Search Results (3,094)

Low-cost miniaturized gas sensors are increasingly considered for outdoor air quality monitoring, yet their performance under real-world environmental conditions remains insufficiently characterized. This work evaluates the dynamic gas response of the Bosch BME688 sensor, whose metal oxide sensing layer is based on tin dioxide (SnO₂) material, focusing on its sensitivity, selectivity, and dynamic response to four representative air pollutants: nitrogen dioxide (NO₂), carbon monoxide (CO), sulfur dioxide (SO₂), and isobutylene. This study provides both quantitative performance metrics and a physicochemical interpretation of the sensing mechanism. Controlled experiments were conducted in a custom test chamber to facilitate the precise regulation of temperature, humidity, and gas concentrations in the ppm to sub-ppm range. Despite large variability in the baseline resistance across devices, normalization yields consistent behavior, enabling cross-sensor comparability. The results show that the optimum operating temperatures fall in the range of 360–400 °C, where response and recovery times are reduced to a few minutes, compatible with mobile sensing requirements. Moreover, humidity strongly influences sensor behavior: it generally decreases sensitivity but improves kinetics, and in the case of CO, it enables enhanced responses through additional hydroxyl-mediated pathways. These findings confirm the feasibility of deploying BME688 sensors in distributed outdoor monitoring platforms, provided that humidity and temperature effects are properly addressed through calibration or compensation strategies. In addition, the variability observed in baseline resistance highlights the need for normalization and, consequently, individual calibration steps for each sensor under reference conditions in order to ensure cross-sensor comparability. The findings provided in this study provide support for the design of robust, low-cost air monitoring networks. Full article

Data augmentation is one of the effective solutions to improve the performance of machine learning models in general and deep learning in particular. Data augmentation techniques bring different effects to each model, but very few studies have considered this issue. This study investigated the effect of five distinct data augmentation strategies on a custom-built Convolutional Neural Network (CNN) and nine pre-trained CNN models for crack detection. All ten models were initially trained on a reference dataset of unaugmented images, followed by separate experiments using the augmented datasets. The results show that the pre-trained models, especially VGG-16, EfficientNet-B7, Xception, DenseNet-201, and EfficientNet-B0, consistently achieved greater than 98% in accuracy across all augmentation techniques. Meanwhile, the custom-built CNN was very sensitive to illumination changes and noise. Image rotation and cropping have minimal negative impact and sometimes improve performance. The findings demonstrate that combining data augmentation with state-of-the-art pre-trained models offers a powerful and efficient alternative to the reliance on large-scale datasets for accurate crack detection using CNNs. Full article

Tertiary buildings, characterized by temporary uses and frequent renovations of internal spaces, present some criticalities in terms of the consumption of materials that quickly become waste, despite their high residual value, not exploited for further use. The goal of rethinking the life cycle of building products, and related construction systems, enabling multiple cycles of use and extending the life span of the products, presupposes new Organizational Models and changes throughout the whole building process. This paper presents two Pilot Projects (developed within Re-NetTA research), which experiment with innovative Organizational Models and disassembly construction solutions in the tertiary building sector with the goal of extending the life cycle of materials and products through reusing and remanufacturing. The Pilot Projects involve two key operators: a manufacturer and a Third Sector organization. The paper highlights the fundamental key role of digital technologies by analyzing the following: (i) the development of virtual models to understand the technical feasibility for disassembly and to foresee reuse and remanufacturing scenarios; and (ii) the use of digital twin, augmented reality, and web-based platforms as a support tools, to put the products on a virtual market to reach customers before the activities of remanufacturing. Finally, the enabling conditions for improving circularity are discussed in terms of design process, environmental and economic sustainability assessment, and operator networking. Full article

Since service robots serving as salespersons are expected to be deployed efficiently and sustainably in retail environments, this paper explores the impacts of their interaction cues on customer experiences within small-scale self-service shops. The corresponding customer experiences are discussed in terms of fluency, comfort and likability. We analyzed customers’ shopping behaviors and designed fourteen body gestures for the robots, giving them the ability to select appropriate movements for different stages in shopping. Two experimental scenarios with and without robots were designed. For the scenario involving robots, eight cases with distinct interaction cues were implemented. Participants were recruited to measure their experiences, and statistical methods including repeated-measures ANOVA, regression analysis, etc., were used to analyze the data. The results indicate that robots solely reliant on voice interaction are unable to significantly enhance the fluency, comfort and likability effects experienced by customers. Combining a robot’s voice with the ability to imitate a human salesperson’s body movements is a feasible way to truly improve these customer experiences, and a robot’s body movements can positively influence these customer experiences in human–robot interactions (HRIs) while the use of colored light cannot. We also compiled design strategies for robot interaction cues from the perspectives of cost and controllable design. Furthermore, the relationships between fluency, comfort and likability were discussed, thereby providing meaningful insights for HRIs aimed at enhancing customer experiences. Full article

Wave soldering is a technological process that allows for the simultaneous soldering of multiple locations on the same circuit board. Its major defects, such as tin bridging and insufficient tin filling, continue to challenge manufacturers, resulting in increased rework, labor, and operational costs. Therefore, reducing errors in wave soldering is crucial to ensure the best quality for customers and achieve cost savings for the company. This study aims to enhance wave soldering performance by using an integrated approach that combines Teoriya Resheniya Izobreatatelskikh Zadatch (TRIZ) and Design of Experiment (DOE) for empirical improvement in an Original Equipment Manufacturer (OEM) factory, a subsidiary of a global OEM company. The results are sound: we eliminated tin till bridge defects by 88%, achieved a 33% reduction in manpower, and increased production volumes by 6%. This proposed framework can be utilized in other electronics manufacturing factories and related industries. Full article

E-learning systems that support personalized learning require sophisticated decision-making methods to adapt content to students optimally. This paper deals with applying multi-criteria decision-making methods in assigning learning objects in an e-learning system to students based on relevant customization criteria. The novelty of this study lies in the application of ANP and DEMATEL to improve content adaptation for students. Structuring the decision-making problem according to the DEMATEL and using ANP for prioritization has made the entire selection of learning objects better with respect to cognitive and learning styles and Bloom’s taxonomy levels. The method consists of various forms. In the first, DEMATEL has identified dependencies between criteria and clusters, mentioning their influence values on a 0–4 scale. A linear transformation model quantified the compatibility level of a student profile to a learning material. The transformed DEMATEL results were incorporated in all the interdependencies among criteria. The unweighted supermatrix was normalized by cluster weights assigned by experts before the iterative computation led to the converging weighted supermatrix. The outcome was that the individual students made these final priority rankings for learning materials. A pilot experiment was carried out to validate the system, and the results revealed that in the experimental group, the personalized learning environment showed the maximum statistical improvement over the control group. The research was conducted in Croatia, and the participants were students (N = 77) from two public universities and one polytechnic. Ultimately, the newly developed combined ANP-DEMATEL approach was effective in an instantaneous result-optimized dynamic learning path generation, ensuring knowledge acquisition. This research further contributes to developing intelligent educational systems by demonstrating how ANP and DEMATEL can be used synergistically to improve e-learning personalization. Future work could include optimizing weight assignment strategies or using new learning contexts to further adaptivity. Full article

Salinization is one of the main environmental challenges affecting agriculture in semi-arid regions. We evaluated the feasibility of foliar magnesium and its effects at different doses on the acclimation of cowpea varieties under salt stress. The experiment occurred in a greenhouse using a randomized block design in a 2 × 3 × 4 factorial scheme, with five replicates. Two cowpea varieties—‘Pingo de Ouro’ and ‘Costela de Vaca’—were subjected to three salinity levels in irrigation water (0.54, 3.50, and 5.00 dS m⁻¹) and four foliar magnesium (Mg) doses (0, 1, 2, and 3 mL L⁻¹). Under 3.50 dS m⁻¹ salinity, the 1 mL L⁻¹ dose resulted in the highest yield per plant (18.29 g). CO₂ assimilation was highest with 2 mL L⁻¹ Mg at 3.50 dS m⁻¹ for ‘Costela de Vaca’, and with 1 mL L⁻¹ Mg at 5.00 dS m⁻¹ for ‘Pingo de Ouro’. The ‘Pingo de Ouro’ variety was more tolerant to ‘Costela de Vaca’. Foliar Mg fertilization proved to be a promising strategy to mitigate the effects of salt stress in cowpea, especially for ‘Pingo de Ouro’. Magnesium effectively reduces salt stress, but its effect varies by plant variety and irrigation salinity, necessitating customized dose adjustments. Full article

Special steel manufacturing involves both discrete processing events and continuous physical flows, forming a representative discrete–continuous hybrid production system. However, due to the visually homogeneous surfaces of steel products, the highly dynamic production environment, and frequent disturbances or anomalies, traditional single-source tracking approaches struggle to maintain accurate and consistent material identification. To address these challenges, this paper proposes a multi-source fusion-based material tracking method tailored for discrete–continuous hybrid scenarios. First, a state–event system (SES) is constructed based on process rules, enabling interpretable reasoning of material states through event streams and logical constraints. Second, on the visual perception side, a YOLOv8-SE detection network embedded with the squeeze-and-excitation (SE) channel attention mechanism is designed, while the DeepSORT tracking framework is improved to enhance weak feature extraction and dynamic matching for visually similar targets. Finally, to handle information conflicts and cooperation in multi-source fusion, an improved Dempster–Shafer (D-S) evidence fusion strategy is developed, integrating customized anomaly handling and fault-tolerance mechanisms to boost decision reliability in conflict-prone regions. Experiments conducted on real special steel production lines demonstrate that the proposed method significantly improves detection accuracy, ID consistency, and trajectory integrity under complex operating conditions, while enhancing robustness against modal conflicts and abnormal scenarios. This work provides an interpretable and engineering-feasible solution for end-to-end material tracking in hybrid manufacturing systems, offering theoretical and methodological insights for the practical deployment of multi-source collaborative perception in industrial environments. Full article

Traffic sign recognition is still one of the challenging aspects of intelligent vehicle systems, mainly when processor or memory resources are limited. In this work, real-time traffic sign detection was evaluated using five YOLO model variants—Nano, Small, Medium, Large, and XLarge—across versions 8 to 11. All models were trained and validated with a custom dataset collected in a simulated urban environment designed to replicate FIRA competition tracks. The models were then deployed and tested on a 1:10 scale autonomous vehicle equipped with a mini PC running the detector in real time. Performance was compared using mAP@50–95, F1-score, inference latency, and preprocessing and postprocessing times. The authors also analyzed training behavior, focusing on convergence speed and stopping criteria. The experiments showed that YOLOv10 B achieved the highest performance across varying conditions, while YOLOv8 M provided a better balance between speed and accuracy. These results can help practitioners select appropriate YOLO architectures for embedded traffic sign recognition systems that must operate in real time on resource-constrained autonomous vehicles. Full article

This study examines key bioprocess parameters influencing the reduction in hydraulic conductivity in porous media via Microbially-Induced Calcite Precipitation (MICP), highlighting its relevance to environmental engineering applications such as bio-barriers and landfill liners. Sporosarcina pasteurii was utilized as the ureolytic bacterium to induce calcium carbonate precipitation under controlled laboratory conditions. Experimental variables included bacterial cell density (OD₆₀₀), diameter of glass beads, concentrations of precipitation solution, bentonite, and yeast extract. A total of 42 experimental runs were conducted based on a custom design in Design-Expert software. Hydraulic conductivity was selected as the response variable to evaluate treatment performance. Response surface methodology (RSM) was applied to develop a second-order polynomial model, with statistical analyses indicating a strong model fit (R² = 0.948, adjusted R² = 0.929, predicted R² = 0.868). ANOVA confirmed the significance of the main effects and interactions, particularly those involving glass bead diameter and OD₆₀₀. Among the tested factors, the precipitation solution exhibited the strongest individual effect, while bentonite and yeast extract demonstrated supportive roles. Optimization revealed that a balanced combination of microbial density and chemical inputs minimized hydraulic conductivity to 0.0399 cm/s (≈95% reduction), with an overall desirability score of 1.000. Laboratory-scale experiments demonstrated field-scale applicability, underscoring the potential of biotechnological soil treatment and empirical modeling for developing sustainable low-permeability barriers. Full article

The General AntiParticle Spectrometer is a balloon-borne experiment designed to search for low-energy cosmic-ray antinuclei as a potential indirect signature of dark matter. Over the course of at least three long-duration flights over Antarctica, it will explore the sub-$250 \mathrm{M}\mathrm{e}\mathrm{V}/\mathit{n}$ energy range with sensitivity to antideuterons and antihelium, while also extending antiproton measurements below 100 $\mathrm{M}$$\mathrm{e}\mathrm{V}$. The instrument features a tracker built from more than one thousand lithium-drifted silicon detectors, each read out by a dedicated custom integrated circuit. With the first flight scheduled for the austral summer of 2025, a new prototype chip, ANTARES4, has been developed using a commercial 65 $\mathrm{n}$$\mathrm{m}$ complementary metal-oxide semiconductor process for use in the second flight. It integrates eight independent analog channels, each incorporating a low-noise charge-sensitive amplifier with dynamic signal compression, a CR–RC shaping stage with eight selectable peaking times, and on-chip calibration circuitry. The charge-sensitive amplifier uses metal-oxide semiconductor feedback elements with voltage-dependent capacitance to support the wide input energy range from 10 $\mathrm{k}$$\mathrm{e}\mathrm{V}$ to 100 $\mathrm{M}$$\mathrm{e}\mathrm{V}$. Four alternative feedback implementations are included to compare performance and design trade-offs. Leakage current compensation up to 200 $\mathrm{n}$$\mathrm{A}$ per detector strip is provided by a Krummenacher current–feedback network. This paper presents the design and architecture of ANTARES4, highlighting the motivations, design drivers, and performance requirements that guided its development. Full article

The chain-of-thought (CoT) approach in large language models (LLMs) has markedly enhanced their performance on complex tasks; however, effectively distilling this capability into LLMs with smaller parameter scales remains a challenge. Studies have found that small LLMs do not always benefit from CoT distillation. Inspired by the concept of teaching students in accordance with their aptitude, we propose an adaptive chain-of-thought distillation (ACoTD) framework. The core idea is to dynamically and adaptively customize distillation data and supervision signals for student models based on their performance on the original problems. Specifically, ACoTD initially evaluates and categorizes the original problems according to the capabilities of the student model. Subsequently, for Easy- and Medium-level problems, a short CoT distillation is employed for a brief lecture to reinforce knowledge and enhance training efficiency, for high-difficulty problems where the student model underperforms, and a detailed long CoT distillation is utilized for in-depth explanation to infuse richer reasoning logic. This differentiated distillation strategy ensures that student models achieve a better grasp of learning. We conducted experiments on multiple benchmark datasets. The results indicate that, compared to the baseline, our method can significantly improve the inference performance of small LLMs. Our method provides a new student-centered paradigm for knowledge distillation, demonstrating that adaptive adjustment of teaching strategies based on student feedback is an effective way to enhance small LLMs’ reasoning ability. Full article

Plant disease recognition is a critical technology for ensuring food security and advancing precision agriculture. However, challenges such as class imbalance, heterogeneous image quality, and limited model interpretability remain unresolved. In this study, we propose a Synergistic Dual-Augmentation and Class-Aware Hybrid (SDA-CAH) model designed to achieve robust and interpretable recognition of plant diseases. Our approach introduces two innovative augmentation strategies: (1) an optimized MixUp method that dynamically integrates class-specific features to enhance the representation of minority classes; and (2) a customized augmentation pipeline that combines geometric transformations with photometric perturbations to strengthen the model’s resilience against image variability. To address class imbalance, we further design a class-aware hybrid sampling mechanism that incorporates weighted random sampling, effectively improving the learning of underrepresented categories and optimizing feature distribution. Additionally, a Grad-CAM–based visualization module is integrated to explicitly localize lesion regions, thereby enhancing the transparency and trustworthiness of the predictions. We evaluate SDA-CAH on the PlantVillage dataset using a pretrained EfficientNet-B0 as the backbone network. Systematic experiments demonstrate that our model achieves 99.95% accuracy, 99.89% F1-score, and 99.89% recall, outperforming several strong baselines, including an optimized Xception (99.42% accuracy, 99.39% F1-score, 99.41% recall), standard EfficientNet-B0 (99.35%, 99.32%, 99.33%), and MobileNetV2 (95.77%, 94.52%, 94.77%). For practical deployment, we developed a web-based diagnostic system that integrates automated recognition with treatment recommendations, offering user-friendly access for farmers. Experimental evaluations indicate that SDA-CAH outperforms existing approaches in predictive accuracy and simultaneously defines a new paradigm for interpretable and scalable plant disease recognition, paving the way for next-generation precision agriculture. Full article

In immersive digital devices, high environmental complexity can lead to rendering delays and loss of interactive details, resulting in a fragmented experience. This paper proposes a lightweight NeRF (Neural Radiance Fields) modeling and multimodal perception fusion method. First, a sparse hash code is constructed based on Instant-NGP (Instant Neural Graphics Primitives) to accelerate scene radiance field generation. Second, parameter distillation and channel pruning are used to reduce the model’s size and reduce computational overheads. Next, multimodal data from a depth camera and an IMU (Inertial Measurement Unit) is fused, and Kalman filtering is used to improve pose tracking accuracy. Finally, the optimized NeRF model is integrated into the Unity engine, utilizing custom shaders and asynchronous rendering to achieve low-latency viewpoint responsiveness. Experiments show that the file size of this method in high-complexity scenes is only 79.5 MB ± 5.3 MB, and the first loading time is only 2.9 s ± 0.4 s, effectively reducing rendering latency. The SSIM is 0.951 ± 0.016 at 1.5 m/s, and the GME is 7.68 ± 0.15 at 1.5 m/s. It can stably restore texture details and edge sharpness under dynamic viewing angles. In scenarios that support 3–5 people interacting simultaneously, the average interaction response delay is only 16.3 ms, and the average jitter error is controlled at 0.12°, significantly improving spatial interaction performance. In conclusion, this study provides effective technical solutions for high-quality immersive interaction in complex public scenarios. Future work will explore the framework’s adaptability in larger-scale dynamic environments and further optimize the network synchronization mechanism for multi-user concurrency. Full article

Background/Objectives: The aim of this study was to establish the relationship between target size and the required diagnostic PET maximum standard uptake value (SUVmax) thresholds needed for successful Biology-guided Radiotherapy (BgRT) delivery on RefleXion X1 PET-linac. The current clinical eligibility recommendation is an SUVmax ≥ 6 at simulation, but the RefleXion system subsequently evaluates Activity Concentration (AC), which must exceed 5 kBq/mL for successful BgRT planning. Methods: A custom 3D-printed phantom containing six spherical targets (8 to 20 mm diameter) was used with varying target-to-background ratios (5:1 to 20:1) of 18F-FDG to systematically achieve a range of SUVmax values for each target size. Images were acquired on Siemens Biograph mCT for SUVmax quantification and RefleXion X1 for AC measurements. Twenty-four BgRT plans were evaluated, and delivery accuracy was validated using ArcCHECK. Additionally, retrospective data from 18 patients across four institutions were analyzed to validate the phantom-derived findings. Results: The PET-linac successfully planned treatments for 13/24 experiments, all achieving an AC > 5 kBq/mL. SUVmax requirements varied by target size: 16–20 mm targets required an SUVmax > 6, consistent with current recommendations, while smaller targets required higher thresholds (e.g., 13 mm: SUVmax > 10, and 11 mm: SUVmax > 15). 8 and 9 mm targets failed to meet AC requirements even at SUVmax 14. Successful deliveries maintained acceptable accuracy, with gamma passing rates of 92.4% ± 5.0% (3%/2 mm) and 97.6% ± 1.9% (3%/3 mm). Analysis revealed that Volume (cc) × SUVmax > 11 consistently predicted successful BgRT planning across all target sizes. This threshold was validated using multi-institutional PET-CT patient data (mean: 11.36, 95% CI: 9.1–12.9), correctly predicting treatment eligibility in 15 of 18 cases. Conclusions: Target size significantly influences BgRT eligibility. We derived a new criterion, Volume(cc) × SUVmax > 11 (95% CI: 9.1–12). Full article