As an institute under the National Science and Technology Council (NSTC), NARLabs is committed to working alongside NSTC to promote scientific and technological development nationwide, upholding four major missions: establish R&D platforms, support academic research, promote frontier science and technology, and foster high-tech manpower. These missions aim to pursue the vision of "Global Excellence, Local Impact" with the objective of innovating technology and safeguarding Taiwan. NARLabs' seven research centers are equipped with valuable hardware and software that is difficult for domestic universities to acquire on their own. In addition, they offer a variety of specialized service platforms to assist academic and research communities in conducting scientific research and developing cutting-edge technology.
In recognition of the top scientific achievements made by teams in the industrial, government, academic, and research sectors using NARLabs' R&D service platforms, NARLabs annually holds the R&D Service Platform Achievement Awards. This year's top award, the High Distinction award, was conferred to a research team lead by Distinguished Professor Pau-Choo Chung of the Department of Electrical Engineering at National Cheng Kung University (NCKU). The team was awarded a prize of NT$300,000 for their achievements. Using the Taiwania 2 AI supercomputer of the National Center for High‑performance Computing (NCHC), the team created an open AI pathology analysis platform.
Prof. Chung's team collaborated with NCHC and doctors from NCKU Hospital, Kaohsiung Veterans General Hospital, and Taichung Veterans General Hospital to develop AI analysis models for digital pathology, which can be used to analyze images, detect human tissue and diseased areas, and calculate quantitative indicators for the diagnosis of disease, so as to pROVide doctors with objective results to assist in the diagnosis process. At present, they have developed a liver tumor detection model, liver fibrosis AI analysis technology, and a liver lipid droplet AI analysis model. A number of their research results have already been published in international journals and at seminars. In addition, the team used the results of their AI analysis to construct the "ALOVAS" digital pathology cloud analysis platform in NCHC's cloud space. The platform integrates digital pathology image analysis, tagging, AI model training, and AI analysis functions to pROVide physicians with a one-stop digital pathology image analysis service. Later, for an AI Capstone project, project leader and former NCKU President Huey-Jen Su collaborated clinically with current NCKU President Meng-Ru Shen (formerly President of NCKU Hospital) to expand the platform's functions and develop a world-first federated learning agent (FLAg) mechanism. The duo also worked with the National Institutes of Health (NIH) of the United States, becoming NIH's first partners in pathology image analysis.
Two teams won Excellence awards: one led by Professor Peichen Yu of the National Yang Ming Chiao Tung University Department of Photonics, and the other led by Professor Horng-Chih Lin of the National Yang Ming Chiao Tung University Institute of Electronics. Employing the "Semiconductor Fabrication Service Platform Lithography and Etching Module Validation Service" of the Taiwan Semiconductor Research Institute (TSRI), Prof. Yu's team developed IPC (intelligent proximity correction) for large-area and high-efficiency metalenses. Metalenses and metasurfaces are optical components composed of subwavelength nanostructures, which can be leveled and made small to reduce the overall size of an optical system. They are used in medical components, optical inspection, and AR/VR headsets. The team developed fast and accurate IPC photomask correction technology by using deep learning for microimage modeling along with parallel computing resources from NCHC, pushing the resolution to the limit of the photoresist. In addition to publishing a number of academic papers, the team has also worked with and carried out technology transfers with TSMC, SUMCO (Japan), which is the world's third largest manufacturer of silicon wafers, Himax, AUO, and TDK (Japan).
Meanwhile, Prof. Lin's team used TSRI's "Advanced Semiconductor Fabrication Service Platform and High-Frequency Measurement Technology Service" to develop T-shaped gate transistor technology. This type of new process T-shaped gate polycrystalline silicon thin-film transistor can be fabricated without additional etching procedures, which can increase the transconductance, cut-off frequency (fT), and maximum oscillation frequency (fmax) of the device and set a world record for poly-Si transistors with an fT/fmax at 62.5/30 GHz. The T-shaped gate structure can also be used to fabricate LDD (lightly doped drain) components for logic circuits, which not only greatly simplifies the fabrication process, but also impROVes off-state leakage current and increases the on-current. In addition to presenting their research at major international conferences, the team has also applied for patents in the United States and Taiwan.
Two Honorable Mention awards were conferred: one to Associate Professor Hsin-Yi Chen of the National Tsing Hua University Department of Engineering and System Science, who used NCHC's Taiwania 1 supercomputer to create computational catalysis tools for the design, optimization, and descriptor development of materials for green energy, and the other to a team led by Associate Professor Syu-Jyun Peng of the Taipei Medical University Professional Master Program in Artificial Intelligence in Medicine, who used NCHC's "LIONS Data Framework" to develop a machine learning quantitative localization system and operation method for positron emission tomography.
NARLabs President Faa-Jeng Lin stated that under the support and guidance of NSTC, NARLabs has established a variety of R&D service platforms to assist academic and research communities in developing top-notch results, which in turn contribute to people's well-being. Nowadays, research on new technologies is becoming more and more complex and challenging, and much future-oriented scientific and technological research relies on teamwork as well as advanced software and hardware, which are the specialty of the various R&D service platforms set up by NARLabs. NARLabs hopes that the R&D Service Platform Achievement Awards will encourage Taiwan's academic and research communities to further collaborate with NARLabs and use its platforms developed through the strength of the nation to attain globally competitive R&D achievements.