• Design Convergence Study
• /
• v.16 no.6
• /
• pp.123-135
• /
• 2017

In recent mobile health care service, health management using number of steps is becoming popular. In addition, a variety of activity trackers have made it possible to measure the number of steps more accurately and easily. Nevertheless, the activity tracker is not popularized, and it is a trend to use the pedometer sensor of the smartphone as an alternative. In this study, we tried to find out how much the number of steps collected by the smartphone versus the actual number of steps in actual situations, and what factors make the difference. We conducted an experiment to collect number of steps data of 21 people using the smartphone and wearable device simultaneously for 7 days. As a result, we found that the average number of steps of the smartphone is 62% compared to the actual number of steps, and that there is a large variation among users. We derived a regression model in which the accuracy of smartphone increases with the degree of awareness of smartphone. We expect that this can be used as a factor to correct the difference from the actual number of steps in the smartphone alone healthcare service.

• The Journal of the Korea Contents Association
• /
• v.19 no.2
• /
• pp.36-48
• /
• 2019

The purpose of this study is to investigate case studies and industry needs for convergence education in the fields of electronics, software, and automobiles, which are the technical foundations of the Fourth Industrial Revolution. Through the literature review, the convergence education programs focusing on electronics, software, and automobile were derived. The areas were validated by the experts review who consisted of three industry experts and professors in the relevant fields. Domestic and foreign curriculum were investigated to understand the current state of technical convergence education in each field. Industry needs survey for technical convergence education was conducted in cooperation with the Sector Council of Industrial Resources. Research instruments consisted of three parts: needs for technical convergence education, needs for the specific convergence education in the field of electronics, software, and automobile, and opinions on convergence education. A total of 341 participants responded to the questionnaires: 132 in the electronic field, 100 in the software field, and 109 in the automobile field. The industry needs for convergence education were analyzed and implications were suggested. The results of this study are expected to provide a guideline for developing convergence education programs in higher education.

• Journal of the Korean Society of Manufacturing Process Engineers
• /
• v.20 no.10
• /
• pp.102-111
• /
• 2021

Maintainability is a major design parameter that includes availability as well as reliability in a RAM (reliability, availability, maintainability) analysis, and is an index that must be considered when developing a system. There is a lack of awareness of the importance of predicting and analyzing maintainability; therefore, it is dependent on past-experience data. To improve the utilization rate, maintainability must be managed as a key indicator to meet the user's requirements for failure maintenance time and to reduce life-cycle costs. To improve the maintainability-prediction accuracy in the detailed design stage, we present a maintainability-prediction method that applies Method B of the Military Standardization Handbook (MIL-HDBK-472) Procedure V, as well as a Korean maintainability-prediction software package that reflects the system complexity.

• Journal of the Korean Society of Manufacturing Process Engineers
• /
• v.20 no.7
• /
• pp.80-88
• /
• 2021

The behavior of the injection system was determined through FFT and PSD analysis of the pressure data of the common rail, and when the diesel fuel is mixed with water, the pressure data of the common rail, depending on the water content and engine rotation speed, represent a different frequency component distribution. Recently, a theory has been suggested that mixing diesel fuel with water controls engine overheating, fuel efficiency, NOx, CO, etc., but if water content exceeds 10%, it can have a fatal adverse effect on the engine's injection system. In the future, it is necessary to promote fault diagnosis and prediction studies of diesel engines using FFT and PSD results from common rail pressure data.

• Journal of the Korean Society of Manufacturing Process Engineers
• /
• v.20 no.3
• /
• pp.1-7
• /
• 2021

With the advent of industry 4.0, big data and machine learning techniques are being widely adopted in the maintenance domain. Inverters are widely used in many engineering applications. However, overloading and complex operation conditions may lead to various failures in inverters. In this study, failure mode effect analysis was performed on inverters and voltages collected to investigate the over-voltage effect on capacitors. Several features were extracted from the collected sensor data, which indicated the health state of the inverter. Based on this correlation, the best features were selected for classification. Moreover, random forest classifiers were used to classify the healthy and faulty states of inverters. Different performance metrics were computed, and the classifiers' performance was evaluated in terms of various health features.

• Journal of the Korean Society of Manufacturing Process Engineers
• /
• v.21 no.7
• /
• pp.91-97
• /
• 2022

Pipes transport and supply fuel in various categories; however, corrosion occurs because of the external environment, impurities are mixed in the fuel, and substances leak to the outside, which can lead to serious accidents. Therefore, in this study, inspection equipment using a laser scanner was manufactured to classify conditions according to the degree of corrosion of the outer wall of the pipe, and the corrosion height and maximum value of the pipe were obtained from the surface information. Using the k-means method, it was classified into four states, and the standard of the average height and maximum height of corrosion for each state was derived.

• Journal of the Korea Safety Management & Science
• /
• v.24 no.4
• /
• pp.61-70
• /
• 2022

In spite of lab safety act for over 10 years, over 100 safety accidents in the laboratory have been constantly occurring. The ideal safety management system is to prevent accidents by differential classifying and managing laboratory regulatory materials according to the risk level. In order to approach this system, in-depth interviews with safety managers were first conducted to identify the current status of safety management in domestic university laboratories. And then through comparative analysis of safety management systems in domestic and foreign laboratories, a new regulatory substance classification standard based on the analysis of the hazards and the classification of risk grades, and a safety management system are proposed. From this study, it will contribute to the creation of a safe laboratory environment by differential classification and management laboratory regulatory materials based on the risk level.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2022.07a
• /
• pp.61-62
• /
• 2022

알츠하이머성 치매는 현존하는 치료법이 없어 경도인지장애 단계에서의 예방이 중요하다. 지금까지의 알츠하이머 연구는 대부분이 뇌영상 마커와 뇌척수액 마커에 집중되어 있었으며, 경도 인지 장애 단계에서의 탐색은 더욱 적었다. 이러한 점에서 혈액 유전자 발현을 이용한 경도 인지장애 단계 예측은 인지 능력에 따른 관련 유전자 식별과 접근 가능한 진단 및 치료 바이오 마커 탐색에 기여할 수 있다. 그러나 유전자 발현 데이터의 경우 환자 수에 비해 높은 차원을 가지기 때문에 과적합을 막고 질병 관련 유전자를 식별하기 위해서는 데이터에서의 의미 있는 차원만을 뽑아내는 차원 축소가 선행되야 한다. 본 연구는 유전자 발현데이터에서의 인지장애 분류를 위해 차원 축소기법과 신경망을 적용하여 인지 장애 정도를 예측하였다. 그 결과, Lasso 이용 차원축소와 신경망을 이용하여 97%의 정확도로 정상과 조기 경도 인지장애, 후기 경도 인지장애 환자를 분류 할 수 있었으며, 더 적은 차원에서도 분류가 가능했다. 이는 혈액 유전자 발현을 이용해 경도 인지장애 단계를 예측한 첫 번째 연구이며, 인지능력 저하에 따른 혈액 유전자 발현의 연관성을 확인하고 향후 조기 진단, 치료 표적 탐색에 기여한다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2024.05a
• /
• pp.749-752
• /
• 2024

연합학습은 클라이언트가 중앙 서버에 원본 데이터를 주지 않고도 학습할 수 있도록 설계된 분산된 머신러닝 방법이다. 그러나 클라이언트와 중앙 서버 사이에 모델 업데이트 정보를 공유한다는 점에서 여전히 추론 공격(Inference Attack)과 오염 공격(Poisoning Attack)의 위험에 노출되어 있다. 이러한 공격을 방어하기 위해 연합학습에 차분프라이버시(Differential Privacy)를 적용하는 방안이 연구되고 있다. 차분 프라이버시는 데이터에 노이즈를 추가하여 민감한 정보를 보호하면서도 유의미한 통계적 정보 쿼리는 공유할 수 있도록 하는 기법으로, 노이즈를 추가하는 위치에 따라 전역적 차분프라이버시(Global Differential Privacy)와 국소적 차분 프라이버시(Local Differential Privacy)로 나뉜다. 이에 본 논문에서는 차분 프라이버시를 적용한 연합학습의 최신 연구 동향을 전역적 차분 프라이버시를 적용한 방향과 국소적 차분 프라이버시를 적용한 방향으로 나누어 검토한다. 또한 이를 세분화하여 차분 프라이버시를 발전시킨 방식인 적응형 차분 프라이버시(Adaptive Differential Privacy)와 개인화된 차분 프라이버시(Personalized Differential Privacy)를 응용하여 연합학습에 적용한 방식들에 대하여 특징과 장점 및 한계점을 분석하고 향후 연구방향을 제안한다.

• Journal of the Korean Society for Heat Treatment
• /
• v.37 no.2
• /
• pp.58-65
• /
• 2024

In the present study, the effect of annealing condition on the microstructures and mechanical properties of the cold-rolled CoCrFeMnNi high entropy alloys were studied. Annealing treatment was performed under six different temperatures. Microstructural analyses confirmed that annealing below 800℃ resulted in the formation of intermetallic sigma (σ) phase within face-centered cubic (FCC) matrix, and this σ phase has beneficial effects on the formation of fine-grained structures through retardation of grain growth and recrystallization due to Zener pinning effect. This led to the enhanced yield strength and tensile strength of ~646 and ~855 MPa, respectively. The microstructures annealed above 800℃ demonstrated single FCC phase, and fully-recrystallized single FCC microstructure resulted in a slight increase in ductility with a considerable decrease in strength. The evolution of mechanical properties, such as strength, ductility, and strain hardening exponent, will be discussed.