Applied Sciences

21 pages, 4169 KB

Open AccessArticle

Dynamic Security Configuration in China Railway Cloud Center: A Combination of Mobile Target Defense and Q-Learning

by Honglei Yao, Yijie Yang and Wenjia Niu

Appl. Sci. 2026, 16(5), 2626; https://doi.org/10.3390/app16052626 - 9 Mar 2026

Viewed by 341

The China Railway Cloud Center, as critical infrastructure, faces escalating cyber threats that demand proactive defense strategies beyond static mechanisms. This paper proposes a dynamic security configuration framework that combines Moving Target Defense (MTD) with Q-learning. We first model the cloud application as [...] Read more.

The China Railway Cloud Center, as critical infrastructure, faces escalating cyber threats that demand proactive defense strategies beyond static mechanisms. This paper proposes a dynamic security configuration framework that combines Moving Target Defense (MTD) with Q-learning. We first model the cloud application as a stack of configurable layers (e.g., OS, Middleware) and formalize vulnerabilities and attackers using Common Vulnerability Scoring System (CVSS) metrics. A transition reward function is then derived, integrating attack probability, security entropy, switching cost, and CVSS-based impact. Finally, the configuration switching problem is formulated as a Markov Decision Process and solved using Q-learning to find the optimal policy. Simulation results demonstrate that the Q-learning strategy converges to configurations yielding higher cumulative rewards and lower costs compared to random, periodic, and greedy baselines, effectively reducing the simulated attack success rate from 34% (random) to 12%. This work provides an adaptive, learning-driven defense framework for railway critical information infrastructure, offering a foundation for automated security orchestration. Full article

(This article belongs to the Section Computing and Artificial Intelligence)

► Show Figures

Figure 1

28 pages, 3372 KB

Open AccessArticle

An FPGA-Based Time-Domain Waveform Recognition Method Using Multi-Feature Voting Fusion

by Yiqi Tang, Zheng Li and Lin Zheng

Appl. Sci. 2026, 16(5), 2625; https://doi.org/10.3390/app16052625 - 9 Mar 2026

Viewed by 403

Abstract

Identifying the time-domain waveform type under broadband conditions is a basic but very challenging task. Traditional methods based on frequency domain or training models generally have the problems of high resource consumption, large delay, and unsuitability for hardware. This paper proposes a time-domain [...] Read more.

Identifying the time-domain waveform type under broadband conditions is a basic but very challenging task. Traditional methods based on frequency domain or training models generally have the problems of high resource consumption, large delay, and unsuitability for hardware. This paper proposes a time-domain waveform recognition architecture based on an FPGA, which is integrated with multi-feature voting. Several lightweight time domain characteristics, such as high amplitude ratio, symmetry, slope uniformity, slope change rate, and flat-top characteristics, are extracted and directly used for waveform classification. Then classify sine waves, square waves, triangular waves, and noise in the time domain according to the decision-making mechanism of voting. In order to improve reliability under non-ideal conditions, adaptive thresholds and noise perception decision-making logic are used to suppress misclassifications caused by random fluctuations and jitter. The whole engineering design focuses on resource consumption and hardware efficiency, using a fully pipeline FPGA architecture. The experimental results prove that the system has the ability of high-precision identification, low power consumption, and real-time processing in the wide frequency band, providing an efficient and practical solution for embedded waveform recognition applications. Full article

► Show Figures

Figure 1

38 pages, 1443 KB

Open AccessArticle

A Systematic Evaluation Method of Graph-Derived Signals for Tabular Machine Learning

by Mario Heidrich, Jeffrey Heidemann, Rüdiger Buchkremer and Gonzalo Wandosell Fernández de Bobadilla

Appl. Sci. 2026, 16(5), 2624; https://doi.org/10.3390/app16052624 - 9 Mar 2026

Viewed by 647

Abstract

While graph-derived signals are widely used in tabular learning, existing studies typically rely on limited experimental setups and average performance comparisons, leaving the statistical reliability and robustness of observed gains largely unexplored. Consequently, it remains unclear which signals provide consistent and robust improvements. [...] Read more.

While graph-derived signals are widely used in tabular learning, existing studies typically rely on limited experimental setups and average performance comparisons, leaving the statistical reliability and robustness of observed gains largely unexplored. Consequently, it remains unclear which signals provide consistent and robust improvements. This paper presents a taxonomy-driven empirical analysis of graph-derived signals for tabular machine learning. We propose a unified and reproducible evaluation method to systematically assess which categories of graph-derived signals yield statistically significant and robust performance improvements. The method provides an extensible setup for the controlled integration of diverse graph-derived signals into tabular learning pipelines. To ensure a fair and rigorous comparison, it incorporates automated hyperparameter optimization, multi-seed statistical evaluation, formal significance testing, and robustness analysis under graph perturbations. We demonstrate the applicability of the method through an extensive case study on a large-scale, imbalanced cryptocurrency fraud detection dataset. The analysis identifies signal categories providing consistently reliable performance gains and offers interpretable insights into which graph-derived signals indicate fraud-discriminative structural patterns. Furthermore, robustness analyses reveal pronounced differences in how various signals handle missing or corrupted relational data. These findings demonstrate the proposed taxonomy-driven evaluation method’s practical utility for fraud detection and illustrate how it can be applied in other application domains. Full article

► Show Figures

Figure 1

12 pages, 3606 KB

Open AccessArticle

Feasibility Study of Plate Inhomogeneities Estimation Using Lamb Wave A₀ Mode Signals Time-of-Flight

by Olgirdas Tumšys

Appl. Sci. 2026, 16(5), 2623; https://doi.org/10.3390/app16052623 - 9 Mar 2026

Viewed by 248

Abstract

Structural health monitoring (SHM) technology enables the monitoring and assessment of the condition of various materials and structures. Lamb-guided waves (LW) are widely used to detect damage in large-scale plate structures. One of the parameters used for these purposes is the time-of-flight (ToF) [...] Read more.

Structural health monitoring (SHM) technology enables the monitoring and assessment of the condition of various materials and structures. Lamb-guided waves (LW) are widely used to detect damage in large-scale plate structures. One of the parameters used for these purposes is the time-of-flight (ToF) of ultrasonic LW signals. In the presented feasibility study, the ToF was determined based on the idea that the zero-crossings of this signal, filtered by several filters, are concentrated around the maximum of the signal envelope. This ToF detection method, unlike threshold- and peak-based methods, avoids uncertainties in signal and noise levels and does not require a signal detection threshold. Compared to the correlation method, no reference signal is required. It has been established that the curves of signal propagation times with varying distance depend on the group and phase velocities of signal propagation and have phase jumps. The proposed methodology for assessing plate inhomogeneities involves comparing signal propagation time curves with and without damage. This methodology has been verified both through theoretical modeling and experimental research. The experimental studies used a 6 mm thick steel specimen with artificial defects of various diameters (10–35 mm). The A₀ mode of Lamb waves with a central frequency of 150 kHz was excited in the steel plate. For experimentally obtained B-scans, the ToF distributions of signals along the scan trajectories were calculated. By comparing the defective and defect-free ToF curves, critical points of the experimental curves were determined, which were used to estimate the dimensions of the defects. Both in the case of theoretical modeling and in the result of experimental measurements, it was determined that the proposed methodology can be used to determine the inhomogeneities of plates. Full article

(This article belongs to the Special Issue Advances in and Research on Ultrasonic Non-Destructive Testing)

► Show Figures

Figure 1

16 pages, 1350 KB

Open AccessArticle

Effect of Water Quality Produced at Each Stage of the Seawater Desalination Process on Hydrogen Production in Water Electrolysis

by Pyae Pyae Shwe Sin, Tomohiro Yadai, Hiroshi Yamamura, Yoshihiro Suzuki, Yasuyuki Ota and Kensuke Nishioka

Appl. Sci. 2026, 16(5), 2622; https://doi.org/10.3390/app16052622 - 9 Mar 2026

Viewed by 363

Abstract

Hydrogen production via water electrolysis using desalinated seawater offers a sustainable energy solution and has attracted considerable attention in recent years. However, its efficiency depends heavily on the quality of water. Many studies have not explored the relationship between treated water quality and [...] Read more.

Hydrogen production via water electrolysis using desalinated seawater offers a sustainable energy solution and has attracted considerable attention in recent years. However, its efficiency depends heavily on the quality of water. Many studies have not explored the relationship between treated water quality and hydrogen generation efficiency at each stage of the seawater desalination process. This study examines a three-step seawater desalination process comprising softening with ballasted flocculation (SBF) as a pretreatment, reverse osmosis (RO) as the main desalination step, and ion exchange as a polishing step to provide high-quality water for electrolysis. Water from each purification stage was supplied to the electrolyzer to compare the impact on water quality and hydrogen generation efficiency. The SBF process removed magnesium (Mg) and calcium (Ca) from seawater, as well as turbidity and bacteria, but hydrogen production via water electrolysis continued for no more than 10 h. However, when feeding RO water and RO water processed by ion exchange after the SBF process, hydrogen was generated stably and continuously for 70 h, achieving high efficiency comparable to that of commercial pure water. High production of green hydrogen by water electrolysis is possible through RO seawater desalination combined with SBF pretreatment. Full article

(This article belongs to the Section Energy Science and Technology)

► Show Figures

Figure 1

15 pages, 1448 KB

Open AccessArticle

Agronomic Potential of Digestates from Pig Slurry and Wine Vinasse Co-Digestion Under Temperature-Phased Anaerobic Digestion

by Belén Cañadas, José Luis Millar, Juan José Iglesias, Juana Fernández-Rodríguez and Montserrat Pérez

Appl. Sci. 2026, 16(5), 2621; https://doi.org/10.3390/app16052621 - 9 Mar 2026

Viewed by 265

Abstract

The management of Pig Slurry (PS) and Wine Vinasse (WV) poses major environmental and economic challenges, Anaerobic co-digestion (AcoD) offers a promising approach, producing both renewable energy and nutrient-rich digestates with agronomic potential. This study evaluated digestates obtained from the AcoD of a [...] Read more.

The management of Pig Slurry (PS) and Wine Vinasse (WV) poses major environmental and economic challenges, Anaerobic co-digestion (AcoD) offers a promising approach, producing both renewable energy and nutrient-rich digestates with agronomic potential. This study evaluated digestates obtained from the AcoD of a 50:50 mixture of pig slurry and wine vinasse under Temperature-Phased Anaerobic Digestion (TPAD) conditions. The acidogenic reactor reached stability at a hydraulic retention time (HRT) of 5 days, achieving 51.34 ± 3.08% of tCOD removal and approximately 0.5 L of daily green hydrogen production, whereas the methanogenic stage reached stability at an HRT of 10 days with 89.14 ± 2.33% tCOD removal and recording daily biomethane production of up to 1 L. Digestates were tested in germination assays using Lepidium sativum (garden cress), Lactuca sativa (lettuce), and Raphanus sativus (radish) seeds to assess phytotoxicity, and pathogen analyses were conducted to confirm sanitary safety (contains 0.8 × 10³ MPN/gTS E. coli). Results showed that agronomic performance was primarily influenced by dilution level, at 10D–15D% dilutions, germination and root growth remained stable, with Germination Index (GI) values above 80%. In contrast, concentrations above 25D% led to marked inhibition, with GI values below 50%. These findings demonstrate that the TPAD system operates effectively when treating pig slurry and winery vinasse, producing digestates that are safe and effective organic amendments. Moreover, given their compliance with sanitary standards, these digestates can be classified as Class A biosolids suitable for agricultural application, provided that adequate dilution is ensured. Full article

(This article belongs to the Special Issue Soil Pollution and Remediation: Recent Developments and Future Perspectives, 2nd Edition)

► Show Figures

Figure 1

16 pages, 844 KB

Open AccessReview

Plant-Derived Nanomaterials and Protein Misfolding Disorders: Green Production Approaches, Biological Interactions, and Research Trends (2015–2025)

by Corina Dalia Toderescu, Iulia Cresneac, Alexandru Oancea, Svetlana Trifunschi, Melania Florina Munteanu and Casiana Boru

Appl. Sci. 2026, 16(5), 2620; https://doi.org/10.3390/app16052620 - 9 Mar 2026

Viewed by 328

Abstract

Protein misfolding and aggregation represent key pathological mechanisms in neurodegenerative and systemic amyloid disorders, yet disease-modifying therapeutic strategies remain limited. In recent years, plant-derived nanomaterials have attracted increasing attention as multifunctional platforms capable of interacting with misfolded proteins and modulating aggregation-related pathways. This [...] Read more.

Protein misfolding and aggregation represent key pathological mechanisms in neurodegenerative and systemic amyloid disorders, yet disease-modifying therapeutic strategies remain limited. In recent years, plant-derived nanomaterials have attracted increasing attention as multifunctional platforms capable of interacting with misfolded proteins and modulating aggregation-related pathways. This review examines the evolution of research between 2015 and 2025 on plant-derived nanomaterials—including green-synthesized metallic nanoparticles, plant extracellular vesicles, and phytochemical-based nano-delivery systems—in the context of protein misfolding disorders. The available literature was analyzed to identify principal mechanisms of action, experimental models, and emerging therapeutic perspectives. Current evidence suggests that these nanomaterials may influence protein aggregation through direct molecular interactions, modulation of oxidative stress and neuroinflammatory responses, and enhancement of cellular protein clearance processes. However, the field remains characterized by methodological heterogeneity, limited standardization, and insufficient translational validation. By synthesizing recent developments, this review highlights key research trends, mechanistic gaps, and future directions necessary for advancing plant-derived nanomaterials toward biomedical applications targeting protein misfolding diseases. Full article

(This article belongs to the Special Issue Green Synthesis of Plant-Based Nanoparticles: Influence of Physicochemical Factors on Formation and Stability)

► Show Figures

Figure 1

13 pages, 514 KB

Open AccessArticle

Dance-Specific Patterns of Relative Oxygen Uptake in Elite Slovak Standard and Latin DanceSport Dancers

by Matej Chren, Milan Špánik, Viktor Plačko, Adéla Chlapcová, Peter Olej and Szymon Kuliś

Appl. Sci. 2026, 16(5), 2619; https://doi.org/10.3390/app16052619 - 9 Mar 2026

Viewed by 352

Abstract

Background: DanceSport involves intermittent high-intensity efforts that may differ between styles and partners within a dance couple. However, dance-specific relative oxygen uptake (%VO₂max) in elite Standard and Latin dancers remains insufficiently described. Objective: This study aimed to characterize relative oxygen uptake [...] Read more.

Background: DanceSport involves intermittent high-intensity efforts that may differ between styles and partners within a dance couple. However, dance-specific relative oxygen uptake (%VO₂max) in elite Standard and Latin dancers remains insufficiently described. Objective: This study aimed to characterize relative oxygen uptake during simulated competition in elite Slovak national team dancers and to examine (i) differences between Latin and Standard styles, (ii) variability across individual dances, and (iii) sex-specific patterns. Methods: Twenty elite dancers (10 couples) participated in the study. Five couples (n = 10 dancers; 5 females and 5 males) specialized in Latin dances, and five couples (n = 10 dancers; 5 females and 5 males) specialized in Standard dances. VO₂max was determined via an incremental treadmill test. During a simulated final round, breath-by-breath gas exchange was recorded using portable spirometry. Style-level differences were analyzed using a two-way ANOVA (Style × Sex), and dance-specific effects were examined using repeated-measures ANOVAs. Results: No significant difference in mean %VO₂max was observed between styles (F_{(1, 16)} = 1.31, p = 0.269, η²_p = 0.076). In the Latin group, relative oxygen uptake differed significantly between dances (F_{(4, 32)} = 22.45, p < 0.001, η²_p = 0.737), with Jive eliciting the highest values (~103–105% VO₂max in males) and Rumba eliciting the lowest values (~88–89% VO₂max). No Dance × Sex interaction was detected in Latin dances (p = 0.526). In the Standard group, a significant Dance × Sex interaction was observed (F_{(4, 32)} = 8.80, p < 0.001, η²_p = 0.524), with male dancers demonstrating higher %VO₂max during Quickstep (~96%) compared with other dances, whereas females showed a more uniform intensity profile (~80–86%). Conclusions: Relative oxygen uptake in DanceSport is highly dance-dependent and shows sex-specific metabolic patterns in Standard dances. Conditioning programs in elite DanceSport should therefore be structured according to individual dance demands and partnership-specific physiological roles. Full article

► Show Figures

Figure 1

20 pages, 9101 KB

Open AccessArticle

Automatic Defect Detection for Concrete Bridge Decks Using Geometric Feature Augmentation and Robust Point Cloud Learning Strategy

by Zhe Sun, Siqi Li, Minghui Huang and Qinglei Meng

Appl. Sci. 2026, 16(5), 2618; https://doi.org/10.3390/app16052618 - 9 Mar 2026

Viewed by 247

Abstract

Surface defects such as depressions, heaving, and irregular undulations frequently develop on aging concrete bridge decks under repeated traffic loading and environmental effects. Accurate and objective identification of such defects is essential for structural serviceability and safety, yet manual inspection remains labor-intensive and [...] Read more.

Surface defects such as depressions, heaving, and irregular undulations frequently develop on aging concrete bridge decks under repeated traffic loading and environmental effects. Accurate and objective identification of such defects is essential for structural serviceability and safety, yet manual inspection remains labor-intensive and subjective. This study develops a systematic framework for surface defect identification through geometric feature augmentation with a streamlined point cloud learning strategy. In practical engineering scenarios, point cloud data of concrete bridge decks can be periodically acquired via vehicle-mounted mobile laser scanning (MLS) systems and subsequently streamlined for analysis. The proposed method heightens defect sensitivity by extracting interpretable geometric descriptors, further integrating multi-scale representations to capture surface defects across varying spatial extents. Evaluated on a public point-level annotated benchmark, the proposed method clearly outperforms the same network trained with geometric coordinates only. To improve result reliability, all experiments were repeated four times with different random seeds, and the performance is reported as mean ± standard deviation. Results show that the proposed method achieves a precision of 0.597 ± 0.021 and an accuracy of 0.933 ± 0.009 under the benchmark protocol. Overall, these results demonstrate a reproducible proof of concept under controlled benchmark conditions for bridge deck surface defect segmentation, while broader cross-site and cross-sensor validation will be pursued in future work. Full article

(This article belongs to the Special Issue Advances in Performance Monitoring and Anomaly Detection Across Structures and Industrial Processes)

► Show Figures

Figure 1

6 pages, 155 KB

Open AccessEditorial

Intelligent Logistics and Supply Chain Systems Based on Industry 4.0/5.0

by Panagiotis Tsarouhas

Appl. Sci. 2026, 16(5), 2617; https://doi.org/10.3390/app16052617 - 9 Mar 2026

Viewed by 457

Abstract

The rapid evolution of global markets, customer expectations, and technological capabilities has fundamentally transformed the way logistics and supply chain systems are designed and managed [...] Full article

(This article belongs to the Special Issue Intelligent Logistics and Supply Chain Systems)

19 pages, 35815 KB

Open AccessArticle

YOLOv10-TWD: An Improved YOLOv10n for Terracotta Warrior Recognition

by Yalin Li, Liang Wang, Xinyuan Zhang, Sijie Dong and Xinjuan Zhu

Appl. Sci. 2026, 16(5), 2616; https://doi.org/10.3390/app16052616 - 9 Mar 2026

Viewed by 235

Abstract

To address challenges such as complex backgrounds, partial occlusion, and high similarity of details in Terracotta Warrior image recognition, this paper proposes a lightweight detection method, YOLOv10-TWD, based on an improved YOLOv10n. Specifically, a lightweight Convolution-Attention Fusion Module (CAFMAttention) and a dual-branch feature [...] Read more.

To address challenges such as complex backgrounds, partial occlusion, and high similarity of details in Terracotta Warrior image recognition, this paper proposes a lightweight detection method, YOLOv10-TWD, based on an improved YOLOv10n. Specifically, a lightweight Convolution-Attention Fusion Module (CAFMAttention) and a dual-branch feature extraction structure (DualConv) are integrated into the detection head to enhance the model’s focus on fine-grained features and its discriminative robustness under partial damage conditions. In the Neck network, Ghost-Shuffle Convolution (GSConv) is introduced to compress the computational cost of multi-scale feature fusion while strengthening context-aware capabilities. Experimental results on a self-built Terracotta Warrior dataset demonstrate that the proposed method achieves a 7.63% improvement in mAP@0.5 compared to the baseline YOLOv10n, while simultaneously achieving a 6.66% increase in inference speed. The model achieves high precision alongside significant optimization in inference efficiency, making it well-suited for rapid recognition tasks in cultural heritage and museum scenarios. Full article

(This article belongs to the Section Computing and Artificial Intelligence)

► Show Figures

Figure 1

45 pages, 5567 KB

Open AccessArticle

Analysis of Tracking Stability and Performance Variations in Multi-Class Structural Damage Objects Under Viewpoint Changes in Disaster Environments

by Sung Min Hong, Hwa Seok Kim, Chang Ho Kang, Soohee Han, Seong Sam Kim and Sun Young Kim

Appl. Sci. 2026, 16(5), 2615; https://doi.org/10.3390/app16052615 - 9 Mar 2026

Viewed by 251

Abstract

This study evaluates the tracking performance of structural damages in disaster environments by combining YOLOv8 detection with the BoT-SORT tracker. Cracks and exposed rebar, characterized by fine and irregular structures, showed high sensitivity to viewpoint changes, with camera motion compensation (CMC) improving [...] Read more.

This study evaluates the tracking performance of structural damages in disaster environments by combining YOLOv8 detection with the BoT-SORT tracker. Cracks and exposed rebar, characterized by fine and irregular structures, showed high sensitivity to viewpoint changes, with camera motion compensation (CMC) improving

I o U

by +19.63% and +20.23%. For exposed rebar, the joint use of CMC and re-identification (Re-ID) further increased

I D F 1

by +37.73%, emphasizing the effectiveness of appearance-based matching. In contrast, delamination and concrete debris, with stable morphology and clear boundaries, exhibited limited benefits from CMC, improving

I o U

by +11.17% and +3.28%. Analysis of

M O T A

,

I D F 1

, and

H O T A

confirms that fine-grained damages require motion- and appearance-based strategies, while stable types maintain high performance through detection consistency. These results highlight the importance of tailored tracking strategies for enhancing disaster-response robots and structural monitoring systems. Full article

► Show Figures

Figure 1

23 pages, 20795 KB

Open AccessArticle

Re-Evaluation of the Source Rocks of the Upper Triassic Xujiahe Formation in the Sichuan Basin

by Chao Zheng, Min Wang, Junfeng Cui, Wei Yang, Xiaojuan Wang, Shuangling Chen, Nan Li, Guiru Yang, Min Jia, Dongmei Bo and Tianya Liu

Appl. Sci. 2026, 16(5), 2614; https://doi.org/10.3390/app16052614 - 9 Mar 2026

Viewed by 337

Abstract

The Upper Triassic Xujiahe Formation (T₃x) represents a critical terrestrial source rock system in the Sichuan Basin, exhibiting pronounced vertical and lateral heterogeneity. Previous stratigraphic subdivisions relied primarily on lithological correlations rather than a systematic sequence stratigraphic framework. This approach has [...] Read more.

The Upper Triassic Xujiahe Formation (T₃x) represents a critical terrestrial source rock system in the Sichuan Basin, exhibiting pronounced vertical and lateral heterogeneity. Previous stratigraphic subdivisions relied primarily on lithological correlations rather than a systematic sequence stratigraphic framework. This approach has led to significant inconsistencies in source rock evaluation. Furthermore, recent discoveries of large gas fields, coupled with data from newly drilled wells, necessitate a comprehensive reassessment of this system. In this study, we re-evaluate the geochemical characteristics and spatial distribution of these source rocks within a newly established sequence stratigraphic framework. This assessment utilizes a robust dataset comprising total organic carbon (TOC) content, Rock-Eval pyrolysis, and vitrinite reflectance (Ro) measurements. The results indicate that the source rocks of the New Member 5 (T₃x⁵) in the slope belt of Central Sichuan exhibit the highest hydrocarbon generative potential. These rocks are characterized by high organic abundance (with 40% of samples showing TOC ≥ 2.0 wt.%), are dominated by Type III and II₂ kerogen (humic–sapropelic), and have reached the mature to high-maturity stage (Ro ranging from 1.0% to 1.7%). Notably, the cumulative thickness of these high-quality source rocks reaches 100~150 m. Specifically, the T₃x⁵ intervals in the Qiulin and Tianfu areas are identified as the most favorable hydrocarbon-generating centers. This reassessment under the new stratigraphic division provides a refined theoretical basis for future exploration targeting the Xujiahe Formation in the Sichuan Basin. Full article

(This article belongs to the Special Issue Advances and Technologies in Rock Mechanics and Rock Engineering)

► Show Figures

Figure 1

13 pages, 755 KB

Open AccessArticle

Development and Validation of a Stability-Indicating HPTLC Method for the Analysis of Gentamicin Sulphate in Pharmaceutical Ointments

by K. M. Yasif Kayes Sikdar, Md Khairul Islam, Edith Kai Yan Tang, Tomislav Sostaric, Lee Yong Lim and Cornelia Locher

Appl. Sci. 2026, 16(5), 2613; https://doi.org/10.3390/app16052613 - 9 Mar 2026

Viewed by 316

Abstract

This study developed and validated a stability-indicating high-performance thin-layer chromatography (HPTLC) method for the identification and quantification of gentamicin sulphate in an ointment formulation using silica gel 60 F254 HPTLC plates as the stationary phase and methanol: chloroform: ammonia solution (25%) (1:1:1, v [...] Read more.

This study developed and validated a stability-indicating high-performance thin-layer chromatography (HPTLC) method for the identification and quantification of gentamicin sulphate in an ointment formulation using silica gel 60 F254 HPTLC plates as the stationary phase and methanol: chloroform: ammonia solution (25%) (1:1:1, v/v/v) as the mobile phase. An ideal solvent ratio, chloroform: methanol (9:1, v/v), was used to dissolve the ointment sample before analysis. According to the guidelines of the International Council for Harmonisation (ICH), the HPTLC method was validated, demonstrating specificity by separating all three bands of gentamicin sulphate without interference from ointment excipients and/or degradation products resulting from photolytic, photolytic and oxidative, oxidative, acidic, and alkaline stress conditions. The findings of the study also revealed that the method has high levels of linearity within the range of 50–300 ng/band (R² ≥ 0.99), with detection and quantification limits of 7.10 ng, and 21.53 ng, respectively. Additionally, the method does not require any sample pre-treatment, such as extraction from the ointment base, making it simple and convenient for the quality control of gentamicin ointments. Full article

► Show Figures

Figure 1

17 pages, 840 KB

Open AccessArticle

Attention-Enhanced LSTM for Real-Time Curling Stone Trajectory Prediction on Resource-Constrained Devices

by Guanyu Chen, Shimpei Aihara and Yoshinari Takegawa

Appl. Sci. 2026, 16(5), 2612; https://doi.org/10.3390/app16052612 - 9 Mar 2026

Viewed by 294

Abstract

Real-time trajectory forecasting for curling stones is essential for on-ice decision support, yet prior work often emphasizes offline analysis, fixed-window predictors, or physics-driven models that require additional measurements, and it rarely reports end-to-end feasibility under edge-computing constraints (latency and memory). This leaves a [...] Read more.

Real-time trajectory forecasting for curling stones is essential for on-ice decision support, yet prior work often emphasizes offline analysis, fixed-window predictors, or physics-driven models that require additional measurements, and it rarely reports end-to-end feasibility under edge-computing constraints (latency and memory). This leaves a practical gap between accurate trajectory reconstruction and deployable rink-side guidance. To bridge this gap, we propose an online forecaster based on low-dimensional

(x, y)

coordinate streams and a lightweight attention-enhanced Long Short-Term Memory (LSTM) architecture optimized for edge devices. The model uses a four-second sliding window (240 frames at 59.94 Hz) to predict fifteen seconds of future positions (900 frames) in a single multi-step forward pass, and an overlapping publication scheme is adopted to retain longer temporal context and stabilize continuous updates. We further provide a TensorFlow Lite (TFLite) conversion and quantization workflow to support on-device inference. Quantitatively, experiments on the CurlTracer dataset (1033 throws at 59.94 Hz) show that the proposed attention–LSTM achieves trajectory-level MAE/MdAE of 0.25/0.22 m over the full prediction horizon, improving over a plain LSTM (0.30/0.24 m) and a physics-based pivot-slide baseline (3.52/3.54 m). At two checkpoints, the first-step MAE/MdAE are 0.14/0.11 m and the mid-step MAE/MdAE are 0.21/0.18 m. For real-time feasibility, on a Raspberry Pi 4B the per-window latency is approximately 0.25 s (including I/O and post-processing), while CPU benchmarks show that TFLite variants provide 7–8× speedups over the original Keras runtime with only minor accuracy loss (e.g., window-level MAE 0.30–0.41 m across FP32/DRQ/FP16/INT8). Qualitatively, representative trajectory visualizations show good agreement in near/mid horizons and reasonable stopping-region guidance, supporting integration with a stone-mounted interface for actionable feedback. Full article

(This article belongs to the Special Issue Advances in Winter Sports and Data Science)

► Show Figures

Figure 1

27 pages, 2003 KB

Open AccessReview

The Convergence of Federated Learning, Knowledge Graphs, and Large Language Models for Language Learning: A Scoping Review

by Michael Kenteris and Konstantinos Kotis

Appl. Sci. 2026, 16(5), 2611; https://doi.org/10.3390/app16052611 - 9 Mar 2026

Cited by 1 | Viewed by 756

Abstract

Large Language Models (LLMs) in Intelligent Computer-Assisted Language Learning enable highly personalized learning, yet raise significant challenges related to pedagogical grounding, data privacy, and instructional validity. Although Knowledge Graphs (KGs) and Federated Learning (FL) can mitigate these issues in isolation, evidence on systematic [...] Read more.

Large Language Models (LLMs) in Intelligent Computer-Assisted Language Learning enable highly personalized learning, yet raise significant challenges related to pedagogical grounding, data privacy, and instructional validity. Although Knowledge Graphs (KGs) and Federated Learning (FL) can mitigate these issues in isolation, evidence on systematic FL–KG–LLM integration for educational language learning remains limited. This scoping review maps the FL–KG–LLM convergence landscape. Following PRISMA-ScR guidelines, we searched six databases and screened 51 papers (2019–2025) using automated extraction. Our findings indicate limited convergence: no papers integrate all three domains, and 58.8% of approaches remain confined to isolated technological silos. Reporting is also uneven across the corpus, with an average “Not Reported” (NR) rate of 84.5%, most notably for privacy mechanisms (92.2%), validation metrics (90.2%), and Common European Framework of Reference for Languages (CEFR) alignment (88.2%). Domain-specific analysis reveals two distinct patterns: inter-domain gaps (disciplinary silos resulting in expected CEFR absence in single-domain papers) and intra-domain gaps (failure to report domain-critical variables, including 100% parameter NR in FL studies, 86.7% validation NR in KG studies, and 100% CEFR NR in convergence papers). Taken together, these gaps suggest that pedagogical grounding is treated as optional rather than structural. We therefore identify two pillars of pedagogical grounding: a Grounding Pillar, which constrains LLM outputs via Knowledge Graph rules, and a Validation Pillar, which concerns how authoritative frameworks (e.g., CEFR) are mapped onto Knowledge Graph schemas and evaluated. The near-universal absence of CEFR alignment and validation reporting suggests that this second pillar is currently missing, which we term the Integrity Gap—a systematic disconnection between technological innovation and pedagogical grounding inin Intelligent Computer-Assisted Language Learning. By reframing the problem as upstream control and validation, this review informs the design of user-facing automated systems where trust, transparency, and human oversight are critical. Full article

(This article belongs to the Special Issue Enhancing User Experience in Automation and Control Systems)

► Show Figures

Figure 1

22 pages, 3583 KB

Open AccessArticle

SSFF-DETR: A Surface Contaminant Detection Transformer for Microsystem Devices with Scale Sequence Feature Fusion

by Mengxiao Cui, Liping Lu and Hanshan Li

Appl. Sci. 2026, 16(5), 2610; https://doi.org/10.3390/app16052610 - 9 Mar 2026

Viewed by 248

Abstract

Microsystem devices are widely used in key fields such as aerospace. The various contaminants generated during their manufacturing process have the characteristics of diverse forms and are easily affected by background interference, making them difficult to detect. To solve this problem, this paper [...] Read more.

Microsystem devices are widely used in key fields such as aerospace. The various contaminants generated during their manufacturing process have the characteristics of diverse forms and are easily affected by background interference, making them difficult to detect. To solve this problem, this paper proposes a surface contaminant detection transformer for microsystem devices with scale sequence feature fusion (SSFF-DETR). This model is based on the real-time detection transformer (RT-DETR) framework. The faster efficient channel attention (Faster-ECA) was constructed as the backbone network, enhancing the extraction ability and computational efficiency of key features of contaminants. By introducing the dynamic feature region collaborative attention (DFRCA) at the end of the backbone network, the contrast between contaminant features and the background was effectively enhanced, thereby improving the model’s ability to identify contaminants. An Encoder based on scale sequence feature (SSF) and triple-branch feature fusion (TFF) is designed. By enhancing multi-scale representation, it effectively retains the detailed features of contaminants in complex backgrounds and alleviates the problem of feature loss during transmission in deep networks. The experimental results show that compared with the RT-DETR model, the SFFE-DETR model has achieved an increase of 2.6% in mean average precision (mAP). At the same time, the Giga Floating-Point Operations Per Second (GFLOPs) have decreased by 2G, and the params have reduced by 0.8 M. This provides a feasible solution for the high-precision and high-efficiency automated detection of surface contaminants in microsystem devices. Full article

► Show Figures

Figure 1

23 pages, 1154 KB

Open AccessReview

Challenges and Optimization Strategies in the Traditional A²/O Wastewater Treatment Process: A Review

by Yong Wang, Xin Jin and Guobiao Zhou

Appl. Sci. 2026, 16(5), 2609; https://doi.org/10.3390/app16052609 - 9 Mar 2026

Viewed by 460

Abstract

Developed by Marais and Rabinowitz, the A²/O process is a pivotal biotechnology for biological nitrogen and phosphorus removal, developed by optimizing the five-stage Phoredox protocol. Renowned for its efficient configuration and straightforward operation, it has been extensively adopted in municipal and [...] Read more.

Developed by Marais and Rabinowitz, the A²/O process is a pivotal biotechnology for biological nitrogen and phosphorus removal, developed by optimizing the five-stage Phoredox protocol. Renowned for its efficient configuration and straightforward operation, it has been extensively adopted in municipal and industrial wastewater treatment projects globally, including numerous facilities in China. However, the conventional A²/O process faces inherent operational challenges: the conflicting SRT requirements between autotrophic nitrifying bacteria (needing long SRT for stable nitrification) and PAOs, intense competition for carbon sources among PAOs and denitrifying bacteria, and the inhibitory effects of residual nitrate and DO on phosphorus release and denitrification. To address these issues, a range of optimization strategies has been developed, including SRT adjustment, carbon source distribution optimization, the integration of biofilm carriers, the addition of external carbon sources, and innovative modified configurations such as the Reversed A²/O, JHB, UCT, and MUCT. These approaches synergistically mitigate nitrate interference and enhance nutrient removal efficiency by decoupling microbial SRT demands, supplementing readily biodegradable carbon sources, and optimizing hydraulic flow paths. Future research should focus on deepening the understanding of the metabolic mechanisms underlying nitrogen and phosphorus removal, developing sustainable and efficient external carbon source systems, refining multi-mode reactor design for engineering scalability, optimizing combined processes for ultra-low C/N ratio wastewater treatment, and advancing low-temperature adaptation technologies. These efforts aim to further improve the process’s efficacy, stability, and sustainability, enabling it to meet increasingly stringent environmental discharge standards. Full article

► Show Figures

Figure 1

21 pages, 4581 KB

Open AccessArticle

Beyond the Floodplain: A Multi-Criteria Framework for Emergency Shelter Placement in Buncombe County, NC

by Kibri Hutchison Everett, Srijana Raut, Tung Le, Sodiq M. Balogun, Shen-En Chen and Jay Wu

Appl. Sci. 2026, 16(5), 2608; https://doi.org/10.3390/app16052608 - 9 Mar 2026

Viewed by 289

Abstract

The catastrophic impact of Hurricane Helene proved that standard FEMA flood maps are often inadequate for assessing risk in complex mountainous terrain. Using Buncombe County, North Carolina, as a case study, this research introduces a replicable framework for siting emergency shelters based on [...] Read more.

The catastrophic impact of Hurricane Helene proved that standard FEMA flood maps are often inadequate for assessing risk in complex mountainous terrain. Using Buncombe County, North Carolina, as a case study, this research introduces a replicable framework for siting emergency shelters based on a multi-dimensional Flood Risk Index. By synthesizing HAND-derived inundation data, land-use intensity, and a machine learning-based Socio-Economic Vulnerability Index (SEVI), we mapped the intersection of hazard and vulnerability. Our analysis reveals a significant misalignment—a large portion of the current shelter network sits in high-risk zones, while safer upland corridors in the north and west remain underutilized. This study delivers a data-driven roadmap for disaster preparedness, ensuring that future shelter placement is not only safe from terrain-driven floods but also strategically and equitably located. Full article

(This article belongs to the Section Environmental Sciences)

► Show Figures

Figure 1

26 pages, 2897 KB

Open AccessArticle

Development and Physicochemical Characterization of Oil-in-Water Cosmetic Creams Containing Vaccinium vitis-idaea (Lingonberry) Fruit Extract

by Daniela Gitea, Manuela Bianca Pasca, Laura Maria Endres, Simona Ioana Vicas, Mirela Marioara Toma, Manuel Alexandru Gitea and Mirela-Liliana Moldovan

Appl. Sci. 2026, 16(5), 2607; https://doi.org/10.3390/app16052607 - 9 Mar 2026

Viewed by 272

Abstract

The purpose of this investigation was to develop and physicochemically characterize two natural O/W cosmetic cream prototypes (LC1, LC2) containing 5% (w/w) of a Vaccinium vitis-idaea (lingonberry) fruit extract (LE) together with their corresponding blank formulations (LC1-BL, LC2-BL). The [...] Read more.

The purpose of this investigation was to develop and physicochemically characterize two natural O/W cosmetic cream prototypes (LC1, LC2) containing 5% (w/w) of a Vaccinium vitis-idaea (lingonberry) fruit extract (LE) together with their corresponding blank formulations (LC1-BL, LC2-BL). The extract was obtained by hydroalcoholic maceration followed by solvent removal and was characterized for total phenolic, flavonoid, and monomeric anthocyanin content. Its antioxidant capacity was evaluated using DPPH, FRAP, CUPRAC, and ABTS assays. The phenolic profile was further explored by HPLC–DAD–ESI(+), enabling tentative identification of phenolic subclasses previously reported in the literature to be associated with antioxidant properties. The prepared creams were evaluated for qualitative organoleptic properties, pH, texture (hardness, adhesiveness, and spreadability), viscosity, and accelerated conditions of stability. All formulations were stable, and no phase separation occurred; however, the addition of the extract modified their color and odor and decreased the pH to values within the physiological skin pH range. An in-silico safety evaluation of the constituents (MoS and TTC) found a good toxicological profile at concentrations employed. Overall, the results support the feasibility of incorporating lingonberry fruit extract into O/W cosmetic cream systems and demonstrate that appropriate formulation design allows the development of stable products with defined physicochemical and mechanical characteristics. Full article

(This article belongs to the Special Issue Development of Innovative Cosmetics—2nd Edition)

► Show Figures

Figure 1

15 pages, 2566 KB

Open AccessArticle

Cytocompatibility and Antibacterial Evaluation of Plant-Mediated Copper Oxide Nanoparticles Synthesized from Ginger, Garlic, and Red Onion Extracts Versus Synthetic Copper Oxide for Biomedical Applications

by Muna M. Kareem, Hussain A. Jaber and Basma A. Al-Ghali

Appl. Sci. 2026, 16(5), 2606; https://doi.org/10.3390/app16052606 - 9 Mar 2026

Viewed by 266

Abstract

Green-synthesis routes for producing CuO nanoparticles offer a simplified, sustainable, and low-cost replacement for conventional chemical methods, eliminating the need for harsh chemicals and providing an easily scalable process for industrial-level production. Although numerous studies have investigated synthesizing CuO nanoparticles from single plant [...] Read more.

Green-synthesis routes for producing CuO nanoparticles offer a simplified, sustainable, and low-cost replacement for conventional chemical methods, eliminating the need for harsh chemicals and providing an easily scalable process for industrial-level production. Although numerous studies have investigated synthesizing CuO nanoparticles from single plant extracts, comparative assessments of multi-plant-mediated CuO nanoparticles alongside synthetic CuO remain limited. In this work, CuO nanoparticles were green-synthesized from three different plant sources, namely ginger, red onion peels, and garlic, and their physicochemical and biological properties were tested against the synthetic CuO. All plant extracts produced pure-phased monoclinic CuO nanoparticles as confirmed by UV–Vis, XRD, FTIR, and SEM/EDX analyses. SEM showed distinct nanoparticle morphologies, with CuO from ginger extract exhibiting uniform nanocubes, while nanoparticles from red onion and garlic extracts exhibited more aggregated and irregular structures. Their crystallite sizes were 8–9 nm lower than the ~11 nm observed for the synthetic CuO, highlighting the phytochemical role in shaping the nanoparticles’ morphology. The antibacterial efficacy against S. aureus and E. coli showed that ginger-derived and synthetic CuO had the strongest bacterial inhibition and bactericidal potency compared to onion- and garlic-derived CuO samples. However, synthetic CuO had the highest cytotoxicity risk, hindering its suitability for biological uses, while CuO-ginger maintained good cell viability at moderate concentrations. CuO-onion and CuO-garlic gave lower antibacterial cytocompatibility performance due to their thicker capping layers, which led to decreased Cu²⁺ release and ROS production. Ginger-derived CuO achieved an optimal trade-off between antibacterial and cytotoxic efficiency, highlighting its prospects as a candidate for biomedical applications. Full article

(This article belongs to the Section Biomedical Engineering)

► Show Figures

Graphical abstract

29 pages, 8371 KB

Open AccessArticle

A Novel Inlet Guiding Structure for Pressure-Loss Reduction in Gas–Liquid Cyclone Separators

by Dongjing Chen, Jin Zhang, Yujie Cheng, Jihui Wang, Zhiyuan Wang, Ying Li and Xiangdong Kong

Appl. Sci. 2026, 16(5), 2605; https://doi.org/10.3390/app16052605 - 9 Mar 2026

Viewed by 309

Abstract

Gas–liquid cyclone separators are an efficient and emerging method for air removal in hydraulic systems, yet often suffer from excessive pressure loss. A novel contracting inlet guiding structure is proposed to minimize hydraulic losses. This study adopts a comprehensive methodology combining theoretical modeling, [...] Read more.

Gas–liquid cyclone separators are an efficient and emerging method for air removal in hydraulic systems, yet often suffer from excessive pressure loss. A novel contracting inlet guiding structure is proposed to minimize hydraulic losses. This study adopts a comprehensive methodology combining theoretical modeling, computational fluid dynamics (CFD) using the Reynolds Stress Model (RSM), and experimental validation. A theoretical pressure-loss model incorporating the diminishing-returns effect of the contraction angle was established. Simulations revealed that increasing the contraction angle reduces energy dissipation by improving the uniformity of the tangential-velocity field. Based on the balance between pressure-loss reduction and degassing potential, a contraction angle of 11° was identified as the optimal design and experimental tests on a prototype confirmed the validity of the numerical model. The results demonstrate that, compared to the conventional straight tangential inlet, the optimized inlet reduces the pressure loss by approximately 30% under rated conditions. The experimental–numerical discrepancy decreases significantly with flow rate, achieving a relative error of approximate 10% at the design flow rate. These findings provide a theoretical basis and practical guidance for the low-energy design of hydraulic cyclone separators. Full article

► Show Figures

Figure 1

31 pages, 11710 KB

Open AccessArticle

Geology, Alteration, Geochemistry, and Regional Sulfur Isotope Constraints on Pb–Zn ± Cu Mineralization in the Biga Peninsula (NW Türkiye): Insights from the Kocayayla Deposit

by Sinan Akıska and Gökhan Demirela

Appl. Sci. 2026, 16(5), 2604; https://doi.org/10.3390/app16052604 - 9 Mar 2026

Viewed by 328

Abstract

The Kocayayla Pb–Zn ± Cu vein-type mineralization is located in the Biga Peninsula, northwestern Türkiye. This study aims to constrain the geological, geochemical, and isotopic characteristics of the mineralization and to clarify its genetic classification. The deposit is hosted mainly by andesitic and [...] Read more.

The Kocayayla Pb–Zn ± Cu vein-type mineralization is located in the Biga Peninsula, northwestern Türkiye. This study aims to constrain the geological, geochemical, and isotopic characteristics of the mineralization and to clarify its genetic classification. The deposit is hosted mainly by andesitic and basaltic andesitic rocks as well as schists and is structurally controlled by E–W-trending strike-slip faults. Mineralogical and petrographic identifications, XRD analyses, whole-rock geochemistry, and sulfur isotope data were integrated to evaluate ore-forming processes. Mineralization is temporally and spatially associated with propylitic and phyllic to argillic alteration and is concentrated within zones of intense silicification and chloritization, accompanied by quartz, sericite, kaolinite/nacrite, chlorite, and carbonate assemblages. The ore assemblage is dominated by galena, sphalerite, and subordinate chalcopyrite, with minor fahlore-group minerals. Rare earth element patterns of ore samples (whole rock) overlap with those of the wall rocks, whereas Pb–Zn enrichment reflects selective hydrothermal metal transport. Sulfur isotope compositions show limited internal variation and indicate sulfur derived predominantly from H₂S-dominated magmatic–hydrothermal fluids. Regional comparison of δ³⁴S datasets and reported Au contents across the Biga Peninsula indicates that Au-rich intermediate-sulfidation epithermal systems exhibit broader and more variable sulfur isotope ranges, whereas Au-poor intermediate-sulfidation epithermal systems show relatively restricted and near-zero δ³⁴S values. These features collectively support the classification of the Kocayayla mineralization as an Au-poor intermediate-sulfidation epithermal Pb–Zn system. Full article

(This article belongs to the Section Earth Sciences)

► Show Figures

Figure 1

29 pages, 4023 KB

Open AccessArticle

IoT Technology and Augmented Reality Integrated into Urban Furniture for Tourism 4.0

by Ana Pamela Castro-Martin, Christian Morales Guanga, Josue Rafael Carrera Barrionuevo, Mayra Paucar Samaniego, Martin Monar Naranjo, Jorge Santamaría Aguirre and Andrés López Vaca

Appl. Sci. 2026, 16(5), 2603; https://doi.org/10.3390/app16052603 - 9 Mar 2026

Viewed by 315

Abstract

Tourism 4.0 integrates Industry 4.0 technologies into tourism services to enhance visitor experiences and improve destination management. This study presents the design, implementation, and pilot validation of an integrated IoT–Augmented Reality (IoT–AR) cyber-physical urban node developed for smart tourism infrastructure in Baños de [...] Read more.

Tourism 4.0 integrates Industry 4.0 technologies into tourism services to enhance visitor experiences and improve destination management. This study presents the design, implementation, and pilot validation of an integrated IoT–Augmented Reality (IoT–AR) cyber-physical urban node developed for smart tourism infrastructure in Baños de Agua Santa, Ecuador. The system combines distributed environmental sensing, LoRa-based communication, edge-level preprocessing, cloud data management via RESTful services, and immersive visualization through a cross-platform augmented reality mobile interface. The development followed the TDDM4IoTS methodology, adapted into five phases covering requirements analysis, technological design, modeling, validation, and deployment. The architecture supports contextual real-time information delivery while maintaining low power consumption and robustness under heterogeneous connectivity conditions. Field tests confirmed stable communication between sensor nodes and the gateway, as well as reliable AR marker recognition under varying light and distance conditions. Usability evaluation using the System Usability Scale (SUS) yielded a mean score of 84.38, classified as excellent, with high internal consistency (α ≈ 0.89). The results demonstrate technical feasibility and strong user acceptance, providing a scalable and replicable model for interactive IoT–AR urban systems in smart tourism environments. Full article

(This article belongs to the Special Issue Application of IoT and Cybersecurity Technologies)

► Show Figures

Graphical abstract

14 pages, 256 KB

Open AccessArticle

An Approach to Developing Likert Scale Survey Results Based on the Example of a Research Study Involving a Limited Number of Students

by Marek Gaworski and Aleksandra Daśko

Appl. Sci. 2026, 16(5), 2602; https://doi.org/10.3390/app16052602 - 9 Mar 2026

Viewed by 638

Abstract

Surveys are important tools for collecting knowledge, including student knowledge, and assessing their opinions and behavior. Survey results inspire information processing and selection of a processing method for further knowledge management. In this study, an improved approach to presenting survey results was developed, [...] Read more.

Surveys are important tools for collecting knowledge, including student knowledge, and assessing their opinions and behavior. Survey results inspire information processing and selection of a processing method for further knowledge management. In this study, an improved approach to presenting survey results was developed, utilizing a Likert scale. In the survey, 20 students answered 10 questions (issues) that examined their opinions on the impact of modern technical equipment on dairy production assessment. The feature significance index (FSI) was utilized to inform the development of the survey study results. The FSI is the ratio of the percentage share of the highest to the lowest ratings on a Likert scale. In the case of four issues, none of the students indicated the options had very little impact and little impact. Therefore, the FSI could not be calculated, so a modified version was proposed. After ranking the issues in the survey based on the FSI, the difference in FSI between the best-rated and worst-rated issues was more than 13 times. This difference was less than two times in the modified version of the FSI. A larger difference allows for a more comprehensive interpretation of the survey results. The study confirmed that the small number of survey participants is a key limitation in developing the survey results. Full article

(This article belongs to the Special Issue New Trends in Model-Based Systems Engineering)

31 pages, 2010 KB

Open AccessArticle

Factors’ Influence on Human–Computer Negotiation Results—A Systematic Evaluation

by Yushan Liu, Rustam Vahidov and Raafat Saade

Appl. Sci. 2026, 16(5), 2601; https://doi.org/10.3390/app16052601 - 9 Mar 2026

Viewed by 257

Abstract

Artificial intelligence (AI) and computer agents are increasingly shaping daily decision-making and commercial interactions. This study investigates the influence of computer agents’ attributes on negotiation results and proposed a systematic method to evaluate the negotiation outcomes. Specifically, it examines the effects of negotiation [...] Read more.

Artificial intelligence (AI) and computer agents are increasingly shaping daily decision-making and commercial interactions. This study investigates the influence of computer agents’ attributes on negotiation results and proposed a systematic method to evaluate the negotiation outcomes. Specifically, it examines the effects of negotiation timespan (synchronous vs. asynchronous), concession tactics, and issue-search mechanisms on both economic and perceptual results in human-agent negotiation. In an experiment, human buyers negotiated purchase of mobile plan contracts with computer agents programmed with one of three concession tactics (conceding, neutral, or competitive) and one of two issue search mechanisms (breadth-first or depth-first). Negotiations occurred under either synchronous or asynchronous timeframes. The experimental results suggest that on the group (dyad) level, timespan has marginal effects on agreement rate, while tactic has a significant impact. On the individual level, agents’ tactics have significant effects on the objective outcomes, while search mechanisms have a significant influence on the subjective outcomes. Full article

(This article belongs to the Special Issue Human-Computer Interaction: Advances, Challenges and Opportunities)

► Show Figures

Figure 1

25 pages, 6915 KB

Open AccessArticle

EXAONE-VLA: A Unified Vision–Language Framework for Mobile Manipulation via Semantic Topology and Hierarchical LLM Reasoning

by Jeong-Seop Park, Yong-Jun Lee, Jong-Chan Park, Sung-Gil Park, Jong-Jin Woo and Myo-Taeg Lim

Appl. Sci. 2026, 16(5), 2600; https://doi.org/10.3390/app16052600 - 9 Mar 2026

Viewed by 597

Abstract

This paper proposes a unified vision–language framework that translates user instructions into navigation for the mobile base and actions for the manipulator in indoor environments. In general, occupancy grid maps constructed via SLAM capture solely the geometric layout of the environment. This renders [...] Read more.

This paper proposes a unified vision–language framework that translates user instructions into navigation for the mobile base and actions for the manipulator in indoor environments. In general, occupancy grid maps constructed via SLAM capture solely the geometric layout of the environment. This renders the robot incapable of leveraging the semantic information required for object distinction. The proposed method encodes semantic information from vision–language models and the robot’s pose in a textual format, referred to as a semantic topological graph. Specifically, the models including GroundingDINO, LG EXAONE, and SAM2 extract object-level semantic information, which is subsequently used to identify room characteristics. A large language model then interprets user instructions to identify the final destination for navigation within the semantic topological graph, followed by reasoning to determine the suitable action network. Notably, the proposed text-based representation facilitates a substantial reduction in inference time, and its effectiveness is validated through real-world experiments. Full article

(This article belongs to the Special Issue Deep Reinforcement Learning for Multiagent Systems)

► Show Figures

Figure 1

23 pages, 4103 KB

Open AccessArticle

Anchorage Strength Model for Large-Diameter Headed Bars Anchored at the Cutoff Point

by Hyung-Suk Jung

Appl. Sci. 2026, 16(5), 2599; https://doi.org/10.3390/app16052599 - 9 Mar 2026

Viewed by 203

Abstract

Design guidance for headed-bar development remains uncertain for large-diameter bars at cutoff points, where bar termination increases anchorage demand and confinement is often limited. This study quantified the anchorage behavior of 43 and 57 mm headed bars and established a regression-based strength model [...] Read more.

Design guidance for headed-bar development remains uncertain for large-diameter bars at cutoff points, where bar termination increases anchorage demand and confinement is often limited. This study quantified the anchorage behavior of 43 and 57 mm headed bars and established a regression-based strength model grounded in a splitting-controlled bond–bearing mechanism. Nineteen reinforced concrete beam specimens were tested under four-point loading configured to place the bending-moment inflection point at the head location. The primary variables were the development length (l_dt = 12–28d_b), concrete compressive strength (f_c′ = 42 and 70 MPa), clear side cover, clear spacing, and transverse reinforcement index (K_tr/d_b = 0–2.0). All the specimens failed by splitting prior to bar yielding, characterized by longitudinal cracking along the development region and cover spalling near the head. The anchorage strength increased with concrete compressive strength and development length and was most strongly enhanced by transverse reinforcement (up to ~60%). At failure, the bond contributed 70–86% of the developed stress, while the head-bearing contribution increased with confinement. Existing ACI 318-19 and KDS-2021 provisions were generally unconservative, particularly for unconfined specimens. The proposed bond–bearing model showed a close agreement with the test database (mean test/prediction = 0.99; COV = 4.72%) within stated parameter limits. Full article

(This article belongs to the Section Civil Engineering)

► Show Figures

Figure 1

34 pages, 10321 KB

Open AccessArticle

Multi-Strategy Enhanced NSGA-III Algorithm and Its Application in the Variable-Thickness Design of Morphing Leading Edges

by Fan Yang, Guang Yang, Hong Xiao, Runchao Zhao, Rongqiang Liu and Hongwei Guo

Appl. Sci. 2026, 16(5), 2598; https://doi.org/10.3390/app16052598 - 9 Mar 2026

Viewed by 293

Abstract

To address the strongly coupled and highly nonlinear optimization problems arising from the increasing system complexity, optimization objectives, and variable dimensions in practical engineering applications, this paper proposes a multi-strategy enhanced NSGA-III algorithm (MSNSGA-III) by introducing K-means clustering, an adaptive hybrid operator, and [...] Read more.

To address the strongly coupled and highly nonlinear optimization problems arising from the increasing system complexity, optimization objectives, and variable dimensions in practical engineering applications, this paper proposes a multi-strategy enhanced NSGA-III algorithm (MSNSGA-III) by introducing K-means clustering, an adaptive hybrid operator, and an assistant evolutionary population strategy on the basis of the NSGA-III algorithm. This algorithm overcomes the performance limitations of the original algorithm in large-scale search with multiple variables. By employing the DTLZ test functions with different variable dimensions and conducting comparisons with six other representative algorithms, the proposed algorithm is proven to have strong competitiveness in terms of diversity and convergence speed. To reflect the superiority of the algorithm in practical applications, this paper establishes a variable-thickness optimization model for the morphing leading edge. By adopting the spline curve-based optimization variable control strategy and the MSNSGA-III algorithm, the optimal thickness distribution of the leading edge skin is obtained. The results show that, compared with the leading edge with a fixed skin thickness of 1.5 mm, the optimized variable thickness skin leading edge achieves 43.6% improvement in shape maintaining accuracy, 40.9% improvement in deformation accuracy, and 17.5% reduction in driving force. Full article

(This article belongs to the Section Mechanical Engineering)

► Show Figures

Figure 1

17 pages, 776 KB

Open AccessArticle

A Two-Stage Adversarial Training Method Based on Stability Contrastive Learning to Enhance Adversarial Robustness

by Wenjuan Ren, Zhanpeng Yang and Guangzuo Li

Appl. Sci. 2026, 16(5), 2597; https://doi.org/10.3390/app16052597 - 9 Mar 2026

Viewed by 266

Abstract

Neural network models are highly susceptible to adversarial sample attacks, causing significant differences in model predictions with even minor perturbations to the samples. Adversarial training is a kind of effective technique for resisting sample adversarial attacks. Traditional adversarial training methods are all single-stage [...] Read more.

Neural network models are highly susceptible to adversarial sample attacks, causing significant differences in model predictions with even minor perturbations to the samples. Adversarial training is a kind of effective technique for resisting sample adversarial attacks. Traditional adversarial training methods are all single-stage training, and in the middle and later stages of training, there is a serious issue of robust overfitting. The accuracy of the adversarial training model does not increase and may even experience severe degradation. For this issue, we propose a multi-stage stability contrastive adversarial training method based on Ulam stability, which performs adversarial training in an optimized space with stability constraints to improve the robustness accuracy and training stability of the model. In the first stage, a stability adversarial training strategy is adopted to enable the model to rapidly improve robust accuracy before overfitting. In the second stage, a stability contrastive learning strategy is employed, focusing on suppressing overfitting of the model and further enhancing robust accuracy. This two-stage adversarial training method can not only improve the robustness accuracy of the model, but also effectively suppress overfitting phenomena. This method has plug and play pendant coupling capability, which can be combined with existing multi-class models to further improve the robustness accuracy of the benchmark model. In addition, this method has the characteristic of stable convergence, which can reduce the dependence on early stopping conditions and make the model training more flexible. Comparative experiments on multiple datasets have also validated the effectiveness of the proposed method. Full article

► Show Figures

Figure 1

Journal Menu

Journal Browser

Appl. Sci., Volume 16, Issue 5 (March-1 2026) – 465 articles

Further Information

Guidelines

MDPI Initiatives

Follow MDPI