Search Results (369)

Automated surface defect detection in steel manufacturing is pivotal for ensuring product quality, yet it remains an open challenge owing to the extreme heterogeneity of defect morphologies—ranging from hairline cracks and microscopic pores to elongated scratches and shallow dents. Existing approaches, whether classical vision pipelines or recent deep-learning paradigms, struggle to simultaneously satisfy the stringent demands of industrial scenarios: high accuracy on sub-millimeter flaws, insensitivity to texture-rich backgrounds, and real-time throughput on resource-constrained hardware. Although contemporary detectors have narrowed the gap, they still exhibit pronounced sensitivity–robustness trade-offs, particularly in the presence of scale-varying defects and cluttered surfaces. To address these limitations, we introduce MBY (MBDNet-Attention-YOLO), a lightweight yet powerful framework that synergistically couples the MBDNet backbone with the YOLO detection head. Specifically, the backbone embeds three novel components: (1) HGStem, a hierarchical stem block that enriches low-level representations while suppressing redundant activations; (2) Dynamic Align Fusion (DAF), an adaptive cross-scale fusion mechanism that dynamically re-weights feature contributions according to defect saliency; and (3) C2f-DWR, a depth-wise residual variant that progressively expands receptive fields without incurring prohibitive computational costs. Building upon this enriched feature hierarchy, the neck employs our proposed MultiSEAM module—a cascaded squeeze-and-excitation attention mechanism operating at multiple granularities—to harmonize fine-grained and semantic cues, thereby amplifying weak defect signals against complex textures. Finally, we integrate the Inner-SIoU loss, which refines the geometric alignment between predicted and ground-truth boxes by jointly optimizing center distance, aspect ratio consistency, and IoU overlap, leading to faster convergence and tighter localization. Extensive experiments on two publicly available steel-defect benchmarks—NEU-DET and PVEL-AD—demonstrate the superiority of MBY. Without bells and whistles, our model achieves 85.8% mAP@0.5 on NEU-DET and 75.9% mAP@0.5 on PVEL-AD, surpassing the best-reported results by significant margins while maintaining real-time inference on an NVIDIA Jetson Xavier. Ablation studies corroborate the complementary roles of each component, underscoring MBY’s robustness across defect scales and surface conditions. These results suggest that MBY strikes an appealing balance between accuracy, efficiency, and deployability, offering a pragmatic solution for next-generation industrial quality-control systems. Full article

In real-world applications, user demands for new functionalities and activities constantly evolve, requiring action recognition systems to incrementally incorporate new action classes without retraining from scratch. This class-incremental learning (CIL) paradigm is essential for enabling adaptive and scalable systems that can grow over time. However, Wi-Fi-based indoor action recognition under incremental learning faces two major challenges: catastrophic forgetting of previously learned knowledge and uncontrolled model expansion as new classes are added. To address these issues, we propose CAREC, a class-incremental framework that balances dynamic model expansion with efficient compression. CAREC adopts a multi-branch architecture to incorporate new classes without compromising previously learned features and leverages balanced knowledge distillation to compress the model by 80% while preserving performance. A data replay strategy retains representative samples of old classes, and a super-feature extractor enhances inter-class discrimination. Evaluated on the large-scale XRF55 dataset, CAREC reduces performance degradation by 51.82% over four incremental stages and achieves 67.84% accuracy with only 21.08 M parameters, 20% parameters compared to conventional approaches. Full article

While the explosive growth in e-commerce stresses urban logistics systems, city planners lack of fine-grained data in order to anticipate and manage the resulting freight flows. Using a three-stage analytical approach combining descriptive zonal statistics, hotspot analysis and different regression modeling from univariate to geographically weighted regression, this study integrates one year of parcel deliveries from a leading courier with open spatial layers of land-use zoning, census population, mobile-signal activity and household income to model last-mile demand across different land use types. A baseline linear regression shows that residential population alone accounts for roughly 30% of the variance in annual parcel volumes (2.5–3.0 deliveries per resident) while adding daytime workforce and income increases the prediction accuracy to 39%. In a similar approach where coefficients vary geographically with Geographically Weighted Regression to capture the local heterogeneity achieves a significant raise of the overall R² to 0.54 and surpassing 0.70 in residential and institutional districts. Hot-spot analysis reveals a highly fragmented pattern where fewer than 5% of blocks generate more than 8.5% of all deliveries with no apparent correlation to the broaden land-use classes. Commercial and administrative areas exhibit the greatest intensity (1149 deliveries per ha) yet remain the hardest to explain (global R² = 0.21) underscoring the importance of additional variables such as retail mix, street-network design and tourism flows. Through this approach, the calibrated models can be used to predict city-wide last-mile demand using only public inputs and offers a transferable, privacy-preserving template for evidence-based freight planning. By pinpointing the location and the land uses where demand concentrates, it supports targeted interventions such as micro-depots, locker allocation and dynamic curb-space management towards more sustainable and resilient urban-logistics networks. Full article

A rising demand for low-gluten beer fuels research into enzymatic solutions. This study optimized Aspergillus niger prolyl endopeptidase (AN-PEP) application timing during brewing to reduce gluten while preserving physicochemical quality. Ale-type beers were produced with AN-PEP (2% v/v) added at mashing, boiling, post-boiling, or post-fermentation, plus a control. Three mashing profiles (Mash A, B, C) were also tested. Gluten was quantified by R5 ELISA (LOQ > 270 mg/L). Color, bitterness, ABV, and foam stability were assessed. Statistical analysis involved ANOVA and Tukey’s HSD (p < 0.05). Enzyme activity and thermal inactivation were also evaluated. Initial gluten levels consistently exceeded LOQ. Significant gluten reduction occurred only post-fermentation. Mashing, boiling, and post-boiling additions effectively lowered gluten to below 20 mg/L. Post-fermentation addition resulted in significantly higher residual gluten (136.5 mg/L). Different mashing profiles (A, B, C) with early enzyme addition achieved similar low-gluten levels. AN-PEP showed optimal activity at 60–65 °C, inactivating rapidly at 100 °C. Physicochemical attributes (color, extract, bitterness, ABV) were largely unaffected. However, foam stability was significantly compromised by mashing and post-fermentation additions, while preserved with boiling and post-boiling additions. AN-PEP effectively produces low-gluten beers. Enzyme addition timing is critical: while mashing, boiling, or post-boiling additions reduce gluten to regulatory levels, only the beginning of boiling or post-boiling additions maintain desirable foam stability. These findings offer practical strategies for optimizing low-gluten beer production. Full article

As the global demand for eco-friendly food ingredients grows, marine macroalgae emerge as a valuable resource for multiple applications using a circular bioeconomy approach. In this study, green macroalgae Ulva flexuosa, naturally accumulated in aquaculture ponds as a residual biomass (by-product) of shrimp and oyster farming, were investigated regarding their bioactivity, chemical composition, and antioxidant properties. The use of aquaculture by-products as raw materials not only reduces waste accumulation but also makes better use of natural resources and adds value to underutilized biomass, contributing to sustainable production systems. For this, a comprehensive approach including the evaluation of its composition and environmentally friendly extraction of bioactive compounds was conducted and discussed. Green macroalgae exhibited high fiber (37.63% dry weight, DW) and mineral (30.45% DW) contents. Among the identified compounds, palmitic acid and linoleic acid (ω-6) were identified in the highest concentrations. Pigment analysis revealed a high concentration of chlorophylls (73.95 mg/g) and carotenoids (17.75 mg/g). To evaluate the bioactivity of Ulva flexuosa, ultrasound-assisted solid–liquid extraction was performed using water, ethanol, and methanol. Methanolic extracts showed the highest flavonoid content (59.33 mg QE/100 g), while aqueous extracts had the highest total phenolic content (41.50 mg GAE/100 g). Ethanolic and methanolic extracts had the most potent DPPH scavenging activity, whereas aqueous and ethanolic extracts performed best at the ABTS assay. Overall, we show the upcycling of Ulva flexuosa, an underexplored aquaculture by-product, as a sustainable and sensible strategy for multiple value-added applications. Full article

This paper presents platforms as a solution to rethink how we build, addressing the pressing paradox between meeting growing housing demands. The construction sector has not fully grasped the advantages of platforms beyond standardisation and efficiency. In contrast, other sectors have begun acknowledging that platforms can capture increased value through interactions among firms within a networked ecosystem. Learning from other sectors, this paper investigates platforms in the construction context, aiming to define, classify, and assess their impact on the construction value chain. The research approach was abductive, involving a cross-sectoral review of 190 platforms across 16 Australian and New Zealand Standard Industrial Classification (ANZSIC) industries and semi-structured interviews with stakeholder groups of the construction value chain in Australia. The findings categorise platforms as physical, digital, or hybrid, highlighting their potential to move value-added activities upstream, facilitate collaboration, and foster innovation through data-driven insights. The paper’s novelty lies in the exhaustive cross-sectoral review, the classification of platforms in the construction context, and the proposition of a platform approach as a versatile framework tailored to diverse needs and circumstances that offers a fresh perspective on sustainable building practices. The practical contribution of this study lies in offering guidelines for industry practitioners aiming to develop or refine a platform-based approach tailored to the construction context. Full article

The growing demand for sustainable materials has intensified the search for biodegradable polymers. Poly(ε-caprolactone) (PCL), though biodegradable, is fossil-derived. In this study, a novel lignin extracted from pine wood using a green solvent was incorporated into PCL and compared with commercial lignins (dealkaline, alkaline, and lignosulfonate). The lignin additions imparted antioxidant properties, enhanced thermal stability, and promoted circular economy goals through lignin valorization. Notably, the green-extracted lignin showed superior compatibility with PCL when compared with commercial lignins, as evidenced by lower water uptake and solubility, and improved surface hydrophobicity (higher contact angle). Although the addition of lignin reduced the tensile strength and elongation at break, it greatly increased the PCL radical scavenging activity (DPPH) from 8 ± 1% of neat PCL to 94.8 ± 0.3% when 20 wt% of lignin-LA was added. Among the tested lignins, lignin-LA stands out as the most promising candidate to be applied as a functional additive in biodegradable polymer blends and composites for advanced sustainable applications. Not only given its intrinsically higher sustainability but also due to its capacity for improving the thermal properties of PCL–lignin blends. Full article

Consumer demand for low-alcohol acidic beers is driving the use of non-conventional yeasts in the brewing process. In this study, the addition of mixed berries and fermentation with L. thermotolerans L31 are performed in crafting a low-alcohol acidic beer. Four different beers were brewed in the primary stage with either Saccharomyces cerevisiae or L. thermotolerans and with or without added berry mixture. Beer was fermented for 8 days at 20 °C, stored, and bottled. pH, density, alcoholic content, bitterness, and color of final beer were analyzed for all samples using analytical methods. Volatile compounds, anthocyanin content, and antioxidant activity were also evaluated. Sensory analysis was performed and correlated (PCA) with the analytical results. The obtained data indicated that beers brewed with L. thermotolerans were significantly more acidic and less bitter than S. cerevisiae beers. No difference in alcoholic content was found. Fruity aroma-associated compounds were present in L. thermotolerans beers, which correlated with the sensory analysis. Fruit beers were also redder and showed higher anthocyanin content and stronger antioxidant activity due to the presence of anthocyanins such as cyanidin, delphinidin, and malvidin from fruit, and other antioxidant compounds. Full article

The large-scale disposal of loquat (Eriobotrya japonica Lindl.) flowers during fruit thinning represents a significant waste of bioactive resources. This study systematically evaluated how three processing methods—fresh (FS), heat-dried (HD), and freeze-dried (FD) treatments—affect the flavonoid composition and antioxidant capacity of loquat flower extracts, with the aim of developing value-added, sugar-free functional tea ingredients. Using UPLC-MS/MS and DPPH assays, we analyzed both pre-(FS/HD/FD) and post-extraction samples (FSP/HDP/FDP) to assess processing-specific metabolic signatures and extraction efficiency. The results revealed that heat-dried powder (HDP) exhibited the highest total flavonoid content and DPPH scavenging capacity (615.24 µg Trolox/g), attributed to enhanced release of stable compounds like quercetin. Freeze-dried powder (FDP) better preserved heat-sensitive flavonoids, such as catechin-(4α→8)-gallocatechin and naringenin, but showed lower overall antioxidant activity. Multivariate analysis confirmed distinct clustering patterns, with heat-drying favoring flavonoid extractability while freeze-drying maintained metabolic diversity. These findings demonstrate that processing methods significantly influence bioactive compound retention and functionality, with heat-drying offering optimal balance between yield and practicality for industrial applications. This work provides a scientific foundation for upcycling loquat flowers into standardized nutraceutical ingredients, addressing both agricultural waste reduction and the growing demand for natural functional foods. Full article

Increasing awareness of the negative health effects associated with high sugar intake has led to a growing demand for reducing added sugar in food products. In this study, the antioxidant properties of commercial yogurts containing pasteurized fruits were evaluated and compared with natural yogurts freshly enriched with 3–20% thawed fruits (bilberries, cherries and strawberries). Additionally, yogurts enriched with cornelian cherry were analyzed. Antioxidant activity was assessed using the ABTS and DPPH methods, along with measurements of total polyphenol content and reducing power. The effect of fruit addition on the number of yogurt bacteria was also investigated. The results showed that the addition of fruits significantly increased the yogurts’ ability to neutralize free radicals, attributed to the presence of natural antioxidants and polyphenols. The addition of fruits helped maintain the vitality of lactic acid bacteria, with bacterial counts remaining well above the minimum threshold of 10⁷ cfu/g. The findings demonstrated that cornelian cherry has great potential as a source of polyphenols with antioxidant properties. These results confirm the high nutritional value of yogurts enriched with thawed fruit, which may serve as a valuable component of a healthy diet and a healthier alternative to sweetened yogurts commonly available in stores. Full article

The rising demand for vegan products calls for new plant-based antimicrobial preservation methods. This study evaluates an antifungal ingredient obtained by fermenting oat drink with lactic acid bacteria to extend vegan mortadella’s shelf life. In vitro tests showed antimicrobial effects against Aspergillus flavus, Penicillium commune, and Listeria monocytogenes (inhibition zones: 2–5 mm). The enrichment of the oat drink culture medium with additional nutrients enhanced fermentation performance and increased antifungal activity. The fermented culture medium with the highest antimicrobial activity was used to develop a bioactive ingredient for the preservation of vegan mortadella conservation. Adding 3% of this ingredient to vegan mortadella improved microbial stability, reducing mesophilic bacteria by 2.5 Log₁₀ CFU/g and increasing lactic acid bacteria. Lower pH and water activity changes were observed but remained within quality standards. Contamination assays showed a consistent reduction of A. flavus over 7 days, while P. commune and L. monocytogenes dropped below detection within 2 days. In contrast, control samples maintained contamination levels near 3.0 Log₁₀ CFU/g. These findings support the potential of fermented oat-based ingredients as effective, natural preservatives for vegan foods. Full article

Large-scale traffic simulations at a microscopic level can mimic the physical reality in great detail so that innovative transport services can be evaluated. However, the simulation times of such scenarios is currently too long to be practical. (1) Background: With the availability of Graphical Processing Units (GPUs), is it possible to exploit parallel computing to reduce the simulation times of large microscopic simulations, such that they can run on normal PCs at reasonable runtimes?; (2) Methods: ParSim, a microsimulator with a monolithic microsimulation kernel, has been developed for CUDA-compatible GPUs, with the aim to efficiently parallelize the simulation processes; particular care has been taken regarding the memory usage and thread synchronization, and visualization software has been optionally added; (3) Results: The parallelized simulations have been performed by a GPU with an average performance, a 24 h microsimulation scenario for Bologna with 1 million trips was completed in 40 s. The average speeds and waiting times are similar to the results from an established microsimulator (SUMO), but the execution time is up to 5000 times faster with respect to SUMO; the 28 million trips of the 24 h San Francisco Bay Area scenario was completed in 26 min. With cutting-edge GPUs, the simulation speed can possibly be further reduced by a factor of seven; (4) Conclusions: The parallelized simulator presented in this paper can perform large-scale microsimulations in a reasonable time on readily available and inexpensive computer hardware. This means microsimulations could now be used in new application fields such as activity-based demand generation, reinforced AI learning, traffic forecasting, or crisis response management. Full article

Natural deep eutectic solvents (NADESs) have emerged as a promising eco-friendly alternative to petrochemicals for extracting plant metabolites. Considering that the demand for sustainable “green” ingredients for industrial applications is growing, those solvents are purported to develop extracts with interesting phytochemical fingerprints and biological activities. Given the interest in flavonoids from Chenopodium quinoa Willd. leaves, an efficient “green” extraction method was developed by investigating eight NADESs with defined molar ratios, i.e., malic acid-choline chloride (chcl)-water (w) (1:1:2, N1), chcl-glucose-w (5:2:5, N2), proline-malic acid-w (1:1:3, N3), glucose-fructose-sucrose-w (1:1:1:11, N4), 1,2-propanediol-chcl-w (1:1:1, N5), lactic acid-glucose-w (5:1:3, N6), glycerol-chcl-w (2:1:1, N7), and xylitol-chcl-w (1:2:3, N8). Rheological measurements of all NADESs confirmed their pseudoplastic behaviors. To improve the extraction processes, differential scanning calorimetry (DSC) allowed us to determine the maximum amount of water that could be added to the most stable NADES (N1, N2, N3, and N4; 17.5%, 20%, 10%, and 10% w/w, respectively) to lower their viscosities without disturbing their eutectic environments. The phytochemical compositions of NADES extracts were analyzed using high-performance thin-layer chromatography (HPTLC), and their free radical scavenging and α-amylase inhibitory properties were assessed using HPTLC-bioautography. N2, diluted with 20% of water, and N7 presented the best potential for replacing methanol for an eco-friendly extraction of flavonoids, radical scavengers, and α-amylase inhibitors from quinoa leaves. Their biological properties, combined with a good understanding of both thermal behavior and viscosity, make the obtained quinoa leaf NADES extracts good candidates for direct incorporation in nutraceutical formulations. Full article

Despite the convenience of solar potential and the magnitude of energy received by the Earth from the sun, solar photovoltaic systems have failed to meet the growing energy demand. This can be attributed to various factors such as low cell efficiency, environmental conditions, and improper tracking of operating points, which further worsen the system’s performance. Various advanced metaheuristic-based Maximum Power Point Tracking (MPPT) techniques were reported in the literature. Most available techniques were designed and tested in subscription-based/paid software such as MATLAB/Simulink, PSIM simulator, etc. Due to this, the simulation and analysis of these MPPT algorithms for developing and underdeveloped countries added an extra economic burden. Many open-source PV libraries are computationally intensive, lack active support, and prove impractical for MPPT testing on resource-constrained hardware. Their complexity and absence of optimization for edge devices limit their viability for the edge device. This issue is addressed in this research by designing a robust framework using an open-source programming language i.e., Python. For demonstration purposes, we simulated and analyzed a solar PV system and benchmarked its performance against the JAP6 solar panel. We implemented multiple metaheuristic MPPT algorithms including Artificial Bee Colony (ABC) and Particle Swarm Optimization (PSO), evaluating their efficacy under both Standard Test Conditions (STC) and complex partial shading scenarios. The results obtained validate the feasibility of the implementation in Python. Therefore, this research provides a comprehensive framework that can be utilized to implement sophisticated designs in a cost-effective manner for developing and underdeveloped nations. Full article

Robust indoor robot navigation typically demands either costly sensors or extensive training data. We propose a cost-effective RGB-D navigation pipeline that couples feature-based relative pose estimation with a lightweight multi-layer-perceptron (MLP) policy. RGB-D keyframes extracted from human-driven traversals form nodes of a topological map; edges are added when visual similarity and geometric–kinematic constraints are jointly satisfied. During autonomy, LightGlue features and SVD give six-DoF relative pose to the active keyframe, and the MLP predicts one of four discrete actions. Low visual similarity or detected obstacles trigger graph editing and Dijkstra replanning in real time. Across eight tasks in four Habitat-Sim environments, the agent covered 190.44 m, replanning when required, and consistently stopped within 0.1 m of the goal while running on commodity hardware. An information-theoretic analysis over the Multi-Illumination dataset shows that LightGlue maximizes per-second information gain under lighting changes, motivating its selection. The modular design attains reliable navigation without metric SLAM or large-scale learning, and seamlessly accommodates future perception or policy upgrades. Full article