Search Results (2,117)

The construction sector is undergoing a rapid transition toward more resilient, sustainable, and digitally connected systems, creating increasing demand for materials capable of providing functions beyond conventional structural performance. In this context, smart materials have emerged as promising solutions due to their ability to respond to mechanical, thermal, chemical, or electromagnetic stimuli through adaptive behaviors such as self-healing, structural sensing, energy regulation, vibration control, and reversible deformation. Despite growing scientific interest, available knowledge remains fragmented across specific material families and isolated application domains. Therefore, this study presents a PRISMA-based systematic review of smart materials in construction using peer-reviewed journal literature indexed in Scopus during the 2021–2026 period. The review examines the principal smart material families currently applied in construction, including self-healing concretes, self-sensing cementitious systems, Shape Memory Alloys (SMA), piezoelectric materials, phase change materials, adaptive coatings, conductive nanocomposites, and multifunctional geopolymers. Their engineering functions, structural and architectural applications, reported performance characteristics, sustainability contributions, digital integration potential, and implementation barriers are comparatively discussed and qualitatively synthesized based on the reviewed literature. The findings indicate that smart materials can improve durability, structural health monitoring, seismic resilience, thermal efficiency, lifecycle performance, and carbon reduction when properly integrated into buildings and infrastructure. However, large-scale adoption remains constrained by high initial costs, manufacturing scalability, regulatory uncertainty, long-term durability validation, and limited market confidence. The review further shows that the greatest future potential lies in combining material intelligence with IoT platforms, artificial intelligence, BIM environments, and digital twins. Overall, smart materials are positioned as strategic enablers of next-generation low-carbon, adaptive, and intelligent construction systems. Full article

Intrusion detection serves as a core technology for securing heterogeneous edge networks, including IoT, industrial edges, and 5G networks. However, existing federated learning-based intrusion detection systems suffer from environmental heterogeneity, limited sample availability, and severe class imbalance—issues that result in inefficient resource allocation and compromised detection performance against rare attacks. In this paper, we propose a novel lightweight intrusion detection model for heterogeneous edge networks, named FedNIDS-CNN, which is based on dynamic distillation-aided federated learning with a CNN backbone. In the data preprocessing phase, a two-level class balancing strategy integrating nearest-neighbor interpolation augmentation and adaptive synthetic sampling is employed to ensure distortion-free sample synthesis. For feature and model optimization, principal component analysis (PCA) is used to reduce the dimensionality of traffic features, while a lightweight 1D-CNN is adopted as the base model to alleviate computational overhead on edge devices. During federated training and knowledge aggregation, a dynamic weight distillation loss mechanism is designed to enhance the model’s ability to recognize minority-class attacks. Meanwhile, the federated framework supports client-side local training and server-side weighted soft-label aggregation, enabling effective knowledge fusion across heterogeneous models. Experimental results on the CICIDS2017 dataset demonstrate that the proposed method achieves an accuracy of 98.55% and an F1-score of 98.40%. Benefiting from the soft-label transmission and parameter-free aggregation design, the framework gets rid of the constraint of homogeneous model architecture and natively supports heterogeneous network models and edge devices with different computing capabilities. It also significantly reduces communication traffic and per-round training latency, confirming its excellent real-time performance and applicability in resource-constrained edge environments. Full article

Plants must continuously balance the trade-offs between growth and defense, a constraint that is exacerbated by biotic and abiotic stresses, particularly when they occur together. Ethylene (ET) serves as a central, integrative regulatory node controlling this by linking developmental programs to stress-responsive signaling networks. Advances at the molecular and systems levels have revealed that ET mediates the redistribution of metabolic resources via coordinated regulation of its synthesis, perception, and downstream signaling. The ETR (Ethylene Receptor)-CTR1 (Constitutive Triple Response 1)-EIN2 (Ethylene Insensitive 2)-EIN3(Ethylene Insensitive 3) signaling module lies at the core of this network, integrating multiple hormonal pathways. Through dynamic crosstalk with jasmonic acid (JA), salicylic acid (SA), abscisic acid (ABA), auxin (AUX), and gibberellins (GA), ET enables the fine-tuned coordination of growth inhibition, immune activation, and stress acclimation in response to environmental fluctuations. Processes such as induced systemic resistance, programmed cell death, and architectural plasticity further reinforce this regulatory framework, with ethylene-responsive transcription factors, including ERFs (ethylene responsive factor gene family) and WRKYs, acting as critical convergence points. Emerging insights into ACC (1-aminocyclopropane-1-carboxylic acid) -dependent signaling, chromatin remodeling, and tissue-specific regulation expand the functional scope of ET beyond traditional hormone paradigms. At the same time, the ability of pathogens to manipulate ET signaling underscores its dual role in both promoting immunity and facilitating susceptibility. By integrating molecular, physiological, and ecological perspectives, this review highlights ET as a central coordinator of plant stress resilience and growth optimization, providing a unifying framework for understanding how plants adapt to complex and dynamic environments. Full article

With the increasing complexity of unmanned aerial vehicle (UAV) missions in complex obstacle environments, cooperative hunting of maneuvering ground targets by UAV swarms has become an important problem for multi-agent autonomous decision-making. This paper focuses on a simulated three-UAV hunting scenario in a two-dimensional obstructed environment, where UAVs must search for, approach, encircle, and continuously track a target while avoiding static obstacles under local observation. To address the problem of multi-UAV cooperative hunting of dynamic targets in complex obstacle environments, this paper proposes a curriculum learning (CL)-based Multi-Agent Proximal Policy Optimization algorithm, termed CL-MAPPO. Specifically, a three-stage progressive training curriculum is designed to overcome the challenges of low exploration efficiency, slow environmental adaptation, and difficult convergence of cooperative hunting policies faced by multi-agent deep reinforcement learning in hunting tasks, thereby gradually enhancing the cooperative hunting capability of UAVs in complex environments. Curriculum I employs fixed obstacles and a stationary target position to train the UAVs’ basic obstacle avoidance and target search abilities. Curriculum II introduces randomly generated obstacles and target positions to improve the UAVs’ adaptability to varying environments. Curriculum III further incorporates a dynamic target, prompting the UAVs to learn effective hunting strategies against maneuvering targets. The simulation experiment includes ablation experiments against MAPPO without curriculum learning and comparative simulations against MADDPG and MADQN, using reward convergence curves and trajectory visualizations to evaluate the training results. The results show that, under the same training episodes in the ablation experiment, CL-MAPPO reaches a higher and more stable reward level than vanilla MAPPO, indicating improved learning efficiency without increasing model complexity. In the comparative experiment, the CL-MAPPO algorithm achieved a higher success rate in cooperative hunting. These simulation experiments verify the effectiveness and superiority of the CL-MAPPO algorithm in multi-agent cooperative hunting tasks. Full article

The economic viability of second-generation (2G) bioethanol production depends on the availability of robust, multistress-tolerant yeast strains capable of withstanding harsh industrial conditions. This study investigates animal manure as a novel ecological niche for discovering such strains, as microbes in these environments naturally adapt to high organic loading and fluctuating temperatures. From eighty-six initial isolates, twenty-nine demonstrated superior xylose fermentation at 37 °C. Eight high-performing isolates (C2-1, B1-2, B1-6, B2-6, B2-8, G1-4, G1-5, and G2-4) exhibited exceptional tolerance to ethanol, high temperatures, and lignocellulosic-derived inhibitors (acetic acid, formic acid, furfural, and vanillic acid). Molecular identification classified isolate C2-1 as Pichia kudriavzevii and the remaining seven as Candida tropicalis. In synthetic media, C. tropicalis B2-8 produced up to 16.33 g/L of ethanol using xylose (60 g/L) as the sole carbon source. While the undetoxified, highly acidic sugarcane bagasse hydrolysate completely inhibited yeast growth, the industrial potential of these strains was successfully validated using the concentrated, undetoxified enzymatic hydrolysate derived from the acid-pretreated sugarcane bagasse solids, which contained 30.15 g/L glucose and 25.58 g/L xylose. P. kudriavzevii C2-1 achieved ethanol titers of 6.02 g/L and 5.71 g/L at 37 °C and 40 °C, respectively. The C. tropicalis strains outperformed P. kudriavzevii, yielding 6.12–6.35 g/L at 37 °C and maintaining 5.75–6.19 g/L at 40 °C. These findings underscore the potential of manure-derived yeasts as resilient biocatalysts. Although their fermentation yields remain relatively low and require further metabolic optimization, their ability to survive and ferment in this concentrated, undetoxified enzymatic hydrolysate at elevated temperatures makes them promising candidates for further development in high-temperature ethanol fermentation (HTEF), offering a potential pathway toward reducing cooling costs associated with 2G biorefineries. Full article

In order to improve the speed regulation accuracy, dynamic response and operation robustness of an automatic guided vehicle (AGV) in a complex road disturbance environment, this paper studies an adaptive AGV speed regulation system based on EKF state estimation on the basis of AGV dynamics modeling and adaptive control. Firstly, through the electrical-mechanical coupling modeling of the AGV drive system, state space construction and external unknown disturbance equivalent design, a unified electromechanical coupling simulation and physical verification environment is built, which lays a model foundation for the research of the speed control algorithm. Secondly, based on the optimal control model of PID and LQR with first-order lead compensation, an EKF adaptive speed regulation model is constructed by combining the extended Kalman filter and adaptive control to realize the online estimation and dynamic compensation of unknown disturbances. Finally, based on MATLAB/Simulink simulation platform and the STM32 embedded experimental platform, the rationality and robustness of the proposed speed control strategy are verified by speed-mutation conditions, load-disturbance condition and a physical verification experiment. The results show that the overshoot of the EKF adaptive control strategy is only 1.8%, which is 84.1% lower than that of PID control and 61.7% lower than that of LQR control. The rise time is 42% shorter than PID and 23% shorter than LQR. The recovery time under load disturbance is 58% shorter than that of PID and 31% shorter than that of LQR. EKF adaptive control is significantly better than PID and LQR in overshoot, rise time and control stability. The disturbance rejection ability and dynamic recovery speed are greatly improved, which can ensure the high robustness and smooth operation of the AGV speed control system under complex working conditions, effectively enhance the response and compensation ability of the system to sudden disturbances, and better meet the actual needs of AGV speed control in complex engineering scenarios. Full article

Particle Image Velocimetry (PIV) has recently evolved from a costly, specialized technique into an accessible method thanks to affordable hardware and open-source software. This work introduces a time calibration validation method tailored for low-cost or Do-It-Yourself (DIY) PIV systems. By utilizing inexpensive components such as light-dependent resistors (LDRs), basic resistors, and data acquisition devices or microcontrollers, the study enables accurate timing analysis of light pulses from synchronized lasers or LEDs. Experimental data obtained in real time using a National Instruments USB-6003 DAQ device confirm the system’s ability to detect light pulses with high temporal resolution. Through voltage signal interpretation, the synchronization accuracy of light sources is validated across different sampling rates. Moreover, the study demonstrates how the internal frequency settings of PIVlab, an open-source PIV software package, can be customized to enhance acquisition flexibility. Timing deviations of up to 20% were identified across selected default frequency settings. The proposed method ensures that low-cost systems maintain sufficient accuracy for phase-sensitive flow measurements, such as oscillatory flow or wave action, contributing to the broader adoption of PIV in resource-limited environments. It presents a low-cost method for validating timing accuracy in PIV systems, employs widely available components and is adaptable to multiple platforms, and enables precise synchronization checks critical for flow visualization. Full article

Resistance training (RT) is commonly used to enhance neuromuscular performance and sprint swimming outcomes. However, the optimal relative load for elite junior swimmers remains unclear. In particular, little is known about whether very low relative loads can elicit meaningful adaptations while minimizing neuromuscular fatigue in athletes exposed to high concurrent training demands. Therefore, the aim of this study was to compare the effects of two land-based RT programs differing only in relative load intensity (40–50% vs. 55–65% 1RM), performed with maximal intended concentric velocity, on strength, jumping ability, and 50 m freestyle swimming performance in elite junior swimmers. Eighteen elite junior swimmers (15.6 ± 0.9 years) from a national high-performance program were randomly assigned to a low-load (40–50% 1RM; n = 9) or moderate-load (55–65% 1RM; n = 9) group. Both groups completed an 8-week RT program (2 sessions·week⁻¹) with identical exercise selection, volume, execution velocity, and in-water training load. Neuromuscular performance (countermovement jump, squat, bench press, and pull-up strength) and swimming performance (50 m freestyle from the starting block and in-water start) were assessed pre- and post-intervention. Both RT protocols improved squat and bench press strength and 50 m freestyle performance, whereas significant improvements in countermovement jump, pull-up strength, and maximal pull-up repetitions were observed only in the low-load group. Significant group × time interactions were found for countermovement jump, maximal number of pull-up repetitions, and 50 m freestyle performance from the starting block, indicating more favorable changes over time in the low-load group. In conclusion, both low- and moderate-load high-velocity RT improved neuromuscular and 50 m freestyle performance outcomes in elite junior swimmers. However, the low-load RT (40–50% 1RM) appeared to provide additional benefits in specific outcomes (i.e., jumping, pull-ups, and 50 m performance from the starting block). These findings suggest that relatively low loads may be a practical alternative to moderate-load RT in high-volume swimming training environments. Full article

Background: In increasingly dynamic and uncertain organizational environments, employees’ proactive behavior—characterized by self-initiation, future orientation, and change orientation—is critical for organizational adaptability and long-term competitiveness. Prior research has primarily examined how externally provided job resources stimulate proactive behavior. More recent work has begun to consider employees’ personal resources, but it largely adopts a capability level-based view, conceptualizing them as self-evaluations of individuals’ ability to control and influence their environment. This focus overlooks capability malleability-based personal resources that shape more fundamental beliefs about the malleability of human capability. Objective: Drawing on the job demands–resources (JD–R) model, this study investigates how employees’ growth mindset—reflecting beliefs that human capability can be developed—promotes proactive behavior by alleviating workplace anxiety, an anticipatory emotional state rooted in concerns about future work-related threats. We further examine time pressure as a key boundary condition. Method: A three-wave, multisource survey design was employed, collecting data from 326 employee–supervisor dyads. Results: The results show that employees’ growth mindset is negatively associated with workplace anxiety, which in turn positively predicts proactive behavior. Moreover, time pressure strengthens both the anxiety-buffering effect of growth mindset and the indirect effect of growth mindset on proactive behavior via workplace anxiety. Conclusions: By incorporating capability malleability-based personal resources into the JD–R model, this study advances understanding of the antecedents of proactive behavior beyond capability level-based self-evaluations toward deeper beliefs about the malleability of human capability. Applications: This study offers practical implications for managers seeking to cultivate employee proactivity. Full article

Tropical rainforests are characterized by high species diversity and complex canopy structure, making accurate forest stand classification important for ecosystem assessment, biodiversity monitoring, and forest carbon estimation. However, single-source remote sensing data lacks sufficient discrimination ability to address the issue of spectral similarity among classes, and conventional convolutional neural networks often struggle to extract discriminative features and integrate heterogeneous data in highly complex forests. To address these challenges, this study developed a dual-branch deep learning framework that integrates DenseNet and ConvNeXt for classification in Hainan Tropical Rainforest National Park. The framework combines sub-meter Google Earth imagery to capture spatial–textural detail with multi-temporal Sentinel-2 imagery to represent phenological variation. The results showed that multi-temporal Sentinel-2 data outperformed single-date imagery by capturing phenological patterns, and that the fusion of high-resolution spatial information and multi-temporal spectral information yielded higher accuracy than either data source alone. The dual-branch model achieved an overall accuracy of 94.47% and a Kappa coefficient of 0.94, outperforming all benchmark models. These findings indicate that branch-specific feature extraction and adaptive fusion can improve fine-scale classification in complex tropical rainforest environments. The proposed framework provides a practical approach for fine-scale forest stand mapping and may support biodiversity monitoring, ecological assessment, and sustainable forest management. Full article

This study addresses the limitation of single-factor environmental assessment in autonomous sweet potato farming under open-field conditions. An IoT-based sweet potato growth environment monitoring and comprehensive evaluation system was developed by integrating an STM32-based acquisition terminal, multi-sensor data collection, Narrow Band Internet of Things (NB-IoT) transmission, and cloud-based visualization. Five key environmental variables, namely soil temperature, soil moisture, soil available nitrogen, photosynthetically active radiation (PAR), and CO₂, were continuously monitored. To improve the evaluation of heterogeneous and uncertain environmental information, a multi-factor environmental quality assessment method combining fuzzy membership functions and an improved D-S evidence theory was proposed. Field experiments were conducted in Danzhou, Hainan, China, and 600 valid synchronized samples were obtained for analysis. The results showed that most samples were classified as Suitable (63.5%), followed by Normal (30.8%) and Poor (5.7%), with a mean comprehensive environmental score of 0.802. Among the monitored variables, PAR and soil temperature showed relatively high adaptive weights, indicating their important roles in environmental quality discrimination. Furthermore, the comprehensive environmental evaluation result exhibited a significant positive correlation with sweet potato yield (r = 0.6501, p = 2.3724 × 10⁻⁷³), demonstrating good explanatory ability for yield variation. The proposed system provides an effective technical framework for real-time environmental monitoring, quantitative suitability evaluation, and precision management in autonomous sweet potato farming. Full article

Deinterleaving of radar signals is an important issue in processing radiolocation data as a part of the analysis of complex electromagnetic signal environments. The introduction contains a draft of radiolocation signal deinterleaving. The authors began with a high-level view of contemporary challenges in radar signal processing and concluded with the genesis of the signals deinterleaving problem and its technical details. A chronicle of the Hypergeometrical Divide (HypGD) algorithm, describing important stages of its development, and a synthesis of knowledge from scientific reports on the examined method were also presented. The aim of this article is to define the performance of the HypGD algorithm adapted to the deinterleaving of radiolocation signals. The research focused on evaluating the clustering performance of the adapted HypGD algorithm, including its ability to determine groups corresponding to emission source types and to support the analysis of radar pulses in complex signal environments. The authors referred to recent publications in the field of radar signal deinterleaving to systematize the current state of knowledge in this area. A detailed, systematic review of significant works on the HypGD method and its applications was provided. The research used real, anonymized data. The results allowed the formulation of conclusions that contribute to the current state of knowledge. For the first time, the effectiveness of the HypGD algorithm adapted to the deinterleaving of radiolocation signals through pulse clustering has been demonstrated. Full article

Listeria monocytogenes is a persistent foodborne pathogen capable of forming biofilms and surviving in food-processing environments. This study investigated the impact of high-pressure processing (HPP) at 200 and 400 MPa/5 min on biofilm viability, biomass, and expression of nine virulence-associated genes in L. monocytogenes strains (n = 6) belonging to the serogroups IIa (LM8, LM40, LM41) and IVb (LM14, LM47, LM48). The pressure levels applied were selected to represent sublethal HPP conditions (below 600 MPa) that allowed the survival of the strains and thus enabled the investigation of adaptive responses in cells that escape complete inactivation. Biofilms were cultivated on stainless-steel 304, polyethylene terephthalate, and polypropylene coupons under static conditions at 25 °C for 72 h and 168 h. Biofilm viability [log₁₀(CFU/cm²)] was assessed by plate count method and biomass quantified via the biofilm production index (BPI). The cultures were subjected to HPP treatment and their ability to form biofilms was re-evaluated. HPP significantly (p < 0.05) reduced biofilm viability and biomass on all types of surfaces tested. Gene expression analysis revealed a pressure-dependent (p < 0.05) modulation of flaA and sigB, while other virulence genes (agrA, agrC, actA, prfA, hly, inlB, and degU) were generally downregulated (gene expression ratio values below 1). Serogroup IVb strains exhibited enhanced stress responses and lower biofilm survival on polyethylene terephthalate and polypropylene surfaces. These findings demonstrate that HPP modulates both phenotypic and genotypic traits linked to L. monocytogenes persistence, emphasizing the need to optimize pressure parameters and surface materials to prevent biofilm formation in HPP-treated food systems. Full article

Human activity recognition—HAR—plays a crucial role in the lives of patients battling neurodegenerative diseases. These debilitating conditions, such as Alzheimer’s or Parkinson’s, affect individuals’ ability to perform daily tasks autonomously and safely. HAR technology offers an invaluable solution by enabling real-time monitoring and assistance, helping to maintain independence and quality of life for patients. Additionally, this technology provides a valuable data source for doctors and caregivers, allowing for more precise and personalized care, which can make a difference in managing and treating these neurodegenerative diseases. The objective of this review is to identify the contribution of Transfer Learning and Reinforcement Learning in supporting the processes of daily activity recognition, thus enhancing the quality of life for patients. As this is a trending topic, the literature surrounding it is quite dispersed, which is why this review aims to present the current line of research in this field. To carry out this analysis, the science tree paradigm was used, which establishes two fundamental stages of analysis. The first is delimited by scientometrics, where the leading countries in the application of such technologies can be identified. This review highlights the evolution in the use of transfer learning and reinforcement learning in HAR in the healthcare field, where these techniques have significantly improved the accuracy and adaptability of real-time monitoring systems. The studies reviewed indicate that transfer learning has allowed models to adapt to data variations without requiring large volumes of manual labeling, which is essential in clinical and patient monitoring contexts. Additionally, reinforcement learning has optimized decision-making in complex scenarios, enabling activity recognition systems to dynamically adjust monitoring parameters, enhancing detection and response to critical or unusual activities in multi-user environments. These advances demonstrate that, by integrating these approaches, greater personalization and robustness can be achieved in human activity recognition, thereby improving the quality of life for patients in clinical settings. Full article

To reduce confusion between adjacent maturity categories, as well as false detections and low detection accuracy caused by complex backgrounds in tomato object detection, this study develops an improved YOLOv7-based model, named YOLO-RCM (Reduce classes misjudgment). First, a stability-enhanced ECANet channel attention module is embedded into the feature pyramid network (FPN) to strengthen discriminative channel responses. Second, a DCNv2-based deformable convolution enhancement module, namely DCNConv with adaptive magnitude constraints, is incorporated into the backbone network to alleviate feature misalignment caused by shape variation, partial occlusion, and fine-grained appearance differences in tomato maturity detection. Third, the WIoU v3 loss function is adopted to refine bounding box regression stability. The model was evaluated on the public Laboro Tomato dataset and TomatOD dataset. Experimental results indicate that YOLO-RCM obtains 83.7% Precision and 89.6% mAP@0.5, exceeding the baseline by 3.3 and 1.2 percentage points, respectively. Its Recall is 80.5%, with a decrease of 0.8 percentage points, whereas GFLOPs are reduced to 96.9, 6.3 lower than the baseline. These results indicate that the proposed method improves detection accuracy and computational efficiency while maintaining an almost unchanged model scale. The confusion matrix and PR curves further show that YOLO-RCM can effectively mitigate misdetections associated with adjacent maturity stages and complex scenes. In the external-dataset robustness test, Precision and mAP@0.5 are improved by 5.8 and 4.0 percentage points over the baseline, respectively, confirming the generalization ability of the proposed model. The main contribution of this study lies in improving tomato maturity detection from three complementary aspects: channel feature discrimination, local geometric perception, and bounding box regression stability. The study offers a practical technical reference for intelligent tomato harvesting systems in complex agricultural environments. Full article