses indicated that acidic conditions and plasma-activated liquids play a key role. Indirect CAP appears to be a promising non-invasive antimicrobial strategy for managing oral and skin biofilms. However, variability among devices and settings, together with the absence of clinical data, underscores the need for further standardized studies and clinical investigations.

Applied Sciences doi: 10.3390/app16147186

Authors: Zhongkai Huang Yiqun Wu Li Guo Nianchen Zeng Haocheng Li Qiang Wang Guochen Zhao

As critical infrastructure supporting sustainable urban development globally, the seismic resilience evaluation of tunnels is paramount to metropolitan safety. This paper proposes a probabilistic seismic resilience analysis framework for tunnel structures in soft soils. By integrating vulnerability models and functionality functions, resilience metrics for shallow, medium-depth, and deep tunnels are obtained, revealing how these metrics evolve with seismic intensity. The research further investigates the impact of different vulnerability models on tunnel resilience, treating occurrence probabilities of different damage states as random variables. Resilience metrics incorporating multiple vulnerability models are derived based on the Dirichlet distribution. The results of the study show that tunnel resilience metrics increase progressively with greater burial depth. At PGA = 0.8 g, for example, the resilience metrics calculated using a single model are 0.907, 0.962, and 0.985 for shallow, medium-depth, and deep tunnels, respectively. When accounting for multiple vulnerability models, the composite resilience metric tends to align with results from models carrying higher weights. This work establishes a scientific decision-making basis for optimizing seismic tunnel design, post-disaster functional recovery, and urban infrastructure resilience management.

Applied Sciences doi: 10.3390/app16147185

Authors: Gabriela Kowalska Kamila Kulbat-Warycha Aleksandra Bilczewska

This study investigated the valorisation potential of dried and frozen pomace obtained as by-products from the processing of Rosa rugosa fruits. The antioxidant activity of the pomace was assessed using electron paramagnetic resonance (EPR) spectroscopy with the TEMPO radical. Total phenolic content was determined using the Folin–Ciocalteu method, and individual phenolic compounds were analysed by UHPLC. Frozen pomace exhibited significantly higher antioxidant activity than dried pomace (2.76 vs. 1.64 mmol TE/100 g d.m.) and contained greater amounts of ellagic acid (19.29 vs. 13.08 mg/100 g d.m.), gallic acid (17.05 vs. 5.07 mg/100 g d.m.), procyanidin B2 (14.92 vs. 11.88 mg/100 g d.m.), and catechin (2.84 vs. 1.12 mg/100 g d.m.). To assess technological applicability, Rosa rugosa pomace was incorporated into an agar-based model fruit gel system at levels of 3%, 5%, and 10% (w/w). The enriched gels were evaluated in terms of antioxidant activity, total phenolic content, and physicochemical parameters, including colour, water activity, and moisture content. The incorporation of pomace into gel systems was found to significantly increase the total phenolic content, raising it from 48.64 mg GAE/100 g in the control sample to 267.56 mg GAE/100 g in gels containing 10% pomace. Antioxidant activity also increased, rising from 335.79 to 1264.37 μmol TE/100 g. The addition of pomace was also found to reduce the moisture content from 71.02% to 64.17%, while the water activity remained stable (0.821–0.840). Sensory evaluation indicated that gels containing 3–5% pomace achieved the highest consumer acceptance. These results demonstrate that Rosa rugosa pomace is a valuable and sustainable source of functional ingredients that can enhance the bioactive properties of food products while maintaining desirable physicochemical characteristics.

Applied Sciences doi: 10.3390/app16147184

Authors: Carlos A. León Omar Sánchez Luis A. Cristancho Karen Castañeda

During the purchase or renovation of housing, a limitation exists in the finish design process for construction projects, as buyers cannot switch between various finish options for their apartments. Instead, they must depend on visual graphics or physical samples to imagine the appearance of these finishes in their home. Combining Building Information Modeling (BIM) with Virtual Reality (VR) presents a promising way to overcome this challenge. This study introduces a BIM-VR tool and a methodological framework designed to support decision-making when selecting finishes, with a focus on cost, waste reduction, and overall project perception. The research method adapts the Design Science Research (DSR) technique for developing an artifact through four key steps: (1) identifying the problem the artifact aims to solve; (2) defining the artifact’s solution goals; (3) designing and developing the artifact; (4) evaluating and fine-tuning the artifact. The tool developed allows for immersive walkthroughs within a digital replica of a residential building, along with the ability to customize and measure finishes. The artifact was assessed through a questionnaire administered to 53 potential residential end users, including potential apartment buyers and individuals interested in residential customization processes. The findings indicate a mean perceived-improvement score of 87.7 out of 100 for the BIM-VR method, reflecting participants’ favorable comparison between the BIM-VR and traditional finish-selection approaches. This innovation benefits the construction industry by enhancing stakeholder communication and supporting a faster design process in building finishes selection.

Applied Sciences doi: 10.3390/app16147183

Authors: Marek Mizerák Jozef Trojan Peter Trebuňa Marek Kliment Štefan Mozol

This paper presents the design, implementation, and validation of a mobile Real-Time Location System (RTLS) based on Ultra-Wideband (UWB) technology for acquiring and transforming production and logistics data in an industrial environment. The aim is to obtain accurate real-time information on the movement of workers and material flows to support data-driven optimization within the Industry 4.0 framework. The proposed solution introduces a mobile RTLS architecture enabling flexible deployment without permanent infrastructure changes. The system was experimentally validated in a manufacturing enterprise, where UWB anchors and wearable tags were used to monitor six operators and handling equipment during a 12 h production shift. The collected data were analyzed using trajectory mapping, heatmaps, and worker activity analysis to identify inefficiencies in the existing production layout. The identified bottlenecks and unnecessary worker movements were subsequently used to redesign the production layout in Tecnomatix Process Simulate. Simulation results demonstrated a 51.9% reduction in worker travel distance, while transportation and waiting activities, which accounted for approximately 24% of the original production lead time, were significantly reduced in the proposed layout. The results confirm that UWB-based RTLS provides reliable input data for simulation-driven manufacturing layout redesign and supports objective decision-making in digital manufacturing environments.

Applied Sciences doi: 10.3390/app16147182

Authors: Marta Bochniak Monika Wereńska Weronika Wójcik

The quality of edible oils is determined not only by the raw material, but also by the applied processing technology. Rapeseed oil is widely valued for its favorable fatty acid profile, however, the effects of combined pressing conditions and low-temperature bleaching on its lipid profile and calculated nutritional quality indices remain insufficiently characterized. This study aimed to evaluate the effects of oil production method and bleaching earth addition on the fatty acid profile, nutritional quality indices, and oxidizability-related indices of rapeseed oil from an applied food processing perspective. Rapeseed oil samples were obtained from a commercial producer using cold pressing, hot pressing, and mixed two-stage pressing. The oils were subjected to low-temperature bleaching with 1–5% bleaching earth. The fatty acid profile was determined by gas chromatography, and selected fatty acid-based indices were calculated and grouped into three categories: fatty acid class ratios, nutritional quality indices, and oxidizability-related indices. These included ratios describing the balance between omega-6 s omega-3 polyunsaturated fatty acids (Σ PUFA n-6/n-3), unsaturated and saturated fatty acids (Σ UFA/Σ SFA), and polyunsaturated and saturated fatty acids (Σ PUFA/Σ SFA). Health-oriented indices such as the Desirable Fatty Acids/Undesirable Fatty Acids ratio (Σ DFA/Σ OFA), hypocholesterolemic/hypercholesterolemic ratio (h/H), Nutritive Value Index (NVI), Atherogenic Index (AI), and Thrombogenic Index (TI), and oxidizability-related indices including the Unsaturation Index (UI), Peroxidizability Index (PI), Oxidizability Index (OI), calculated Oxidizability Value (Cox), and Oxidative Stability Index (OS). The results showed that the production method significantly influenced saturated fatty acid content and selected fatty acid-based nutritional quality indices. Cold-pressed oil was characterized by the lowest saturated fatty acid content and the most beneficial values of selected nutritional quality indices. Hot-pressed oil showed higher saturated fatty acid content and less favorable values for selected indices, whereas mixed oil generally showed values between those of cold-pressed and hot-pressed oils. However, differences in the main unsaturated fatty acid fractions were limited and, for most monounsaturated and polyunsaturated fatty acid parameters, were not statistically significant. Across all processing variants, the nutritional quality indices remained advantageous, with omega-6/omega-3 polyunsaturated fatty acid ratio values of 2.1–2.4, atherogenic index values of approximately 0.04–0.05, and thrombogenic index values of approximately 0.08–0.09. Although significant interactions between production method and bleaching earth addition were observed for selected oxidizability-related indices, the magnitude of bleaching-induced changes was limited. Overall, the findings indicate that low-temperature bleaching may be applied as a mild processing step without substantially compromising the fatty acid-based quality characteristics of rapeseed oil. The study provides practical insight into the relationship between pressing technology, bleaching conditions, and lipid quality assessment in industrial rapeseed oil production.

Applied Sciences doi: 10.3390/app16147181

Authors: Nurdan Bilgin

This paper presents a hybrid control framework that combines State-Dependent Riccati Equation (SDRE) control with Model Reference Adaptive Control (MRAC) for nonlinear systems with parametric uncertainties. The SDRE component provides an optimal state-dependent feedback structure, while MRAC enhances robustness by compensating for model uncertainties through adaptive gain adjustment. To address the strong dependence of controller performance on design parameters, Bayesian Optimization (BO) is employed for systematic and sample-efficient tuning. The proposed approach is evaluated on a nonlinear two-link robotic manipulator under varying initial conditions, external disturbances, and parameter uncertainties. The results show that, although BO improves the SDRE-only controller, its standalone effectiveness remains limited. In contrast, the integration of the adaptive mechanism leads to significant improvements in both tracking accuracy and control efficiency while preserving overall system behavior. The analysis further reveals inherent trade-offs between control effort and tracking performance, highlighting the role of cost function design. Overall, the SDRE–MRAC–BO framework provides a practical and computationally efficient solution for controlling uncertain nonlinear systems and indicates potential for future real-time self-tuning applications.

Applied Sciences doi: 10.3390/app16147180

Authors: İlayda Pehlivan Yağmur Atakav Merve Badem Şeyda Kanbolat

Wound dressings play a crucial role in promoting tissue regeneration while protecting damaged tissue from microbial infections. The aim of this study is to produce antimicrobial PVA-chitosan membranes for wound dressing applications using Stevia rebaudiana Bertoni leaf extract (stevia) as a natural bioactive agent. Although widely recognized as a natural sweetener, stevia contains a high concentration of diterpene glycosides, particularly stevioside and rebaudioside A, together with phenolic compounds that contribute to its biological activities. Therefore, stevia was first evaluated against selected microorganisms and demonstrated effective antimicrobial activity. PVA-chitosan (P/Ch) membranes were prepared by solvent casting method by incorporating different amounts of stevia (P/Ch, P/Ch@20, P/Ch@40, and P/Ch@60). The antimicrobial activity of stevia was also successfully maintained in P/Ch membranes. Good antibacterial activity was observed against Bacillus subtilis and Escherichia coli, whereas the antifungal activity against Candida albicans was comparatively modest. Contact surface analysis showed complete bactericidal activity in Gram-negative and spore-forming bacteria. The cytocompatibility of the membranes was investigated by MTT assay. All formulations maintained a high cell viability (>80%) while the P/Ch@40 membrane increased the metabolic activity to above 100% at higher extract concentrations. In conclusion, stevia-incorporated membranes exhibited high antimicrobial activity, acceptable characteristic properties, and good cytocompatibility, suggesting that they are promising alternative biomaterials for tissue engineering applications.

Applied Sciences doi: 10.3390/app16147179

Authors: Hongkun Chen Xue Hu Qiang Xiu Lixin Zhang Min Lin

With the rapid expansion of ground-mounted photovoltaic power stations, the wind-induced response and reinforcement of existing photovoltaic support structures have become increasingly important. This study investigates an existing ground-mounted photovoltaic support in Yuli County, Xinjiang, China. A fluctuating wind model based on the Davenport spectrum was established, and a one-way fluid–structure interaction method was used to analyze wind pressure distribution, displacement response, and weak components under different wind direction angles. The results indicate that the 0° wind direction is the most unfavorable condition, and that the purlins and their connection regions are the dominant weak components under fluctuating wind loads. To address the reinforcement needs of existing in-service photovoltaic power stations, a low-disturbance in situ reinforcement method was proposed by inserting internal auxiliary members into the original purlins, which can be implemented without dismantling photovoltaic modules, purlins, or existing support components. Parameter analysis showed that a 2.5 mm-thick internal reinforcement member provided an effective balance between stiffness improvement and material use. Array-scale verification demonstrated that the maximum displacement of the first-row support decreased from 5.9738 mm to 3.9377 mm after reinforcement, corresponding to a reduction of approximately 34.1%. The proposed method provides a practical, low-disturbance reinforcement strategy for existing photovoltaic support structures.

Applied Sciences doi: 10.3390/app16147178

Authors: Zhihao Kong Rui Ye Jialin Wang Mingxiong Ou

Air-assist atomization exploits high-velocity gas streams to shear liquid phases, achieving fine droplets at supply pressures significantly lower than those required by conventional pressure atomization. This technology is extensively deployed across diverse sectors, including pesticide spraying, food spray drying, sanitization/disinfection, and combustion atomization. Although specific operational fields dictate vastly contrasting droplet size distributions, velocities, and deposition uniformities—with target diameters spanning from <50 μm to >200 μm—their foundational atomization mechanisms remain inherently unified, governed primarily by toroidal vortex-induced primary breakup and Kelvin–Helmholtz/Rayleigh–Taylor (KH–RT) or Taylor Analogy Breakup (TAB) secondary breakup. This review systematically parses the atomization mechanisms, critical performance metrics, numerical simulation frameworks, and experimental characterization methodologies of air-assist nozzles. Crucially, from a novel “cross-disciplinary custom design” perspective, we contrast the optimal droplet parameter windows across the agricultural, food, sanitization, and combustion sectors (e.g., electrostatic plant protection yields a 203–1350% increase in abaxial leaf deposition; mine wind-assisted misting achieves a >90% collection efficiency for PM10 dust; and SCR air-assisted injectors reduce the Sauter Mean Diameter (SMD) to 25 μm). Synthesized insights reveal that segmented VOF-to-DPM transition frameworks, corner-vortex-induced breakup theories, and closed-loop adaptive control architectures possess substantial cross-domain migration value. Current bottleneck challenges are highlighted, including the mesh dependency of droplet collision–coalescence models in dense spray regimes, the absence of robust atomization constitutive formulations for non-Newtonian fluids, and the ongoing paradigm shift from open-loop presetting to intelligent closed-loop regulation. This work establishes a comprehensive theoretical foundation and technical roadmap for cross-disciplinary integration and next-generation smart nozzle design in air-assist atomization.

Applied Sciences doi: 10.3390/app16147177

Authors: Wenxuan Xu Yong Wang Hairui Zhang

Road crack detection is an important part of the operation and maintenance of intelligent transportation infrastructure. However, the traditional convolution method has the problem of geometric mismatch when dealing with the slender linear structure of cracks, and the multi-scale adaptability and anti-interference ability of the detection model under a complex road background are still insufficient. Aiming at key problems such as the difficulty of linear feature extraction, poor multi-scale adaptability, and complex background interference, this paper proposes a road crack detection algorithm based on an improved RT-DETR multi-dimensional feature fusion method. This method improves the detection performance by introducing three core innovative modules. Firstly, a lightweight directional decoupled dynamic convolution (D3Conv) is designed, which makes the convolution kernel fit the crack direction through direction prediction and adaptive sampling, so as to improve the recall rate by 0.012 (from 0.647 to 0.659) and reduce the computational burden. Secondly, a multi-scale cross-attention enhancement module (MCAA) was proposed to fuse multi-scale convolution and direction-aware strip convolution, and the dual attention mechanism was combined to enhance the perception of crack morphology and scale. Furthermore, a context-guided feature reconstruction (CGFR) module was constructed, which effectively aggregated global semantics and local details through dynamic feature selection and a multi-branch refining mechanism to improve the accuracy of boundary location. Experiments on the public dataset SVRDD2024 show that the proposed algorithm achieves an mAP@50 of 0.724, which is 2.2% higher than the baseline RT-DETR-R18 in terms of mAP@50. Especially, the detection performance is significantly improved for small cracks and in complex environments. It provides a reliable and efficient solution for the automatic inspection of intelligent transportation infrastructure.

Applied Sciences doi: 10.3390/app16147176

Authors: Jianping Xiong Yangpeng Zhang Zhiming Yan Jinsong Pang Zhiyong Liu Youneng Liu Jiming Yang

The accurate and nondestructive quantification of subgrade moisture content is a core demand for highway construction quality control and long-term performance maintenance. In order to study the response relationship between subgrade moisture content and ground-penetrating radar (GPR) signal eigenvalues, this study constructs the volumetric moisture content–dielectric constant relationship of Guangxi high-plasticity clay and carries out gprMax forward numerical simulations. Fourteen radar signal eigenvalues are extracted from preprocessed signals via time-domain waveform analysis, Hilbert transform analysis, and power spectral density analysis. Seven key eigenvalues are screened out through Pearson correlation coefficient-based dimensionality reduction. Three machine learning algorithms—artificial neural network (ANN), random forest (RF), and light gradient boosting machine (LightGBM)—are adopted to optimize the subgrade moisture-content inversion model, which is finally validated through indoor model box tests and field subgrade tests. The results show that: (1) The linear fitting formula is the most suitable for describing the volumetric moisture content–dielectric constant relationship of Guangxi clay, with a coefficient of determination (R2) of 0.979 and a mean absolute error (MAE) of 0.31. (2) The feature matrix after dimensionality reduction effectively alleviates the degradation of model generalization ability and interpretability. (3) The LightGBM model achieves the highest prediction accuracy for clay volumetric moisture content, with an R2 of 0.99926 and an MAE of 0.172%. (4) For gravimetric moisture-content inversion, the maximum relative error is 1.6% in the indoor model box test and 1.7% in the field test, both within the 2% tolerance of engineering requirements. This study verifies the feasibility of the proposed subgrade moisture-content identification method based on GPR signal eigenvalues. The proposed method provides an efficient technical path for the large-area and nondestructive detection of subgrade moisture and has promising application prospects in subgrade construction quality acceptance, daily maintenance monitoring and hidden disease early warning.

Applied Sciences doi: 10.3390/app16147175

Authors: Milena Kostović Zorica Vukadinović Zoran Gligorić Miloš Gligorić

Wet grinding is an important operation in the technological process of ceramic tile production. The properties of the slurry obtained from grinding (slip) are conditioned by the raw materials (the type and characteristics of raw material in mixture, recipes for mixture), and by the operating parameters in grinding (technical characteristics of mill, type of grinding system, mill charge, grinding media body, grinding time, etc.). The optimal selection of these influential parameters results in satisfactory properties of slip, i.e., in efficient grinding as process operation, and, consequently, in smooth and efficient realisation of subsequent operations in the process, particularly spray drying. At the end of the technological process, the final goal is to obtain a ceramic tile of satisfactory quality. Multi-criteria decision-making (MCDM) is an increasingly applied tool for selecting optimal technological parameters for the purpose of optimisation, problem solving and improvement of technological processes. This paper presents the application of the symmetry point of criterion—ranking alternatives by perimeter similarity (SPC-RAPS) as an MCDM hybrid method for the selection of optimal technological parameters in grinding in the ceramic tile production process. The ranking and selection of alternatives (raw materials, grinding balls and grinding time) were performed according to various criteria. In addition to the technological parameters related to the characteristics of the products from the grinding (slip), and to the technical characteristics of the final product (ceramic tiles), the criteria also included economic parameters (the market price of raw material and specific energy consumption in grinding). The developed mathematical model enabled the selection of the best alternative as a solution for this problem.

Applied Sciences doi: 10.3390/app16147174

Authors: Gabriel de Moura Costa Gonçalo Figueira António Paulo Moreira Marcelo R. Petry

Matrix manufacturing requires close coordination between collaborative workcells, mobile robots, and battery management resources to support the execution of heterogeneous human-robot operations in reconfigurable production environments. This paper presents a cyber-physical robotic coordination framework for human-robot teams deployed in an industrial matrix manufacturing system, integrating a collaborative workstation, a fleet of mobile programmable cobots, and an automatic battery changer through ROS/OPC UA communication. The framework coordinates task execution, intra-logistics, and energy management through a decision layer that assigns operations to human and robotic agents, relocates idle mobile robots, and triggers battery swaps. Three coordination modules—a Battery Management Module, a Task Allocation Module, and a Robot Relocation Module—implement this pipeline by computing feasible execution plans at each scheduling cycle, accounting for human and robot capabilities, workstation availability, transport times, and battery state. The approach is validated on a deployed industrial matrix manufacturing platform through a disassembly task comprising human-only, robot-only, and human–robot collaborative operations, demonstrating the feasibility of coordinating heterogeneous robotic and human resources in a physical reconfigurable manufacturing environment.

Applied Sciences doi: 10.3390/app16147173

Authors: Ahmed Abed Gatea Al-Shammary Jesús Fernández-Gálvez Andrés Caballero-Calvo

This study evaluated the individual and interactive effects of straw management, tillage systems, and wheat growth stages on surface energy balance (SEB) dynamics in a semi-arid wheat production system, with particular focus on net radiation (Rn), sensible heat flux (H), latent heat flux (LE), Bowen ratio (β), and energy partitioning (EP). A field experiment was conducted during the 2022–2023 growing season using a split–split plot design with two straw management treatments, four tillage systems, and three growth stages. Surface energy balance components were estimated through field-based micrometeorological measurements. Data were analysed using ANOVA, variance partitioning analysis, and Pearson correlation analysis. All experimental factors significantly affected SEB components, although growth stage represented the dominant source of variability, accounting for 42–58% of total variance. Flowering stage consistently promoted the highest LE values and the lowest β and EP values, indicating enhanced evaporative cooling during maximum crop development. Conservation-oriented tillage systems substantially modified thermal partitioning, with no-tillage (NT) significantly increasing LE and reducing H relative to conventional tillage (CT). The combination of straw incorporation and NT during flowering (IS + NT + S2) produced the highest LE value (129.15 W m−2) and one of the lowest H values (18.35 W m−2). Bowen ratio progressively decreased from CT (8.57) to NT (1.44), confirming a shift from sensible to latent heat exchange under conservation-oriented management. Crop phenology and conservation-oriented soil management jointly regulated thermal partitioning and evaporative cooling in semi-arid wheat systems. NT combined with straw incorporation substantially enhanced latent heat exchange while reducing sensible heating, particularly during flowering. This study provides novel field-based evidence regarding the combined influence of straw management, tillage systems, and wheat phenology on SEB dynamics under semi-arid conditions, contributing to improved understanding of land–atmosphere interactions and climate-adaptive agricultural management strategies.

Applied Sciences doi: 10.3390/app16147172

Authors: Xun Lu Mingyu Guo Ruisheng Diao Peng Rao

With the fast-increasing integration of offshore wind power, coordinating generation resources with demand-side flexibility has become essential for secure and efficient power system operation. This paper proposes a coordinated dispatch approach considering the demand response (DR) of load aggregators (LAs). A bi-level optimization model is developed, where the upper level describes system operation based on optimal power flow, and the lower level captures the demand response behavior of LAs, including curtailable and transferable loads. To improve computational tractability, the lower-level problem is reformulated using Karush–Kuhn–Tucker (KKT) conditions, and the original bi-level optimization problem is transformed into a single-level mixed-integer optimization model. Case studies are conducted on a modified IEEE 33-bus system with offshore wind integration and distributed LAs. The results show that the proposed approach effectively reshapes load profiles through peak shaving and load shifting, resulting in smoother system load characteristics. Meanwhile, the coordination between load response and renewable generation improves the accommodation of offshore wind power. It also reduces wind curtailment and decreases grid import. Compared with conventional dispatch without DR, the proposed method enhances system flexibility and provides an effective operational scheme for power systems with high renewable penetration.

Applied Sciences doi: 10.3390/app16147171

Authors: Joseph L. Hagan

Background: The neonatal intensive care unit (NICU) is one of medicine’s most data-rich Internet of Things (IoT) environments, and monitoring digital twins (DTs) that predict clinical events from continuously acquired physiologic streams are a leading proposed application. Methodological standards for validating the prediction models that underlie such DTs remain poorly defined. A monitoring DT can be no more trustworthy than the validation of the prediction models it contains. NICU physiologic data are characterized by repeated within-subject measurements and clustering by infant and center, structural features that violate the independence assumptions of standard prediction model validation methods. Methods: This methodological review examines validation strategies for clinical prediction models with repeated-measures predictors, with emphasis on cross-validation approaches that preserve validity under within-subject dependence, calibration-aware model evaluation, and feature selection stability. Results: Several methodological gaps are identified across the monitoring DT development pipeline. Recent simulation evidence quantifying the optimism produced by naive cross-validation in this setting is summarized and positioned as the foundation for the validation strategies recommended here. Conclusions: The NICU DT Validation Framework (NICU-DTVF) is proposed, articulating the validation conditions a trustworthy monitoring DT should satisfy, at a level of methodological maturity that is stated explicitly for each component. An open-source R package (cawCV) implementing cluster-aware validation as one component of that framework is under development. The validation principles described are general and extend to clustered, repeated-measures prediction problems beyond the neonatal setting.

Applied Sciences doi: 10.3390/app16147170

Authors: Jesus-Angel del-Hoyo-Gabaldon Antonio Garcia-Cabot Eva Garcia-Lopez David de-Fitero-Dominguez

Fault Localization (FL) has evolved with the emergence of the Transformer architecture. Traditionally, FL relied on generating a ranking of suspiciousness scores for each code unit using information obtained during program execution from a set of test cases. However, Language Models trained with source code have enabled approaches that avoid the need for test suites, which are, sometimes, difficult or time-consuming to obtain. In this work, we present a refined training strategy tested on DeBERTa-v3-base, leveraging an encoder-based approach, to address two key challenges in FL: (1) the accurate classification of lines of code according to their suspiciousness scores, and (2) the class imbalance inherent in this type of problem. To this end, we extract the last hidden state corresponding to the embeddings of the new line tokens to generate a suspiciousness score for each line. This score then classifies each line as defective or not. The evaluation of our approach on the ConDefects dataset, repeated over ten random seeds, shows promising results compared to existing approaches, with ROC AUC gains up to 0.27 in cross-dataset comparisons. Thereby, we offer a competitive baseline for future test-free fault localization research while using language models that are approximately 200 times smaller than those used in prior work.

Applied Sciences doi: 10.3390/app16147168

Authors: Javier V. Juan Rubén Martínez Eduardo Iáñez Mario Ortiz Jesús Tornero José M. Azorín

(1) Background: Motor imagery (MI) analysis via electroencephalography (EEG) is crucial for Brain–Machine Interfaces (BMIs) and neurorehabilitation, though inherent EEG noise challenges robust decoding algorithm development for assistive devices aiding patient recuperation. (2) Methods: This study introduces a deep learning pipeline utilizing the IFNet convolutional neural network with frequency filtering and two scaling methods to decode pedaling MI. To mitigate catastrophic forgetting during transfer learning, a rotative Elastic Weight Consolidation (rEWC)-based weight adjustment procedure was incorporated. The pipeline was validated with 23 non-disabled participants across four sessions each, evaluating four approaches via pseudo-online (open-loop) and real-time online (closed-loop) analyses. (3) Results: The decoding accuracy for MI exceeded 80% (up to 95% individually) in selected pseudo-online approaches, while real-time online classification averaged 70% accuracy with a slight bias towards the relax task, confirming statistical significance for both scenarios. (4) Conclusions: These findings demonstrate that mitigating catastrophic forgetting in transfer learning within deep learning models and applying proper scaling are crucial for optimal performance. Consequently, the proposed deep learning pipeline holds significant promise for effective, adaptive EEG-based neurorehabilitation systems.

Applied Sciences doi: 10.3390/app16147169

Authors: Junzi Zhang Jiaxin He Le Zou Linying Li Bin Zhang

To enhance the atomization and evaporation performance of evaporative tubes, this study employs an optimization design methodology based on a few-sample surrogate model. By modifying the geometric parameters of the evaporative tube, we investigated how structural parameters affect the Sauter Mean Diameter and evaporation rate. The optimization process consisted of 100 initial CFD samples and five rounds of iterative parameter additions, with six additional CFD re-computation samples added per round, resulting in a total of 130 CFD evaluations for surrogate model updates and optimization validation. A baseline configuration was used for relative improvement rate calculations and excluded from the optimization sample count. After five rounds of optimization, the Sauter Mean Diameter surrogate model achieved a maximum error of 4.80%, a mean error of 2.74%, and an R2 modeling score of 0.962 in the fifth iteration. The evaporation rate surrogate model demonstrated a maximum error of 14.27%, a mean error of 6.59%, and an R2 score of 0.988. Based on statistics from 30 CFD re-computation samples generated across five iterations, the SMD showed an average improvement of 24.26%, and the evaporation rate improved by 234.35%, demonstrating that this few-sample optimization approach effectively identifies optimal design solutions with limited CFD evaluation resources.

Applied Sciences doi: 10.3390/app16147166

Authors: Diego Vergara Eduardo García-Sardón Jorge Rafael Díaz-Dumont Pablo Fernández-Arias Antonio del Bosque

This study presents a bibliometric review of Geographic Information Systems (GIS) applications in Industrial Engineering, based on a literature search covering the period 2010–2025. Scopus and Web of Science yielded 2137 initial records, of which 1702 came from Scopus and 435 from Web of Science. After removing duplicates and applying exclusion criteria based on document type, language, and publication period, the final dataset consisted of 247 peer-reviewed journal articles. The analysis was conducted using Bibliometrix 5.3.0 in R 4.6.0, focusing on scientific production, source distribution, authorship patterns, country and institutional contributions, citation impact, and thematic evolution. The results show a sustained increase in scientific production on GIS applications in Industrial Engineering, especially after 2018, with recurring topics related to supply chain management, logistics, facility location, optimization, spatial analysis, sustainability, pollution, and environmental impact. The thematic structure indicates that GIS, supply chain management, and optimization management form the basic themes of the field, while impact appears as a motor theme. Overall, the findings show that GIS has evolved into a strategic tool for operational optimization, sustainable planning, and data-driven industrial decision-making.

Applied Sciences doi: 10.3390/app16147167

Authors: Emma Barelles-Vicente Maria Eugenia Torner-Feltrer Jaime Llinares Millán Carolina Aparicio-Fernández Daniela Besana

Building facades are highly exposed envelope components whose degradation affects durability, habitability, urban image, and maintenance planning. Several studies address facade anomalies and service-life prediction. However, a need remains for integrated, reproducible procedures that combine visual inspection, taxonomic classification, and statistical analysis within a single framework. This research develops and validates a standardized methodology for the assessment of facade pathologies in urban buildings. The proposed framework is structured into sequential phases: documentary research, systematic visual inspection, photographic recording, damage classification, facade mapping, standardized inspection sheets, database generation, statistical analysis, and cluster-based interpretation of damage patterns. The methodology was validated through an urban case study in Valencia, Spain, where 168 building facades were inspected and 1600 damage were identified, classified, mapped, and digitized. The collected data were analysed according to building age, environmental exposure, and affected facade units. Soiling due to differential washing was the most frequent damage type, with 295 cases. Buildings constructed between 1930 and 1960 concentrated the highest number of recorded cases (639), while the wall area near ground level was the most affected facade unit (499 cases). K-means analysis retained a three-cluster solution, with a Silhouette Score of 0.65 and a BSS/TSS ratio of 86.13%. In addition, K-means cluster analysis was applied to classify damage types according to their frequency after Z-score standardization and validation through Silhouette Score and BSS/TSS metrics. The results demonstrate that the proposed framework enables homogeneous data collection, reproducible classification, and diagnostic interpretation of recurrent facade damage. Beyond the specific findings of the Valencia case study, the main contribution of this work is the development of a transferable assessment framework that can support preventive maintenance protocols, inspection planning, and evidence-based conservation strategies in other urban contexts.

Applied Sciences doi: 10.3390/app16147165

Authors: Mingli Ge Runhua Zhang Jinshan Wang Liping Tong Xiaowei Zhou Zhu Liu Weidong Meng Jianwen Gao Hui Li Yang Ren

Yttrium iron garnet (Y3Fe5O12, YIG) films were deposited on gadolinium gallium garnet (Gd3Ga5O12, GGG) through room-temperature RF magnetron sputtering and annealed in flowing oxygen. We examined the effects of annealing temperature and film thickness using X-ray diffraction (XRD), scanning electron microscopy (SEM), longitudinal magneto-optical Kerr effect (MOKE) loops and multi-frequency field-swept ferromagnetic resonance (FMR). Representative SEM images showed laterally uniform surface backgrounds with sparse localized particulates. Gaussian switching-field fits with residual-bootstrap analysis were used to obtain Kerr-derived coercive fields, switching-field standard deviations and their normalized ratios. In the temperature series, the 900 °C sample gave the smallest fitted σSFD/Hc, whereas the 850 °C sample gave the lowest effective damping of α = (4.2 ± 0.27) × 10−3. In the thickness series annealed at 850 °C, the 75 nm film combined the smallest σSFD/Hc (3.11 ± 0.38%) with the lowest measured α, (5.84 ± 0.16) × 10−3. The different temperature optima show that switching uniformity and dynamic loss should be assessed separately. The results provide an empirical comparison for selecting processing conditions within the investigated RF-sputtered YIG/GGG series.

Applied Sciences doi: 10.3390/app16147164

Authors: Sara Fileković Emine Yaman

Network Function Virtualization (NFV) is a novel concept in computer networking that enhances scalability and flexibility in network services. Its increasing complexity and highly dynamic nature challenge reliable and efficient cyberattack detection. This study evaluates the predictive performance and computational efficiency of ensemble machine learning (ML) models, including Random Forest, eXtreme Gradient Boosting (XGBoost), and CatBoost, using three NFV-specific subsets of the VNFCYBERDATA dataset, and compares them with a stacking-based meta learner for classifying benign and malicious network traffic. To enhance the ML workflow, a reproducible preprocessing pipeline and automated hyperparameter optimization with the Optuna framework were employed. Performance was assessed using: precision, recall, F1-score, false positive rate (FPR), Precision–Recall Area Under the Curve (PR-AUC), training and inference speed, memory usage, and serialized model size. SHapley Additive exPlanations (SHAP) was used to interpret model predictions and identify the strongest indicators of cyberattacks. The results demonstrate that Random Forest and the stacked meta learner consistently achieved the highest predictive performance, attaining precision and PR-AUC values of up to 99%, while retaining very low false positive rates. Meanwhile, XGBoost and CatBoost demonstrated higher efficiency, resulting in substantially faster training and inference times (10–12 ms), which makes them better suited for latency-sensitive deployments. These findings support the applicability of boosting models in resource-constrained environments, while bagging and stacking models provide robust detection in high-dimensional and imbalanced datasets. SHAP analysis identified traffic volume, transport-layer protocol flags, and network flow characteristics as the strongest indicators of malicious activity in NFV networks, providing deeper insights into model predictions.

Applied Sciences doi: 10.3390/app16147159

Authors: Yufei Wang Tong Zhang Fan Zhou Fang Liu Bo Qian Jun Liu

In complex and unknown environments, unmanned aerial vehicle (UAV) autonomous navigation still faces issues such as insufficient representation of state characteristics, fixed reward guidance, and low efficiency in utilizing key experience samples. To address these problems, this paper proposes an improved soft actor–critic (SAC) method that integrates dual-path channel attention (DPCA), adaptive reward feedback (ARF), and prioritized experience replay (PER); this method is named DPCA-ARF-PER-SAC. The proposed DPCA module is introduced into the actor network to recalibrate one-dimensional navigation state features and enhance the representation ability of key decision-making information. At the same time, the ARF mechanism can dynamically adjust the reward weights according to the training progress, while PER is used to improve the utilization efficiency of key samples. The experiments are conducted in a two-stage structure, including module-level ablation verification in the two-dimensional (2D) SimpleAvoid scenario and main performance comparison in the three-dimensional (3D) NH_center scenario. The experimental results show that the success rate of this method reaches 1.00 in the 2D scenario, and the collision rate is 0.00. In the 3D scenario, the success rate is 0.76, the collision rate is 0.24, and the average episode length is 232.8 steps. Compared with the baseline SAC, the success rate is increased by 2 percentage points, the collision rate is reduced by 2 percentage points, and the average episode length is reduced by 4.3 steps. Compared with twin delayed deep deterministic policy gradient (TD3), SAC, and representative reinforcement learning methods such as attention-mechanism SAC (AM-SAC) based on attention mechanism enhancement, the proposed DPCA-ARF-PER-SAC method shows more balanced performance in task completion, navigation safety, and path efficiency. These results indicate that DPCA-ARF-PER-SAC provides a more robust navigation strategy for complex 3D UAV autonomous navigation tasks.

Applied Sciences doi: 10.3390/app16147163

Authors: Vladimir Leonidovich Trushko Oleg Igorevich Kolosov

This methodological paper describes the development and preliminary validation of an experimental procedure for testing rock specimens under volumetric compression using the MDS-450 and MTS-815 servo-hydraulic systems. The apparatus allows the simultaneous application of axial load, confining pressure, pore pressure, and temperature to simulate thermobaric conditions corresponding to depths of about 10–12 km. In this study, the system was used only under room-temperature conditions in order to verify the loading path and data acquisition scheme. The procedure provides stepwise reproduction of reservoir conditions and is designed to record triaxial compressive strength, deformation moduli, dilatancy, and acoustic and ultrasonic responses. Results from preliminary tests on sedimentary rock specimens at confining pressures of 90 and 120 MPa and room temperature demonstrate typical stages of volumetric compression and confirm the stability of the experimental method. To support the verification of the methodology, additional tests were also performed at a confining pressure of 50 MPa, incorporating the temperature factor and acoustic emission monitoring.

Applied Sciences doi: 10.3390/app16147162

Authors: Jinxuan Ding Hao Wu Qian Zhang

This study focuses on subseasonal 2-m temperature prediction over East Asia for lead times of 1–6 weeks. Based on ERA5 reanalysis data from January 1982 to December 2024, multiple variables, including temperature, wind fields, and geopotential height, were selected as predictor fields. A SwinUNet-AR subseasonal prediction model integrating SwinUNet, AFNO, and Resize-Conv was constructed. The results indicate that the integrated architecture supports multiscale spatial representation, long-range dependency modeling, and high-resolution spatial reconstruction. Among different prediction strategies, simultaneous multi-lead prediction showed the most stable performance during Weeks 3–6. Multivariable input produced a statistically significant improvement in cumulative ACC during Weeks 3–6. The cumulative ACC increased by 0.07 compared with the scheme using only T2m as input, corresponding to an improvement of approximately 12.1%. The ablation experiments further indicate that the combined use of AFNO and Resize-Conv yields a modest improvement in medium- and extended-range forecast skill, particularly in cumulative ACC. Compared with the baseline SwinUNet, SwinUNet-AR reduced the mean RMSE by 0.02 °C during Weeks 3–4 and 0.03 °C during Weeks 5–6, while increasing the cumulative ACC during Weeks 3–6 by 0.04, corresponding to an improvement of approximately 7.4%. Compared with raw CFSv2 forecasts, SwinUNet-AR reduced the mean RMSE over Weeks 1–6 by 0.84 °C, corresponding to a reduction of approximately 44.9%. During Weeks 4–6, the mean RMSE was approximately 0.83 °C lower, corresponding to a reduction of about 40.8%. In terms of ACC, SwinUNet-AR achieved values comparable to those of CFSv2 during Weeks 4–6 and slightly higher values at some lead times. Overall, SwinUNet-AR shows potential for improving medium- and extended-range subseasonal 2-m temperature prediction over East Asia and provides a useful data-driven framework for regional subseasonal forecasting.

Applied Sciences doi: 10.3390/app16147161

Authors: Pavel Kic Jiri Hascyn

This study evaluates winter in-car microclimate and aerosol conditions in a Prague Metro line C M1.3 car under real operation. Five Letňany–Háje runs were divided into 19 inter-station segments. TDC, PM10, PM4, PM2.5, PM1, CO2, air temperature, relative humidity, A-weighted sound pressure level, and occupancy were recorded. Aerosol fractions were measured in separate runs and treated as operational optical estimates, not as reference-equivalent or simultaneous size-distribution data. Temperature, relative humidity, CO2, sound level, and occupancy were aggregated to one value per run and segment and analyzed by randomized-block ANOVA with run as the block. Significant segment effects were found for occupancy, CO2, relative humidity, and air temperature; LA showed no stable segment effect. Central segments showed passenger-related CO2 and heat accumulation, whereas aerosol maxima did not coincide with maximum occupancy. Larger cut-point PM runs showed greater dispersion and episodic peaks, consistent with surface–tunnel exchange, resuspension, and station/ventilation pathways. Across 19 segment means, occupancy correlated more strongly with CO2 than with air temperature (r = 0.837 vs. 0.633); after Run adjustment, the slopes were 8.494 ppm/passenger and 0.011 °C/passenger, respectively. An overnight Vltavská–Florenc transect provided qualitative evidence of local ventilation influence near Štvanice. The study is a winter, vehicle-specific segmental case study.

Applied Sciences doi: 10.3390/app16147160

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

To address the safety, efficiency, and adaptability challenges of autonomous path planning for lunar rovers in complex and unknown terrain, this study proposes a path planning method that integrates deep reinforcement learning with Hybrid A*. Multi-level obstacle maps are generated from lunar digital elevation models and local terrain reconstruction, and an environment model integrating terrain features and rover states is constructed. A Deep Deterministic Policy Gradient framework is introduced, including a state encoder, a policy network, and dual Q-networks, to dynamically adapt the search parameters of Hybrid A* and optimize the path-cost function online. Simulation experiments are conducted under simple, moderate, and complex scenarios. The results show that the proposed method achieves a 100% path planning success rate in the evaluated baseline scenarios, with comprehensive accuracy scores of 0.931, 0.907, and 0.886, respectively. It outperforms conventional A*, RRT, and Hybrid A* algorithms in path length, smoothness, and safety. The method can support safe and efficient autonomous exploration by lunar rovers under the tested dynamic-illumination and multi-obstacle conditions and provides technical support for future deep-space exploration missions such as Chang’e and Tianwen.

Applied Sciences doi: 10.3390/app16147158

Authors: Gozde Basak Ozturk Fatih Soygazi Busra Ozen

This study develops an ontology-driven framework for the semantic enrichment of the Ephesus Ancient City within a Heritage Digital Twins (HDT) context. The proposed approach integrates heterogeneous cultural heritage data into a unified semantic structure, enabling consistent representation of complex relationships. A domain-specific ontology is developed, followed by the construction of a knowledge graph in Neo4j to support structured querying. The framework moves beyond visualization-oriented digital twins by introducing a knowledge-centered semantic layer intended to support future reasoning and decision-making capabilities. Rather than constituting a complete digital twins system, the proposed approach establishes a semantic backbone that can serve as a foundation for future Heritage Digital Twins implementations. The Ephesus case suggests the practical applicability of the framework on a complex CH site. The results suggest a scalable and semantic foundation for developing interoperable Heritage Digital Twins.

Applied Sciences doi: 10.3390/app16147157

Authors: Raquel López María J. San José Sonia Alvarez Francisco J. Peñas

This study investigated the adsorption of pharmaceutical formulations (aspirin, enantyum, ibuprofen, metamizol, nolotil, paracetamol, and termalgin) and simultaneous binary mixtures of them onto activated carbons derived from olive stone (OSPC) and pine nut shell (PNSPC). Both biocarbons exhibited superior properties without prior washing, exhibiting the highest adsorption capacity (45 mg/g) with 1.0 g OSPC/L and 50 mg ENA/L at pH 2. FTIR analysis identified oxygenated and phosphate-related surface groups in both carbons, with OSPC exhibiting stronger functionalization, and revealed minimal differences in adsorption between commercial formulations and pure active ingredients. Raman spectroscopy showed small increases in the ID/IG ratio after adsorption, attributed to slight increases in defect density, without significantly altering the carbon structure. SEM analysis revealed that OSPC exhibited a more heterogeneous and porous structure than PNSPC, consistent with its larger BET surface area. The presence of surface phosphate groups was detected using XRD, and they were found to be more pronounced in OSPC; this was attributed to the fact that the adsorbents were not washed beforehand. Finally, OSPC was used to adsorb each of the six drugs independently in 14–15 consecutive cycles, with removal efficiency exceeding 85%; thus, its use could be considered in the treatment of wastewater containing the drugs from this study, in line with the circular economy, which aims to achieve zero waste.

Applied Sciences doi: 10.3390/app16147156

Authors: Fahad Ebrahim Mike Joy

Source code plagiarism is a significant challenge in software engineering and computer science education, affecting academic integrity, intellectual property rights, and software quality assurance. However, Source Code Plagiarism Classification (SCPC) remains difficult because labelled training data are limited, mainly due to the sensitivity of plagiarism cases. This restricts the effective use of machine learning (ML) and deep learning (DL) methods, especially in low-resource settings. This work investigates low-resource SCPC using Pre-trained Code Models (PCMs). We first examine Full Fine-Tuning (FFT), where all model parameters are updated, across multiple public datasets. We then evaluate Parameter-Efficient Fine-Tuning (PEFT), where only a small subset of parameters is trained. Specifically, we apply three adapter-based PEFT methods and compare them with FFT in terms of classification performance, training time, inference time, GPU usage, trainable parameter percentage, and model size. The results show that, when labelled training data are available, fine-tuned PCMs achieve strong SCPC performance and higher F1 scores than the unsupervised open-source plagiarism-detection tools in our evaluation, JPlag and Dolos. Overall, PEFT achieves a performance that is similar, comparable to, or slightly lower than that of FFT, while requiring fewer trainable parameters and lower GPU usage, at the cost of slightly higher inference time.

Applied Sciences doi: 10.3390/app16147155

Authors: Bochuan Geng Peidong Su Xiao Quan Xinhua Tao Xinghao Lu

This research addresses the critical safety issue of harmful gas influx during tunnel excavation through unconventional gas-bearing structures. It focuses on six tunnels affected by shale gas influx along the Zhangnan Expressway of northwestern Hunan, China. The research reveals the geochemical characteristics, distribution patterns, and accumulation mechanisms of harmful gas in the shale formations of northwestern Hunan. The research adopts an integrated approach of “geological background analysis—multi-parameter testing—comprehensive evaluation”. It is based on geological and borehole data, field geological surveys, as well as laboratory and field tests. The research systematically analyzes the gas-bearing structural characteristics, geochemical parameters, and reservoir physical properties of the shale gas area in the Zhangnan Expressway. The geological regularities are summarized. The results show that the Longmaxi Formation of the Silurian system and the Qixia Formation of the Permian system serve as source rocks in the tunnel sites. The reservoirs are characterized by ultra-low porosity and permeability, with limited late-stage hydrocarbon generation potential. The gas-related hazard during tunnel construction and operation is primarily associated with the release of existing free and adsorbed gas. According to the calculation standards for absolute gas emission rates during construction, three tunnels are classified as micro-gas tunnels and three as non-gas tunnels. Two accumulation patterns are proposed: the self-sourcing composite accumulation pattern with micro-scale migration, and the accumulation pattern of self-generated and self-storage type of water pressure confinement. Enhanced monitoring, ventilation, and grouting sealing are recommended. This study develops an integrated “geology—testing—evaluation” assessment method for shale-gas-bearing tunnels. It provides important guidance for controlling harmful gas hazards in such tunnels and guaranteeing the sustainable development of tunnel construction.

Applied Sciences doi: 10.3390/app16147153

Authors: Han Hu Yongcun Feng Guangyu Wang Jiecheng Yan Xiaorong Li

Pump start-up and shutdown, flow-rate adjustment, and tripping operations during drilling can induce bottomhole pressure fluctuations. These fluctuations may alter fracture aperture and change the development of lost-circulation pathways. To investigate the dynamic evolution of fracture aperture under fluctuating pressure, a thermo-hydro-mechanical (THM) coupled numerical model was established using ABAQUS. Bottomhole pressure fluctuations were induced by applying periodic perturbations to the inlet flow rate. The effects of fluctuation amplitude, fluctuation duration, and drilling-fluid temperature were then analyzed. The results indicate that fracture aperture exhibits a transient response before reaching a stable state. The fluctuation amplitude has a significant effect on the maximum transient fracture aperture. When the fluctuation amplitude increases to 30%, the maximum fracture aperture increases by 44%. In contrast, the fracture that has already formed may undergo reclosure during the low-pressure stage. The fluctuation duration mainly affects the persistence of the fracture opening and reclosure process, but has a relatively weak effect on the maximum fracture aperture. A decrease in drilling-fluid temperature promotes fracture opening and tip propagation. When the formation temperature is 100 °C, low-temperature drilling fluid increases the maximum fracture aperture by 3.06% and the fracture length by 13.89% compared with the isothermal reference case. These findings indicate that fracture aperture under fluctuating pressure cannot be characterized only by its stabilized value. The maximum transient fracture aperture, minimum fracture aperture, and temperature-induced changes in fracture morphology should also be considered. This study provides a numerical insight into the transient response of fracture aperture to bottomhole pressure fluctuations and drilling-fluid temperature during drilling in stress-sensitive fractured formations.

Applied Sciences doi: 10.3390/app16147154

Authors: Rongyong Zhao Da Pu Cuiling Li Zeyi Zhang Xingzhu Peng Yunlong Ma

Multimodal video understanding (MVU) has emerged as a fast-growing research frontier, driven by major advances in video-language pre-training and large multimodal models over the past decade. MVU aims to synergistically integrate visual, audio and textual modalities to interpret complex video semantics, supporting widespread downstream tasks including cross-modal retrieval, dense captioning, video question answering, event analysis and intelligent assistance. Despite the rapid proliferation of specialized MVU models, the community still lacks a unified capability-centric framework to systematically clarify the hierarchical competency architecture and evolutionary trajectory of state-of-the-art approaches. To address this issue, this paper presents a structured, comprehensive survey of the latest MVU progress, establishing a novel three-tier taxonomy that categorizes existing studies into cross-modal alignment, multi-granularity semantic expression and multimodal reasoning. Along this pipeline, we further systematically synthesize core modality fusion strategies, mainstream benchmark datasets and standardized evaluation protocols. Through a fine-grained analysis of representative published results, we highlight the critical impact of inconsistent evaluation settings, cross-experiment comparability bottlenecks and inherent methodological trade-offs between performance and efficiency. Finally, we identify and dissect three key open challenges: ultra-long video scalability, performance degradation from modality noise and missing data, and factual reliability risks in generative MVU systems. This capability-oriented systematic reference clarifies the methodological evolution logic of MVU, and provides actionable guidance for developing next-generation robust, high-performance multimodal video understanding systems.

Applied Sciences doi: 10.3390/app16147152

Authors: Yashuang Zheng Chuangxin Du

Plasma-catalytic ammonia synthesis can operate without high temperatures and pressures, and its performance is closely tied to discharge modes. However, research on its microscopic discharge mechanisms remains limited. A 2D fluid model incorporating plasma-activated heterogeneous reactions was developed to investigate voltage polarity effects on discharge dynamics in an N2/H2 planar dielectric barrier discharge (DBD) reactor with catalyst-coated barriers. Under both polarities, discharge starts as a gas-phase streamer at the catalyst apex due to local field enhancement and then evolves into a surface ionization wave (SIW) with an order-of-magnitude higher electron density. Positive voltage restricts the SIW to the catalyst surface, whereas negative voltage induces SIWs on both the upper bare dielectric and the catalyst, driven by distinct charge accumulation patterns. During the short discharge pulse, the gas phase primarily functions as a radical generator, while surface reactions dominate NH3 synthesis. Because positive voltage effectively targets plasma energy to the catalyst surface, it yields a higher peak NH3 density near the catalyst (2.85 × 1019 m−3) compared to negative polarity (8.10 × 1018 m−3).

Applied Sciences doi: 10.3390/app16147151

Authors: Zheng Hu Jin Liu Zhengjie Wang Wenyue Che Yong Xia Bing Zhang Shuyu Wu Kexun Zheng Feng Huang Bo Zhang

Tunnel construction in tectonically active mountainous regions is frequently hampered by elevated geothermal conditions, threatening construction safety and long-term infrastructure performance. In the Yunnan–Guizhou Plateau, characterized by complex fault systems and intense hydrothermal activity, rigorous assessment of geothermal hazard susceptibility along tunnel corridors is of critical engineering importance. However, the sparsity of geothermal observational data renders conventional assessment approaches insufficient, as they fail to quantify predictive uncertainty, which is essential for reliable risk decision-making. To address this gap, this study proposes a three-stage framework that integrates multi-source data fusion with Monte Carlo-based uncertainty quantification, using the Nige Tunnel as a case study. Ten conditioning factors were incorporated, with temperature-weighted positive samples constructed from field-surveyed hot springs. Gaussian noise injection and fractal buffer randomization were applied across 500 Monte Carlo iterations of an L2-regularized logistic regression model, evaluated by leave-one-out cross-validation. The three-stage assessment framework achieved robust predictive performance (mean leave-one-out cross-validation area under the curve (LOO-AUC) = 0.824) under sparse-sample conditions. High and Very High susceptibility zones account for 14.2% of the study area, concentrated along fault traces and collocated with hydrothermal discharge locations, with the Nige Tunnel traversing predominantly High to Very High susceptibility zones. Fault distance emerges as the dominant predictive factor, surpassing heat flow and Moho depth, indicating that structural permeability is the rate-limiting control on geothermal fluid enrichment in fault-dominated systems. The findings offer scientific support and methodological insights for risk zoning and hazard mitigation design in tunnel engineering projects in comparable geological settings.

Applied Sciences doi: 10.3390/app16147148

Authors: Anna Guzanová Dagmar Draganovská Štefan Novotný Miroslav Tomáš Gabriela Ižaríková Teodor Tóth Petr Szelag Miroslav Džupon Marek Vojtko

The aim of the presented study is to verify the possibility of forming hybrid adhesive-mechanical joints between aluminum sheet and PP matrix composites reinforced with carbon and glass fibers using thermal drilling technology. The research responds to weight reduction trends in the automotive and aerospace industries, seeking joining methods that preserve the continuity of reinforcing fibers. The methodology included applying an experimental organosilicate agent to the aluminum, sequential thermal drilling, and an innovative modification of bushing geometry using a 9.3 mm diameter tool. Joint quality was evaluated via tensile shear testing and non-destructive analysis using computed tomography (CT). Results showed that the organosilicate layer significantly increased the load-bearing capacity and adhesion of glass fiber joints. Hybrid joints exhibited higher energy absorption than purely adhesive joints. The proposed bushing geometry modification led to a statistically significant increase in total dissipated energy (by 28% to 37%) and a desired change in the failure mechanism from composite pull-out to bushing shear. CT analysis confirmed the preservation of fiber integrity through radial deflection. Consequently, hybrid joining via thermal drilling with modified geometry effectively utilizes the mechanical properties of metallic materials in multi-material structures.

Applied Sciences doi: 10.3390/app16147143

Authors: Supakan Matawong Sitthisak Audomsi Chonlatee Photong Taweesak Thongsan Worawat Sa-Ngiamvibool

This paper introduces an enhanced speed control strategy for permanent magnet synchronous motors (PMSMs) through the application of sliding mode control (SMC) optimized by the Chess Optimization Algorithm (COA). Additionally, the proposed method is compared with other validated algorithms, such as Harris Hawk Optimization (HHO) and the Whale Optimization Algorithm (WOA). The SMC parameters are optimized using Integral of Time-weighted Absolute Error (ITAE) and Integral of Time-weighted Square Error (ITSE) as objectives function, and the controllers are validated in MATLAB/Simulink analyzing transient response, speed variation, and load disturbance conditions. The results show that COA-tuned SMC provides better performance, with the lowest overshoot of 0.1929% for ITAE and 3.4800% for ITSE, the fastest settling times of 0.00307 s for ITSE, and the lowest MAE and RMSE in most evaluated scenarios, offering the most balanced overall performance among the optimization algorithms tested. Comparison to PI and PID controllers demonstrates that SMC provides enhanced accuracy, robustness, and disturbance rejection. COA-SMC provides an efficient and stable solution for PMSM applications that demand quick and accurate speed regulation.

Applied Sciences doi: 10.3390/app16147150

Authors: Agnieszka Leśniak Olga Skrzypczak Dominik Ożóg Bartosz Leśniak Michał Pietrzak

Renovation projects in existing university buildings involve considerable uncertainty due to incomplete documentation, aging building systems, phased execution, and the need to maintain ongoing educational activities during construction. This study investigates multilayer perceptron (MLP) architectures as an exploratory proof-of-concept for mapping an expert-based renovation risk index expressed as a continuous value between 0 and 1. The analysis was based on 122 real renovation cases described by 13 input variables representing quantitative factors and encoded qualitative characteristics related to technical and organizational project conditions. Data preprocessing included qualitative data encoding and min–max normalization. The models were trained and evaluated using an 80/10/10 hold-out split with validation-based early stopping. Two MLP architectures developed in MATLAB R2025b were compared to assess the effect of network depth on predictive performance. Model performance was evaluated using the coefficient of determination (R2) and mean squared error (MSE). The model with two hidden layers achieved R2 ≈ 0.58 and MSE ≈ 0.065, whereas the model with four hidden layers achieved R2 ≈ 0.86 and MSE ≈ 0.010. An ordinary multiple linear regression model, fitted to the full dataset as a reference linear analysis, showed weak explanatory power (R2 = 0.151). The results suggest that, within this exploratory dataset and the adopted hold-out procedure, the deeper MLP architecture achieved a closer fit to the available data than the shallower architecture. The findings provide a feasibility-oriented contribution to risk-informed planning in public building renovation projects, while requiring confirmation through repeated resampling and external validation.

Applied Sciences doi: 10.3390/app16147149

Authors: Xingyu Liu Yinzilong Zhang Kang Fei

Biomimetic snake-scale interfaces exhibit pronounced shear anisotropy, demonstrating promising application potential in the optimal design of soil–structure interfaces. To elucidate the underlying mechanisms of this anisotropy, this study employs the discrete element method (DEM) to develop a ring shear model of the glass bead–biomimetic snake-scale interface. Based on the simulation results, the shear response and its micromechanical evolution under constant normal stress are analyzed, and the effect of normalized scale height H/d and its underlying mechanism are discussed. The results indicate that the biomimetic snake-scale interface exhibits significant directional shear differences, with cranial shearing yielding a higher peak shear stress than caudal shearing. Moreover, cranial shearing forms a thicker shear band, where particles adjacent to the interface are more prone to interlocking at the steep slopes of the scales, thereby establishing a stronger force chain structure. Furthermore, the scale height exerts a nonlinear influence on the interfacial strength. For the particle-size and scale-length conditions considered in this study, cranial shear strength exhibits an apparent threshold response with increasing H/d, which is associated with particle climbing and trapping mechanisms, whereas caudal shear strength increases gradually. These findings may provide a useful reference for the design and potential engineering application of biomimetic snake-scale interfaces.

Applied Sciences doi: 10.3390/app16147134

Authors: Jinru Luo Jun Hu

High-rise glass curtain wall cleaning has long relied on dangerous manual high-altitude operations, which suffer from low efficiency, high labour costs and prominent safety hazards. Existing cleaning robots generally lack a systematic user demand-driven design framework, making it difficult to balance safety, obstacle-crossing capability and maintenance performance. To fill this gap, this study proposes a four-stage integrated sustainable design methodology integrating the KANO model, Analytic Hierarchy Process (AHP), Quality Function Deployment (QFD) and TRIZ. This methodology covers requirement attribute classification, priority weight quantification, demand-technology mapping and engineering conflict resolution, which can assist in sorting out multi-constraint design contradictions for the studied high-altitude cleaning equipment. Applying this methodology, the study develops the modular NetWing curtain wall cleaning drone with a vertically separable structure, retractable obstacle-crossing cleaning arm and quick-release cleaning head, and the feasibility of the conceptual scheme is preliminarily verified through numerical simulation. This work provides a quantifiable and replicable design paradigm for high-altitude service equipment, and the proposed solution can effectively reduce occupational risks and support the sustainable development of the facade cleaning industry.

Applied Sciences doi: 10.3390/app16147147

Authors: Argyro Kontogianni Antreas Kantaros Theodore Ganetsos Panagiotis Kousoulis Melina G. Mouzala Evangelos Papakitsos

Coptic represents the final stage of the Ancient Egyptian language and remains an important component of Christian and Mediterranean cultural heritage. Although several digital resources exist for Coptic textual corpora, Greek-oriented computational tools for the interpretation of Coptic inscriptions on artifacts remain limited. This study presents the design and early implementation of a semi-automated software tool for the computer-assisted translation of Coptic inscriptions into Greek, with optional English support. The tool combines a Coptic–Greek digital dictionary, an interactive character-selection interface, and two dictionary-search strategies: a length/alphabetically structured linear search and a weighted linear search based on expected word frequency. The application is intended to support scholars working with inscriptions on fragile or fragmented cultural heritage objects, where full automation is not realistic and human supervision remains essential. The paper describes the linguistic and material challenges of Coptic inscriptions, the structure of the lexical database, the interface design, and the planned use of Coptic corpora for improving retrieval efficiency. The proposed approach contributes to cultural heritage digitization by offering a practical, expandable, and user-oriented framework for supporting the study, interpretation, and preservation of Coptic inscriptions.

Applied Sciences doi: 10.3390/app16147145

Authors: Magdalena Walasek-Janusz Rafał Papliński Magdalena Kowalska Milena Kaczmarczyk Gabriela Tomulik Marlena Kokoszka Agata Szabat Gabriela Bernat Gabriela Śrótwa Oliwia Łatecka Zofia Przybyła Jakub Laurentowski Angelika Latkowska Barbara Mysiak Renata Nurzyńska-Wierdak

Common basil is a plant with high environmental requirements and is of great importance both in amateur and large-scale cultivation. One method of obtaining fresh basil is container cultivation. The study aimed to determine the feasibility of growing Greek basil in autumn under temperate climate conditions in containers with diameters of 9, 10, and 11 cm, under cover (polytunnel and unheated greenhouse), as well as in the risky field cultivation due to the risk of frost. The highest yield (23.8 g·plant−1) was obtained in greenhouse cultivation in the smallest containers. The plants growing in the greenhouse were characterized by the highest content of chlorophyll a (72.77 mg·100 g−1 FW), chlorophyll a + b (105.1 mg·100 g−1 FW), and carotenoids (20.87 mg·100 g−1 FW). Fifty-three compounds of basil essential oil were identified and characterized, with growth conditions (location/pot) modifying its chemical composition. The highest essential oil content (0.30%) was found in plants grown in the greenhouse in the largest pots. Methyl eugenol (35.96–69.85%) and eugenol (7.29–28.12%) dominated in the tested essential oil, followed by linalool (1.90–9.77%), α-trans-bergamotene (3.78–7.20%), (E)-β-farnesene (2.39–3.74%), and 1,8-cineole (1.49–5.22%). The conducted studies confirm the possibility of harvesting aromatic basil raw material rich in biocomponents from autumn cultivation in containers in the field and under cover.

Applied Sciences doi: 10.3390/app16147146

Authors: Anahit Khlghatyan Andrea Bergamaschi Andrey Medvedev Vahagn Muradyan Shushanik Asmaryan Fabio Dell’Acqua

Continuous monitoring of vineyard dynamics is essential for optimizing viticultural practices and assessing plant health. While the seasonal behaviors of satellite-derived vegetation indices are widely studied, robust parametric modeling of these temporal trends remains underexplored. Building upon initial clues derived from Italian vineyards, this study proposes a novel analytical framework based on the consistent parabolic temporal signature of optical and Synthetic Aperture Radar (SAR) indices. Focusing on the elevated Trinity Canyon Vineyards in Armenia, we model the yearly evolution and temporal aggregations of these indices using a parabolic fitting approach. Our results suggest that the parabola vertex, which we hypothesize corresponds to the absolute maximum of vegetative activity, remains remarkably stable across diverse vine types, satellite orbits, and years. While this stable behavior suggests an underlying phenological or structural consistency, distinct exceptions to this trend have also been identified and considered. Furthermore, to bridge the gap between remote sensing observables and agronomic traits, we investigated the relationship between the fitted parabolic parameters and the Winkler index, which is used here as an estimator of above-ground biomass (AGB). By correlating the vegetation indices’ temporal dynamics with biomass growth and by isolating specific anomalies driven by environmental or anthropogenic factors, this work offers a basis for a predictive methodology that enables tracking vineyard structural development.

Applied Sciences doi: 10.3390/app16147144

Authors: Ruhi Taş

Large-scale field video archiving under real-world operational conditions presents four compounding challenges: heterogeneous edge devices with inconsistent hardware capabilities, low-light and high-variance illumination environments, intermittent satellite and mobile network connectivity, and arbitrary camera mounting orientations. Existing QR-based video ingestion pipelines fail to address these challenges jointly, resulting in high decode failure rates and unreliable cloud archiving in distributed field deployments. This paper presents a framework that directly addresses all four challenges through a unified edge-to-cloud pipeline. Rather than introducing new computer-vision primitives, the contribution lies in the robust system architecture that integrates and orchestrates established techniques, and in its engineering validation at scale. The pipeline combines four engineered components: (i) a visibility-aware adaptive transcoding strategy; (ii) a priority-weighted non-uniform temporal sampling scheme; (iii) a 23-angle rotation ensemble decoder; and (iv) a checkpoint-resumable block-staged cloud synchronization mechanism with deferred reclassification. The framework is validated on 132 independent edge deployments across 849 operational regions, processing 87,873 production videos totaling 1.05 TB over a 10-day observation window (27 May–5 June 2026). The evaluation demonstrates 92.1% QR recall at 26.8% of full-scan CPU cost, 98.0% upload reliability, and 33.5% deferred recovery of initially unresolvable codes (p < 0.001, Cohen’s d > 0.97). To the best of our knowledge, this constitutes the largest reported real-world evaluation of an edge QR-based video ingestion framework and the first to characterize recall variance across more than 100 independent field deployments.

Applied Sciences doi: 10.3390/app16147142

Authors: Daoyun Tang Gulshat Amirkhanova Yanwei Fu

Robust multimodal sentiment analysis requires models that can integrate linguistic, acoustic, and visual cues while avoiding over-reliance on noisy nonverbal signals. This study proposes UCMR-Net, a text-anchored residual fusion framework for continuous multimodal sentiment intensity prediction. The model uses contextual textual representations as the primary semantic backbone and introduces acoustic and visual representations as adaptive residual correction signals. Instead of treating the learned positive residual coefficient as a direct estimate of aleatoric or epistemic uncertainty, the proposed framework interprets it as a residual reliability score for regulating nonverbal contribution. Under a unified five-seed evaluation protocol on CMU-MOSI, UCMR-Net achieves MAE = 0.699 ± 0.009, RMSE = 0.997 ± 0.011, Pearson correlation = 0.802 ± 0.006, Acc-2 = 85.82 ± 0.62%, and F1 = 85.60 ± 0.62% (mean ± SD). Controlled ablation results show that text anchoring is the dominant contributor to regression improvement, while residual fusion, adaptive residual weighting, counterfactual distillation, and multi-task supervision provide secondary stabilization effects. Unified missing-modality evaluation further indicates that performance remains relatively stable when audio or visual streams are removed, but degrades substantially when text is unavailable, confirming that the model is text-anchored rather than modality-symmetric. Calibration analysis shows that raw UCMR-Net only modestly improves calibration-related metrics, whereas post-hoc temperature scaling reduces ECE from 0.121 to 0.064 and NLL from 2.337 to 2.286. Additional CMU-MOSEI validation suggests that the proposed fusion strategy generalizes beyond CMU-MOSI, although cross-dataset transfer remains more challenging. Overall, UCMR-Net provides an effective and empirically validated framework for complete-modality sentiment intensity prediction, with moderate robustness under nonverbal missing or corrupted conditions.

Applied Sciences doi: 10.3390/app16147140

Authors: Elena Mirela Stetco Teodora Minzat Ovidiu Aurel Pop Mircea Nasui Ramona Bianca Sonher

The properties of solution-derived nickel oxide (NiO) thin films are strongly influenced by precursor chemistry and post-deposition thermal treatment. In this work, NiO thin films were deposited on indium tin oxide (ITO)-coated glass substrates by chemical solution deposition using a spin-coating process. A 0.6 M nickel acetylacetonate-derived precursor solution prepared in a methanol–propionic acid solvent system was used as the starting solution. The films were thermally treated at 600 °C under continuous oxygen flow, with annealing holding times of 2, 4, and 6 h, to evaluate the effect of thermal-treatment duration on phase formation, surface morphology, and electrochemical redox behavior. Thermal and infrared analyses confirmed precursor decomposition and the formation of nickel–oxygen bonding after thermal conversion. X-ray diffraction confirmed the formation of single-phase cubic NiO at all holding times, while quantitative peak analysis indicated a modest increase in crystallite size from 12.4 to 13.9 nm with prolonged annealing. Atomic force microscopy revealed continuous and homogeneous film surfaces, with low root-mean-square roughness values in the range of 2.5–2.7 nm. Cyclic voltammetry in a LiClO4/propylene carbonate electrolyte showed a reversible redox response associated with the Ni2+/Ni3+ transition, accompanied by visible switching between bleached and colored states. The results indicate that annealing holding time affects the structural and morphological characteristics of spin-coated NiO thin films, while the electrochemical measurements confirm reversible redox activity in the investigated sample. These findings provide a processing-oriented basis for evaluating thermal-treatment conditions in solution-derived NiO films intended for electrochemical and optical modulation applications.

Applied Sciences doi: 10.3390/app16147139

Authors: Daniela Munteanu Jean-Luc Autran Soilihi Moindjie Matteo Cecchetto Martin Dentan

The present study investigates the effects of neutron shielding by concrete slabs in the context of nuclear fusion (14 MeV neutrons). MCNP and Geant4 numerical simulations have been performed to characterize neutron transport through concrete slabs with thicknesses ranging from 1 to 160 cm. Using a monoenergetic, normally incident 14 MeV neutron source, we examine neutron-flux attenuation and spectral evolution after propagation through concrete. The results show that multiple elastic and inelastic scattering processes dominate the transport, leading to a strong spectral redistribution and a significant build-up of scattered neutrons. For intermediate thicknesses (50–100 cm), the transmitted flux is largely composed of scattered and secondary neutrons, with a pronounced epithermal slowing-down component approximately following a 1/E behavior, a thermal peak resulting from hydrogen moderation, and a residual high-energy tail corresponding to partially degraded primary neutrons. At larger thicknesses (≥100 cm), absorption and capture of thermalized neutrons become dominant, resulting in a sharp decrease in the transmitted flux. The impact of these spectral modifications on the soft-error rate (SER) of microelectronic devices is evaluated by folding the transmitted neutron spectra with energy-dependent single-event upset (SEU) cross-sections representative of advanced semiconductor technologies. The results indicate that, although total neutron flux decreases significantly with increasing concrete thickness, the spectral redistribution and the persistence of intermediate-energy neutrons can maintain a non-negligible contribution to SER in certain configurations. These findings highlight the importance of using full transport simulations, including spectral effects, when assessing radiation-induced reliability risks in fusion facilities.

Applied Sciences doi: 10.3390/app16147141

Authors: Muhammad Faizan Qazi Waqas Khan Murad Ali Khan Syed Shehryar Ali Naqvi Ji-Eun Kim SeungMyeong Jeong Il-yeop Ahn Do Hyeun Kim

Smart sensor networks for environmental monitoring require accurate and continuous estimation of key variables such as temperature, humidity, and wind speed; however, physical sensor deployments are frequently limited by high costs, hardware failures, and data quality degradation, while existing virtual sensor approaches rely on single-model architectures that lack explicit uncertainty modeling and fail to capture the complex non-linear dynamics of real-world IoT time-series data. This paper proposes VAS-DPFF, a virtual augmented sensor framework based on deterministic and probabilistic feature fusion, which contributes a principled integration of well-established deterministic and probabilistic techniques within a unified AIoT-compatible virtual sensing architecture. The framework integrates: (i) a deterministic pipeline comprising temporal encoding, rolling statistics, and mutual information-based feature selection; (ii) a probabilistic pipeline employing Bayesian Ridge Regression (BRR) and Gaussian Process Regression (GPR) to generate uncertainty-aware synthetic features; and (iii) an early feature-level fusion strategy feeding an XGBoost regression model augmented with Gaussian noise injection. Experiments on 84,582 time-series records from a nine-station IoT environmental monitoring network in Gwacheon City, South Korea, demonstrate strong multi-target prediction performance: temperature RMSE = 0.811 °C, R2 = 0.973; humidity RMSE =4.113%, R2 = 0.964; and wind speed RMSE =0.602 m/s, R2 = 0.798, representing RMSE reductions of 61.2%, 60.7%, and 62.3% over the existing method, respectively. Comprehensive ablation studies, sensitivity analysis, and augmentation validation confirm that the proposed integration of deterministic and probabilistic features yields consistent and practically valuable improvements in virtual sensing performance suitable for AIoT-enabled smart sensor network deployments across multiple environmental monitoring targets.

Applied Sciences doi: 10.3390/app16147138

Authors: Katarzyna Chojnacka Marcin Mikulewicz

Introduction: Genetic susceptibility shapes paediatric oral health, yet candidate variants remain scattered across phenotype-specific studies. Aim: To synthesise genetic association evidence for oral and craniofacial phenotypes with paediatric onset, and to propose a candidate single-nucleotide polymorphism (SNP) panel to inform future risk-stratification research supporting precision prevention in paediatric dentistry. Methodology: A PRISMA 2020-informed systematic search of PubMed, Web of Science, and Scopus (2015–2024) identified genome-wide association studies (GWASs), meta-analyses, and replicated candidate-gene studies reporting SNP-phenotype associations for paediatric oral traits. Variants were assigned to five trait clusters and graded Strong, Moderate, or Weak against pre-specified replication, study design, and effect-estimate criteria. Phenotypic, population, and methodological heterogeneity precluded quantitative pooling; narrative synthesis was performed. Results: The candidate panel comprises 22 SNPs across five trait clusters: eruption and number; morphology, agenesis, and craniofacial growth; enamel structure and caries; periodontal inflammation; and pain sensitivity. Twelve variants (Tier 1) meet genome-wide significance or meta-analytic replication thresholds; ten (Tier 2) have moderate or exploratory evidence, designated for research-grade use pending prospective paediatric validation. Conclusion: This evidence-graded 22-SNP panel offers a reproducible candidate framework for future genetic risk-stratification research in paediatric dentistry, not a validated clinical test. Prospective multicentre validation, population-level calibration, and implementation studies are required before clinical adoption. The article reports a systematic evidence synthesis and panel design rationale, not assay or clinical validation data.

Applied Sciences doi: 10.3390/app16147137

Authors: Krzysztof Dobosz Dawid Hanak Przemysław Kudłacik

Graphical representations of mathematical functions remain difficult to access for blind and visually impaired users, particularly in educational contexts where understanding the overall shape and behavior of a function is essential. This paper proposes an adaptive sonification method for mathematical function graphs designed for touchscreen-based exploration supported by continuous auditory feedback and voice interaction. The study focuses on balancing perceptual accessibility and auditory saturation by analyzing the influence of key sonification parameters, including auditory plateau width, slope length, and slope curvature. Two user experiments involving participants with different levels of visual impairment were conducted to evaluate the perceived usability of the proposed approach. The results indicate that usability improves with increasing auditory coverage only up to a certain threshold, beyond which excessive auditory density reduces perceptual clarity. The experimental results suggest that the preferred auditory plateau width is approximately 52 dp (about 9.8 mm), while the optimal slope length is approximately 84 dp (about 15.8 mm). In contrast, slope curvature has relatively little influence on subjective usability. The paper also introduces the Sonification Index (SI) and Auditory Saturation (AS), two quantitative measures that support the analysis and adaptive selection of sonification parameters based on graph complexity and device characteristics.

Applied Sciences doi: 10.3390/app16147136

Authors: Xue Li Stefano Capizzano Mirko Frappa Enrico Drioli Francesca Macedonio

Solutions with NaCl and Na2SO4 mixture simulating three real industrial RO waste brines were treated by membrane crystallization (MCr) with the aim of recovering the ions as crystals. In particular, the three different solutions have NaCl and Na2SO4 molar ratios of 1:2, 2:1 and 1:1, respectively. Experiments at two different feed temperatures were performed and analyzed. Results showed that the separation of the NaCl and Na2SO4 mixture can be realized by the MCr process and the Na2SO4 salts were successfully recovered from the mixture. Different morphologies of Na2SO4 crystals were obtained by selecting an appropriate feed temperature. Crystal size distribution (CSD) and cumulative fraction of the crystals were calculated to determine the distribution of the crystals’ size and their variation around the average size. The Na2SO4 salts obtained from all experiments were of high purity, indicating promising salt separation, recovery, and reuse from RO brine from a sustainable perspective.

Applied Sciences doi: 10.3390/app16147135

Authors: Kairat Sarsembayev Amoussou Coffi Adoko Rashid Afshar Candan Gokceoglu

Rockburst risk in seismically active mines poses a significant threat to underground safety. This study aims to improve the prediction of rockburst-induced damage by addressing the challenge of class imbalance, which is commonly encountered in rockburst datasets. Various data-balancing techniques, including the Synthetic Minority Oversampling Technique (SMOTE) and its variants, namely SMOTE-Tomek, KM-SMOTE, SMOTE-ENN, SVM-SMOTE, Borderline-SMOTE, and Adaptive Synthetic (ADASYN) sampling, were applied to a dataset containing rockburst damage scales and selected influencing parameters. The data were collected from Canadian and Australian underground mining operations affected by mining-induced seismicity. Three machine learning classifiers, namely Random Forest (RF), CatBoost (CB), and Gradient Boosting (GB), were trained and evaluated using accuracy, recall, precision, and F1-score metrics. The five best-performing models were SMOTE-ENN-GB, SMOTE-ENN-CB, SMOTE-ENN-RF, KM-SMOTE-CB, and Borderline-SMOTE-CB, achieving testing accuracies ranging from 70% to 88%. SHAP analysis further revealed that stress conditions and peak particle velocity (PPV) are the dominant factors controlling rockburst severity, while geological factors and support conditions act as secondary contributing factors. Compared with previous studies using the same dataset, the proposed approach achieved substantial improvements in predictive performance, particularly for the minority and severe rockburst classes. It is concluded that SMOTE-based balancing techniques, when combined with ensemble learning algorithms, can significantly improve rockburst damage prediction and contribute to safer and more effective risk management in deep, seismically active mining environments.

Applied Sciences doi: 10.3390/app16147133

Authors: Vincent Majanga Ernest Mnkandla Yifan Luo Daniel Oladele

Deep learning methods have exemplified performance in various dental image segmentation tasks. However, their interpretability inadequacy remains a hindrance to clinical adoption. This study introduces a quantitative explainability-driven evaluation framework to systematically assess and compare convolutional and transformer-based segmentation models in dental imaging. Specifically, we evaluate a convolutional Watershed Encoder–Decoder Neural Network (WEDN) and the hybrid Vision Transformer U-Net (ViTUNet) model using both qualitative and quantitative explainability metrics. The qualitative analysis employs Grad-CAM, occlusion sensitivity, and attribution-based visualization techniques to assess model decision patterns. Quantitatively, we define and apply explainability metrics, including Dice coefficient, boundary Dice, sparsity, and faithfulness metric scores, to independently evaluate segmentation performance and explanation quality. The experimental results show that ViTUNet achieves superior spatial localization, with higher Dice (0.765) and boundary Dice (0.421), compared to the WEDN model (0.664 and 0.221), indicating improved lesion delineation in dental images. Moreover, the ViTUNet model exhibits higher sparsity scores in attribution maps, indicating more concentrated and structured interpretability regions, while the WEDN model demonstrates higher faithfulness scores under perturbation-based evaluations, thus reflecting stronger dependence on localized feature representations. These findings highlight a key distinction between convolutional and transformer-based architectures with predictive behavior and explanation characteristics. Notably, the results further indicate that gradient-based explanation methods are more stable for transformer-based models, while perturbation-based methods better capture the decision logic of convolutional networks. In general, this study establishes an integrated framework for collectively evaluating segmentation performance and explainability, allowing more transparent and clinically reliable deployment of deep learning models in dental imaging.

Applied Sciences doi: 10.3390/app16147132

Authors: Lintao Chen Jun Wang Xiaojun Peng Xueshen Chen Minna Wang Xiangwei Mou Minghui Jiang Xu Ma Huanyu Jiang

Affected by moisture cohesion, wet rice seeds exhibit poor flowability and frequent arching blockage during silo discharging, which seriously restricts stable grain storage and conveying. To address this issue and reveal its intrinsic flow mechanism, this study establishes a discrete element method (DEM) model for wet rice seed-silo systems adopting the Hertz-Mindlin with Johnson-Kendall-Roberts (JKR) contact model, which incorporates surface energy to reflect moisture-induced cohesive effects. The model is verified via physical silo discharge tests, with consistent flow patterns, wall pressure error below 3.7% and discharge time error of 2.14%. EDEM parametric simulations are conducted to analyze velocity fluctuation at different silo heights. Coordination number and normalized contact force distribution are adopted to assess micro-contact force distribution in discharge areas, and a force chain extraction algorithm is used to explore variations in force chain length and orientation. Results demonstrate that during discharge, average particle velocity drops from silo bottom to top with growing fluctuation amplitude, presenting obvious stratified flow and intense upper-layer velocity pulsation. Weak contacts dominate wet rice seed groups and conform to exponential decay distribution. The force chain network undergoes three evolution phases: formation, force arch generation and collapse. Quantitative analysis reveals long force chain proportion falls steadily from 81.40% to 3.68% throughout discharge. Short force chains rise to 60.54% in the arch-forming stage and reach 96.32% after arch collapse. Horizontal force chains account for a maximum of 71.09% during arch formation, while vertical ones decline from 99.52% initially to 61.57% post collapse. This research offers mechanical references and quantitative parameters for the design and operation of silos for wet granular farm grains like rice seeds, and is particularly relevant to post-harvest engineering, grain storage safety, and agricultural machinery design.

Applied Sciences doi: 10.3390/app16147131

Authors: Kenneth Kaunda Maria Jerzykiewicz

The objective of this study was to introduce Diplorhynchus condylocarpon as a potential latex source. In addition, the physicochemical properties of the polymer as a trap for NO2 were investigated. Advanced spectroscopic methods, such as FTIR-ATR, EPR, and 1H and 13C NMR, were used in addition to traditional elemental analysis to evaluate the properties of the latex. Analysis of latex from the Diplorhynchus condylocarpon tree revealed the presence of rubber. In addition to rubber, analyses revealed the presence of other aromatic and aliphatic compounds in the latex sample. The presence of semiquinone radicals indicates the presence of antioxidant phenolic structures in Diplorhynchus condylocarpon. The reaction of latex with NO2 resulted in the formation of a stable nitroxyl radical, as proven by the characteristic EPR parameters.

Applied Sciences doi: 10.3390/app16147130

Authors: Nagima Karabassova Gulbarshin Shambilova Saule Bukanova Igor Makarov Ivan Levin Georgy Makarov Abzal Taltenov Junlong Song Zhanar Kadasheva

The article addresses the challenges associated with processing heavy petroleum feedstock with high metal content in the Resid Fluid Catalytic Cracking (RFCC) unit of the Atyrau Oil Refinery. Elevated concentrations of iron, nickel, and vanadium were shown to cause rapid catalyst deactivation, deterioration of fluidization properties, and instability of the reactor–regenerator system. Analysis of industrial data revealed that the iron content in the feed significantly exceeded the design value, reaching 150 mg/kg, which resulted in increased catalyst consumption and reduced process efficiency. The industrial application of metal-passivating chemical reagents reduced the iron content by 48–62%, thereby improving catalyst performance and increasing the yield of target products. However, the use of these reagents is associated with high operating costs. Therefore, additional demetallization methods were investigated, particularly solvent deasphalting. Experimental studies on a pilot unit demonstrated that deasphalting reduced the concentrations of Fe, Ni, and V by 7–9 times and produced a deasphalted oil yield of 81.1 wt.%. The obtained product exhibited low Conradson carbon residue and improved quality, making it suitable as a feed component for catalytic cracking. An integrated approach combining chemical treatment and solvent deasphalting is proposed to improve refining efficiency and catalyst stability.

Applied Sciences doi: 10.3390/app16147129

Authors: Adrian Kampa Damian Krenczyk Piotr Michalski Iwona Paprocka Bożena Skołud

Designing modern disassembly systems requires the integration of industrial automation equipment. Due to the support of various communication protocols, PLCs not only perform control tasks but also act as intelligent data centers in distributed production systems. PLC solutions increasingly combine traditional approaches to automation with modern digital technologies, enabling predictive maintenance, real-time data analysis, as well as remote process management and integration with digital twin simulation. The takt time of manual disassembly may vary due to human and technical factors; therefore, its estimation is a problem in many processes including, for example, Bluetooth speakers. This article discusses the issue of PLC-based control systems for a sorting process of dismantled parts, and the methodology of a digital twin framework in FlexSim software. A prototype of a sorting line based on a conveyor belt with an S7-1200 series PLC controller and a full digital twin development cycle were presented. The explicit assessment of takt-related temporal fidelity in PLC-mediated event streams remains less developed. Therefore, this article addresses this gap by using a Digital-Twin-in-the-Loop (DTiL) configuration as a digital twin validation setup in which a source process model generates PLC-mediated events and a separate resulting digital twin model is evaluated against this source. The article focuses on temporal fidelity, PLC-mediated event transfer, and takt-time estimation. Thus, the gathered empirical time data were then fed into the digital twin model and analyzed to obtain information about the time delay of the PLC signals. This article separates the general digital twin architecture from one specific validation scenario implemented in a digital twin in-the-loop configuration with FlexSim, Siemens TIA Portal, PLCSim Advanced, and a local network communication chain. Delay analysis is based on photocell event timestamps and inter-event time differences, which reduce the effect of initial clock mismatch. The results indicate that, under the tested local-network DTiL configuration, absolute event delays are visible, while inter-event timing and aggregated takt statistics remain highly consistent between the source and resulting models. These findings support the preliminary feasibility of PLC-mediated takt-oriented monitoring for long manual operations. Nevertheless, broader validation under different controller configurations, communication conditions, and operating scenarios is required before generalizing the proposed approach.