Search Results (2,919)

Shop-floor weld inspection may appear to be a solved problem until a camera is deployed near a galvanized-sheet MIG welding line. The seam reflects light, the texture changes from frame to frame, and the defects of interest are often small and visually subtle. Additionally, the hardware near the line is rarely a data-center GPU. With those constraints in mind, this paper presents YOLO-MIG, a compact detector built on YOLOv10n for weld-seam inspection in practical production conditions. We make three focused changes to the baseline: a C2f-EMSCP backbone block to better preserve weak defect cues with modest parameter growth, a BiFPN neck to keep small-target information alive during feature fusion, and a C2fCIB head to clean up predictions that otherwise get distracted by seam edges and illumination artifacts. On a workshop-collected dataset containing 326 original images, with the training subset expanded through augmentation to 2608 labeled samples in total, YOLO-MIG achieves 98.4% mAP@0.5 and 56.29% mAP@0.5:0.95 on the test set while remaining lightweight (1.83 M parameters, 3.87 MB FP16 weights). Compared with YOLOv10n, the proposed model improves mAP@0.5 by 9.36 points and mAP@0.5:0.95 by 4.89 points, while reducing parameters, GFLOPs, and model size by 43.4%, 19.9%, and 29.9%, respectively. The results suggest that YOLO-MIG is not only accurate but also realistic to deploy at the edge for intelligent weld quality control. Full article

Automated video-based detection of cognitive disorders can enable a scalable non-invasive health monitoring. However, existing methods focus on a single disease and provide limited interpretability, whereas real-world videos often contain co-occurring conditions. We propose a novel unified multi-task method to detect depression and Parkinson’s disease (PD) from in-the-wild video data called DEPART (DEpression and PArkinson’s Recognition Technique). It performs body region extraction, Contrastive Language-Image Pre-training (CLIP)-based visual encoding, Transformer-based temporal modeling, and prototype-aware classification with a gated fusion technique. Gradient-based attention maps are used to visualize task-specific regions that drive predictions. Experiments on the In-the-Wild Speech Medical (WSM) corpus demonstrate competitive performance: the multi-task model achieves Recall of 82.39% for depression and 78.20% for PD, compared with 87.76% and 78.20%, for the best single-task models. The multi-task learning initially increases false positives for healthy persons in the PD subset, mainly due to annotation–modality mismatches, static visual content misinterpreted as motor impairments, and occasional body detection failures. After cleaning the test data, Recall for healthy individuals becomes comparable across models; the multi-task model improves Recall for both depression (from 82.39% to 87.50%) and PD (from 78.20% to 86.14%), suggesting better robustness for real-life clinical applications. Full article

The sustainable management of balneological resources is vital for the development of eco-friendly health tourism and regional economic stability. This study presents a comprehensive physicochemical and eco-toxicological characterization of the therapeutic peloids (mud) from Lake Maly Akkol, which is located in the Zhambyl region of Kazakhstan. Utilizing an integrated approach of laboratory analysis and Python-based statistical modeling, we evaluated the resource’s clinical potential and environmental safety. The results identify the deposit as a high-quality sulfide–silt peloid with a mean humidity of 66.91% (95% CI: [65.21, 68.60]) and a mineralization level of 11.21 g/dm³ (95% CI: [10.84, 11.57]). Statistical validation using one-sample t-tests confirmed that critical therapeutic indicators, including shear strength ($\mu$ = 2593.72 dyne/cm²) and total sulfide content ($\mu$ = 0.079%), are significantly aligned with international balneological standards (p < 0.05). Eco-toxicological screening for heavy metals revealed that Lead (37.03 mg/kg) and Cadmium (0.06 mg/kg) remain well below safety thresholds, ensuring the resource’s “clean” environmental profile. These findings establish a statistically robust “Digital Quality Passport” for the Lake Maly Akkol deposit, providing the scientific foundation necessary for its sustainable industrial utilization and long-term ecological preservation. Full article

Operationally, energy transition takes place at the local level, that is, in cities and rural municipalities. Its effectiveness is, therefore, dependent on individual actions undertaken in enterprises and households. It also constitutes a particularly challenging task for industrial regions with centuries-old mining traditions. Meanwhile, the opinions of residents living in mining cities receive little attention in the literature. For these reasons, this study used survey research conducted in 19 Silesian cities with county rights and on a representative sample of 1863 residents. In this way, answers were sought to the following research questions: (1) How do urban residents in a developing economy in a post-mining region assess their knowledge regarding environmental protection and energy transition? (2) How do they evaluate local authorities’ actions concerning the replacement of non-ecological heating sources in households? The analysis of results employed descriptive statistics and non-parametric statistical tests, identifying differences in respondents’ assessments according to gender, age, education, and place of residence. The analyses conducted indicate that residents assess their environmental awareness as average. They also rate their knowledge of the energy transition below average, despite being in the midst of it. The assessments of men, older individuals, and those with vocational and secondary education are higher in both cases than the assessments of women, younger generations, and respondents with primary, post-secondary, and higher education. Respondents also rate financial and informational–educational support for heating source replacement as average. Importantly, however, these actions are noticed and appreciated. They meet the expectations of less formally educated individuals (formal education level: primary, vocational, and secondary). However, they do not generate enthusiasm among those with post-secondary and higher education, whose environmental needs and expectations may be higher. The level of financial support, and to a lesser extent informational–educational support, differs significantly among the studied cities, indicating the absence of a coherent regional policy. This may also result in deepening environmental disparities and inequalities in quality of life among the studied urban centers. The two-dimensional assessment reveals that the majority of the examined cities fall into the stagnator category, exhibiting average levels of both environmental awareness and institutional support for energy transition. The most favorable prospects for effective energy transition are observed in Gliwice and Żory, while Zabrze, Świętochłowice, and Jastrzębie-Zdrój—post-mining cities burdened by limited development potential and financial constraints—demonstrate the least promising outlook. The conclusions and recommendations derived from this article directly align with the implementation of Sustainable Development Goal 7—Affordable and Clean Energy (SDG 7)—which addresses energy transition, including the adoption of clean heat sources. They also support the development of sustainable cities, thereby contributing to the achievement of Sustainable Development Goal 11—Sustainable Cities and Communities (SDG 11). Full article

Electric vehicle headlamp illuminance directly affects the driver’s visibility. Accurately predicting electric vehicle headlamp illuminance is crucial to enhancing driving safety. Existing deep learning models are trained using data collected from real-world road testing, yet external factors may compromise its reliability. Electric vehicle headlamp illuminance prediction primarily relies on data fitting, and such models are prone to overfitting when input data are affected by external disturbances. To solve the problem, we propose a luminancxel properties physical information neural network (LumiPINN) prediction model. Test conditions are designed in accordance with standard. The data was collected in an indoor laboratory to eliminate the influence of external factors, then underwent cleaning and pre-processing to ensure data quality. During the modelling process, the physical model is treated as a constraint, with the loss function to jointly optimise the prediction model. Compared with Deep Neural Network and Artificial Neural Network prediction models, the Mean Absolute Error, Mean Square Error, Root Mean Square Error, Mean Relative Error were reduced by 60.2%, 83.6%, 59.6%, 61.3%, and 71.7%, 90.7%, 69.5%, 71.4%. The Coefficient of Determination improved by 0.0015 and 0.0029. The results show that the LumiPINN prediction model demonstrates higher accuracy in prediction outcomes. Full article

This paper presents the analysis and results of the numerical simulation of the biofuel combustion process: namely, the volumetric mixture of diesel oil (ON) and camelina seed oil methyl ester (CSME) in a diesel engine. The mathematical model used in the simulation is based on a four-stroke diesel engine acting as a power generator. To enable simulations depending on the type of biofuel, a model algorithm was developed in the MATLAB/Simulink environment that allowed for the conditions and parameters to be adjusted according to specific test requirements. The numerical simulation was built on the basis of a real stand, in order to confirm the results of previous research both theoretically and in real applications. The calculation approach starts with the elemental composition of the fuel and goes through the intake, compression, combustion, and expansion stages, culminating in the thermal balance of the engine. The mathematical model confirmed the obtained results, which are comparable to the results from the research station. The obtained results confirm the legitimacy of using CSME as an additive to diesel and show its impact on engine performance that can be optimized to achieve the desired results. The use of pure CSME (100%) resulted in an increase in engine power and torque, probably due to the oxygen content of the biofuel molecules and its higher cetane number, which improves its ignition characteristics. However, an increase in unit fuel consumption has been observed, indicating lower energy efficiency compared to clean diesel, which is partially offset by the higher density of biofuel. The model takes into account the physicochemical properties of the fuel, such as the viscosity, cetane number and density, which significantly affect the fuel injection and atomization processes. Although the simulation is based on simplified assumptions, its results highlight the potential of biofuels in heavy transport and their cost-effectiveness as an alternative to fossil fuels. The developed model is used not only to evaluate the engine performance, but also as a tool for assessing the thermal efficiency, and optimizing the composition of the fuel mixture. Full article

Titanium dioxide nanoparticles (TiO₂ NPs) are incorporated into automotive coatings to enhance durability, corrosion, UV resistance, and, in some formulations, photocatalytic self-cleaning. While the toxicology of pristine TiO₂ is well studied, the behavior of TiO₂ NPs embedded in polymer matrices and subjected to real-world aging, maintenance, and removal remains poorly characterized. This narrative review synthesizes 24 publications spanning the lifecycle of TiO₂ nano-enabled automotive coatings, from synthesis and formulation through application, in-service weathering, repair, refinishing, and end-of-life environmental fate. Upstream properties, such as coating functionality and performance, have been examined as determinants of later-life release, exposure, and fate. Across studies, dispersion state, interfacial compatibility, and surface modification—together with transformations such as agglomeration, photocatalysis, weathering, and eco-corona formation—shape particle stability, release, exposure relevance, and toxicological risk. Evidence indicates that sanding and accelerated weathering predominantly generate matrix-associated, polymer-fragment-dominated aerosols rather than pristine TiO₂ NPs, while NP-specific exposure measurements during spray application remain limited. Hazard data suggest matrix embedding may attenuate, but does not eliminate, biological responses relative to pure particles. Wastewater treatment plants and biosolids have been shown to act as sinks with potential for soil accumulation following sludge application. Regulatory frameworks rarely account for aging, transformation, and release, stressing the need for synchronized testing of aged materials and nano-specific exposure metrics to support safer-by-design coatings and risk governance. Full article

The chicken farm is a specific type of agricultural site with high electricity and heat consumption, which makes it ideal for the implementation of green energy. The specificity of the farm (need for continuous ventilation, lighting, and heating) allows achieving energy independence and reducing costs. Small farms can meet their own electricity needs using clean energy through the application of photovoltaics and converting waste biomass to usable energy. These two ways of power production could also reduce carbon footprints. In this study, the feasibility of using renewable energy for energy management in a poultry farm by consecutively involving solar and biomass energy was revealed. A biotechnological process for the production of biogas from chicken litter in a continuously stirred system of tank bioreactors was performed. It was supplied by electricity from a photovoltaic system. To obtain the maximum amount of solar energy, a photovoltaic system consisting of four panels, invertor and a battery with smart control was designed to collect, store, and bring energy to the reactor system collector and connected to the laboratory bioreactor, conveying the biogas production process. Several hydraulic retention times (HRT) were tested for optimizing biogas (biomethane) production, reaching a maximum of 575.49 NmL CH₄/dm³ at an HRT of 13.3 days for the first bioreactor and 278.7 NmL CH₄/g VS_add at an HRT of 120 days for the whole system. The energy balance made, reporting meteorological data, showed the economic feasibility for small farms to meet their own electricity needs. Involving renewable energy technologies could solve the problem of fossil fuel dependency and waste management for environmental protection and profit increase. It would permit a transition toward sustainable energy practices in agriculture and food production. Full article

Background/Objectives: To make a dental office more environmentally conscious, it is essential to use eco-friendly disinfectants without compromising patient safety. Our aim was to examine and compare the effectiveness of the commonly used conventional disinfectants in Romania with the effectiveness of available eco-friendly disinfectants. Methods: Two traditional (Gigasept, Zeta 1 Ultra), and two eco-friendly (IDactiv, Sekusept Aktiv) disinfectants and an eco-friendly (gigazyme) cleaning agent were compared. For a thorough evaluation, minimal inhibitory and bactericidal concentration (MIC/MBC) tests were conducted, along with a Bradford assay to measure the concentration of residual proteins on instruments contaminated by controlled contamination and dental office use after disinfection. Results: During the MIC test, with the exception of Gigazyme^®, which is an enzymatic cleaner, all of the tested disinfectants were effective in the case of the bacterial reference strains used at the concentration prescribed by the manufacturers. After disinfection, the amount of protein remaining on the dental instruments did not exceed the maximum concentration according to the ISO 15883-5:2005 protocol, and the protein removal efficiency was between 96.04% and 99.03% for all disinfectants. In the MIC/MBC test, the peracetic acid-based Sekusept Aktiv was the most effective, while Gigazyme^® and BossKlein IDactiv^® showed the highest protein removal efficiency. Conclusions: The tested eco-friendly disinfectants meet the necessary performance criteria and thus offer a sustainable alternative to traditional disinfectants. Full article

Background: This study compares the effects of relativized barbell loads (% of one-repetition maximum or 1RM) versus standardized prescribed loads on High-Intensity Functional Training (HIFT) performance, strength adaptations, physiological response, and perceived effort. Methods: In total, 22 experienced HIFT athletes (12 males, 10 females) were randomly assigned to either a standardized load (SL) or relativized load (RL) group. Both groups completed an 8-week HIFT program with benchmark workouts. Performance was assessed using a local muscle endurance test, maximal strength through 1RM testing (back squat, clean, and clean and jerk), and neuromuscular performance via countermovement jump (CMJ). Cardiopulmonary response (VO²peak, VO²mean, heart rate, and blood lactate levels) and perceived effort (Borg CR-10) were also evaluated. Results: RL participants did not show a difference in the interaction between group and time in TT performance but differences were founded for strength gains in back squat (p = 0.005, 95% CI [3.1, 8.6]) and clean (p = 0.027, 95% CI [1.2, 5.7]) compared to the SL group. No significant differences were found in clean and jerk performance or CMJ height. Cardiopulmonary responses were similar between groups, indicating comparable physiological stress. RL participants reported significantly lower perceived exertion (p < 0.001, 95% CI [6.3, 9.8]), suggesting more efficient load management and recovery. Conclusions: Use of individualized loads based on 1RM enhanced HIFT performance and strength adaptations, without increasing physiological stress, enabling more efficient training with reduced fatigue. Full article

Organic solvents are known to affect the nervous system, but neurotoxicity testing is not routinely required for industrial chemicals under current European regulations. Glycol ethers are widely used in consumer and industrial products. They can cross skin and lung barriers, distribute systemically, and penetrate the blood–brain barrier due to their physicochemical properties, while their neurotoxic potential remains poorly characterized. P-series glycol ethers now dominate the European market, making exposure assessment critical for public health. We compiled and integrated data from five authoritative sources to build an inventory of glycol ethers currently used in Europe and performed a structured descriptive analysis of high-volume propylene glycol ether compounds. Six high-volume compounds (≥1000 t/year) were selected for analysis. Production volumes, Swiss product registrations, occupational exposure limits, and product categories were compiled. Propylene glycol methyl ether (PGME) showed the highest tonnage (100,000–1,000,000 t/year) and was present in 9497 registered products, followed by propylene glycol ethyl ether (PGEE) (10,000–100,000 t/year; 1333 products). Paints/coatings and cleaning agents were the most frequent product categories, while additional presence in personal care and indoor-use products was observed. These products may lead to exposure depending on use conditions, such as spraying or inadequate ventilation, which can increase inhalation and skin contact. Their presence in diverse products suggests potential for both occupational and chronic low-level exposures. By providing an integrated overview of market presence, use patterns, and available neurotoxicity evidence for propylene glycol ethers, our findings highlight a critical gap in chemical risk assessment: the absence of neurotoxicity testing despite high production volumes and widespread use. Integrating neurotoxicity endpoints and new approach methodologies into regulatory frameworks is essential to strengthen public health protection. Full article

The effectiveness of deep learning methods in image segmentation has led to interest in their deployment for 3D point cloud segmentation, particularly in the context of pre-grasp identification of a unique object amongst distractors. However, existing 3D object datasets are not ideal for training and evaluation of these methods. Datasets developed for grasp planning are often CAD models that are too clean for sim-to-real transfer. Real-world datasets can lack texture information or have been collected using sets of objects and/or specialized sensor setups that are hard to reproduce. In this work, we introduce the MiniMarket80 dataset to address this gap.The dataset consists of 1200 colored point cloud partial views, each of 80 standard grocery objects, collected with widely used Realsense RGB-D cameras (D415 and D435) under variable lighting conditions. We also provide a complete pipeline to generate a per-object segmentation dataset from these partial views suitable for use in training. We use this dataset to evaluate 11 state-of-the-art point cloud segmentation methods. Only four of these are able to (partially) segment the target object in a real-world test, still producing significant false positives and false negatives. Full article

Non-typhoidal Salmonella (NTS) are one of the most common foodborne pathogens worldwide, and pork is a major food vehicle together with eggs and poultry meat. Contamination of pork within food processing facilities, such as slaughterhouses, can be associated with persistence of Salmonella in the environment due to biocide resistance. In this study, we assessed the susceptibility of NTS isolates from pig carcasses to a QAC-based commercial formulation according to the EN 1656/2009 standard and the presence of QAC resistance genes through PCR. The qacEΔ1 and qacF genes were found in 31.8% and 29.5% of the isolates respectively, while qacE was absent. All isolates were found to be susceptible at a tested concentration 10 times lower (0.1%) than the minimum in-use recommended concentration, with MIC values below 0.1% (≈70 mg/L of Benzalkonium Chloride). Our findings point towards the importance of correct cleaning and disinfection protocols and the role of good hygiene practices as corrective and/or preventive measures to avoid cross-contamination. Full article

Aiming at the problems of inter-row straw congestion, soil accumulation, and consequent uneven seeding depth during high-speed sowing with narrow row spacing under the summer maize no-tillage sowing mode in the Huang-Huai-Hai region, this study proposed a maize seedling belt pre-sowing treatment device suitable for narrow row spacing operation by analyzing the physical properties of straw and soil in the region. Dynamic analysis of the mechanical device was carried out, and the key factors affecting the straw removal effect of the seedling belt and the degree of soil disturbance were identified as machine offset distance, traction speed, and straw-cleaning wheel angle. Discrete element method simulation experiments were conducted via EDEM-ADAMS coupling; the key factors were simulated and optimized, and the optimal parameter combination of the device was determined as follows: machine offset distance of 165 cm (the relative distance between the front and rear positions of the right wheel of adjacent unit cleaning components), traction speed of 11 km/h, and straw-cleaning wheel angle of 44°. Field validation tests of the prototype were performed. The test results showed that the overall straw removal rate of the seedling belt reached 95%, and no large-scale straw and soil accumulation caused by pushing was observed between rows. Compared with the simulation results, the error of straw removal rate was only 0.5%. Sowing comparison tests were conducted, and the results indicated that the device could significantly improve the uniformity of seeding depth and meet the seedling belt quality requirements for high-speed sowing with narrow row spacing of summer maize. This study provides new ideas and methods for the design of straw-cleaning mechanisms in no-till seeding systems. Full article

Chemical monitoring of pollutants and hazardous materials in water supply systems traditionally depends on centralized laboratories, advanced instrumentation, and trained personnel, limiting accessibility and preventing real-time, on-site analysis. This work presents an alternative cost-effective, field-deployable approach that uses genetically engineered bioluminescent bioreporters, encapsulated in self-sufficient alginate capsules and integrated with an optoelectronic detection circuit, to detect and quantify target materials in water. We have developed a scalable single-channel prototype featuring four sensing tracks—two for sample measurement, one for clean water, and one for a standard reference solution. The latter employs the standard ratio (SR) method to ensure robust quantification, compensating for batch variability and environmental effects. System characterization showed high uniformity across tracks. Validation with nalidixic acid (NA) demonstrated reliable quantitative performance, with a blind test estimation of 5.6 mg/L for a true concentration of 5 mg/L, well within the calibration error range. Additional sensitivity testing confirmed detection of mitomycin C (MMC) at concentrations as low as 50 µg/L. Overall, the results highlight the potential of bacterial chemical sensing as a practical and scalable tool for real-time, in situ water quality monitoring networks. Full article