• Journal of the Korea Safety Management & Science
• /
• v.24 no.4
• /
• pp.1-8
• /
• 2022

Elevators are the main means of transport in buildings. A malfunction of an elevator in operation may cause in convenience to users. Furthermore, fatal accidents, such as injuries and death, may occur to the passengers also. Therefore, it is important to prevent failure before accidents happen. In related studies, preventive measures are proposed through analyzing failures, and the lifespan of elevator components. However, these methods are limited to existing an elevator model and its surroundings, including operating conditions and installed environments. Vibration occurs when the elevator is operated. Experts have classified types of faults, which are symptoms for malfunctions (failures), via analyzing vibration. This study proposes an artificial intelligent model for classifying faults automatically with deep learning algorithms through elevator vibration data, hereby preventing failures before they occur. In this study, the vibration data of six elevators are collected. The proposed methodology in this paper removes "the measurement error data" with incorrect measurements and extracts operating sections from the input datasets for proceeding deep learning models. As a result of comparing the performance of training five deep learning models, the maximum performance indicates Accuracy 97% and F1 Score 97%, respectively. This paper presents an artificial intelligent model for detecting elevator fault automatically. The users' safety and convenience may increase by detecting fault prior to the fatal malfunctions. In addition, it is possible to reduce manpower and time by assisting experts who have previously classified faults.

• Nuclear Engineering and Technology
• /
• v.55 no.2
• /
• pp.506-515
• /
• 2023

This study is to find a method to learn engine sound after the start-up of a diesel generator installed in nuclear power plant with an unsupervised deep learning algorithm (CNN autoencoder) and a new method to predict the failure of a diesel generator using it. In order to learn the sound of a diesel generator with a deep learning algorithm, sound data recorded before and after the start-up of two diesel generators was used. The sound data of 20 min and 2 h were cut into 7 s, and the split sound was converted into a spectrogram image. 1200 and 7200 spectrogram images were created from sound data of 20 min and 2 h, respectively. Using two different deep learning algorithms (CNN autoencoder and binary classification), it was investigated whether the diesel generator post-start sounds were learned as normal. It was possible to accurately determine the post-start sounds as normal and the pre-start sounds as abnormal. It was also confirmed that the deep learning algorithm could detect the virtual abnormal sounds created by mixing the unusual sounds with the post-start sounds. This study showed that the unsupervised anomaly detection algorithm has a good accuracy increased about 3% with comparing to the binary classification algorithm.

• Korean Journal of Construction Engineering and Management
• /
• v.9 no.6
• /
• pp.257-267
• /
• 2008

In order to encourage construction practitioners to acknowledge failures and disseminate the information, the failure information must be documented and accumulated with a well-structured format, which contains not only the fact and result but also the circumstance and cause of the failure. In the Korean construction industry, many failures are not explained clearly and often not even reported publicly, partly because due to the lack of understanding positive aspects of failures, which can improve construction practices as a result of learning from failures. The purpose of this study is to develop a web-based construction failure information system using the case-based reasoning techniques, which can systematically accumulate, manage, and share the valuable failure information using a structured failure cases database. It can be utilized for planning proactive solutions on future failures by searching the very similar past failure cases.

• Structural Engineering and Mechanics
• /
• v.78 no.2
• /
• pp.209-218
• /
• 2021

For engineering, there are two major challenges in reliability analysis. First, to ensure the accuracy of simulation results, mechanical products are usually defined implicitly by complex numerical models that require time-consuming. Second, the mechanical products are fortunately designed with a large safety margin, which leads to a low failure probability. This paper proposes an efficient and high-precision adaptive active learning algorithm based on the Kriging surrogate model to deal with the problems with low failure probability and time-consuming numerical models. In order to solve the problem with multiple failure regions, the adaptive kernel-density estimation is introduced and improved. Meanwhile, a new criterion for selecting points based on the current Kriging model is proposed to improve the computational efficiency. The criterion for choosing the best sampling points considers not only the probability of misjudging the sign of the response value at a point by the Kriging model but also the distribution information at that point. In order to prevent the distance between the selected training points from too close, the correlation between training points is limited to avoid information redundancy and improve the computation efficiency of the algorithm. Finally, the efficiency and accuracy of the proposed method are verified compared with other algorithms through two academic examples and one engineering application.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.17 no.8
• /
• pp.2016-2029
• /
• 2023

Accurate prediction of critical illness is significant for ensuring the lives and health of patients. The selection of indicators affects the real-time capability and accuracy of the prediction for critical illness. However, the diversity and complexity of these indicators make it difficult to find potential connections between them and critical illnesses. For the first time, this study proposes an indicator analysis model to extract key indicators from the preoperative and intraoperative clinical indicators and laboratory results of critical illnesses. In this study, preoperative and intraoperative data of heart failure and respiratory failure are used to verify the model. The proposed model processes the datum and extracts key indicators through four parts. To test the effectiveness of the proposed model, the key indicators are used to predict the two critical illnesses. The classifiers used in the prediction are light gradient boosting machine (LightGBM) and eXtreme Gradient Boosting (XGBoost). The predictive performance using key indicators is better than that using all indicators. In the prediction of heart failure, LightGBM and XGBoost have sensitivities of 0.889 and 0.892, and specificities of 0.939 and 0.937, respectively. For respiratory failure, LightGBM and XGBoost have sensitivities of 0.709 and 0.689, and specificity of 0.936 and 0.940, respectively. The proposed model can effectively analyze the correlation between indicators and postoperative critical illness. The analytical results make it possible to find the key indicators for postoperative critical illnesses. This model is meaningful to assist doctors in extracting key indicators in time and improving the reliability and efficiency of prediction.

• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.30B no.11
• /
• pp.74-83
• /
• 1993

This paper presents methodology which enables the derivation of goal ordering rules from the analysis of problem failures. We examine all the possible ways of taking actions that lead to failures. If there are restrictions imposed by a problem state on possible actions to be taken, the restrictions manifest themselves in the form of a restricted set of possible operator bindings. Our method makes use of this observation to derive general control rules which are guaranteed to be correct. The overhead involved in learning is very low because this methodology needs only small amount of data to learn from namely, the goal stacks from the leaf nodes of a failure search tree. Empirical tests show that the rules derived by our system PAL couperform those derived by other systems such as PRODIGY and STATIC.

• Proceedings of the Korea Contents Association Conference
• /
• 2005.05a
• /
• pp.36-50
• /
• 2005

e-Learning has been mushrooming with wide range of learning groups from pedagogy to andragogy. Despite of increasing e-learning opportunities, many people doubt whether e-learning learners really learn something. The related research papers emphasized that e-Learning would be a failure in terms of understanding of e-learners and intuitive learning activities for activating learner's long-term memory span. The current learning strategies in e-Learning may be based on the traditional classroom, and this results in boring and ineffective learning outcomes. This paper analyzed that how learners have received e-Learning for the last few years from the research and explained what could be the failing aspects of e-Learning. To be successful, e-Learning should consider the e-Learner's individualized teaming style and thinking patterns. When considering of various e-Learning components, the quality of e-learning should not be focused on any specific single factor, but develop every individual factor to the high level of quality. In conclusion, this paper suggest that we need new understand of e-Learning and e-Learner. Also the e-Learning strategies should be examined throughly whether they are on the side of learners and realized how they learn from e-Learning. Finally, we should add enormous imagination into e-Learning for next generation because their teaming patterns significantly differ from their parent's generation.

• Journal of the Korea Society of Computer and Information
• /
• v.25 no.4
• /
• pp.19-27
• /
• 2020

In this paper, we propose a machine learning method for diagnosing the failure of a gas pressure regulator. Originally, when implementing a machine learning model for detecting abnormal operation of a facility, it is common to install sensors to collect data. However, failure of a gas pressure regulator can lead to fatal safety problems, so that installing an additional sensor on a gas pressure regulator is not simple. In this paper, we propose various machine learning approach for diagnosing the abnormal operation of a gas pressure regulator with only the flow rate and gas pressure data collected from a gas pressure regulator itself. Since the fault data of a gas pressure regulator is not enough, the model is trained in all classes by applying the over-sampling method. The classification model was implemented using Gradient boosting, 1D Convolutional Neural Networks, and LSTM algorithm, and gradient boosting model showed the best performance among classification models with 99.975% accuracy.

• Journal of Technology Innovation
• /
• v.29 no.2
• /
• pp.1-30
• /
• 2021

Recent years have witnessed various attempts of firms to acquire new knowledge. Purchasing intellectual property or merger and acquisition (M&A) can be such attempts, but joint venture can also be an effective way internalizing new complementary assets from external partners. However, due to difficulties in the formation and implementation of learning strategies, many joint ventures have failed to acquire necessary knowledge. In this respect, based on contingency theory and dynamic capability, the current research aims to investigate the failure case of knowledge transfer in an international joint venture - KEFICO established by Hyundai motors and BOSCH. Case firm optimized for hardware technology but did not establish a differentiated learning strategy and organizational structure to acquire software skills, which are intellectuals of different natures. Due to this inconsistency, it was not able for KEFICO to absorb new type of knowledge (skills related to engine control system). This study suggests the theoretical framework illustrating the case and provides some important implications for organizational learning.

• International journal of thyroidology
• /
• v.11 no.2
• /
• pp.130-136
• /
• 2018

Background and Objectives: Intraoperative neural monitoring (IONM) of recurrent laryngeal nerve (RLN) in thyroid surgery has been employed worldwide to identify and preserve the nerve as an adjunct to visual identification. The aims of this study was to evaluate the efficacy of IONM and difficulties in the learning curve. Materials and Methods: We studied 63 patients who underwent thyroidectomy with IONM during last 2 years. The standard IONM procedure was performed using NIM 3.0 or C2 Nerve Monitoring System. Patients were divided into two chronological groups based on the success rate of IONM (33 cases in the early period and 30 cases in the late period), and the outcomes were compared between the two groups. Results: Of 63 patients, 32 underwent total thyroidectomy and 31 thyroid lobectomy. Failure of IONM occurred in 9 cases: 8 cases in the early period and 1 case in the late period. Loss of signal occurred in 8 nerves of 82 nerves at risk. The positive predictive value increased from 16.7% in the early period to 50% in the late period. The mean amplitude of the late period was higher than that of the early period (p<0.001). Conclusion: IONM in thyroid surgery is effective to preserve the RLN and to predict postoperative nerve function. However, failure of IONM and high false positive rate can occur in the learning curve, and the learning curve was about 30 cases based on the results of this study.