Search Results (3,765)

In the context of global food security and sustainable agricultural development, the efficient recognition and precise management of agricultural insect pests and their predators have become critical challenges in the domain of smart agriculture. To address the limitations of traditional models that overly rely on single-modal inputs and suffer from poor recognition stability under complex field conditions, a multimodal recognition framework has been proposed. This framework integrates RGB imagery, thermal infrared imaging, and environmental sensor data. A cross-modal attention mechanism, environment-guided modality weighting strategy, and decoupled recognition heads are incorporated to enhance the model’s robustness against small targets, intermodal variations, and environmental disturbances. Evaluated on a high-complexity multimodal field dataset, the proposed model significantly outperforms mainstream methods across four key metrics, precision, recall, F1-score, and mAP@50, achieving 91.5% precision, 89.2% recall, 90.3% F1-score, and 88.0% mAP@50. These results represent an improvement of over 6% compared to representative models such as YOLOv8 and DETR. Additional ablation studies confirm the critical contributions of key modules, particularly under challenging scenarios such as low light, strong reflections, and sensor data noise. Moreover, deployment tests conducted on the Jetson Xavier edge device demonstrate the feasibility of real-world application, with the model achieving a 25.7 FPS inference speed and a compact size of 48.3 MB, thus balancing accuracy and lightweight design. This study provides an efficient, intelligent, and scalable AI solution for pest surveillance and biological control, contributing to precision pest management in agricultural ecosystems. Full article

This study investigates the flexural performance, tensile splitting strength, and fracture behaviour of self-compacting concrete (SCC) reinforced with polypropylene (PP) and basalt (BF) fibres. A total of eleven SCC mixtures with varying fibre types and volume fractions (0.025–0.25%) were tested at 7 and 28 days. In this study, the term high-performance concrete (HPC) refers to SCC mixtures with a 28-day compressive strength exceeding 60 MPa, as commonly accepted in European standards and literature. The control SCC achieved 68.2 MPa at 28 days. While fibre addition enhanced the tensile and flexural properties, it reduced workability, demonstrating the trade-off between mechanical performance and flowability in high-performance SCC. The experimental results demonstrate that both fibre types improve the tensile behaviour of SCC, with distinct performance patterns. PP fibres, owing to their flexibility and crack-bridging capability, were particularly effective at early ages, enhancing the splitting tensile strength by up to 45% and flexural toughness by over 300% at an optimal dosage of 0.125%. In contrast, BF fibres significantly increased the 28-day toughness (up to 15.7 J) and post-cracking resistance due to their superior stiffness and bonding with the matrix. However, high fibre contents adversely affected workability, particularly in BF-reinforced mixes. The findings highlight a dosage-sensitive behaviour, with optimum performance observed at 0.05–0.125% for PP and 0.125–0.25% for BF. While PP fibres improve crack distribution and early-age ductility, BF fibres offer higher stiffness and energy absorption in post-peak regimes. Statistical analysis (ANOVA and Tukey’s test) confirmed significant differences in the mechanical performance among fibre-reinforced mixes. The study provides insights into selecting appropriate fibre types and dosages for SCC structural applications. Further research on hybrid fibre systems and long-term durability is recommended. The results contribute to sustainable concrete design by promoting enhanced performance with low-volume, non-metallic fibres. Full article

Understanding the intrinsic relationship between microscopic structures and macroscopic mechanical properties of rock under freeze–thaw (F-T) conditions is essential for ensuring the safety and stability of geotechnical engineering in cold regions. In this study, a series of F-T cycle tests, nuclear magnetic resonance (NMR) measurements, and uniaxial compression tests were conducted on sandstone samples. The mechanisms by which F-T cycles influence pore structure and mechanical behavior were analyzed, revealing their internal correlation. A degradation model for peak strength was developed using mesopore porosity as the key influencing parameter. The results showed that with increasing F-T cycles, the total porosity and mesopore and macropore porosities all exhibited increasing trends, whereas the micropore and different fractal dimensions decreased. The compaction stage in the stress–strain curves became increasingly prominent with more F-T cycles. Meanwhile, the peak strength and secant modulus decreased, while the peak strain increased. When the frost heave pressure induced by water–ice phase transitions exceeded the ultimate bearing capacity of pore walls, smaller pores progressively evolved into larger ones, leading to an increase in the mesopores and macropores. Notably, mesopores and macropores demonstrated significant fractal characteristics. The transformation in pore size disrupted the power-law distribution of pore radii and reduced fractal dimensions. A strong correlation was observed between peak strength and both the mesopore and mesopore fractal dimensions. The increase in mesopores and macropores enhanced the compaction stage of the stress–strain curve. Moreover, the expansion and interconnection of mesopores under loading conditions degraded the deformation resistance and load-bearing capacity, thereby reducing both the secant modulus and peak strength. The degradation model for peak strength, developed based on changes in mesopore ratio, proved effective for evaluating the mechanical strength when subjected to different numbers of F-T cycles. Full article

Scene Knowledge-guided Visual Grounding (SK-VG) is a multi-modal detection task built upon conventional visual grounding (VG) for human–computer interaction scenarios. It utilizes an additional passage of scene knowledge apart from the image and context-dependent textual query for referred object localization. Due to the inherent difficulty in directly establishing correlations between the given query and the image without leveraging scene knowledge, this task imposes significant demands on a multi-step knowledge reasoning process to achieve accurate grounding. Off-the-shelf VG models underperform under such a setting due to the requirement of detailed description in the query and a lack of knowledge inference based on implicit narratives of the visual scene. Recent Vision–Language Models (VLMs) exhibit improved cross-modal reasoning capabilities. However, their monolithic architectures, particularly in lightweight implementations, struggle to maintain coherent reasoning chains across sequential logical deductions, leading to error accumulation in knowledge integration and object localization. To address the above-mentioned challenges, we propose SplitGround—a collaborative framework that strategically decomposes complex reasoning processes by fusing the input query and image with knowledge through two auxiliary modules. Specifically, it implements an Agentic Annotation Workflow (AAW) for explicit image annotation and a Synonymous Conversion Mechanism (SCM) for semantic query transformation. This hierarchical decomposition enables VLMs to focus on essential reasoning steps while offloading auxiliary cognitive tasks to specialized modules, effectively splitting long reasoning chains into manageable subtasks with reduced complexity. Comprehensive evaluations on the SK-VG benchmark demonstrate the significant advancements of our method. Remarkably, SplitGround attains an accuracy improvement of 15.71% on the hard split of the test set over the previous training-required SOTA, using only a compact VLM backbone without fine-tuning, which provides new insights for knowledge-intensive visual grounding tasks. Full article

This paper discusses the potential of using steel fiber to produce self-compacting high-strength concrete. The effects of water–binder ratio and mortar and steel fiber content on the workability and mechanical properties of high-performance concrete were studied. The working performance of cementitious materials was evaluated by a slump expansion test, T₅₀₀, L-shaped instrument, U-shaped instrument, and V-shaped funnel. The mechanical properties were evaluated by compressive strength and flexural strength. The results show that when the compressive strength of self-compacting high-strength concrete with steel fiber is 90 MPa, the optimum mix ratio is a water–binder ratio of 0.22, sand ratio of 46%, and steel fiber content of 0.3%. When the steel fiber content is 0.3%, the compressive strength of the time can be increased by more than 4%, and the flexural strength can be increased by more than 5%. When the steel fiber content is 0.6% to 0.9%, the compressive strength of the specimen can be increased by more than 10%, and the flexural strength can be increased by more than 7%. However, with the increase in steel fiber content, self-compacting concrete becomes less and less dense, and the bond strength becomes lower and lower. When the water–binder ratio is 0.20, the fluidity of self-compacting concrete is poor, and the forming effect is not good. When the water–binder ratio is 0.24, the working performance of self-compacting concrete is better, but the cohesion is poor, and it can easily produce segregation. When the water–binder ratio is 0.22, the working performance of self-compacting concrete can be the best, and the strength of concrete is higher and more stable. The optimum sand ratio is 46%. At this time, the compressive strength and flexural strength of self-compacting concrete are the largest, and the working performance is also the best. When the sand ratio is lower than the optimum sand ratio, the self-compacting concrete will produce segregation. When the sand ratio is higher than the optimum sand ratio, the fluidity of self-compacting concrete is poor. This study provides insights into the potential for large-scale and high-value utilization of steel fibers and the development of cost-effective ways to reduce the carbon footprint of self-compacting concrete production. Full article

This paper investigates and characterizes a tunable inductor structure based on coupled-line configurations, referred to as a coupled-line tunable inductor (CLTI). By integrating switches along the coupled-line paths, the mutual inductance can be selectively enabled or disabled, providing a means for active inductance modulation. Spiral inductors with one-turn and two-turn cores were used in conjunction with inner-coupled-line placements to explore different coupling configurations. The test structures were implemented using printed circuit board (PCB) technology, and their performance was analyzed through electromagnetic simulations and vector network analyzer (VNA) measurements. The results confirm that switch-controlled coupled lines enable effective inductance tuning, with a measurable reduction in inductance when the coupled-line path is activated. In the switch-OFF state, only minimal performance degradation was observed due to parasitic effects. These findings provide useful insights into the practical behavior of coupled-line tunable inductors and suggest their applicability in RF circuits and adaptive analog systems, particularly where integration and compact tunability are desired. Full article

This paper presents a lightweight and cost-effective computer vision solution for automated industrial inspection using You Only Look Once (YOLO) v8 models deployed on embedded systems. The YOLOv8 Nano model, trained for 200 epochs, achieved a precision of 0.932, an mAP@0.5 of 0.938, and an F1-score of 0.914, with an average inference time of ~470 ms on a Raspberry Pi 500, confirming its feasibility for real-time edge applications. The proposed system aims to replace physical jigs used for the dimensional verification of extruded polyamide tubes in the automotive sector. The YOLOv8 Nano and YOLOv8 Small models were trained on a Graphics Processing Unit (GPU) workstation and subsequently tested on a Central Processing Unit (CPU)-only Raspberry Pi 500 to evaluate their performance in constrained environments. The experimental results show that the Small model achieved higher accuracy (a precision of 0.951 and an mAP@0.5 of 0.941) but required a significantly longer inference time (~1315 ms), while the Nano model achieved faster execution (~470 ms) with stable metrics (precision of 0.932 and mAP@0.5 of 0.938), therefore making it more suitable for real-time applications. The system was validated using authentic images in an industrial setting, confirming its feasibility for edge artificial intelligence (AI) scenarios. These findings reinforce the feasibility of embedded AI in smart manufacturing, demonstrating that compact models can deliver reliable performance without requiring high-end computing infrastructure. Full article

Based on the G85 high-fill subgrade project in east Gansu Province, this study conducts one-dimensional compression tests in the laboratory on both disturbed and in situ-compacted loess. Through the combination of the test results of remolded soil, compaction standards for each layer of the subgrade fill are established, and quality inspections of the compacted subgrade are performed. The experimental results demonstrate that the compression deformation of remolded loess exhibits a positive correlation with compaction degree and a negative correlation with moisture content. Under constant compaction degree conditions, axial pressure and deformation follow a linear relationship, whereas under fixed conditions, the relationship adheres to a quadratic trend. Specimen void ratios show minimal variation within the 25–100 kPa stress range but undergo significant reduction between 100 and 400 kPa. Under an axial compressive load of 100–200 kPa, the compression coefficient at a height of 10 m within the subgrade ranges from 0.163 to 0.171 MPa⁻¹. At a height of 6 m, it ranges from 0.177 to 0.183 MPa⁻¹, and at 1 m, from 0.183 to 0.186 MPa⁻¹. These values indicate that the compaction quality throughout the subgrade corresponds to a low compressibility level. However, the compaction quality near the slopes on both sides is slightly lower than that along the centerline of the subgrade. Overall, the compaction quality meets the required standards. Full article

Heavy machinery degrades agricultural soils, with severity influenced by wheel type, contact area, and soil moisture. Tropical agriculture is characterized by the constant maintenance of straw on the ground. This permanent cover, among other benefits, can mitigate the stress imposed by wheels on the physical structure of the soil. This study aimed to evaluate the effect of tire types and straw amounts on soil stresses. Static studies were carried out under controlled conditions in a static tire test unit (STTU), equipped with standardized sensors and systems that simulated real farming conditions. Three tire models were tested: road truck double wheelset—2 × 275/80R22.5 (p1); agricultural radial tire—600/50R22.5 (p2); and bias-ply tire—600/50-22.5 (p3) on four contact surfaces (rigid surface; bare soil; soil with 15 and 30 Mg ha⁻¹ straw cover). We performed comparative statistical tests and subsurface stress simulations for each tire and surface condition. On the hard surface, the contact areas were 4.7 to 6.8 times smaller than on bare soil. Straw increased the tire’s contact area, reducing compaction and subsoil stresses. Highest pressure was imposed by the road tire (p1) and lowest by the radial tire (p2). Adding 15 Mg ha⁻¹ of straw reduced soil SPR by 18%, while increasing it to 30 Mg ha⁻¹ led to an additional 8% reduction. Tire selection and effective straw management improve soil conservation and agriculture sustainability. Full article

A new method of diagnostics of technological parameters of the material agglomeration process and a design of a new matrix for compacting fine-grained materials were presented. Two different dies were used in the agglomeration process to carry out the tests. One of them is a new prototype design of an opening matrix, which allows for its quick opening and closing and easier extraction of the briquette and measurement. Comparative tests were carried out for four materials with different chemical compositions. The experimental part showed changes in the temperature of samples subjected to agglomeration. Analysis of the results for the Electric Arc Furnace Dust (EAFD) mixture showed that the average temperature in the middle of the material was higher by 1.3 °C than the highest temperature on the external surface. Comparing the test results for the hydrated lime obtained in the opening matrix with the results obtained for the classic closed matrix, it can be seen that the results are very similar, which indicates the advantage of the prototype solution of the opening matrix. The prototype solution of the opening matrix significantly speeds up the agglomeration process. The obtained results confirm the possibility of using thermography diagnostics in the agglomeration processes of fine-grained materials and the usefulness of the newly designed die. Thermography examination can be practically used to control the agglomeration process, product quality, and, indirectly, roller wear. Full article

The mechanical behavior of dental composites depends on the sample size and stress configuration. This makes it difficult to extrapolate laboratory data to clinical restorations with significant variations in size and geometry. Intrinsic parameters, such as fracture toughness, are therefore of great importance, even if they are less common and more difficult to measure. The aim of this study was to apply principles of fractography and fracture mechanics to exploit the results obtained from a three-point bending test. The objectives include calculating a material-specific constant, validating the experimental findings, and establishing a correlation with fracture toughness. Forty representative composites with wide variation in filler quantity (65–83% by weight and 46.4–64% by volume), type (compact glasses and pre-polymerized), and composition were examined. Fracture toughness/K_Ic was evaluated in a notchless triangular prism test. Fracture type, origin, and mirror size were determined on 280 flexural fracture specimens (n = 20). The amount of filler strongly influences all measured parameters, with the effect strength varying in the sequence: mechanical work (η_P² = 0.995), modulus of elasticity (η_P² = 0.991), flexural strength (η_P² = 0.988), fracture toughness (η_P² = 0.979), and mirror constant (η_P² = 0.965). Fracture surfaces allowed the delineation of the fracture mirror and the application of fracture mechanics approaches. The mirror constant was derived from the radius of the fracture mirror, measured in the direction of constant stress, using Orr’s equation, and correlates well with K_Ic (0.81). Larger confidence intervals were observed for the mirror constant data, while for 5 of 14 materials, the mirror constant was overestimated compared to K_Ic. The overestimation was attributed to the lower refractive index of the urethane methacrylate composition. Full article

This study investigates the operational behavior and voltage stability of a 1 kW-class AIP PEMFC stack under high-pressure H₂ and O₂ conditions. AIP PEMFCs, unlike conventional air-based systems, operate in enclosed environments using stored O₂, requiring designs that minimize parasitic power losses while ensuring stable operation. To establish a performance baseline, single cell tests were conducted to isolate the effects of in-plane components, including the MEA, GDL, and flow field geometry. Results indicated that temperature and pressure significantly influenced performance, whereas humidity and flow rate had minimal effects under the tested conditions. A 27-cell stack was then assembled and evaluated under various current densities, flow rates, and humidity levels. Time-resolved voltage measurements revealed that low flow rates (stoichiometry ≤ 1.5) led to voltage instability, particularly at high humidity and current density. Instability was more pronounced in cells positioned farthest from the inlet and outlet ports. These findings underscore the importance of optimizing operational parameters and stack architecture to achieve stable AIP PEMFC performance under reduced flow conditions. The results provide key insights for developing compact, efficient, and durable AIP fuel cell systems for use in enclosed or submerged environments such as submarines or unmanned underwater vehicles, while highlighting key challenges associated with AIP-targeted applications. Full article

This study presents a wireless, non-invasive sensing system for monitoring the dielectric permittivity of materials, with a particular focus on applications in cultural heritage conservation. The system integrates a passive split-ring resonator tag, electromagnetically coupled to a compact antipodal Vivaldi antenna, operating in the reactive near-field region. Both numerical simulations and experimental measurements demonstrate that shifts in the antenna’s reflection coefficient resonance frequency correlate with variations in the dielectric permittivity of the material under test. A calibration curve was established using reference materials—including low-density polyvinylchloride, polytetrafluoroethylene, polymethyl methacrylate, and polycarbonate—and validated through precise permittivity measurements. The system was subsequently applied to wood samples (fir, poplar, beech, and oak) at different humidity levels, revealing a sigmoidal relationship between moisture content and permittivity. The behavior was also confirmed using a portable and low-cost setup, consisting of a point-like coaxial sensor that could be easily moved and positioned as needed, enabling localized measurements on specific areas of interest of the sample, together with a miniaturized Vector Network Analyzer. These results underscore the potential of this portable, contactless, and scalable sensing platform for real-world monitoring of cultural heritage materials, enabling minimally invasive assessment of their structural and historical integrity. Moreover, by enabling the estimation of moisture content through dielectric permittivity, the system provides an effective method for early detection of water-induced deterioration in wood-based heritage items. This capability is particularly valuable for preventive conservation, as excessive moisture—often indicated by permittivity values above critical thresholds—can trigger biological or structural degradation. Full article

Moisture-induced instability in rock masses presents a significant threat to the safety and sustainability of underground infrastructure. This study proposes a nonlinear energy signal fusion framework to predict failure in moisture-affected limestone by integrating acoustic emission data with energy dissipation metrics. Uniaxial compression tests were carried out under controlled moisture conditions, with real-time monitoring of AE signals and strain energy evolution. The results reveal that increasing moisture content reduces the compressive strength and elastic modulus, prolongs the compaction phase, and induces a transition in failure mode from brittle shear to ductile tensile–shear behavior. An energy partitioning analysis shows a clear shift from storage-dominated to dissipation-dominated failure. A dissipation factor (η) is introduced to characterize the failure process, with critical thresholds η_min and η_f identified. A nonlinear AE-energy coupling model incorporating water-sensitive parameters is proposed. Furthermore, an energy-based instability criterion integrating multiple indicators is established to quantify failure transitions. The proposed method offers a robust tool for intelligent monitoring and predictive stability assessment. By integrating data-driven indicators with environmental sensitivity, the study provides engineering insights that support adaptive support design, long-term resilience, and sustainable decision making in groundwater-rich rock environments. Full article

Fracture toughness anisotropy is a key concern in IN718 components produced by Laser Powder Bed Fusion (LPBF), due to their strong crystallographic texture and characteristic lamellar microstructure. In this study, the effect of grain orientation on fracture toughness was evaluated by testing two LPBF IN718 builds with the same laser scanning strategy (R0), but with two different orientations: vertical (R0-0) and 45° inclined (R0-45) relative to the build direction. The mechanical response was assessed through compact tension (CT) tests following ASTM E399 and ASTM E1820 standards. Results show that the R0-45 specimens exhibited a fracture toughness nearly 2.5 times higher than R0-0 specimens. Detailed microstructural analysis, supported by EBSD and SEM, reveals that the higher toughness in the R0-45 orientation is linked to a combination of smaller effective grain size along the crack path, higher levels of geometrically necessary dislocations (GND), and increased kernel average misorientation (KAM), which collectively enhance plastic accommodation and crack-tip shielding. These findings support and reinforce the established understanding of the relationship between microstructure and anisotropic fracture behavior in LPBF IN718, facilitating its practical application in the design and orientation of additively manufactured components. Full article