• Journal of Korean Institute of Industrial Engineers
• /
• v.40 no.4
• /
• pp.404-414
• /
• 2014

Yield prediction is important to manage semiconductor quality. Many researches with machine learning algorithms such as SVM (support vector machine) are conducted to predict yield precisely. However, yield prediction using SVM is hard because extremely imbalanced and big data are generated by final test procedure in semiconductor manufacturing process. Using SVM algorithm with imbalanced data sometimes cause unnecessary support vectors from major class because of unselected support vectors from minor class. So, decision boundary at target class can be overwhelmed by effect of observations in major class. For this reason, we propose a under-sampling method with minor class based SVM (MCSVM) which overcomes the limitations of ordinary SVM algorithm. MCSVM constructs the model that fixes some of data from minor class as support vectors, and they can be good samples representing the nature of target class. Several experimental studies with using the data sets from UCI and real manufacturing process represent that our proposed method performs better than existing sampling methods.

• The Journal of Fisheries Business Administration
• /
• v.48 no.1
• /
• pp.1-16
• /
• 2017

The fisheries industry has led the Korean economy, and has been achieving high-level position in the world. However, this industry meets aging, low growth and profit. In order to overcome this critical situation, it is needed to understand the overall status of industry. In industry level, most of previous researches focused on ocean industry rather than fisheries. In addition, scholars have been getting a lot of attention about fisheries cooperatives, fishing-ports, methods of fishery, and manufacturing process in fisheries sector. The aim of this research is analysis of domestic fisheries industry's managerial performance using data envelopment analysis(DEA) considering operating and scale view. Furthermore, the comparative analysis is performed by firm size, and industry type. In results, fisheries industry's managerial performance is not high, overall. In more detail, most of big size firms are under decreasing returns to scale(DRS) status. Fishery processing industry's performance is low, and fishery distribution industry has the best performance. This paper suggests that transferring operating capability from big firms to small firms, and policy supports and firm's activities should be accompanied for high-value added in fisher, and fishery processing industries.

• Journal of the Korean Society of Manufacturing Process Engineers
• /
• v.20 no.12
• /
• pp.85-91
• /
• 2021

In our daily life, artificial intelligence performs simple and complicated tasks like us, including operating mobile phones and working at homes and workplaces. Artificial intelligence is used in industrial technology for diagnosing various types of equipment using the machine learning technology. This study presents a fault mode effect analysis (FMEA) of start motors using machine learning and big data. Through multiple data collection, we observed that the primary failure of the start motor was caused by the melting of the magnetic switch inside the start motor causing it to fail. Long-short-term memory (LSTM) was used to diagnose the condition of the magnetic locations, and synthetic data were generated using the synthetic minority oversampling technique (SMOTE). This technique has the advantage of increasing the data accuracy. LSTM can also predict a start motor failure.

• Design & Manufacturing
• /
• v.15 no.3
• /
• pp.1-6
• /
• 2021

An artificial neural network model based on a deep learning algorithm is known to be more accurate than humans in image classification, but there is still a limit in the sense that there needs to be a lot of training data that can be called big data. Therefore, various techniques are being studied to build an artificial neural network model with high precision, even with small data. The transfer learning technique is assessed as an excellent alternative. As a result, the purpose of this study is to develop an artificial neural network system that can classify burr images of light guide plate products with 99% accuracy using transfer learning technique. Specifically, for the light guide plate product, 150 images of the normal product and the burr were taken at various angles, heights, positions, etc., respectively. Then, after the preprocessing of images such as thresholding and image augmentation, for a total of 3,300 images were generated. 2,970 images were separated for training, while the remaining 330 images were separated for model accuracy testing. For the transfer learning, a base model was developed using the NASNet-Large model that pre-trained 14 million ImageNet data. According to the final model accuracy test, the 99% accuracy in the image classification for training and test images was confirmed. Consequently, based on the results of this study, it is expected to help develop an integrated AI production management system by training not only the burr but also various defective images.

• Journal of the Semiconductor & Display Technology
• /
• v.23 no.2
• /
• pp.123-128
• /
• 2024

Major countries are trying to expand the construction of smart factories by introducing ICT such as the Internet of Things, cloud, and big data into the manufacturing sector to secure national-level manufacturing competitiveness in the era of the 4th industrial revolution. In addition, Germany is pushing for Industry 4.0 to build a fully automatic production system through the Internet of Things, and China is pushing for the expansion of smart factories to enhance the country's industrial competitiveness through Made in China 2025, Japan's intelligent manufacturing system, and the Korean government's manufacturing innovation 3.0. In this study, considering the increasing security connectivity of smart factories, we would like to identify security threats in the external connection part of smart factories and suggest security enhancement measures based on domestic and international standard security models to respond to the identified security threats. Eventually the proposed method can be applied by accurately identifying the smart factory security status, diagnosing vulnerabilities, establishing appropriate improvement plans, and expanding security strategies to respond to security threats.

• Journal of Digital Contents Society
• /
• v.19 no.4
• /
• pp.789-799
• /
• 2018

This article presents machine learning based approach on Big data to analyzing time series data for anomaly detection in such industrial complex system. Long Short-Term Memory (LSTM) network have been demonstrated to be improved version of RNN and have become a useful aid for many tasks. This LSTM based model learn the higher level temporal features as well as temporal pattern, then such predictor is used to prediction stage to estimate future data. The prediction error is the difference between predicted output made by predictor and actual in-coming values. An error-distribution estimation model is built using a Gaussian distribution to calculate the anomaly in the score of the observation. In this manner, we move from the concept of a single anomaly to the idea of the collective anomaly. This work can assist the monitoring and management of Smart Factory in minimizing failure and improving manufacturing quality.

• Journal of the Korea Society of Computer and Information
• /
• v.27 no.12
• /
• pp.111-119
• /
• 2022

The purpose of this paper is to examine the change in perception of 3D virtual fashion before and after COVID-19 using big data analysis. The data collection period is from January 1, 2017, before the outbreak of COVID-19, to October 30, 2022, after the outbreak. Big data was collected for key words related to 3D virtual fashion extracted from social media such as Naver, Daum, Google, and YouTube using Textom. After the collected words were refined, word cloud, word frequency, connection centrality, network visualization, and CONCOR analysis were performed. As a result of extracting and analyzing 32,461 words with 3D virtual fashion as a keyword, the frequency and centrality of fashion, virtual, and technology appeared the highest, and the frequency of appearance of digital, design, clothing, utilization, and manufacturing was also high. Through this, it was found that 3D virtual fashion is being used throughout the industry along with the development of technology. In particular, the key words that stand out the most after COVID-19 are metaverse and 3D education, which are in high demand in the fashion industry.

• Journal of ILASS-Korea
• /
• v.20 no.4
• /
• pp.223-229
• /
• 2015

This paper was designed to investigate emission characteristics of regulated pollutants (CO, HC, NOx and PM) from 134 diesel and gasoline passenger cars based on emission standards according to fuel additives. The experiments using chassis dynamometer were conducted under NEDC and CVS-75 modes. Comparison for fuel additive management and test between Korea, USA, EU and Japan, Korea was more strict than others. The fuel additives of this study was satisfied within fuel manufacturing standards. For with/without fuel additives according to diesel emission standards, NOx of EURO 4 and EURO 5 showed a relatively similar tendency. In the case of PM reduction rate, EURO 5 was over 20% increased than EURO 4. In the case of standard deviation/average ratio for gasoline vehicles, variation interval was big for LEV 23.3~58% and ULEV 31.6~56.4%. Following the imposition of stricter regulations (EURO 5 and ULEV), difference rate for standard deviation was big. Especially, in the case of diesel vehicles, difference rate for NOx 68% and PM 48% was most big. The results of present study will be of assistance in completing the legislative process and will provide basic data to set up emission standards for fuel additives in Korea.

• Design & Manufacturing
• /
• v.17 no.1
• /
• pp.27-32
• /
• 2023

In this paper, we intend to predict the mass of the spiral using CAE and machine learning. First, We generated 125 data for the experiment through a complete factor design of 3 factors and 5 levels. Next, the data were derived by performing a molding analysis through CAE, and the machine learning process was performed using a machine learning tool. To select the optimal model among the models learned using the learning data, accuracy was evaluated using RMSE. The evaluation results confirmed that the Support Vector Machine had a good predictive performance. To evaluate the predictive performance of the predictive model, We randomly generated 10 non-overlapping data within the existing injection molding condition level. We compared the CAE and support vector machine results by applying random data. As a result, good performance was confirmed with a MAPE value of 0.48%.

• Journal of Convergence for Information Technology
• /
• v.9 no.3
• /
• pp.16-21
• /
• 2019

Small and medium-sized Korean companies are currently changing their industrial structure faster than in the past due to various environmental factors (such as securing competitiveness and developing excellent products). In particular, the importance of collecting and utilizing data produced in smart factory environments is increasing as diverse devices related to artificial intelligence are put into manufacturing sites. This paper proposes an artificial intelligence-based smart factory model to improve the process of products produced at the manufacturing site with the recent smart factory. The proposed model aims to ensure the increasingly competitive manufacturing environment and minimize production costs. The proposed model is managed by considering not only information on products produced at the site of smart factory based on artificial intelligence, but also labour force consumed in the production of products, working hours and operating plant machinery. In addition, data produced in the proposed model can be linked with similar companies and share information, enabling strategic cooperation between enterprises in manufacturing site operations.