Hasil untuk "Production capacity. Manufacturing capacity"

Wei Jie Ng, Mun Kou Lai, Ching Hao Lee et al.

Material extrusion (MEX) 3D-printed parts are primarily used for prototyping rather than functional components due to lower mechanical strength. To address this limitation and promote sustainability, current work explores the reinforcement of plant-based polylactic acid (PLA) with pineapple leaf fibre (PALF). An in situ approach was proposed to embed continuous PALF within the middle layer of a 3D-printed component during the MEX process. An experimental investigation was conducted to evaluate the impact performance of composites produced via this new fabrication method. To optimize the fibre–matrix interface, an alkaline treatment was applied to the natural fibre, enhancing interfacial adhesion. Neat PLA, along with two types of PALF-reinforced PLA composite, were printed with both single-strand and three-strand fibre configurations. Fracture surfaces were analyzed under a digital microscope and a scanning electron microscope (SEM) to correlate morphological characteristics with the impact strength. The results showed that the impact strength of the three-strand treated PALF-PLA composite (3 PALF-PLA) surpassed that of neat PLA by 2.71% due to reduced porosity. In contrast, the one-strand PALF-PLA composites exhibited lower performance compared to neat PLA due to the presence of the fibre gap caused by the mid-print pause. Treated fibres consistently outperformed untreated ones due to their rougher surface morphology resulting from alkaline treatment. The results demonstrate that the combination of alkaline treatment and continuous fibre reinforcement significantly enhances energy absorption of 3D-printed MEX parts and offers a sustainable pathway for 3D-printed MEX parts.

Ibrahim Mousad, Moustafa Gadalla, Wael Salaheddin Moughith et al.

In the competitive shoe manufacturing industry, efficient production scheduling often struggles to balance minimizing makespan (MS) and reducing the number of cells, leading to increased complexity. This study introduces a biobjective optimization model to minimize the MS and the number of cells in a cellular manufacturing (CM) system. The model includes three phases: (1) a capacity phase calculates the required number of cells to meet the production demand within a suitable production rate; (2) a scheduling phase calculates the start and finish times for each product within a machine; (3) an optimization phase uses genetic algorithm (GA) for generating a set of optimal solutions considering the workforce distribution and product sequencing. The performance of the developed model was evaluated using a case study discussed in the literature to demonstrate its capabilities and highlight its superiority over similar models. These results showed a significant reduction of 0.4% in MS while maintaining the same number of cells compared to other models. A sensitivity analysis is conducted by varying the crossover and mutation probabilities. The findings indicate that the model achieves the most stable and consistent performance for paired crossover–mutation settings with crossover probabilities between 70% and 90% (corresponding to mutation probabilities between 30% and 10%), resulting in the lowest average MS and the most balanced cell utilization. The capabilities of the developed model provide production managers with reliable insights for identifying the production requirements of the shoe manufacturing company. The algorithm was also benchmarked on 12 published instances using 30 independent runs per instance to assess solution quality and stability.

Donatella Cimini, Sergio D’ambrosio, Odile Francesca Restaino et al.

Abstract Melanins are pigments widely distributed in microbial, plant, and animal kingdoms. Their UV–visible light shielding capacity, metal chelation ability, antioxidant, and antimicrobial properties make these pigments suitable for different industrial applications like in cosmetic and bioremediation fields. The actual manufacturing process relies on the extraction from animal tissues like the ink of Sepia officinalis and/or on synthetic chemical procedures. Streptomycetes might be the ideal candidates for the development of biotechnological processes of melanin production due to their ability to produce pigments as secondary metabolites, extracellularly released. Here, a new strain of Streptomyces nigra, capable of efficiently producing eumelanin, was isolated from soil samples in Messina, Sicily, Italy, and characterized first by 16S rRNA analysis and then by whole genome sequencing, with a complete gene clusters analysis. The strain ability of growing and producing melanin was tested on four media, including newly formulated ones, and by also optimizing temperature and pH conditions of growth, a melanin production of 2.45 ± 0.01 g/L was reached. The pigment, once produced under the optimal conditions, was purified and characterized by UV–visible, FT-IR, NMR, and EPR spectroscopy, revealing an eumelanin-like structure. Key points • A new Streptomyces nigra strain, MT6, was isolated and identified • A new formulated medium boosted melanin production up to 2.45 g/L • The extracellular pigment was characterized as eumelanin

Domingos F. Oliveira, Domingos F. Oliveira, Miguel A. Brito et al.

IntroductionAccurate prediction of steps required to address machine faults is critical for minimizing downtime and enhancing production efficiency in modern manufacturing. This study utilizes machine failure data and Failure Mode and Effects Analysis to demonstrate how machine learning supports maintenance teams in selecting optimal repair methods.MethodsThe research adopts the Design Science Research paradigm, which emphasizes the creation of artifacts to address practical challenges. For the practical component, quality assurance and control frameworks in data science projects were implemented by integrating two widely used methodologies: CRISP-DM and PDCA, to ensure rigorous quality assurance and control in data science initiatives.ResultsRepair actions serve as the target variables, while the input comprises ten multivariate time-series machine parameters. The prediction task is formulated as a classification problem. Two modeling approaches are evaluated. The first approach merges multiple time series into a single sequence, facilitating the application of Multi-Layer Perceptron, Convolutional Neural Networks, and Fully Convolutional Networks. The second approach preserves the time series as three-dimensional arrays, enabling advanced applications of MLP, CNN, Multi-Head CNN, and FCN models.DiscussionThe models are assessed based on their capacity to predict repair actions, with particular emphasis on the impact of time-series processing and model architecture on classification accuracy. The findings highlight effective strategies for predicting machine repairs and advancing prescriptive maintenance in manufacturing environments.

Rumi Sawada, Shinji Kusakawa, Mika Kusuhara et al.

Introduction: The Quality by Design (QbD) approach for developing cell therapy products using mesenchymal stromal/stem cells (MSCs) is a promising method for designing manufacturing processes to improve the quality of MSC products. It is crucial to ensure the reproducibility and robustness of the test system for evaluating critical quality attributes (CQAs) in the QbD approach for manufacturing of pharmaceutical products. In this study, we explored the key factors involved in establishing a robust evaluation system for the immunosuppressive effect of MSCs, which can be an example of a CQA in developing and manufacturing therapeutic MSCs for treating graft-versus-host disease, etc, and we have identified method attributes to increase the robustness of a simple in vitro assay to assess the immunosuppressive effects of MSCs. Methods: We evaluated the performance of an assay system to examine the proliferation of peripheral blood mononuclear cells (PBMCs) activated with the mitogen phytohemagglutinin (PHA) when co-cultured with MSCs, the so-called one-way mixed lymphocyte reaction (MLR) assay. The MLR assay was performed on the same MSCs using 10 PBMC lots from different donors. In addition, 13 cytokine production levels in PHA-stimulated PBMCs were assessed. Results: The PHA-stimulated proliferation response of PBMCs, the action of MSCs in the MLR test, and the cytokine release of the respective PBMCs significantly differed among the PBMC lots (p < 0.05). A correlation analysis between the amounts of cytokines released by PBMCs and the immunosuppressive potency of MSCs showed that IFNγ, TNFα, CXCL10, PD-L1, HGF, and CCL5 production in PBMCs was significantly correlated with the MSC-mediated inhibition of PBMC proliferation (p < 0.05). Therefore, we selected two PBMC lots with high PBMC proliferation and PHA-stimulated cytokine (such as IFNγ and TNFα) release for the subsequent one-way MLR assay. The robustness of the established test system was confirmed by repeating the assay several times on different days for the same MSCs (coefficient of variation <0.2). Conclusions: To make robust the MSC immunosuppressive potency assay system, controlling the quality of PBMCs used for the assay is essential. Evaluating the inflammatory cytokine production capacity of PBMCs is effective in assessing the quality of the MLR assay system.

Peter Pak, Achuth Chandrasekhar, Amir Barati Farimani

Agentic systems enable the intelligent use of research tooling, augmenting a researcher's ability to investigate and propose novel solutions to existing problems. Within Additive Manufacturing (AM), alloy selection and evaluation remains a complex challenge, often requiring expertise in the various domains of materials science, thermodynamic simulations, and experimental analysis. Large Language Model (LLM) enabled agents can facilitate this endeavor by utilizing their extensive knowledge base to dispatch tool calls via Model Context Protocol (MCP) to perform actions such as thermophysical property diagram calculations and lack of fusion process map generation. In addition, the multi-agent system can effectively reason through complex user prompts and provide analysis on the lack of fusion process window of common alloys such as SS316L and IN718 along with proposed composition variants of known alloys. These agents can dynamically adjust their task trajectory to the outcomes of tool call results, effectively enabling autonomous decision-making in practical environments. This work aims to showcase the benefits of adopting a LLM enabled multi-agent system to automate and accelerate the task of evaluating proposed additive manufacturing alloys, both novel and known.

András Horváth, Dénes Kollár

This paper investigates the residual stresses induced by metal inert/active gas (MIG/MAG) welding in normal strength steel box sections, focusing on the validity of uniaxial residual stress assumption. Advanced manufacturing simulations are conducted using deterministic uncoupled transient thermomechanical analysis with a double-ellipsoidal heat source model, employing 8-node solid elements and material models calibrated for extreme temperatures per EN 1993-1-2. A comprehensive parametric analysis investigates the effects of primary welding variables, such as heat source power and welding speed, alongside geometric parameters of the heat source model using random Latin hypercube sampling technique in the analyzed parameter set. The relationship between the size and shape of the characteristic isotherms, i.e., the aspect ratio and the Rosenthal number, underscores that the analyzed welding heat sources are in the fast regime with the validity of uniaxial residual stresses based on the analytical assumption (minimal values are AR = 9.94 and Ro = 30.47). The validity and limitations of uniaxial residual stress assumptions for 59 welded and 51 heated box sections are critically evaluated by using the finite element model-based stress triaxiality parameter. Results confirm that longitudinal residual stresses dominate typical MIG/MAG welding applications, supporting the application of uniaxial residual stress models in advanced structural design by neglecting in-plane and through-thickness residual stresses. Conversely, three-dimensional residual stress state dominates under conditions such as preheating or thermal straightening.

MATTHEW ONYEMA AGU, Umar Kyakyaidi, Collins Chibuzor Odidika et al.

The properties of high-quality soap is determined by its cleansing, foam formation, surface tension, emulsion capacity, pH, etc. To produce a generally better product, soap manufacturing companies take these properties into account during production which necessitated the physical assessment properties of some selected made-in-Nigeria synthetic surfactant for surface activity. The samples were evaluated based on tests for saponins, foam formation and stability, emulsion formation/capacity, surface tension, and pH at different concentrations. The percentage of each of the synthetic surfactants (Ariel=DA, Bonus=DB, Good mama=DG, Sunlight=DS and Zip=DZ) concerning foam formation at five minutes and stability after 30 minutes were as follows; DZ(6.7) > DS(6.3) > DG(5.3) > DA(4.2) > DB(4.1) and DZ(2.6) > DG(1.9) > DA(1.1)/DS(1.1) > DB(1.0) in cm, respectively. The study of the emulsion formation/capacity of the different samples at different concentrations in terms of creamy behaviour was observed as DS > DZ > DA > DG > DB. All the samples studied showed a good reduction in surface tension. At 2g and 10g of the different samples, the following results were obtained: DS(2.87), DZ(20.25), DB(21.55), DA(21.65), DG(21.85) Nm-1 and DS(22.45), DZ(22.25), DB(22.45), DA(26.65), DG(22.65) N/m, respectively. The result of the pH values of the different samples varied within the range of 10.80 to 10.91 as moderately weak base with an increase in concentration. Detergents studied conform to the minimum quality examinations but regular assessment check should be carried out by enforcement/regulatory agencies to ensure quality of detergents produced in the country meets the minimum required standard.

Dongyue Qu, Wenchao Liang, Yuting Zhang et al.

Machining quality prediction is the critical link of quality control in parts machining. With the advent of the Industry 4.0 era, intelligent manufacturing and data-driven technologies bring new ideas for quality control in complex machining processes. Quality control is complicated for multi-process, multi-condition, small-batch, and high-precision parts processing requirements. To solve this problem, this paper proposes a machining quality prediction method based on the machining error transfer network and the grey neural network. Initially, by constructing a processing error transfer network, the error transfer law in part processing is described, and the PageRank algorithm and the influence degree of the nodes are used to determine the critical quality features. Additionally, the problem of low prediction accuracy due to small sample data and multiple coupling relationships is solved using the grey neural network algorithm, and a high accuracy prediction of critical quality features is achieved. Finally, the effectiveness and reliability of the method are verified by the case of medium-speed marine diesel engine fuselage processing. The results indicate that this method not only effectively identifies critical quality features in the machining process of complex parts, but it also maintains a high predictive accuracy for these features, even with small samples and limited data.

Giacomo Baruffa, Andrea Pieressa, Marco Sorgato et al.

Optimizing process parameters to minimize defects remains an important challenge in injection molding (IM). Machine learning (ML) techniques offer promise in this regard, but their application often requires extensive datasets. Transfer learning (TL) emerges as a solution to this problem, leveraging knowledge from related tasks to enhance model training and performance. This study explores TL’s viability in predicting weld line visibility in injection-molded components using artificial neural networks (ANNs). TL techniques are employed to transfer knowledge between datasets related to different components. Furthermore, both source datasets obtained from simulations and experimental tests are used during the study. In order to use process simulations to obtain data regarding the presence of surface defects, it was necessary to correlate an output variable of the simulations with the experimental observations. The results demonstrate TL’s efficacy in reducing the data required for training predictive models, with simulations proving to be a cost-effective alternative to experimental data. TL from simulations achieves comparable predictive metric values to those of the non-pre-trained network, but with an 83% reduction in the required data for the target dataset. Overall, transfer learning shows promise in streamlining injection molding optimization and reducing manufacturing costs.

Boqun Zhao, Chongjun Ouyang, Xingqi Zhang et al.

The channel capacity of near-field (NF) communications is characterized by considering three types of line-of-sight multiuser channels: i) multiple access channel (MAC), ii) broadcast channel (BC), and iii) multicast channel (MC). For NF MAC and BC, closed-form expressions are derived for the sum-rate capacity as well as the capacity region under a two-user scenario. These results are further extended to scenarios with an arbitrary number of users. For NF MC, closed-form expressions are derived for the two-user channel capacity and the capacity upper bound with more users. Further insights are gleaned by exploring special cases, including scenarios with infinitely large array apertures, co-directional users, and linear arrays. For comparison, the MAC and BC sum-rates achieved by typical linear combiners and precoders are also analyzed. Theoretical and numerical results are presented and compared with far-field communications to demonstrate that: i) the NF capacity of these three channels converges to finite values rather than growing unboundedly as the number of array elements increases; ii) the capacity of the MAC and BC with co-directional users can be improved by using the additional range dimensions in NF channels to reduce inter-user interference (IUI); and iii) the MC capacity benefits less from the NF effect compared to the MAC and BC, as multicasting is less sensitive to IUI.

Mostafa Ebrahimpour, Mahmoud Moradi, Aida Fallahpour

Introduction: In recent decades, with the scarcity of world resources and the change of customer perspective beyond products and services, manufacturing and service companies must think beyond the economic benefits of their products and services and make every effort to maintain these resources and improve the country and the world. Considering the sustainability performance and its improvement, it provides the opportunity for firms to consider the economic sector, the environmental sector of the firm and the social sector, because the success of today's organizations mixed with the concept of sustainability and provides an opportunity for small and medium-sized companies to move towards sustainable economic development. Organizations need to leverage their creativity and innovation in order to maintain their sustainablility performance in today's complex and dynamic environment. The supply chain dynamics may be able to help. Another dynamic concept, supply chain resilience, introduced as a capacity for survival, adaptation and growth in turbulent times. In addition, dynamic capabilities are organizational processes that intentionally modify and change the organization's resource base and known as an emerging strategy to achieve the sustainable goals of an organization. The present study aimed to investigate the effect of supply chain dynamics on firm sustainability performance with mediating role of remanufacturing capability and supply chain resilience.  Methodology: The statistical population of the study is small and medium-sized companies active in Sepidrood industrial Town. Data collected from 48 companies through a questionnaire. In each company, one manager (selected by available sampling method) was distributed a questionnaire. Cronbach's alpha coefficient and Composite reliability was also used to measure reliability. Given that for all variables, this value is above 0.7 and the value of CR is greater than 0.7. For validity, content validity (using experts' opinions) and construct validity with convergent validity (Average Variance Extracted) and Discriminant validity (The Fornell-Larcker Criterion) have been done. For all variables, the value of AVE is greater than 0.5 . The first order latent variable of this article, it was calculated by the software and its output is presented. But regarding the sustainability firm performance, due to its second order latent variable, it was manually calculated and reported. Data analysis in this study was done by structural equation modeling( SEM) with partial least squares approach using Smart PLS .  Results and Discussion: The purpose of study was to investigate the effect of supply chain dynamics on firm sustainability performance with mediating role of remanufacturing capability and supply chain resilience. Therefore, by using a questionnaire, the necessary information collected from the Sepidrood industrial Town of Rasht, and Smart PLS used to analyze the data. The results show that supply chain dynamics affect remanufacturing capability and supply chain resilience directly. In addition, the remanufacturing capability has a positive effect on sustainability Firm performance and supply chain resilience. Supply chain resilience has a positive effect on the sustainability Firm performance directly and indirectly through the remanufacturing capability and supply chain dynamics. This paper did not confirm the role of supply chain dynamics and sustainability Firm performance.  Conclusion: It is concluded that small and medium-sized companies in order to take advantage of sustainable performance opportunities, should consider the drivers of supply chain dynamics and provide the basis for the development of firm sustainable performance by using supply chain dynamics indirectly through remanufacturing capability and supply chain resilience. With strategies such as creating an opportunity by managers to study the environment and learn from it to increase the speed of adapting to the environment and responding to its changes, strengthening the role of innovation in the organization's products and services and operational processes, passing laws regarding the collection of obsolete products, strengthening this insight. that recycled products do not mean second-hand products, the use of reverse logistics to re-enter parts into the production flow can reduce the cost of product production for both the company and the customer to some extent, improve the risk management culture in the organization and etc, can improve sustainability performance according to the results of this study. This article provides relevant and useful information related to the sustainability performance of small and medium-sized companies and the factors affecting it. Therefore, in order to improve their sustainability performance, small and medium-sized companies can take advantage of the effect of the variables of the implemented model to achieve sustainable performance.

Abid Shah, Rezo Aliyev, Henning Zeidler et al.

Wire arc additive manufacturing (WAAM) is an emerging and promising technology for producing medium-to-large-scale metallic components/structures for different industries, i.e., aerospace, automotive, shipbuilding, etc. It is now a feasible alternative to traditional manufacturing processes due to its shorter lead time, low material waste, and cost-effectiveness. WAAM has been widely used to produce components using different materials, including copper-based alloy wires, in the past decades. This review paper highlights the critical aspects of WAAM process in terms of technology, various challenges faced during WAAM process, different in-process and post-process operations, process monitoring methods, various gases, and different types of materials used in WAAM process. Furthermore, it briefly overviews recent developments in depositing different copper-based alloys via WAAM process.

Gesine Meyer‐Rath, Lise Jamieson, Yogan Pillay

Abstract Introduction The proven effectiveness of injectable cabotegravir (CAB‐LA) is higher than that of any other HIV prevention intervention ever trialled or implemented, surpassing medical male circumcision, condoms and combination antiretroviral treatment. Based on our own analyses and experience with the South African oral pre‐exposure prophylaxis (PrEP) programme, we review the supply and demand side factors that would need to be in place for a successful rollout of CAB‐LA, and delineate lessons for the launch of other long‐acting and extended delivery (LAED) antiretroviral drugs. Discussion On the supply side, CAB‐LA will have to be offered at a price that makes the drug affordable and cost‐effective to low‐ and middle‐income countries, especially those with high HIV prevalence. An important factor in lowering prices is a guaranteed market volume, which in turn necessitates the involvement of large funders, such as PEPFAR and the Global Fund, and a fairly rapid scale‐up of the drug. Such a scale‐up would have to involve speedy regulatory approval and WHO pre‐qualification, swift integration of CAB‐LA into national guidelines and planning for large enough manufacturing capacity, including the enabling of local manufacture. On the demand side, existing demand for HIV prevention products has to be harnessed and additional demand created, which will be aided by designing CAB‐LA programmes at the primary healthcare or community level, and involving non‐traditional outlets, such as private pharmacies and doctors’ practices. Conclusions CAB‐LA could be the game changer for HIV prevention that we have been hoping for, and serve as a useful pilot for other LAEDs. A successful rollout would involve building markets of a guaranteed size; lowering the drug's price to a level possibly below the cost of production, while also lowering the cost of production altogether; harnessing, creating and sustaining demand for the product over the long term, wherever possible, in national programmes rather than single demonstration sites; and establishing and maintaining manufacturing capacity and supply chains. For this, all parties have to work together—including originator and generic manufacturers, donor organizations and other large funders, and the governments of low‐ and middle‐income countries, in particular those with high HIV prevalence.

Robert Setter, Jan Hafenecker, Richard Rothfelder et al.

Multi-material additive manufacturing (AM) attempts to utilize the full benefits of complex part production with a comprehensive and complementary material spectrum. In this context, this research article presents new processing strategies in the field of polymer- and metal-based multi-material AM. The investigation highlights the current progress in powder-based multi-material AM based on three successfully utilized technological approaches: additive and formative manufacturing of hybrid metal parts with locally adapted and tailored properties, material-efficient AM of multi-material polymer parts through electrophotography, and the implementation of UV-curable thermosets within the laser-based powder bed fusion of plastics. Owing to the complex requirements for the mechanical testing of multi-material parts with an emphasis on the transition area, this research targets an experimental shear testing set-up as a universal method for both metal- and polymer-based processes. The method was selected based on the common need of all technologies for the sufficient characterization of the bonding behavior between the individual materials.

Seongyoon Kim, Hangsoon Jung, Minho Lee et al.

Early degradation prediction of lithium-ion batteries is crucial for ensuring safety and preventing unexpected failure in manufacturing and diagnostic processes. Long-term capacity trajectory predictions can fail due to cumulative errors and noise. To address this issue, this study proposes a data-centric method that uses early single-cycle data to predict the capacity degradation trajectory of lithium-ion cells. The method involves predicting a few knots at specific retention levels using a deep learning-based model and interpolating them to reconstruct the trajectory. Two approaches are used to identify the retention levels of two to four knots: uniformly dividing the retention up to the end of life and finding optimal locations using Bayesian optimization. The proposed model is validated with experimental data from 169 cells using five-fold cross-validation. The results show that mean absolute percentage errors in trajectory prediction are less than 1.60% for all cases of knots. By predicting only the cycle numbers of at least two knots based on early single-cycle charge and discharge data, the model can directly estimate the overall capacity degradation trajectory. Further experiments suggest using three-cycle input data to achieve robust and efficient predictions, even in the presence of noise. The method is then applied to predict various shapes of capacity degradation patterns using additional experimental data from 82 cells. The study demonstrates that collecting only the cycle information of a few knots during model training and a few early cycle data points for predictions is sufficient for predicting capacity degradation. This can help establish appropriate warranties or replacement cycles in battery manufacturing and diagnosis processes.

Jiehua Chen, Gergely Csáji

We consider many-to-one matching problems, where one side consists of students and the other side of schools with capacity constraints. We study how to optimally increase the capacities of the schools so as to obtain a stable and perfect matching (i.e., every student is matched) or a matching that is stable and Pareto-efficient for the students. We consider two common optimality criteria, one aiming to minimize the sum of capacity increases of all schools (abbrv. as MinSum) and the other aiming to minimize the maximum capacity increase of any school (abbrv. as MinMax). We obtain a complete picture in terms of computational complexity: Except for stable and perfect matchings using the MinMax criteria which is polynomial-time solvable, all three remaining problems are NP-hard. We further investigate the parameterized complexity and approximability and find that achieving stable and Pareto-efficient matchings via minimal capacity increases is much harder than achieving stable and perfect matchings.

Kelin Chen, Adrian J. Carter, Yannis P. Korkolis

Flange wrinkling is often seen in deep-drawing process when the applied blankholding force is too small. This paper investigates the plastic wrinkling of flange under a constant blankholding force. A series of deep-drawing experiments of AA1100-O blanks are conducted with different blankholding forces. The critical cup height and wrinkling wave numbers for each case is established. A reduced-order model of flange wrinkling is developed using the energy method, which is implemented to predict the flange wrinkling of AA1100-O sheet by incrementally updating the flange geometry and material hardening parameters during the drawing process. A deep-drawing finite element model is developed in ABAQUS/standard to simulate the flange wrinkling of AA1100-O blanks under constant blankholding force. The predicted cup height and wave numbers from the finite element model and reduced-order model are compared with the experimental results, which demonstrates the accuracy of the reduced-order model, and its potential application in fast prediction of wrinkling in deep-drawing process.

Javier Canto, Lizaveta Ihnatsyeva, Juha Lehrbäck et al.

We examine the relations between different capacities in the setting of a metric measure space. First, we prove a comparability result for the Riesz $(β,p)$-capacity and the relative Hajlasz $(β,p)$-capacity, for $1<p<\infty$ and $0<β\le 1$, under a suitable kernel estimate related to the Riesz potential. Then we show that in geodesic spaces the corresponding capacity density conditions are equivalent even without assuming the kernel estimate. In the last part of the paper, we compare the relative Hajlasz $(1,p)$-capacity to the relative variational $p$-capacity.

Xin Guo, Zheng-Chu Guo, Lei Shi

This article provides convergence analysis of online stochastic gradient descent algorithms for functional linear models. Adopting the characterizations of the slope function regularity, the kernel space capacity, and the capacity of the sampling process covariance operator, significant improvement on the convergence rates is achieved. Both prediction problems and estimation problems are studied, where we show that capacity assumption can alleviate the saturation of the convergence rate as the regularity of the target function increases. We show that with properly selected kernel, capacity assumptions can fully compensate for the regularity assumptions for prediction problems (but not for estimation problems). This demonstrates the significant difference between the prediction problems and the estimation problems in functional data analysis.