Receive KAIST news by email!
Type your e-mail address here.
by recently order
by view order
Deep Learning Framework to Enable Material Design in Unseen Domain
Researchers propose a deep neural network-based forward design space exploration using active transfer learning and data augmentation A new study proposed a deep neural network-based forward design approach that enables an efficient search for superior materials far beyond the domain of the initial training set. This approach compensates for the weak predictive power of neural networks on an unseen domain through gradual updates of the neural network with active transfer learning and data augmentation methods. Professor Seungwha Ryu believes that this study will help address a variety of optimization problems that have an astronomical number of possible design configurations. For the grid composite optimization problem, the proposed framework was able to provide excellent designs close to the global optima, even with the addition of a very small dataset corresponding to less than 0.5% of the initial training data-set size. This study was reported in npj Computational Materials last month. “We wanted to mitigate the limitation of the neural network, weak predictive power beyond the training set domain for the material or structure design,” said Professor Ryu from the Department of Mechanical Engineering. Neural network-based generative models have been actively investigated as an inverse design method for finding novel materials in a vast design space. However, the applicability of conventional generative models is limited because they cannot access data outside the range of training sets. Advanced generative models that were devised to overcome this limitation also suffer from weak predictive power for the unseen domain. Professor Ryu’s team, in collaboration with researchers from Professor Grace Gu’s group at UC Berkeley, devised a design method that simultaneously expands the domain using the strong predictive power of a deep neural network and searches for the optimal design by repetitively performing three key steps. First, it searches for few candidates with improved properties located close to the training set via genetic algorithms, by mixing superior designs within the training set. Then, it checks to see if the candidates really have improved properties, and expands the training set by duplicating the validated designs via a data augmentation method. Finally, they can expand the reliable prediction domain by updating the neural network with the new superior designs via transfer learning. Because the expansion proceeds along relatively narrow but correct routes toward the optimal design (depicted in the schematic of Fig. 1), the framework enables an efficient search. As a data-hungry method, a deep neural network model tends to have reliable predictive power only within and near the domain of the training set. When the optimal configuration of materials and structures lies far beyond the initial training set, which frequently is the case, neural network-based design methods suffer from weak predictive power and become inefficient. Researchers expect that the framework will be applicable for a wide range of optimization problems in other science and engineering disciplines with astronomically large design space, because it provides an efficient way of gradually expanding the reliable prediction domain toward the target design while avoiding the risk of being stuck in local minima. Especially, being a less-data-hungry method, design problems in which data generation is time-consuming and expensive will benefit most from this new framework. The research team is currently applying the optimization framework for the design task of metamaterial structures, segmented thermoelectric generators, and optimal sensor distributions. “From these sets of on-going studies, we expect to better recognize the pros and cons, and the potential of the suggested algorithm. Ultimately, we want to devise more efficient machine learning-based design approaches,” explained Professor Ryu.This study was funded by the National Research Foundation of Korea and the KAIST Global Singularity Research Project. -Publication Yongtae Kim, Youngsoo, Charles Yang, Kundo Park, Grace X. Gu, and Seunghwa Ryu, “Deep learning framework for material design space exploration using active transfer learning and data augmentation,” npj Computational Materials (https://doi.org/10.1038/s41524-021-00609-2) -Profile Professor Seunghwa Ryu Mechanics & Materials Modeling Lab Department of Mechanical Engineering KAIST
From Dark to Light in a Flash: Smart Film Lets Windows Switch Autonomously
Researchers have developed a new easy-to-use smart optical film technology that allows smart window devices to autonomously switch between transparent and opaque states in response to the surrounding light conditions. The proposed 3D hybrid nanocomposite film with a highly periodic network structure has empirically demonstrated its high speed and performance, enabling the smart window to quantify and self-regulate its high-contrast optical transmittance. As a proof of concept, a mobile-app-enabled smart window device for Internet of Things (IoT) applications has been realized using the proposed smart optical film with successful expansion to the 3-by-3-inch scale. This energy-efficient and cost-effective technology holds great promise for future use in various applications that require active optical transmission modulation. Flexible optical transmission modulation technologies for smart applications including privacy-protection windows, zero-energy buildings, and beam projection screens have been in the spotlight in recent years. Conventional technologies that used external stimuli such as electricity, heat, or light to modulate optical transmission had only limited applications due to their slow response speeds, unnecessary color switching, and low durability, stability, and safety. The optical transmission modulation contrast achieved by controlling the light scattering interfaces on non-periodic 2D surface structures that often have low optical density such as cracks, wrinkles, and pillars is also generally low. In addition, since the light scattering interfaces are exposed and not subject to any passivation, they can be vulnerable to external damage and may lose optical transmission modulation functions. Furthermore, in-plane scattering interfaces that randomly exist on the surface make large-area modulation with uniformity difficult. Inspired by these limitations, a KAIST research team led by Professor Seokwoo Jeon from the Department of Materials Science and Engineering and Professor Jung-Wuk Hong of the Civil and Environmental Engineering Department used proximity-field nanopatterning (PnP) technology that effectively produces highly periodic 3D hybrid nanostructures, and an atomic layer deposition (ALD) technique that allows the precise control of oxide deposition and the high-quality fabrication of semiconductor devices. The team then successfully produced a large-scale smart optical film with a size of 3 by 3 inches in which ultrathin alumina nanoshells are inserted between the elastomers in a periodic 3D nanonetwork. This “mechano-responsive” 3D hybrid nanocomposite film with a highly periodic network structure is the largest smart optical transmission modulation film that exists. The film has been shown to have state-of-the-art optical transmission modulation of up to 74% at visible wavelengths from 90% initial transmission to 16% in the scattering state under strain. Its durability and stability were proved by more than 10,000 tests of harsh mechanical deformation including stretching, releasing, bending, and being placed under high temperatures of up to 70°C. When this film was used, the transmittance of the smart window device was adjusted promptly and automatically within one second in response to the surrounding light conditions. Through these experiments, the underlying physics of optical scattering phenomena occurring in the heterogeneous interfaces were identified. Their findings were reported in the online edition of Advanced Science on April 26. KAIST Professor Jong-Hwa Shin’s group and Professor Young-Seok Shim at Silla University also collaborated on this project. Donghwi Cho, a PhD candidate in materials science and engineering at KAIST and co-lead author of the study, said, “Our smart optical film technology can better control high-contrast optical transmittance by relatively simple operating principles and with low energy consumption and costs.” “When this technology is applied by simply attaching the film to a conventional smart window glass surface without replacing the existing window system, fast switching and uniform tinting are possible while also securing durability, stability, and safety. In addition, its wide range of applications for stretchable or rollable devices such as wall-type displays for a beam projection screen will also fulfill aesthetic needs,” he added. This work was supported by the National Research Foundation of Korea (NRF), and the Korean Ministries of Science, ICT and Future Planning (MSIP), and Science and ICT (MSIT). Publication: Cho, D, et al. (2020) ‘High-Contrast Optical Modulation from Strain-Indicated Nanogaps at 3D Heterogeneous Interfaces’ Advanced Science, 1903708. Available online at https://doi.org/10.1002/advs.201903708 Profile: Seokwoo Jeon, PhD Professor email@example.com https://fdml.kaist.ac.kr/ Flexible Device and Metamaterials Lab (FDML) Department of Materials Science and Engineering (MSE) Korea Advanced Institute of Science and Technology (KAIST) https://www.kaist.ac.krDaejeon 34141, Korea Profile: Jung-Wuk Hong, PhD Associate Professor firstname.lastname@example.org http://aaml.kaist.ac.kr Advanced Applied Mechanics Laboratory (AAML) Department of Civil and Environmental Engineering KAIST Profile: Donghwi Cho PhD Candidate email@example.comFDML, MSE, KAIST Profile: Young-Seok Shim, PhD Assistant Professor firstname.lastname@example.org Division of Materials Science and Engineering Silla University https://www.silla.ac.kr Busan 46958, Korea (END)
Professor Sukyung Park Named Presidential Science and Technology Adviser
Professor Sukyung Park from the Department of Mechanical Engineering was appointed as the science and technology adviser to the President Jae-in Moon on May 4. Professor Park, at the age of 47, became the youngest member of the president’s senior aide team at Chong Wa Dae. A Chong Wa Dae spokesman said on May 4 while announcing the appointment, “Professor Park, a talent with a great deal of policymaking participation in science and technology, will contribute to accelerating the government’s push for science and technology innovation, especially in the information and communications technology (ICT) sector.” Professor Park joined KAIST in 2004 as the first female professor of mechanical engineering. She is a biomechanics expert who has conducted extensive research on biometric mechanical behaviors. Professor Park is also a member of the KAIST Board of Trustees. Before that, she served as a senior researcher at the Korea Institute of Machinery and Materials (KIMM) as well as a member of the Presidential Advisory Council on Science and Technology. After graduating from Seoul Science High School as the first ever two-year graduate, Professor Park earned a bachelor and master’s degrees in mechanical engineering at KAIST. She then finished her Ph.D. from the University of Michigan. (END)
First Korean Member of OceanObs' Organizing Committee
Professor Sung Yong Kim from the Department of Mechanical Engineering became the first Korean to be elected as an organizing committee member of the international conference OceanObs’19’, specializing in the ocean observing field. Professor Kim has been actively engaged in advisory panels, technical committees, and working groups for the North Pacific Marine Science Organization (PICES). Through numerous activities, he was recognized for his professionalism and academic achievements, which led him to be appointed as a member of the organizing committee. The organizing committee is comprised of leading scholars and researchers from 20 countries, and Professor Kim will be the first Korean scientist to participate on the committee. Since 1999, the conference has been held every decade. Global experts specializing in oceanic observation gather to discuss research directions for the next ten years by monitoring physical, biological, and chemical variables in regional, national, and global oceans and applying marine engineering. This year, approximately 20 institutes including NASA’s Jet Propulsion Laboratory (JPL), the National Science Foundation, the National Oceanic and Atmospheric Administration, and the European Space Agency will support funds as well as high-tech equipment to the conference. This year’s conference theme is the governance of global ocean observing systems such as underwater gliders, unmanned vehicles, remote sensing, and observatories. The conference will hold discussions on monitoring technology and information systems to ensure human safety as well as to develop and preserve food resources. Additionally, participants will explore ways to expand observational infrastructures and carry out multidisciplinary approaches. There will also be collaborations with the Global Ocean Observing System (GOOS) and the Partnership for Observation of the Global Oceans (POGO) to organize ocean observing programs and discuss priorities. Finally, they will set a long-term plan for solving major scientific issues, such as climate change, ocean acidification, energy, and marine pollution. Professor Kim said, “Based on the outcomes drawn from the conference, I will carry out research on natural disasters and climate change monitoring by using unmanned observing systems. I will also encourage more multidisciplinary research in this field.”
Professor Sung Yong Kim Presents a Keynote Speech at the International Ocean Color Science Meeting (IOCS) 2015
Professor Sung Yong Kim of the Mechanical Engineering Department at KAIST delivered a keynote speech at the International Ocean Color Science Meeting (IOCS) 2015 held in San Francisco on June 15-18, 2015. His speech was entitled “Research and Applications Using Sub-mesoscale GOCI (Geostationary Ocean Color Imager) Data.” The IOCS, organized by the International Ocean Color Coordinating Group (IOCCG), is a community consultation meeting providing communication and collaboration between space agencies and the ocean color community, building strong ties among international representatives of the ocean color communities, and providing a forum for discussion and the evolution of community thinking on a range of issues. Professor Kim was recognized for his contribution towards the development of remote exploration of sub-mesoscale processes including eddies, fronts, and environmental fluid dynamics. He also attended the 26th General Assembly of the International Union of Geodesy and Geophysics (IUGG) in Prague, the Czech Republic, on June 22, 2015 and gave a presentation on the sub-mesoscale eddies circulation research.
Professor Sung-Yong Kim Receives the Young Scientist Award
Professor Sung-Yong Kim of the Department of Ocean Systems Engineering at KAIST received the Young Scientist Award for 2014 conferred by the Korean Society of Oceanography (KSO). The award was presented at the KSO’s fall conference that took place on November 6, 2014 at the campus of the Naval Academy of the Republic of Korea in Jinhae. Professor Kim has been recognized for his outstanding research in coastal oceanography and environmental fluid mechanics. His research papers are frequently published in prestigious journals such as the Journal of Geophysical Research-Oceans by the American Geophysical Union.
Professor Sung Yong Kim Appointed as Committee Member to Serve PICES
The Pacific International Council for the Exploration of the Sea: North Pacific Marine Science Organization (PICES) is an intergovernmental organization, which was established in 1992 to promote and coordinate marine research in the North Pacific and adjacent areas. Currently, the United States, Canada, Japan, China, Russia, and Korea are members of the organization. Professor Sung Yong Kim of Ocean Systems Engineering, KAIST, has been appointed to serve the Scientific and Technical Committees of PICES. He will begin his stint from July 1, 2014. During his assignment, Professor Kim will identify the need for observation of the North Pacific marine environment, develop observation methodology, and publish an annual report on the observation. Professor Kim is an expert in marine physics and environmental fluids, with a focus on coastal circulation and dynamics, mesoscale and submesoscale eddies, integrated coastal ocean observing system, and statistical and dynamic data analysis.
Observation of a water strider led to a new method of measuring properties of Nano films
Even the mechanical properties of Nano films of a few nanometers thick can be measured Posted online Nature Communications on the 3rd of October The joint research team of KAIST’s Department of Mechanical Engineering’s Professor Taek-Soo Kim and Doctor Seung-Min Hyun of the Nano mechanics laboratory of Korea Institute of Machinery and Materials has developed a new method to evaluate mechanical properties of Nano films using the characteristics of water surfaces. The research findings have been posted on the online edition of Nature Communications on the 3rd of October. The technology can obtain accurate results by directly measuring the mechanical properties such as the strength and elasticity of Nano films. Academia and the industry expect the simplicity of the technology to present a new paradigm in the evaluation of mechanical properties of Nano films. Evaluation of the mechanical properties of Nano films is essential not only in predicting the reliability of semiconductors and displays, but also in finding new phenomena in the Nano world. However, mechanical strength was difficult to test since the test demands the falling of objects to the ground to measure their strength, and nano films can easily break in the process. The research team observed insects such as water striders freely floating on the surface of the water. The team used the properties of water, large surface tension and low viscosity, to float a 55 nanometers (nm) gold Nano film to successfully measure its mechanical properties without damaging it. The technology could be used to measure the mechanical properties of not only various types of Nano films but also films only a few nm thick. Professor Taek-Soo Kim said, “We effectively performed an evaluation of the mechanical characteristics of Nano films, which was difficult in the past, by developing a new strength test using the properties of water.” He continued to say, “The team plans to discover the mechanical properties of 2D Nano films such as graphene that could not have been measured with the existing strength test methods.” The research by KAIST’s Department of Mechanical Engineering’s graduate student Jae-Han Kim (lead author) under the supervision of Professor Taek-Soo Kim and Doctor Seung-Min Hyun of Korea Institute of Machinery and Materials was sponsored by the National Research Foundation of Korea. Evaluation process of mechanical properties of Nano films by using the characteristics of water surfaces Dr Seung-Min Hyun, Jae-Han Kim, and Professor Taek-Soo Kim from left to right
Principle behind increasing the catalytic property of nanocatalysts proven
The technology that allows full control of the catalytic property of nanocatalysts using oxide formation on nanocatalysts has been developed by KAIST researchers. The breakthrough opens up the possibility of the development of a new kind of catalysts that maximizes catalytic property and minimizes waste. *nanocatalyst is a material that catalyzes gas reactions on its surface. It is composed of a high surface area oxide scaffold with nano-sized metal particles dispersed. The team was led by Professor Park Jeong Young of the KAIST EEWS Graduate School and consists of Kamran Qadir Ph.D. candidate (1st Author), Professor Joo Sang Hoon of UNIST, Professor Moon Bong Jin of Hanyang University, and Professor Gabor Somorajai of UC Berkeley. Support for the research was provided from Ministry of Education Science and Technology, National Research Foundation, and Ministry of Knowledge Economy. The results were published as the online edition of Nano Letters: “Intrinsic Relation between Catalytic Activity of CO Oxidation on Ru Nanoparticles and Ru Oxides Uncovered with Ambient Pressure XPS”. Catalysts are included in above 80% of all the products used in everyday life and are therefore included in most aspects of our lives. The focus on nanocatalysts is based on finding solutions to increasing the efficiency for application to energy production and for solving environmental issues. Most nanocatalysts are composed of nanoparticles and oxides where the nanoparticles increase the surface area of the catalyst to increase its activity. The efficiency of a nanocatalyst is affected by the surface oxide of the nanoparticles. However the proving of this assumption remained difficult to do as it required in-situ measurement of the oxide state of the nanoparticles in the specific environment. Thus far, the experiments were conducted in a vacuum and therefore did not reflect the actual behavior in real life. The recently developed X-ray Photoelectron Spectroscopy allows for measurement of the oxidization state at standard atmospheric pressure. Professor Park’s research team successfully measured the oxidization state of the nanoparticle using the atmospheric pressure X-ray Photoelectron Spectroscopy in the specified environment. They confirmed the effect the oxidization state on the catalytic effect of the nanoparticles and additionally found that a thin layer of oxide can increase the catalytic effect and the effectiveness of the nanoparticle can controlled by the oxidation state.
KAIST researchers verify and control the mechanical properties of graphene
KAIST researchers have successfully verified and controlled the mechanical properties of graphene, a next-generation material. Professor Park Jung Yong from the EEWS Graduate School and Professor Kim Yong Hyun from the Graduate School of Nanoscience and Technology have succeeded in fluorinating a single atomic-layered graphene sample and controlling its frictional and adhesive properties. This is the first time the frictional properties of graphene have been examined at the atomic level, and the technology is expected to be applied to nano-sized robots and microscopic joints. Graphene is often dubbed “the dream material” because of its ability to conduct high amounts of electricity even when bent, making it the next-generation substitute for silicon semiconductors, paving the way for flexible display and wearable computer technologies. Graphene also has high potential applications in mechanical engineering because of its great material strength, but its mechanical properties remained elusive until now. Professor Park’s research team successfully produced individual graphene samples with fluorine-deficiency at the atomic level by placing the samples in Fluoro-xenon (XeF2) gas and applying heat. The surface of the graphene was scanned using a micro probe and a high vacuum atomic microscope to measure its dynamic properties. The research team found that the fluorinated graphene sample had 6 times more friction and 0.7 times more adhesiveness than the original graphene. Electrical measurements confirmed the fluorination process, and the analysis of the findings helped setup the theory of frictional changes in graphene. Professor Park stated that “graphene can be used for the lubrication of joints in nano-sized devices” and that this research has numerous applications such as the coating of graphene-based microdynamic devices. This research was published in the online June edition of Nano Letters and was supported by the Ministry of Science, Technology, and Education and the National Research Foundation as part of the World Class University (WCU) program.
마지막 페이지 1
KAIST, 291 Daehak-ro, Yuseong-gu, Daejeon 34141, Republic of Korea
Copyright(C) 2020, Korea Advanced Institute of Science and Technology,
All Rights Reserved.