Over more than a century following the discovery of superconductors, extensive research has been conducted to leverage their properties, particularly in fundamental materials science applications. Nonetheless, significant challenges still exist that hinder the comprehensive production of superconductors, including the optimal critical temperature (Tc) and current density (Jc), concerns regarding the purity of superconductor phases, sintering temperature, complexities in crystal growth, and various high-cost fabrication-related issues. Recent significant breakthroughs in artificial intelligence (AI) approaches have provided disruptive solutions, such as machine learning (ML), to address these fundamental issues. Thus, ML approaches can be employed to address the issues associated with superconductivity and serve as a means to achieve optimal conditions for superconductors and their applications. ML methodologies can deliver rapid, efficient, and precise solutions for intricate and nonlinear technological, manufacturing, and economic challenges in the domain of superconductivity. This paper initially presents the notion of AI and the often employed ML techniques. A comprehensive conceptual overview is provided for studies employing ML methods aimed at properties enhancement, condition monitoring, and the structural analysis of existing superconductors, along with other pertinent applications. This subject overview is organized into three primary topics: fundamental application utilizing ML, databases used by ML models, and our main focus which is the optimization techniques used alongside with ML in superconductors. Furthermore, the difficulties associated with using ML methodologies in superconductivity and their applications are presented. Ultimately, prospective developments regarding the integration of ML approaches with superconducting for various applications are examined.
A stent maintains normal blood flow by expanding a stenotic artery from the inside. However, current stents are mechanically sub-optimized and can exert excessive forces on the vascular wall, leading to inflammation, late thrombosis, and in-stent restenosis. Optimizing the mechanical performance of stents requires not only reproducing the mechanical field within the stented vessel but also evaluating endothelialization, which serves as a key biological indicator of vascular neointimal formation. To this end, this study aimed to develop a three-dimensional in vitro stent endothelialization model that enables quantitative evaluation of endothelial responses under physiologically relevant mechanical conditions and to provide detailed fabrication protocols for its construction. Polydimethylsiloxane (PDMS) was used to mimic the adventitial structure of arteries. Human carotid artery endothelial cells (HCtAECs) were then seeded on the luminal surface and cultured for 24 h to form a confluent monolayer (intima). The constructed model was installed in the flow-exposure culturing system, and hemodynamic stimuli (two types of shear stress (SS); 0.5 Pa and 2.3 Pa) were applied to the HCtAECs inside to reproduce the physiological state of blood vessels. A self-expanding stent was then placed in the model during perfusion culture to evaluate in-stent endothelialization under controlled flow conditions. We examined the performance of the developed model based on quantitative evaluations of endothelial morphology in response to SS and in-stent endothelialization. Exposure to SS for 24 and 48 h caused endothelial orientation and elongation in the direction of flow, confirming the physiological responses of blood vessels. Furthermore, spatial and temporal analyses of in-stent endothelialization confirmed that the model can reproduce key biological processes associated with vascular neointimal formation in the presence of mechanical stimulation. The present model successfully integrates the mechanical and biological aspects of stent–vessel interaction, providing a reproducible platform for evaluating in-stent endothelialization under physiologically relevant conditions. This system can serve as a powerful tool not only for the quantitative assessment of endothelial dynamics but also for guiding the optimization of mechanical forces in stented blood vessels. Consequently, it offers a foundation for designing next-generation stents that promote rapid endothelialization and reduce the risk of restenosis and thrombosis.
The combination of aluminum alloys and stainless steels is in high demand across a wide range of applications, as it enables optimization of material properties and cost. When these dissimilar metals are joined by fusion, brittle intermetallic compounds (IMCs) form at the interface, severely degrading joint strength. Yet, identifying process parameters that simultaneously achieve optimal IMC thickness, a high bonding ratio, and a flat cladding surface remains highly challenging due to complex parameter interactions such as heat input, torch path, and wire feed speed. This study establishes the process conditions required for high-quality cladding, i.e., satisfying the three objectives mentioned above. To this end, an A5356 aluminum alloy layer was deposited onto a SUS304 stainless steel substrate while controlling heat input through weaving amplitude, wavelength, and frequency. The deposition behavior was captured with a high-speed camera, and thermal analysis was employed to predict favorable conditions. The results demonstrate that deposition with reduced heat input—when accompanied by a sharp contact angle of the molten aluminum alloy and a bead width exceeding the weaving wavelength λ—consistently yields high-quality cladding. Moreover, thermal analysis revealed the critical thermal window: the substrate surface directly beneath the torch must remain above the liquidus temperature of A5356 ( 640 °C) but below the solidus of SUS304 ( 1400 °C). These findings provide clear design criteria, offering a robust basis for optimizing dissimilar-metal cladding.
The design of closed-die forging processes requires determining process parameters such as the number of forming stages and die-surface geometry at each stage while satisfying evaluation indices such as forging load and shape accuracy. When single-stage forging is not feasible, many combinations of the number of stages and intermediate die-surface geometries must be explored, which are referred to as the “process layout” in this paper. This leads to a time-consuming and iterative trial-and-error process. In a previous study, an automatic design system for process layouts was proposed for disk-shaped products. The functional surface connection method was also introduced to ensure high flexibility in creating die-surface geometry. Subsequently, this method was integrated with the finite element method (FEM) and an optimization technique to establish a framework for process design. However, a major limitation was the high computational cost due to the large number of FEM simulations. In this study, a novel automatic design system was developed, the acceleration of which is achieved with machine learning (ML) models trained in advance instead of evaluations using FEM simulations. ML models were trained using datasets from process layouts generated for a variety of target shapes. By integrating the pre-trained ML models with an optimization algorithm, appropriate process layouts that satisfy the target requirements were generated. Results indicate that the ML-accelerated automatic design system can significantly reduce the computational cost of FEM simulations after the target shape is obtained. This reduction enables more efficient generation of process layouts.
The petunia genome contains an endogenous pararetrovirus, petunia vein clearing virus (PVCV). Previous analyses indicate that PVCV has suppressor activity against RNA silencing, but the suppressor protein has not been identified. Here we tested whether an open reading frame (ORF) of PVCV confers the activity that can suppress cosuppression of the CHS-A genes encoding chalcone synthase, which has a high rate of RNA turnover in the petal tissues of petunia. Petunia transformants that express PVCV ORF under the control of cauliflower mosaic virus 35S promoter were produced. The transgenic plants were crossed with those that have CHS-A cosuppression to produce plants that contain both the PVCV ORF transgene and CHS-A transgene. The coexistence of these transgenes resulted in phenotypic changes: pigmentation of various extents occurred on the originally white petals of CHS-A cosuppression phenotype. The generation of pigmented portions in flower petals coincided with higher transcript levels of CHS-A and PVCV ORF and less CHS-A short interfering RNA. These results indicate that the PVCV ORF can suppress CHS-A cosuppression and change the flower color phenotype when it is expressed as a transgene.
