In CO₂ laser surface machining of plastic films in modern display manufacturing, scattering of fume particles could be a major source of well-recognized film surface contamination. This computational fluid dynamics research investigates the suction air flow patterns over a film surface as well as the dispersion of micron-sized fume particles with low-Reynolds number particle drag model. The numerical results show the recirculatory flow patterns near laser machining point on film surface and also over the surface of vertical suction slot, which may hinder the efficient removal of fume particles from film surface. The dispersion characteristics of fume particles with various particle size have been tested systematically under different levels of suction flow intensity. It is found that suction removal efficiency of fume particles heavily depends on the particle size in highly nonlinear manners and a higher degree of suction does not always results in more efficient particle removal.

The edge ring placed on the outside of the electrostatic chuck (ESC) is a key component for protecting the ESC and controlling the etching uniformity of the edge of the wafer. Therefore, it is very important to understand the etching phenomenon of edge rings for edge ring management and equipment homeostasis. In this study, a specimen with SiO₂ hard mask and underlying Si mold was installed on the edge ring surface and the etching results were measured by varying the edge ring 2MHz RF power. By developing PI-VM model with high prediction accuracy and analyzing the roles of key parameters in the model, we were able to evaluate the effect of plasma and sheath characteristics around the edge ring on edge ring erosion. This analysis method provided information necessary for edge ring maintenance and operation.

A subway station is a prominent multi-purpose facility where the quantitative management of fine dust, generated by various factors, is conducted. Recently, eco-friendly air purification methods using air-purifying plants are being discussed, with the focus on biofiltration through vegetation. Previous research in this field has confirmed the reduction effects of transition metals such as Fe, which have been identified as harmful to human health. This study aimed to identify the sources of fine dust dispersion within subway stations and derive an efficient operational strategy for air-purifying plants that takes into account the behavior characteristics of fine dust within multi-purpose facilities. The experiment monitored regional fine dust levels through IAQ stations established based on prior research. Also, the data was analyzed through time-series and correlation analyses by linking it with passenger counts at subway stations and the frequency of train stops. Furthermore, to consider energy efficiency, we conducted component-specific power consumption monitoring. Through this study, we were able to derive the optimal operational strategy for air-purifying plants based on time-series comprehensive analysis data and confirm significant energy efficiency.

With an attempt to investigate the correlation between the internal pressure distribution of slit nozzle and the thickness uniformity of slot-coated thin films, we have performed computational fluid dynamics (CFD) simulations of slit nozzles and slot coating of high-viscosity (4,800 cPs) polydimethylsiloxane (PDMS) using a gantry slot-die coater. We have calculated the coefficient of variation (CV) to quantify the pressure and velocity distributions inside the slit nozzle and the thickness non-uniformity of slot-coated PDMS films. The pressure distribution inside the cavity and the velocity distribution at the outlet are analyzed by varying the shim thickness and flow rate. We have shown that the cavity pressure uniformity and film thickness uniformity are enhanced by reducing the shim thickness. It is addressed that the CV value of the cavity pressure that can ensure the thickness non-uniformity of less than 5% is equal to and less than 1%, which is achievable with the shim thickness of 150 ㎛. It is also found that as the flow rate increases, the average cavity pressure is increased with the CV value of the pressure unchanged and the maximum coating speed is increased. As the shim thickness is reduced, however, the maximum coating speed and flow rate decrease. The highly uniform PDMS films shows the tensile strain as high as 180%, which can be used as a stretchable substrate.

We investigated the impact of NF₃ / H₂O remote plasma dry cleaning conditions on the SiO₂ etching rate at different preparation states during the fabrication of ultra-large-scale integration (ULSI) devices. This included consideration of factors like Si crystal orientation prior to oxidation and three-dimensional structures. The dry cleaning process were carried out varying the parameters of pressure, NF₃ flow rate, and H₂O flow rate. We found that the pressure had an effective role in controlling anisotropic etching when a thin SiO₂ layer was situated between Si₃N₄ and Si layers in a multilayer trench structure. Based on these observations, we would like to provide further guidelines for implementing the dry cleaning process in the fabrication of semiconductor devices having 3D structures.

For the potential applications in large-area OLED lightings, hydrogen fuel cells, and secondary batteries, we have performed an intermittent coating of high-viscosity polydimethylsiloxane using roll-to-roll slot die coater. During intermittent coating, dead zones inevitably appear where the thickness of PDMS patch films becomes non-uniform, especially at the leading/trailing edge. To reduce it, we have coated the PDMS patches by varying the process parameters such as the installation angle of the slot die head, coating speed, and patch interval. It is observed that the PDMS solution flows down and thus the thickness profile is non-uniform for horizonal intermittent coating, whereas the PDMS solution remaining on the head lip causes an increase in the PDMS thickness at the leading/trailing edges for vertical intermittent coating when the coating velocity is low. As the coating speed increases, however, the dead zone is shown to be reduced. It is addressed that the overall dead zone (the dead zone at the leading edge + the dead zone at the trailing edge) is smaller with horizontal intermittent coating than with vertical intermittent coating.

Three dimensional (3D) packaging is a next-generation packaging technology that vertically stacks chips such as memory devices. The necessity of 3D packaging is driven by the increasing demand for smaller, high-performance electronic devices (HPC, AI, HBM). Also, it facilitates innovative applications across another fields. With growing demand for high-performance devices, companies of semiconductor fields are trying advanced packaging techniques, including 2.5D and 3D packaging, MR-MUF, and hybrid bonding. These techniques are essential for achieving higher chip integration, but challenges in mass production and fine-pitch bump connectivity persist. Advanced bonding technologies are important for advancing the semiconductor industry. In this review, it was described 3D packaging technologies for chip integration including mass reflow, thermal compression bonding, laser assisted bonding, hybrid bonding.

From the past to the present, banners are consistently used as effective advertising means. In the case of Korea, there are frequent situations in which hidden advertisements are installed. As a result, such hidden advertisement materials may damage urban aesthetics and moreover, incur unnecessary manpower consumption and waste of money. The proposed method classifies the detected banners into good banner and bad banner. The classification results are based on whether the relevant banners are installed in compliance with legal guidelines. In the process, YOLO and Open Computer Vision library are used to determine from various perspectives whether banners in CCTV images comply with the guidelines. YOLO is used to detect the banner area in CCTV images, and OpenCV is used to detect the color values in the area for color comparison. If a banner is detected in the video, the proposed method calculates the location of the banner and the distance from the designated bulletin to determine whether it was installed within the designated location, and then compares whether the color used in the banner is complied with local government guidelines.

The storage capacity of 3D-NAND flash memory has been enhanced by the multi-layer dielectrics. The deposition process has become more challenging due to the tight process margin and the demand for accurate process control. To reduce product costs and ensure successful processes, process diagnosis techniques incorporating artificial intelligence (AI) have been adopted in semiconductor manufacturing. Recently there is a growing interest in process diagnosis, and numerous studies have been conducted in this field. For higher model accuracy, various process and sensor data are required, such as optical emission spectroscopy (OES), quadrupole mass spectrometer (QMS), and equipment control state. Among them, OES is usually used for plasma diagnostic. However, OES data can be distorted by viewport contamination, leading to misunderstandings in plasma diagnosis. This issue is particularly emphasized in multi-dielectric deposition processes, such as oxide and nitride (ON) stack. Thus, it is crucial to understand the potential misunderstandings related to OES data distortion due to viewport contamination. This paper explores the potential for misunderstanding OES data due to data distortion in the ON stack process. It suggests the possibility of excessively evaluating process drift through comparisons with a QMS. This understanding can be utilized to develop diagnostic models and identify the effects of viewport contamination in ON stack processes.

Edge detection in image processing and computer vision is one of the most fundamental operations. Canny edge detection algorithm has excellent performance and is currently widely used. However, it is difficult to process the algorithm in real-time because the algorithm is complex. In this study, the equations required in the algorithm were simplified to facilitate hardware implementation, and the calculation speed was increased by using a parallel structure. In particular, the size and management of local buffers were selected in consideration of parallel processing and filter size so that data could be processed without bottlenecks. It was designed in verilog and implemented in FPGA to verify operation and performance.

This paper describes a hybrid cleaning method of silicon wafer combining nano-bubble and ultrasound to remove sub-micron particles and contaminants with minimal damage to the wafer surface. In the megasonic cleaning process of semiconductor manufacturing, the cavitation induced by ultrasound can oscillate and collapse violently often with re-entrant jet formation leading to surface damage. The smaller size of cavitation bubbles leads to more stable oscillations with more thermal and viscous damping, thus to less erosive surface cleaning. In this study, ultrasonic energy was applied to the wafer surface in the DI water to excite nano-bubbles at resonance to remove contaminant particles from the surface. A patented nano-bubble generator was developed for the generation of nano-bubbles with concentration of 1×10⁹ bubbles/ml and nominal nano-bubble diameter of 150 nm. Ultrasonic nano-bubble technology improved a contaminant removal efficiency more than 97% for artificial nano-sized particles of alumina and Latex with significant reduction in cleaning time without damage to the wafer surface.

Carbon neutrality has been emerged as important mission for all the manufacturing industry to reduce energy usage and carbon emission equivalent. Korean semiconductor and display manufacturing industries are also in huge interest by minimize the energy usage as well as to find a less global warming product gases in both etch and cleaning. In addition, Korean government is also investing long term research and development plan for the safe environment in various ways. In this paper, we revisit previous research activities on carbon emission equivalent and current research activities performed in semiconductor process diagnosis research center at Myongji University with respect to the reduction of NF3 usage for the PECVD chamber cleaning, and we present the analytical result of the exhaust gas with residual gas analysis in both 6 inches and 12 inches PECVD equipment. The presented result can be a reference study of the development of new substitution gas in near future to compare the cleaning rate of the silicon oxide deposition chamber.

This study is about a service that allows people to naturally learn to bask in the sun, which is a habit to relieve depression. Modern people do not have enough time to bask in the sun due to their busy lives, and as a result, depression and fatigue are increasing day by day. Therefore, in order to relieve depression, there is a need for the development of technology to manage the schedule of sunlight basking so that users can experience sunlight more naturally. The sunlight service developed through this study can help you easily plan your sunlight schedule by recommending good dates, times, and locations for sunbathing. In addition, users can receive coins as much as they bask in the sun, and they can be motivated by the act of basking in the sun by raising their character with those coins.

Recently, demand for embedded systems requiring low power and high specifications has been increasing, and RISC-V processors are being widely applied. RISC-V, a RISC-based open instruction set architecture (ISA), has been developed and researched by UC Berkeley and other researchers since 2010. RV32I ISA is sufficient to support integer operations such as addition and subtraction instructions, but M-extension should be defined for multiplication and division instructions. This paper proposes an RV32I, RV32IM processor, and indicates benchmark performance scores compared to an existing processor. Additionally, A non-stalling method was proposed to support a 2-stage pipelined DSP multiplier to the 5-stage pipelined RV32IM processor. Proposed RV32I and RV32IM processors satisfied a maximum operating frequency of 50 MHz on Artix-7 FPGA. The performance of the proposed processors was verified using benchmark programs from Dhrystone and Coremark. As a result, the Coremark benchmark results of the proposed processor showed that it outperformed the existing RV32IM processor by 23.91%.

As the elderly population increases, the number of patients with various joint diseases, including degenerative arthritis, is steadily increasing. CPM medical devices are needed to effectively treat degenerative arthritis that is common in the elderly population. Domestic CPM medical devices have limited functions and are highly dependent on imports for expensive imported medical devices. To solve this problem, we designed a ROM measurement function using a current sensor that is not present in existing composite joint CPM medical devices. The algorithm was designed using the fact that the force caused by joint stiffness greatly increases the current flowing through the DC motor. In addition, the need for digital healthcare in the medical field is gradually expanding as the proportion of chronically ill patients increases due to the spread of the non-face-to-face economy due to COVID-19 and the aging population. Therefore, this paper aims to improve the performance of CPM medical devices by allowing real-time confirmation of rehabilitation exercise information and operation range measurement results in accordance with digital healthcare trends through a Bluetooth application developed as an Android studio.

As semiconductor devices become highly integrated and process difficulty increases, the need for highly sensitive sensors that can detect micro leaks is increasing. However, the noise contained in the CCD sensor itself acts as an obstacle to detecting fine leaks. In this study, integration time was changed for each condition, the sensor was cooled to 0℃, and the dark voltage level was measured to confirm through experiment the characteristics of the temporal noise included in the CCD sensor, a component of OES (Optical Emission Spectroscopy). When integration time was reduced from 30msec to 10msec, the dark voltage level decreased by about 20.5 % from an average of 151.5mV to 120.5mV. In the case of cooling device, Peltier elements were selected because of their simple structure and small size. During temperature cooling, the target temperature was controlled to within ±0.5℃ through PID control. When cooled from 20℃ to 0℃ using this cooling device, it was confirmed that the dark voltage level decreased by about 7% from an average of 147.0mV to 137.0mV.

In this paper, the system is proposed and implemented to share the part number, the part name, and the vehicle type through the improvement sharing bulletin board for automobile repair and maintenance. And when photos of damage parts are uploaded to the system, the system analyzes it using a deep learning model to analyze whether it is damaged and automatically classify the type of damage. By providing repair and maintenance quotes for a significant part, the system provides economically repaired by providing comparative adjustment information on repair costs to drivers who are particularly concerned about the market prices of parts and maintenance services. Through the existing bulletin board, you can exchange and share information about parts by sharing various information on repair and maintenance. This paper provides in detail the average market price per type of damage during automobile repair and maintenance, helping drivers who do not know the details of parts and maintenance services to receive reasonable quotes by providing price information.

To solve the reliability problem of organic devices that are often used outdoors, multifunctional gas barriers that block reactive gases such as moisture and oxygen and reflect harmful light such as ultraviolet rays are needed. In this study, ALD nanolaminate-based optically functional n-DBR was developed to overcome the poor gas permeability of polymer substrates and protect organic devices from harmful light. n-DBR not only achieved a WVTR of 8.76 × 10^-6 g·m^-2·day^-1, but also showed a visible light transmittance of 94.3% and an ultraviolet ray blocking ability of 2.67%. In particular, n-DBR based on a nanolaminate structure maintained its permeability characteristics even in a high temperature and high humidity environment despite being used as a layer of Al₂O₃. This functional barrier Structure can not only be used as a functional encapsulation barrier for the reliability of organic devices, but can also be used as a tinting film for vehicles.

Precision manufacturing technology is rapidly developing due to the extreme miniaturization of semiconductor processes to comply with Moore's Law. Accurate and precise alignment, which is one of the key elements of the semiconductor pre-process and post-process, is very important in the semiconductor process. The center detection of wafer align marks plays a key role in improving yield by reducing defects and research on accurate detection methods for this is necessary. Methods for accurate alignment using traditional image sensors can cause problems due to changes in image brightness and noise. To solve this problem, engineers must go directly into the line and perform maintenance work. This paper emphasizes that the development of AI technology can provide innovative solutions in the semiconductor process as high-resolution image and image processing technology also develops. This study proposes a new wafer center detection method through variable thresholding. And this study introduces a method for detecting the center that is less sensitive to the brightness of LEDs by utilizing a high-performance object detection model such as YOLOv8 without relying on existing algorithms. Through this, we aim to enable precise wafer focus detection using artificial intelligence.

Maintenance of semiconductor equipment processes is crucial for the continuous growth of the semiconductor market. The process must always be upheld in optimal condition to ensure a smooth supply of numerous parts. Additionally, it is imperative to monitor the status of the robots that play a central role in the process. Just as many senses of organs judge a person's body condition, robots also have numerous sensors that play a role, and like human joints, they can detect the condition first in the joints, which are the driving parts of the robot. Therefore, a normal state test bed and an abnormal state test bed using an aging reducer were constructed by simulating the joint, which is the driving part of the robot. Various sensors such as vibration, torque, encoder, and temperature were attached to accurately diagnose the robot's failure, and the test bed was built with an integrated system to collect and control data simultaneously in real-time. After configuring the user screen and building a database based on the collected data, the characteristic values of normal and abnormal data were analyzed, and machine learning was performed using the KNN (K-Nearest Neighbors) machine learning algorithm. This approach yielded an impressive 94% accuracy in failure diagnosis, underscoring the reliability of both the test bed and the data it produced.

This study is normal orientation technology of slit coating equipment to improve the quality of curved displays. Currently, the demand for curved displays is increasing significantly due to advantages such as screen immersion or design in various industries. Accordingly, changes in the display coating process are essential. In the curved display coating process, unlike the existing flat coating process, the nozzle must be rotated along the curvature of the curved surface to spray the coating solution. The coating solution must be applied while maintaining a uniform thickness. If the thickness of the coating liquid applied to the target surface is non-uniform, the quality of the product may be degraded such as image quality deterioration and light spreading. This paper presents technology and experimental results for keeping the nozzle of slit coating equipment perpendicular to the curved surface and is expected to contribute to the quality improvement of curved displays.

Recently, fan out wafer level packaging, which enables high integration, miniaturization, and low cost, is being rapidly applied in the semiconductor industry. In particular, FOWLP is attracting attention in the mobile and Internet of Things fields, and is recognized as a core technology that will lead to technological advancements such as 5G, self-driving cars, and artificial intelligence in the future. However, as chip density and package size within the package increase, FOWLP warpage is emerging as a major problem. These problems have a direct impact on the reliability and electrical performance of semiconductor products, and in particular, cause defects such as vacuum leakage in the manufacturing process or lack of focus in the photolithography process, so technical demands for solving them are increasing. In this paper, warpage simulation according to the thickness of FOWLP material was performed using finite element analysis. The thickness range was based on the history of similar packages, and as a factor causing warpage, the curing temperature of the materials undergoing the curing process was applied and the difference in deformation due to the difference in thermal expansion coefficient between materials was used. At this time, the stacking order was reflected to reproduce warpage behavior similar to reality. After performing finite element analysis, the influence of each variable on causing warpage was defined, and based on this, it was confirmed that warpage was controlled as intended through design modifications.

Neural network-based time series models called time series neural networks (TSNNs) are trained by the error backpropagation algorithm and used to predict process shifts of parameters such as gas flow, RF power, and chamber pressure in reactive ion etching (RIE). The training data consists of process conditions, as well as principal components (PCs) of optical emission spectroscopy (OES) data collected in-situ. Data are generated during the etching of benzocyclobutene (BCB) in a SF₆/O₂ plasma. Combinations of baseline and faulty responses for each process parameter are simulated, and a moving average of TSNN predictions successfully identifies process shifts in the recipe parameters for various degrees of faults.

The continuous shrinking of transistors in integrated circuits leads to difficulties in improving performance, resulting in the emerging transistors such as nanosheet field-effect transistors. In this paper, we propose a TCAD-machine learning framework of nanosheet FETs to model the current-voltage characteristics. Sentaurus TCAD simulations of nanosheet FETs are performed to obtain a large amount of device data. A machine learning model of I-V characteristics is trained using the multi-layer perceptron from these TCAD data. The weights and biases obtained from multi-layer perceptron are implemented in a PSPICE netlist to verify the accuracy of I-V and the DC transfer characteristics of a CMOS inverter. It is found that the proposed machine learning model is applicable to the prediction of nanosheet field-effect transistors device and circuit performance.

This paper presents a novel design of a differential patch antenna for 60-GHz millimeter-wave applications. The design process of the back-to-back (BTB) patch antenna is based on the conventional single-patch antenna. The initial design of the BTB patch antenna (Type-I) has a patch size of 0.66 × 0.98 mm² and a substrate size of 0.99 × 1.48 mm². It has a gain of 1.83 dBi and an efficiency of 94.4% with an omni-directional radiation pattern. A 0.4 mm-thick high-resistivity silicon (HRS) is employed for the substrate of the BTB patch antenna. The proposed antenna is further analyzed to investigate the effect of substrate size and resistivity. As the substrate resistivity decreases, the gain and efficiency degrade due to the substrate loss. As the substrate (HRS) size decreases approaching the patch size, the resonant frequency increases with a higher gain and efficiency. The BTB patch antenna has optimal performances when the substrate size matches the patch size on the HRS substrate (Type-II). The antenna is redesigned to have a patch size of 0.81 × 1.18 mm² on the HRS substrate in the same size. It has an efficiency of 94.9% and a gain of 1.97 dBi at the resonant frequency of 60 GHz with an omni-directional radiation pattern. Compared to the initial design of the BTB patch antenna (Type-I), the optimal BTB patch antenna (Type-II) has a slightly higher efficiency and gain with a considerable reduction in antenna area by 34.8%.

A digital controller uses a microprocessor and is a controller implemented as a program. This method has the advantage of being more maintenance-friendly than existing analog controllers. However, it inevitably requires computation time to execute the internal program. Therefore, the digital controller uses a method of controlling the system at a certain cycle by considering this time, and this cycle is very closely related to the performance of the microprocessor used. In other words, in the case of very high performance, this control cycle can be shortened to near real time, but this may result in a disadvantage in terms of cost. In this paper, we propose a method to solve this problem by implementing two processors with slightly lower performance in a control system in a series-parallel structure. For this purpose, we will use a digital distributed control system and implement an experimental system to examine its effectiveness.

Recently, with the activation of eBook services, books are being published simultaneously as physical books and digitized eBooks. Paper books are more expensive than e-books due to printing and distribution costs, so demand for relatively inexpensive e-books is increasing. There are cases where previously published physical books cannot be digitized due to the circumstances of the publisher or author, so there is a movement among individual users to digitize books that have been published for a long time. However, existing research has only studied the advancement of the pre-processing process that can improve text recognition before applying OCR technology, and there are limitations to digitization depending on the condition of the book. Therefore, support for book digitization services depending on the condition of the physical book is needed. need. In this paper, we propose a method to support digitalization services according to the status of physical books held by book owners. Create images by scanning books and extract text information from the images through OCR. We propose a method to recover text that cannot be extracted depending on the state of the book using BERT, a natural language processing deep learning model. As a result, it was confirmed that the recovery method using BERT is superior when compared to RNN, which is widely used in recommendation technology.

This paper designed a voltage feedback driving circuit to compensate for the characteristic deviation of the Active Matrix Organic Light Emitting Diode driving Thin Film Transistor. This paper describes a stable and fast circuit by applying charge sharing and polar stabilization methods. A 12-inch Organic Light Emitting Diode with a Double Wide Ultra eXtended Graphics Array resolution creates a screen distortion problem for line parasitism, and charge sharing and polar stabilization structures were applied to solve the problem. By applying Charge Sharing, all data lines are shorted at the same time and quickly positioned as the average voltage to advance the compensated change time of the gate voltage in the next operation period. A buffer circuit and a current pass circuit were added to lower the Amplifier resistance connected to the line as a polar stabilization method. The advantage of suppressing the Ringing of the driving Thin Film Transistor can be obtained by increasing the stability. As a result, a circuit was designed to supply a stable current to the Organic Light Emitting Diode even if the characteristic deviation of the driving Thin Film Transistor occurs.

This paper presents the synaptic characteristics of IGZO memristors in neuromorphic computing, using MATLAB/Simulink and NeuroSim. In order to investigate the variations in the conductivity of IGZO memristor and the corresponding changes in the hidden layer, simulations are conducted by using the MNIST dataset. It was observed from simulation results that the recognition accuracy could be dependent on various parameters of IGZO memristor, along with the experimental exploration. Moreover, we identified optimal parameters to achieve high accuracy, showing an outstanding accuracy of 96.83% in image classification.

With the expansion of metaverse content and hardware platforms, various interactions in the virtual world have been built, raising expectations for an increase in immersion which is a major element of the metaverse. However, among hardware platforms that increase virtual immersion elements, the typical HMD platform can be a barrier to new user inflows due to its high cost. Thus, this paper focused on improving virtual-to-real interactions by extracting motion data using relatively inexpensive webcam equipment in PC environments, utilizing Unity game engines, Photon unity network, multi-platform implementations, and Barracuda neural network inference libraries.

We propose the modeling methodology of CMOS inverter made of LTPO TFT using a machine learning. LTPO can achieve advantages of LTPS TFT with high electron mobility as a driving TFT and IGZO TFT with low off-current as a switching TFT. However, since the unified model of both LTPS and IGZO TFTs is still lacking, it is necessary to develop a SPICE-compatible compact model to simulate the LTPO current-voltage characteristics. In this work, a generic framework for combining the existing formula of I-V characteristics with artificial neural network is presented. The weight and bias values of ANN for LTPS and IGZO TFTs is obtained and implemented into PSPICE circuit simulator to predict CMOS inverter. This methodology enables efficient modeling for predicting LTPO TFT circuit characteristics.