• Journal of Intelligence and Information Systems
• /
• v.29 no.4
• /
• pp.129-164
• /
• 2023

In a situation where the use and introduction of Large Language Models (LLMs) is expanding due to recent developments in generative AI technology, it is difficult to find actual application cases or implementation methods for the use of internal company data in existing studies. Accordingly, this study presents a method of implementing generative AI services using the LLM application architecture using the most widely used LangChain framework. To this end, we reviewed various ways to overcome the problem of lack of information, focusing on the use of LLM, and presented specific solutions. To this end, we analyze methods of fine-tuning or direct use of document information and look in detail at the main steps of information storage and retrieval methods using the retrieval augmented generation (RAG) model to solve these problems. In particular, similar context recommendation and Question-Answering (QA) systems were utilized as a method to store and search information in a vector store using the RAG model. In addition, the specific operation method, major implementation steps and cases, including implementation source and user interface were presented to enhance understanding of generative AI technology. This has meaning and value in enabling LLM to be actively utilized in implementing services within companies.

• IEMEK Journal of Embedded Systems and Applications
• /
• v.18 no.3
• /
• pp.89-100
• /
• 2023

Recently, on-device artificial intelligence (AI) solutions using mobile devices and embedded edge devices have emerged in various fields, such as computer vision, to address network traffic burdens, low-energy operations, and security problems. Although vision transformer deep learning models have outperformed conventional convolutional neural network (CNN) models in computer vision, they require more computations and parameters than CNN models. Thus, they are not directly applicable to embedded edge devices with limited hardware resources. Many researchers have proposed various model compression methods or lightweight architectures for vision transformers; however, there are only a few studies evaluating the effects of model compression techniques of vision transformers on performance. Regarding this problem, this paper presents a performance evaluation of vision transformers on embedded platforms. We investigated the behaviors of three vision transformers: DeiT, LeViT, and MobileViT. Each model performance was evaluated by accuracy and inference time on edge devices using the ImageNet dataset. We assessed the effects of the quantization method applied to the models on latency enhancement and accuracy degradation by profiling the proportion of response time occupied by major operations. In addition, we evaluated the performance of each model on GPU and EdgeTPU-based edge devices. In our experimental results, LeViT showed the best performance in CPU-based edge devices, and DeiT-small showed the highest performance improvement in GPU-based edge devices. In addition, only MobileViT models showed performance improvement on EdgeTPU. Summarizing the analysis results through profiling, the degree of performance improvement of each vision transformer model was highly dependent on the proportion of parts that could be optimized in the target edge device. In summary, to apply vision transformers to on-device AI solutions, either proper operation composition and optimizations specific to target edge devices must be considered.

• Journal of the Korean Geosynthetics Society
• /
• v.23 no.2
• /
• pp.43-52
• /
• 2024

Recent abnormal climate conditions have increased the risk of slope collapses, which frequently result in significant loss of life and property due to the absence of early prediction and warning dissemination. In this paper, we develop a slope condition analysis system using IoT sensors and AI-based camera to assess the condition of slopes. To develop the system, we conducted hardware and firmware design for measurement sensors considering the ground conditions of slopes, designed AI-based image analysis algorithms, and developed prediction and warning solutions and systems. We aimed to minimize errors in sensor data through the integration of IoT sensor data and AI camera image analysis, ultimately enhancing the reliability of the data. Additionally, we evaluated the accuracy (reliability) by applying it to actual slopes. As a result, sensor measurement errors were maintained within 0.1°, and the data transmission rate exceeded 95%. Moreover, the AI-based image analysis system demonstrated nighttime partial recognition rates of over 99%, indicating excellent performance even in low-light conditions. Through this research, it is anticipated that the analysis of slope conditions and smart maintenance management in various fields of Social Overhead Capital (SOC) facilities can be applied.

• International Journal of Naval Architecture and Ocean Engineering
• /
• v.11 no.2
• /
• pp.952-969
• /
• 2019

This study develops new analytical solutions to water wave diffraction by vertical truncated cylinders in the context of linear potential theory. Three typical truncated surface-piercing cylinders, a submerged bottom-standing cylinder and a submerged floating cylinder are examined. The analytical solutions utilize the multi-term Galerkin method, which is able to model the cube-root singularity of fluid velocity near the edges of the truncated cylinders by expanding the fluid velocity into a set of basis function involving the Gegenbauer polynomials. The convergence of the present analytical solution is rapid, and a few truncated numbers in the series of the basis function can yield results of six-figure accuracy for wave forces and moments. The present solutions are in good agreement with those by a higher-order BEM (boundary element method) model. Comparisons between present results and experimental results in literature and results by Froude-Krylov theory are conducted. The variation of wave forces and moments with different parameters are presented. This study not only gives a new analytical approach to wave diffraction by truncated cylinders but also provides a reliable benchmark for numerical investigations of wave diffraction by structures.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.24 no.2
• /
• pp.23-28
• /
• 2024

This study focuses on investigating and exploring solutions to problems related to the increase in the elderly population in modern society where aging is rapidly progressing. With the advent of an aging society the elderly population requires more attention and support. In response companies are attempting to provide elder care services using AI and robotic technology. These services can assist seniors not only with health management and daily life care but also with emotional health aspects. This paper analyzes various elder-related technologies available in the current market investigates their pros and cons and potential for development. The paper concludes that companies need to develop and provide more AI and robot-based elder care services to solve aging problems. Such services can alleviate social and emotional issues in an aging society enhancing the quality of life for seniors.

• Information Systems Review
• /
• v.25 no.2
• /
• pp.31-56
• /
• 2023

As the COVID-19 pandemic is prolonged, non-contact work has increased, as well as the demand for automation of simple and repetitive questions and tasks with success of using them. Therefore, companies are attempting to expand the area of automated business and apply various technologies such as AI to complex and various business processes of E2E to provide automation of all business. However, the extension to Intelligent Process Automation (IPA) is still in its beginning stage so that it is difficult to find practical use cases and related solutions. In this aspect, it is safe to say that there is insufficient evidence for companies which have various and complex enterprise processes to make a decision about the adoption. In this study, to solve this problem, a Hyper Automation Platform (HAP) that consists of RPA, Chatbot, and AI technology was proposed. Moreover, an implementation method that can bring intelligent process automation using HAP, and practical use-cases were provided so that it makes it possible to review the implementation of the HAP objectively and comprehensively. This study is meaningful and valuable to check the feasibility of the Hyper Automation concept and to actively utilize HAP.

• Electronics and Telecommunications Trends
• /
• v.33 no.1
• /
• pp.11-19
• /
• 2018

IoT is used not only as a technical terminology but also as a paradigm representation. As the number of IoT devices spread tremendously throughout the world, they are able to be located anywhere,recognize their environment, and achieve adaptable reactions. All market investigation agencies expect the number of IoT devices to reach tens to hundreds of billions in number. They also expect various technical problems owing to the huge number of connected things and data that will emerge during the AI era. The decentralization of centralized computing for AI is the one of the technical solutions to such problems, and the computing roles for AI will be soon distributed into the things, which can be located anywhere. In this article, the traditional distributed intelligence and its current research activities are introduced, and the next distributed intelligence target for the IoT 2.0 era is briefly touched upon using the keyword Socio-Things.

• Journal of the Korean Chemical Society
• /
• v.45 no.2
• /
• pp.156-161
• /
• 2001

Various oxidation states of Polyaniline(PANI) were chemically synthesized, and characterized by FT-IR, UV-Vis, GPC, TG-DTA, Single layer light emitting diodes(LED) were perpared by spin coating of LEB-PANI solutions which have various oxidation states onto an ITO substrate and subsequent vacuum deposition of aluminum top electrode and then current-voltage characteristics. EL spectrum was investigated It was found that ${\pi}$-${\pi}$* transition were shifled to longer wavelength and molecular excition transition were decreased in the UV-Vis spectra and the intensity of EL and PL were increased as the contents of fully reduced form LEB increased. The turn-on voltage of ITO/LEB/AI structured LED was 5 V. It was found that the white light was emitted only from the phase with reduced epeat unit.

• Journal of the Korean Society for Aviation and Aeronautics
• /
• v.32 no.1
• /
• pp.39-48
• /
• 2024

Aviation industry is moving towards the third innovative era of AAM with electric power and AI after the JET-powered era following the Wright brothers' first flights. Research on UAM, eVTOL development, certification, and operations is competitively progressing, primarily in aviation-leading countries, aiming to resolve urban traffic saturation and foster the future aviation industries. This study introduces the concept of the pilot's role transition in operational safety as AI autonomous flight advances, comparing K-UAM operational concept with research from FAA, NASA, and EASA. It is to identify and propose solutions for challenges from the pilot's perspective in developing UAM and its safe operation system. To succeed in Advanced Air Mobility National Project, we suggest the collaboration among industry, academia, and institutions, along with the cooperation between civilians, governments, military, and the need for Urban Air Mobility integrated policies.

• Journal of information and communication convergence engineering
• /
• v.17 no.3
• /
• pp.198-204
• /
• 2019

In this study, the author introduces a supply-chain management (SCM) solution that connects suppliers, manufacturers, customers, and other companies within a transactional relationship to enable efficient inventory management and timely product supply, which ultimately maximizes corporate profits. This proposed solution exploits Fourth Industrial Revolution technologies, such as artificial intelligence (AI) and the Internet of Things (IoT), which provide solutions to complex management issues generated by the broader market. The goal of the current study was to develop an advanced and intelligent smart SCM solution that complies with logistics standards, to enhance the visibility, safety, and efficiency of a supply chain made up of manufacturers and suppliers. This smart SCM solution aims at maximizing corporate profits through efficient inventory management and timely supply of products, and solves the complex management problems caused by operating within a wide range of markets.