Search Results (709)

Medical consumable orders are characterized by diverse product types, small batch sizes, frequent orders, and high customization requirements, often leading to inefficient workshop scheduling and difficulties in meeting multiple production constraints. To address these challenges, this study proposes a bi-level optimization model for order splitting and reorganization considering multi-dimensional and multi-scale characteristics. The multi-dimensional characteristics encompass materials, processes, equipment, and work efficiency, while the multi-scale aspects involve finished products, components, assemblies, and parts. At the upper level, the model optimizes order task splitting by refining splitting strategies and preprocessing constraints to generate high-quality input for the reorganization phase. The lower level optimizes sub-task prioritization, batch sizes, and resource scheduling to develop a production plan that balances cost and efficiency. Subsequently, to solve this bi-level optimization problem, a hybrid bi-objective optimization algorithm is designed, integrating a collaborative iterative strategy to enhance solution efficiency and quality. Finally, a case study and comparative experiments validate the practicality and effectiveness of the proposed model and algorithm. Full article

Energy consumption has emerged as a critical design constraint in deploying high-performance neural networks, especially on edge devices with limited power resources. In this paper, a comparative study is conducted for two prevalent deep learning paradigms—convolutional neural networks (CNNs), exemplified by ResNet18, and transformer-based large language models (LLMs), represented by GPT3-small, Llama-7B, and GPT3-175B. By analyzing how the scaling of memory energy versus computing energy affects the energy consumption of neural networks with different batch sizes (1, 4, 8, 16), it is shown that ResNet18 transitions from a memory energy-limited regime at low batch sizes to a computing energy-limited regime at higher batch sizes due to its extensive convolution operations. On the other hand, GPT-like models remain predominantly memory-bound, with large parameter tensors and frequent key–value (KV) cache lookups accounting for most of the total energy usage. Our results reveal that reducing weight-memory energy is particularly effective in transformer architectures, while improving multiply–accumulate (MAC) efficiency significantly benefits CNNs at higher workloads. We further highlight near-memory and in-memory computing approaches as promising strategies to lower data-transfer costs and enhance power efficiency in large-scale deployments. These findings offer actionable insights for architects and system designers aiming to optimize artificial intelligence (AI) performance under stringent energy budgets on battery-powered edge devices. Full article

Detecting metal surface crack defects is of great significance for the safe operation of industrial equipment. However, most existing mainstream deep-object detection models suffer from complex structures, large parameter sizes, and high training costs, which hinder their deployment and application in frontline construction sites. Therefore, this paper optimizes the existing YOLO series head structure and proposes a lightweight detection head structure, ZoomHead, with lower computational complexity and stronger detection performance. First, the GroupNorm2d module replaces the BatchNorm2d module to stabilize the model’s feature distribution and accelerate the training speed. Second, Detail Enhanced Convolution (DEConv) replaces traditional convolution kernels, and shared convolution is adopted to reduce redundant structures, which enhances the ability to capture details and improves the detection performance for small objects. Next, the Zoom scale factor is introduced to achieve proportional scaling of the convolution kernels in the regression branch, minimizing redundant computational complexity. Finally, using the YOLOv10 and YOLO11 series models as baseline models, ZoomHead was used to replace the head structure of the baseline models entirely, and a series of performance comparison experiments were conducted on the rail surface crack dataset and NEU surface defect database. The results showed that the integration of ZoomHead effectively improved the model’s detection accuracy, reduced the number of parameters and computations, and increased the FPS, achieving a good balance between detection accuracy and speed. In the comparative experiment of the SOTA model, the addition of ZoomHead resulted in the model having the smallest number of parameters and the highest FPS, while maintaining the same mAP value as the SOTA model, indicating that the ZoomHead structure proposed in this paper has better comprehensive detection performance. Full article

Traditional manufacturing processes yield plastic containers in large batches, even for minimal production runs, resulting in elevated production costs. Three-dimensional printing has emerged as a viable alternative for very low production volumes, offering properties comparable to traditional methods at significantly reduced costs. To assess the tensile strength, specimens printed with identical geometric parameters to the 3D-printed containers were tested according to ASTM D638 standards, enabling the determination of the stress–strain curve behavior. A compression test was conducted on containers obtained from both manufacturing processes to establish their respective resistance and deformation characteristics. The results revealed a 67% difference in resistance, indicating greater rigidity in the 3D-printed container, and a higher deformation in the blow-molded container, reaching up to 4 mm in height without fracture. Similarly, impact resistance was analyzed using finite element analysis with Ls-Dyna software, showing deformation differences of 0.91% and stress differences of 2.15%. Therefore, 3D printing presents itself as a compelling alternative for the fabrication of plastic containers in small production runs. Full article

Fed-batch cultivation is a widely used strategy for microbial lipid production, offering flexibility in nutrient control and the potential for high lipid productivity. However, optimizing feeding strategies remains a complex challenge, as it depends on multiple factors, including strain-specific metabolism and process limitations. In this study, we developed a computational framework based on dynamic flux balance analysis and small-scale metabolic models to evaluate and optimize lipid production in Rhodosporidium toruloides strains. We proposed equations to estimate both the carbon and energy source mass feed rate ($F_{in}·s_r$) and its concentration in the feed ($s_r$) based on lipid accumulation targets, and defined minimum feeding flow rate ($F_{in}$) according to process duration. We then assessed the impact of these parameters on commonly used bioprocess metrics—lipid yield, titer, productivity, and intracellular accumulation—across wild-type and engineered strains. Our results showed that the selection of $F_{in}·s_r$ was strongly strain-dependent and significantly influenced strain performance. Moreover, for a given $F_{in}·s_r$, the specific values of $s_r$, and the resulting $F_{in}$, had distinct and non-equivalent effects on performance metrics. This methodology enables the rational pre-selection of feeding strategies and strains, improving resource efficiency and reducing the probability of failed experiments. Full article

Ru/TiO₂ catalysts are well known for their high activity in the hydrogenation of CO₂ to CH₄ (the Sabatier reaction). This activity is commonly attributed to strong metal–support interactions (SMSIs), associated with reducible oxide layers partly covering the Ru-metal particles. Moreover, isothermal rates of formation of CH₄ can be significantly enhanced by the exposure of Ru/TiO₂ to light of UV/visible wavelengths, even at relatively low intensities. In this study, we confirm the significant enhancement in the rate of formation of methane in the conversion of CO₂, e.g., at 200 °C from ~1.2 mol g_Ru⁻¹·h⁻¹ to ~1.8 mol g_Ru⁻¹·h⁻¹ by UV/Vis illumination of a hydrogen-treated Ru/TiO_x catalyst. The activation energy does not change upon illumination—the rate enhancement coincides with a temperature increase of approximately 10 °C in steady state (flow) conditions. In-situ DRIFT experiments, performed in batch mode, demonstrate that the Ru–CO absorption frequency is shifted and the intensity reduced by combined UV/Vis illumination in the temperature range of 200–350 °C, which is more significant than can be explained by temperature enhancement alone. Moreover, exposing the catalyst to either UV (predominantly exciting TiO₂) or visible illumination (exclusively exciting Ru) at small intensities leads to very similar effects on Ru–CO IR intensities, formed in situ by exposure to CO₂. This further confirms that the temperature increase is likely not the only explanation for the enhancement in the reaction rates. Rather, as corroborated by photophysical studies reported in the literature, we propose that illumination induces changes in the electron density of Ru partly covered by a thin layer of TiO_x, lowering the CO coverage, and thus enhancing the methane formation rate upon illumination. Full article

Background/Objectives: Medical leech (Hirudo in the Chinese Pharmacopoeia) is renowned in traditional medicine for its significant antithrombin activity. As an animal-derived medicine with complex and incompletely understood composition, its insufficient quality control measures are met with widespread counterfeiting caused by limited animal resources and rising demand. Methods: In this study, an integrated quality evaluation strategy guided by “Totality of the Evidence” (TOE) method is proposed. This strategy combines chemical characterization of small and macromolecular components with bioassays relevant to its clinical functions. A total of 28 batches of samples were analyzed, comprising 23 genuine and 5 counterfeit batches. Species origins were identified by morphology and DNA barcoding. Chemical characterization included TLC, HPLC and UPLC-QTOF-MS/MS for small molecules, and SDS-PAGE with HPLC-Orbitrap Fusion Lumos Tribrid-MS for macromolecules. Antithrombotic activity was assessed by thrombin titration and platelet aggregation assays. Results: Several characteristic components were discovered and identified as key quality control markers, including eight small molecules such as an unreported compound SZ-1, plus seven major differential proteins across species. Based on these markers, accurate and rapid authentication methods were established using SDS-PAGE for macromolecules, and both HPLC and TLC for small molecules. Furthermore, using bioassay methods we established for quality evaluation, Hirudo nipponica exhibits potent anti-thrombin activity and inhibits platelet aggregation, while Whitmania pigra shows weak anti-thrombin activity and promotes platelet aggregation. Conclusions: This quality evaluation strategy is not only applicable for the quality assessment of genuine Hirudo products of different origins, but also for distinguishing medical leeches from their counterfeits. Full article

The COVID-19 pandemic disrupted medical education as the shift from face-to-face to remote teaching raised concerns about learning outcomes and well-being. However, while established schools’ adaptations have been widely studied, the challenges faced by new institutions in the pandemic’s aftermath remain underexplored. This research provides a valuable case study examining the motivations and concerns of faculty and students at a newly established medical school in the Philippines during its inaugural academic year. Employing a mixed-methods design, data were obtained via validated Likert-based questionnaires assessing motivation and survey questions eliciting concerns. Descriptive and inferential approaches were utilized to analyze data. Results show that faculty motivations were primarily intrinsic, whereas students were driven by a mix of intrinsic and extrinsic factors. Subgroup analyses revealed no difference in motivational profiles across demographic characteristics. Faculty were primarily concerned with their readiness for academic roles, use of modern teaching technologies, and ensuring effective student learning. Students’ concerns focused on training quality, institutional capacity, and the uncertainties of being the pioneer batch. Despite the pioneer cohort’s small size, this study highlighted the importance of understanding faculty and student motivations and concerns, already shaped by post-pandemic realities, to provide targeted support for new medical programs in the evolving post-pandemic landscape. Full article

Coal combustion generates significant volumes of ash, a technogenic by-product that poses a serious threat to regional environmental sustainability (environmental chemical contamination and air pollution). This study aims to assess the feasibility of utilizing this type of ash as a raw material component in the fabrication of refractory bricks and to investigate the fundamental properties of the resulting experimental products. Ash was incorporated into the batch composition at concentrations ranging from 10% to 40% by weight, blended with clay and water, then shaped through pressing and subjected to firing at 1000 °C and 1100 °C in an air atmosphere for 2 h. After complete cooling, the samples were subjected to compressive strength testing. Samples containing 40 wt% coal ash exhibited insufficient compressive strength and were therefore excluded from subsequent investigations. For the remaining samples, apparent density, open porosity and slag resistance were determined. The microstructural characterization was performed, and the phase composition of the samples was analyzed. The results revealed that the phase composition of the experimental samples differs significantly from that of the reference sample (ShA-grade chamotte brick in accordance with GOST 390-96, currently used as lining in metallurgical furnaces across the country), exhibiting a higher mullite content and the absence of muscovite. A small amount of kaolinite was detected in the experimental samples even after a 2-h firing process. This observation may be attributed to the effect of kaolinite crystallinity on the transformation process from kaolinite to metakaolinite. The mechanical strength of the experimental samples meets the relevant standards, while slag resistance demonstrated an improvement of approximately 15%. Open porosity was found to decrease in the experimental samples. In addition, a change in the pore size distribution was observed. Notably, the proportion of pores larger than 10,000 nm was significantly reduced. These findings confirm the feasibility of incorporating coal ash as a viable raw material component in the formulation of refractory materials. Full article

This paper aims to describe a method of non-destructive testing in highly variable production using a robotic ultrasound diagnostic system. Highly variable production involves producing products of different shapes and dimensions, which requires flexible positioning and the adaptation of diagnostic technology. Typically, highly variable products are made in small batches. Implementing automation tools, including robotic systems, can meet this requirement, even in diagnostic processes that demonstrate the required quality level. This paper deals with the design, simulation, optimization, and verification of an innovative robotic ultrasound diagnostic system for non-destructive testing in highly variable production environments. The robotic positioning of the diagnosed object and the ultrasound probe achieves the automatic adaptation of the system. Robots can automatically position the ultrasound probe relative to the diagnosed object using laser distance measurements in the water environment of a diagnostic vessel. Computer processing of the data measured by the ultrasound probe enables the evaluation of data and the documentation of quality criterion fulfillment in digital form. Setting the technology parameters, monitoring the technology status, and displaying the quality control results are enabled by a human–machine interface system also described in this paper. Full article

Objective: Current good manufacturing practice (cGMP) guidance for positron emission tomography (PET) drugs has been established in Europe and the United States. In Japan, the Pharmaceuticals and Medical Devices Agency (PMDA) approved the use of radiosynthesizers as medical devices for the in-house manufacturing of PET drugs in hospitals and clinics, regardless of the cGMP environment. Without adequate facilities, equipment, and personnel required by cGMP regulations, the quality assurance (QA) and clinical effectiveness of PET drugs largely depend on the radiosynthesizers themselves. To bridge the gap between radiochemistry standardization and site qualification, the Japanese Society of Nuclear Medicine (JSNM) has issued guidance for the in-house manufacturing of small-scale PET drugs under academic GMP (a-GMP) environments. The goals of cGMP and a-GMP are different: cGMP focuses on process optimization, certification, and commercialization, while a-GMP facilitates the small-scale, in-house production of PET drugs for clinical trials and patient-specific standard of care. Among PET isotopes, N-13 has a short half-life (10 min) and must be synthesized on site. [¹³N]Ammonia ([¹³N]NH₃) is used for myocardial perfusion imaging under the Japan Health Insurance System (JHIS) and was thus selected as a working example for the manufacturing of PET drugs in an a-GMP environment. Methods: A [¹³N]NH₃-radiosynthesizer was installed in a hot cell within an a-GMP-compliant radiopharmacy unit. To comply with a-GMP regulations, the air flow was adjusted through HEPA filters. All cabinets and cells were disinfected to ensure sterility once a month. Standard operating procedures (SOPs) were applied, including analytical methods. Batch records, QA data, and radiation exposure to staff in the synthesis of [¹³N]NH₃ were measured and documented. Results: 2.52 GBq of [¹³N]NH₃ end-of-synthesis (EOS) was obtained in an average of 13.5 min in 15 production runs. The radiochemical purity was more than 99%. Exposure doses were 11 µSv for one production run and 22 µSv for two production runs. The pre-irradiation background dose rate was 0.12 µSv/h. After irradiation, the exposed dosage in the front of the hot cell was 0.15 µSv/h. The leakage dosage measured at the bench was 0.16 µSv/h. The exposure and leakage dosages in the manufacturing of [¹³N]NH₃ were similar to the background level as measured by radiation monitoring systems in an a-GMP environments. All QAs, environmental data, bacteria assays, and particulates met a-GMP compliance standards. Conclusions: In-house a-GMP environments require dedicated radiosynthesizers, documentation for batch records, validation schedules, radiation protection monitoring, air and particulate systems, and accountable personnel. In this study, the in-house manufacturing of [¹³N]NH₃ under a-GMP conditions was successfully demonstrated. These findings support the international harmonization of small-scale PET drug manufacturing in hospitals and clinics for future multi-center clinical trials and the development of a standard of care. Full article

The application of CAD/CAM technologies in modern production has revolutionized manufacturing processes, leading to significant improvements in precision, efficiency, and flexibility. These technologies enable the design and manufacturing of complex geometries with high accuracy, reducing errors and material waste. CAD/CAM integration streamlines workflows, enhances productivity, and facilitates rapid prototyping, accelerating the time-to-market for new products. Additionally, it supports customization and scalability in production, allowing for cost-effective small-batch and large-scale manufacturing. Without a 3D model of the product, it is not possible to use the advantages of applying advanced CAD/CAM technologies. Recognizing 3D models from engineering drawings is essential for modern production, especially for outsourcing companies in fluctuating market conditions, where the production process is organized with 2D workshop drawings on paper. This paper proposes a novel methodology for reconstructing 3D models from 2D engineering drawings, specifically those in DXF file format, leveraging a genetic algorithm. A core component of this approach is the representation of the 2D drawing as a symmetric adjacency matrix. This matrix serves as the foundational data structure for the genetic algorithm, enabling the evolutionary process to effectively optimize the 3D reconstruction. The experimental evaluation, conducted on multiple engineering drawing test cases (including both polyhedral and cylindrical geometries), demonstrated consistent convergence of the proposed GA-based method toward topologically valid and geometrically accurate 3D wireframe models. The approach achieved successful reconstruction in all cases, with fitness scores ranging from 1.1 to 112.2 depending on model complexity, and average execution times from 2 to 100 s. These results confirm the method’s robustness, scalability, and applicability in real-world CAD environments, while establishing a new direction for topology-driven 3D reconstruction using evolutionary computation. Full article

Photosynthetic bacteria exhibit significant bioremediation potential and resource recycling characteristics, rendering them valuable candidates for sustainable wastewater treatment applications. Sea cucumber boiling broth (SCBB) contains high concentrations of organic compounds and nutrient salts, whose indiscriminate discharge poses serious environmental risks. This study aimed to evaluate a photosynthetic bacterium capable of effectively treating SCBB, which was isolated from the intertidal sediment samples. The bacterial strain was identified using 16S rDNA sequencing, and optimal growth conditions, including light, pH, and temperature, were determined. Finally, a small-scale trial was conducted in a fed-batch fermenter. The results showed that 16S rDNA analysis placed this strain in the Chromatiaceae family, forming a distinct lineage from the closest related species Marichromatium purpuratum and M. gracile, and was tentatively named Marichromatium sp. DYYC01. The strain exhibited optimal growth under anaerobic conditions at 30 °C, light intensity of 100 μmol photons/m²/s, and pH 7.0. Batch culture experiments demonstrated maximum biomass accumulation (OD₆₆₀ = 0.831) in SCBB medium with an initial COD loading of 3913 mg L⁻¹, concomitant with significant nutrient removal efficiencies: 76.45% COD, 55.82% total nitrogen (TN), and 56.67% total phosphorus (TP). Scaling up to fed-batch fermentation enhanced bioremediation performance, achieving removal rates of 83.13% COD, 72.17% TN, and 73.07% TP with enhanced growth (OD₆₆₀ = 1.2). This study reveals Marichromatium sp. DYYC01’s exceptional halotolerance in high-salinity organic wastewater treatment. The strain’s capacity for simultaneous biomass production and efficient nutrient recovery from hypersaline processing effluent positions it as a promising candidate for developing circular bioeconomy strategies. Full article

To obtain the complete shape and pose of corn under occlusion, this study proposes a point cloud completion algorithm for completing the fragmented corn point cloud after segmentation. Considering that this work focuses on a single-class crop—corn—the proposals mainly focus on the deep learning model size and the completion of the overall shape of the corn. In this work, the 3D corn models derived from segmentation are employed to systematically output the fragmented point cloud data in batches. The Shape Coding PointAttN (SCPAN) algorithm is also proposed, which is based on PointAttN. The model’s structure is simplified to output sparse point clouds and minimize computational complexity, and a gated multilayer perceptron (MLP) containing 3D position coding is introduced to enhance the model’s spatial awareness. In addition, the prior shape encoder module is initially trained and subsequently integrated into the model to enhance its focus on shape characteristics. Compared to the original model, PointAttN, SCPAN achieves a 34.2% reduction in the number of parameters, and the inference time is reduced by 30 ms while maintaining comparable accuracy. The experimental results show that the proposed method can complete the corn point cloud more effectively, using a small model to help estimate the pose and dimensions of corn accurately. This work supports the precise phenotypic analysis of corn and similar crops, such as citrus and tomatoes, and promotes the development of smart agricultural technology. Full article

Adam (Adaptive Moment Estimation) is a well-known algorithm for the first-order gradient-based optimization of stochastic objective functions, based on adaptive estimates of lower-order moments. As shown by computational experiments, with an increase in the degree of conditionality of the problem and in the presence of interference, Adam is prone to looping, which is associated with difficulties in step adjusting. In this paper, an algorithm for step adaptation for the Adam method is proposed. The principle of the step adaptation scheme used in the paper is based on reproducing the state in which the descent direction and the new gradient are found during one-dimensional descent. In the case of exact one-dimensional descent, the angle between these directions is right. In case of inexact descent, if the angle between the descent direction and the new gradient is obtuse, then the step is large and should be reduced; if the angle is acute, then the step is small and should be increased. For the experimental analysis of the new algorithm, test functions of a certain degree of conditionality with interference on the gradient and learning problems with mini-batches for calculating the gradient were used. As the computational experiment showed, in stochastic optimization problems, the proposed Adam modification with step adaptation turned out to be significantly more efficient than both the standard Adam algorithm and the other methods with step adaptation that are studied in the work. Full article