• Journal of the Korea Society of Computer and Information
• /
• v.28 no.10
• /
• pp.27-35
• /
• 2023

In this paper, we propose a process of increasing productivity by applying a deep learning-based defect detection and classification system to the prepreg fiber manufacturing process, which is in high demand in the field of producing composite materials. In order to apply it to toe prepreg manufacturing equipment that requires a solution due to the occurrence of a large amount of defects in various conditions, the optimal environment was first established by selecting cameras and lights necessary for defect detection and classification model production. In addition, data necessary for the production of multiple classification models were collected and labeled according to normal and defective conditions. The multi-classification model is made based on CNN and applies pre-learning models such as VGGNet, MobileNet, ResNet, etc. to compare performance and identify improvement directions with accuracy and loss graphs. Data augmentation and dropout techniques were applied to identify and improve overfitting problems as major problems. In order to evaluate the performance of the model, a performance evaluation was conducted using the confusion matrix as a performance indicator, and the performance of more than 99% was confirmed. In addition, it checks the classification results for images acquired in real time by applying them to the actual process to check whether the discrimination values are accurately derived.

• Korean Journal of Acupuncture
• /
• v.40 no.4
• /
• pp.184-193
• /
• 2023

Objectives : The purpose of this study is to understand the temperature characteristics depending on the thickness and material of the needle used with the Electronic Warm Acupuncture Device (EWAD). Methods : We controlled experimental environment and measured temperature changes of a silicon phantom in which K-type thermocouples were inserted at depths of 2, 7 mm. EWAD perfomed with acupuncture needles of various thicknesses (0.25×60 mm, 0.40×60 mm, and 0.50×60 mm) and materials (Gold 0.40×50 mm). We set non-needle (only heated with EWAD skin heater) group as a control group. Results : The maximum temperature and temperature changes of 0.40 mm, 0.50 mm needle group were significantly higher than the non-needle group. The highest temperature range in all needle groups was 0.50 mm needle group (41.44±0.31℃). However, the 0.25 mm needle group was not significantly different from the non-needle group. Maximum temperature of gold needle group was significantly higher than stainless steel needle group. Temperature changes of gold needle group were higher than stainless steel group at the depth of 7 mm. Conclusions : It was found that needle thickness and material of acupuncture had an effect on the temperature of the EWAD. When performing EWAD treatment, consideration of thickness and material of acupuncture is needed. Future research is needed using phantoms that can reflect actual clinical situations and better mimic the human body.

• Journal of Environmental Health Sciences
• /
• v.49 no.6
• /
• pp.334-343
• /
• 2023

Background: South Korea's Act on Registration and Evaluation, etc. of Chemicals (known as K-REACH) was established to protect public health and the environment from hazardous chemicals. 4,4'-Methylenedianiline (MDA), which is used as a major intermediate in industrial polymer production and as a vulcanizing agent in South Korea, is classified as a toxic substance under the K-REACH act. Although MDA poses potential ecological risks due to industrial emissions and hazards to aquatic ecosystems, no ecological risk assessment has been conducted. Objectives: The aim of this study is to assess the ecological risk of MDA by identifying the actual exposure status based on the K-REACH act. Methods: Various toxicity data were collected to establish predicted no effect concentrations (PNECs) for water, sediment, and soil. Using the SimpleBox Korea v2.0 model with domestic release statistical data and EU emission factors, predicted environmental concentrations (PECs) were derived for ten sites, each referring to an MDA-using company. Hazard quotient (HQ) was calculated by ratio of the PECs and PNECs to characterize the ecological risk posed by MDA. To validate the results of modeling-based assessment, concentration of MDA was measured using in-site freshwater samples (two to three samples per site). Results: PNECs for water, sediment, and soil were 0.000525 mg/L, 4.36 mg/kg dw, and 0.1 mg/kg dw, respectively. HQ for surface water and sediment at several company sites exceeded 1 due to modeling data showing markedly high PEC in each environmental compartment. However, in the results of validation using in-site surface water samples, MDA was not detected. Conclusions: Through an ecological risk assessment conducted in accordance with the K-REACH act, the risk level of MDA emitted into the environmental compartments in South Korea was found to be low.

• Journal of Venture Innovation
• /
• v.6 no.4
• /
• pp.1-21
• /
• 2023

This study is related to the performance of open innovation collaboration between startups and large corporations and financial institutions. In the life cycle of a typical company, the growth of a startup is difficult to predict. Startups that possess innovative technology but have only recently been established seek to verify their technology and capabilities by participating in open innovation with large corporations and financial institutions, and further strive to lay the foundation for corporate growth. However, if you approach it only as a theoretical coexistence plan, it will be viewed as a vague attempt from the startup's perspective. The purpose of this study is to differentiately verify the benefits of open innovation by analyzing the difference in sales growth of startups for the purpose of sales performance based on the open innovation participation of large companies and small and medium-sized companies(startups). In verifying this, the analysis was based on the sales results of the actual open innovation collaboration B2C model, and the difference was confirmed by comparing before and after collaboration. Here, the differentiation of the study was added by reflecting the corporate growth stage theory, a growth theory. When the corporate growth stage theory was excluded, it was confirmed that sales growth due to open innovation of startups was applied from the third month, and sales growth depending on participation was confirmed to be significant. On the other hand, when the corporate growth stage theory was applied, sales growth was not significant, but the difference in growth could be confirmed from the fourth month, and it was also confirmed in sales growth depending on participation. As a result, this study objectively confirms the effects that can be gained when startups participate in Open-innovation, and it is expected that Open-innovation led by large corporations, financial institutions, and government agencies will develop into a high-quality program environment.

• Journal of the Korea Society of Computer and Information
• /
• v.29 no.2
• /
• pp.31-41
• /
• 2024

Deep learning models show excellent performance in tasks such as image classification and object detection in the field of computer vision, and are used in various ways in actual industrial sites. Recently, research on improving robustness has been actively conducted, along with pointing out that this deep learning model is vulnerable to hostile examples. A hostile example is an image in which small noise is added to induce misclassification, and can pose a significant threat when applying a deep learning model to a real environment. In this paper, we tried to confirm the robustness of the edge-learning classification model and the performance of the adversarial example detection model using it for adversarial examples of various algorithms. As a result of robustness experiments, the basic classification model showed about 17% accuracy for the FGSM algorithm, while the edge-learning models maintained accuracy in the 60-70% range, and the basic classification model showed accuracy in the 0-1% range for the PGD/DeepFool/CW algorithm, while the edge-learning models maintained accuracy in 80-90%. As a result of the adversarial example detection experiment, a high detection rate of 91-95% was confirmed for all algorithms of FGSM/PGD/DeepFool/CW. By presenting the possibility of defending against various hostile algorithms through this study, it is expected to improve the safety and reliability of deep learning models in various industries using computer vision.

• Journal of the Korean Recycled Construction Resources Institute
• /
• v.12 no.2
• /
• pp.229-238
• /
• 2024

This study evaluated the odor removal performance of a bacteria-based odor reduction kit. The bacteria used were Rhodobacter capsulatus, Paracoccus limosus, and Brevibacterium hankyongi, which can remove ammonia (NH₃), hydrogen sulfide (H₂S), total nitrogen (T-P), and total phosphorus (T-N), which are odor pollutants. The materials used were bacteria and porous aggregates (expanded vermiculite, zeolite beads, activated carbon), and the combination of the materials varied depending on the removal mechanism. Materials with a physical adsorption mechanism (zeolite beads and activated carbon) gradually slowed down the concentration reduction rate of odor pollutants (NH₃, H₂S, T-P, and T-N), and had no further effect on reducing the concentration of odor pollutants after 60 hours. Expanded vermiculite, in which bacteria that remove odors through a bio-adsorption mechanism were immobilized, had a continuous decrease in concentration, and the concentration of odor pollutants reached 0 ppm after 108 hours. As a result, the odor removal performance of materials with physical adsorption mechanisms in actual river water did not meet the odor emission standard required by the Ministry of Environment, while the expanded vermiculite immobilized with bacteria satisfied the odor emission permissible standard and achieved water quality grade 1.

• KIPS Transactions on Computer and Communication Systems
• /
• v.13 no.1
• /
• pp.21-30
• /
• 2024

As the self-driving car market continues to grow, the need for charging infrastructure is growing. However, in the case of a wireless charging system, stability issues are being raised because it requires a large amount of power compared with conventional wired charging. SAE J2954 is a standard for building autonomous vehicle wireless charging infrastructure, and the standard defines a communication method between a vehicle and a power transmission system. SAE J2954 recommends using physical media such as Wi-Fi, Bluetooth, and UWB as a wireless charging communication method for autonomous vehicles to enable communication between the vehicle and the charging pad. In particular, UWB is a suitable solution for indoor and outdoor charging environments because it exhibits robust communication capabilities in indoor environments and is not sensitive to interference. In this standard, the process for building a wireless power transmission system is divided into several stages from the start to the completion of charging. In this study, UWB technology is used as a means of fine alignment, a process in the wireless power transmission system. To determine the applicability to an actual autonomous vehicle wireless power transmission system, experiments were conducted based on distance, and the distance information was collected from UWB. To improve the accuracy of the distance data obtained from UWB, we propose a Single Model and Multi Model that apply machine learning and deep learning techniques to the collected data through a three-step preprocessing process.

• Journal of Korean Library and Information Science Society
• /
• v.54 no.4
• /
• pp.229-254
• /
• 2023

The university library's service strategy needs to be established based on users, who are actual service consumers, including an outlook on changes in the social environment. Accordingly, in this study, college and graduate students, instructors, and researchers who are university library users were identified to understand users' perceptions of the university library functions and services currently provided, and the demand for services that need to be improved or developed. Data was collected through an online survey, with 1,216 responses from the student group and 433 responses from the researcher group. The survey results were organized by each group, and implications were drawn from common results. First, it was found that both groups had a continuous demand for strengthening the collection and access to information resources. Second, there was a need to expand information provision services, such as strengthening the sharing of information resources through cooperation with other organizations and wishing to use overseas academic materials in various ways. Third, although the library was recognized as an important institution, it was found that satisfactory use was not achieved due to lack of publicity. Fourth, it was found that university libraries recognize that they must provide open services to everyone without discrimination. The results of this study can be used as basic data when establishing strategies to develop and improve university library services optimized for users.

• Korean Journal of Construction Engineering and Management
• /
• v.25 no.4
• /
• pp.3-14
• /
• 2024

Unlike conventional construction production methods, OSC (Off-Site Construction) have many restrictions depending on the production environment and technology, so it is important to develop a design plan considering these restrictions and select the optimal design alternative considering the overall efficiency of the building production process. Accordingly, the construction industry is paying attention to DfMA (Design for Manufacturing & Assembly) to derive the optimal design plan for the OSC project. Leading OSC countries such as Singapore and the United Kingdom have recognized the need to apply DfMA and presented DfMA guidelines and application strategies suitable for the characteristics of the OSC industry, and several researchers are conducting research to integrate DfMA into the construction industry. However, in the case of Korea, the need for industrial application and industrial application of DfMA is recognized, but there are no methods and tools necessary to implement the concept of DfMA in the design of actual construction projects. Therefore, this study aims to present the basic direction of strategy for applying DfMA and developing tools for the development of the domestic OSC industry by analyzing the development and application of DfMA in overseas construction. The results of this study can be used as basic data for the development of DfMA tools suitable for the domestic OSC industry and the establishment of policies related to DfMA in the future.

• The Journal of Bigdata
• /
• v.9 no.1
• /
• pp.153-168
• /
• 2024

In summer, as chillers are considered the main energy consumer of building, the efficient chiller operation is considered important. However, it is difficult to operate chillers to meet the cooling demand of the building as the demand fluctuates with various factors like the internal, external environment and behavior of the occupants and as chiller's constraint cause the current operation constrains operation in future. To address these problems, this study proposes a multi-chiller operation model based on deep reinforcement learning considering the minimum up-time of the chiller. The proposed model learns the value of the chiller operations according to the state composed of metrological and cooling system information and determines operation that minimizes the difference between the supply load and the cooling demand among feasible operations. The practical applicability was improved by applying the training algorithm considering the minimum up-time constraint and Experiments results using the actual data from a Korean university confirmed that the proposed model complies with the chiller constraints and outperforms the existing chiller operation logic of the university in terms of differences from the building cooling demand.