• Geomechanics and Engineering
• /
• v.37 no.3
• /
• pp.223-241
• /
• 2024

Evaluating the performance of Tunnel Boring Machines (TBMs) stands as a pivotal juncture in the domain of hard rock mechanized tunneling, essential for achieving both a dependable construction timeline and utilization rate. In this investigation, three advanced artificial neural networks namely, gated recurrent unit (GRU), back propagation neural network (BPNN), and simple recurrent neural network (SRNN) were crafted to prognosticate TBM-rate of penetration (ROP). Drawing from a dataset comprising 1125 data points amassed during the construction of the Alborze Service Tunnel, the study commenced. Initially, five geomechanical parameters were scrutinized for their impact on TBM-ROP efficiency. Subsequent statistical analyses narrowed down the effective parameters to three, including uniaxial compressive strength (UCS), peak slope index (PSI), and Brazilian tensile strength (BTS). Among the methodologies employed, GRU emerged as the most robust model, demonstrating exceptional predictive prowess for TBM-ROP with staggering accuracy metrics on the testing subset (R² = 0.87, NRMSE = 6.76E-04, MAD = 2.85E-05). The proposed models present viable solutions for analogous ground and TBM tunneling scenarios, particularly beneficial in routes predominantly composed of volcanic and sedimentary rock formations. Leveraging forecasted parameters holds the promise of enhancing both machine efficiency and construction safety within TBM tunneling endeavors.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.22 no.6
• /
• pp.1-16
• /
• 2023

Accurate traffic information prediction is considered to be one of the most important aspects of intelligent transport systems(ITS), as it can be used to guide users of transportation facilities to avoid congested routes. Various deep learning models have been developed for accurate traffic prediction. Recently, ensemble techniques have been utilized to combine the strengths and weaknesses of various models in various ways to improve prediction accuracy and stability. Therefore, in this study, we developed and evaluated a traffic information prediction model using various deep learning models, and evaluated the performance of the developed deep learning models as a stacking ensemble. The individual models showed error rates within 10% for traffic volume prediction and 3% for speed prediction. The ensemble model showed higher accuracy compared to other models when no cross-validation was performed, and when cross-validation was performed, it showed a uniform error rate in long-term forecasting.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.22 no.6
• /
• pp.49-65
• /
• 2023

In response to the growing global concern for the environment, the international community has recently committed to achieving 'carbon neutrality.' As a result, numerous studies have been conducted on mode choice models that include carbon emissions as a variable. However, few studies have established a correlation between individual preferences and carbon emissions. In this study, a new mode of transportation named sustainable public transit (SPT), incorporating carbon-reducing transport options like electric scooters, is proposed. Analyzing the individual preferences of commuters on carbon emissions through factor analysis, a stated preference (SP) survey was conducted. A mode choice model for SPT was constructed using multinomial logit models. The results of the analysis showed that gender, income, and specific preferences, such as a passion for exploring new routes, a preference for intermodal transfers, knowledge of carbon reduction, and carbon reduction practices, significantly influence latent preferences for SPT. Therefore, this study is significant as it considers carbon emissions as an attribute variable during the construction of mode choice models and reflects the individual preference variables associated with carbon reduction.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.22 no.6
• /
• pp.157-167
• /
• 2023

The use of big data for transportation often involves using data that includes personal information, such as the driver's driving routes and coordinates. This study explores the creation of a route choice prediction model using a large dataset from mobile navigation apps using federated learning. This privacy-focused method used distributed computing and individual device usage. This study established preprocessing and analysis methods for driver data that can be used in route choice modeling and compared the performance and characteristics of widely used learning methods with federated learning methods. The performance of the model through federated learning did not show significantly superior results compared to previous models, but there was no substantial difference in the prediction accuracy. In conclusion, federated learning-based prediction models can be utilized appropriately in areas sensitive to privacy without requiring relatively high predictive accuracy, such as a driver's preferred route choice.

• The Korean Journal of Pain
• /
• v.37 no.1
• /
• pp.41-50
• /
• 2024

Background: Recognizing the seriousness of the misuse and abuse of medical narcotics, the South Korean government introduced the world's first narcotic management system, the Narcotics Information Management System (NIMS). This study aimed to explore the recent one-year opioid prescribing patterns in South Korea using the NIMS database. Methods: This study analyzed opioid prescription records in South Korea for the year 2022, utilizing the dispensing/administration dataset provided by NIMS. Public data from the Korean Statistical Information Service were also utilized to explore prescription trends over the past four years. The examination covered 16 different opioid analgesics, assessed by the total number of units prescribed based on routes of administration, type of institutions, and patients' sex and age group. Additionally, the disposal rate for each ingredient was computed. Results: In total, 206,941 records of 87,792,968 opioid analgesic units were analyzed. Recently, the overall quantity of prescribed opioid analgesic units has remained relatively stable. The most prescribed ingredient was oral oxycodone, followed by tapentadol and sublingual fentanyl. Tertiary hospitals had the highest number of dispensed units (49.4%), followed by community pharmacies (40.2%). The highest number of prescribed units was attributed to male patients in their 60s. The disposal rates of the oral and transdermal formulations were less than 0.1%. Conclusions: Opioid prescription in South Korea features a high proportion of oral formulations, tertiary hospital administration, pharmacy dispensing, and elderly patients. Sustained education and surveillance of patients and healthcare providers is required.

• Journal of Aerospace System Engineering
• /
• v.18 no.2
• /
• pp.40-46
• /
• 2024

This study suggests a method to enhance drone flight path planning and safety evaluation in urban areas using Computational Fluid Dynamics (CFD). As the use of drones in urban environments has been growing rapidly, there is a lack of established methods for path planning and safety evaluation, which leads to a risky approach relying on experimental methods. Therefore, this research takes into account the intricate 3D fluid dynamics between drones and buildings by employing CFD to quantitatively plan flight paths and evaluate their safety. To accomplish this, the study focuses on Gimcheon Innovation City as the target area and collects relevant terrain and building data, and selects prospective flight routes. CFD analysis is then carried out to gather essential data for flight simulations and safety assessment. The safety assessments are conducted based on environmental fluid dynamics when the drone operates along the proposed flight paths

• Safety and Health at Work
• /
• v.14 no.4
• /
• pp.347-357
• /
• 2023

Introduction: The COVID-19 pandemic turned biological hazards in the working environment into a global concern. This systematized review of published reviews aimed to provide a comprehensive overview of the specific jobs and categories of workers exposed to biological hazards with the related prevention. Methods: We extracted reviews published in English and French in PubMed, Embase, and Web of Science. Two authors, working independently, subsequently screened the potentially relevant titles and abstracts recovered (step 1) and then examined relevant full texts (step 2). Disagreements were resolved by consensus. We built tables summarizing populations of exposed workers, types of hazards, types of outcomes (types of health issues, means of prevention), and routes of transmission. Results: Of 1426 studies initially identified, 79 studies by authors from every continent were selected, mostly published after 2010 (n = 63, 79.7%). About half of the reviews dealt with infectious hazards alone (n = 38, 48.1%). The industrial sectors identified involved healthcare alone (n = 16), laboratories (n = 10), agriculture (including the animal, vegetable, and grain sectors, n = 32), waste (n = 10), in addition of 11 studies without specific sectors. The results also highlighted a range of hazards (infectious and noninfectious agents, endotoxins, bioaerosols, organic dust, and emerging agents). Conclusion: This systematized overview allowed to list the populations of workers exposed to biological hazards and underlined how prevention measures in the healthcare and laboratory sectors were usually well defined and controlled, although this was not the case in the agriculture and waste sectors. Further studies are necessary to quantify these risks and implement prevention measures that can be applied in every country.

• Journal of Ginseng Research
• /
• v.47 no.5
• /
• pp.638-644
• /
• 2023

Background: The anti-platelet activity of the saponin fraction of Korean Red Ginseng has been widely studied. The saponin fraction consists of the panaxadiol fraction (PDF) and panaxatriol fraction (PTF); however, their anti-platelet activity is yet to be compared. Our study aimed to investigate the potency of anti-platelet activity of PDF and PTF and to elucidate how well they retain their anti-platelet activity via different administration routes. Methods: For ex vivo studies, Sprague-Dawley rats were orally administered 250 mg/kg PDF and PTF for 7 consecutive days before blood collection via cardiac puncture. Platelet aggregation was conducted after isolation of the washed platelets. For in vitro studies, washed platelets were obtained from Sprague-Dawley rats. Collagen and adenosine diphosphate (ADP) were used to induce platelet aggregation. Collagen was used as an agonist for assaying adenosine triphosphate release, thromboxane B2, serotonin, cyclic adenosine monophosphate, and cyclic guanosine monophosphate (cGMP) release. Results: When treated ex vivo, PDF not only inhibited ADP and collagen-induced platelet aggregation, but also upregulated cGMP levels and reduced platelet adhesion to fibronectin. Furthermore, it also inhibited Akt phosphorylation induced by collagen treatment. Panaxadiol fraction did not exert any antiplatelet activity in vitro, whereas PTF exhibited potent anti-platelet activity, inhibiting ADP, collagen, and thrombin-induced platelet aggregation, but significantly elevated levels of cGMP. Conclusion: Our study showed that in vitro and ex vivo PDF and PTF treatments exhibited different potency levels, indicating possible metabolic conversions of ginsenosides, which altered the content of ginsenosides capable of preventing platelet aggregation.

• Journal of the Korea Computer Graphics Society
• /
• v.30 no.2
• /
• pp.21-29
• /
• 2024

Although there is an increasing demand for capturing various natural motions, collecting climbing motion data is difficult due to technical complexities, related to obscured markers. Additionally, scanning climbing structures and preparing diverse routes further complicate the collection of necessary data. To tackle this challenge, this paper proposes a climbing motion synthesis using reinforcement learning. The method comprises two learning stages. Firstly, the hanging policy is trained to grasp holds in a natural posture. Once the policy is obtained, it is used to extract the positions of the holds, postures, and gripping states, thus forming a dataset of favorable initial poses. Subsequently, the climbing policy is trained to execute actual climbing maneuvers using this initial state dataset. The climbing policy allows the character to move to the target location using limbs more evenly in a natural posture. Experiments have shown that the proposed method can effectively explore the space of good postures for climbing and use limbs more evenly. Experimental results demonstrate the effectiveness of the proposed method in exploring optimal climbing postures and promoting balanced limb utilization.

• Journal of Environmental Health Sciences
• /
• v.50 no.1
• /
• pp.43-53
• /
• 2024

Background: Exposure assessment is an important part of risk assessment for consumer products. Exposure models are used when estimating consumer exposures by considering exposure routes, subjects, and circumstances. These models differ based on their tiers, types, and target populations. Consequently, exposure estimates may vary between models. Objectives: This study aimed to compare the results of different exposure models using identical exposure factors. Methods: Chemical exposure from consumer products was calculated using four consumer exposure assessment models: Targeted Risk Assessment 3.1, Consumer Exposure Model 2.1 (CEM), ConsExpo web 1.1.1, and the Korean Exposure Algorithm (primary and detailed) issued by the Ministry of Environment, No. 972 (MOE). The same exposure factors were used in each model to calculate inhalation and dermal exposures to acetaldehyde, d-limonene, and naphthalene in all-purpose cleaners, leather coating sprays, and sealants. Results: In the results, TRA provided the highest estimate. Generally, MOE (detailed), CEM and ConsExpo showed lower exposures. The inhalation exposure for leather coating spray showed the largest differences between models, with differences reaching up to 1.2×10⁷ times. Since identical inputs were used for the calculations, it is likely that the models significantly influenced the estimated results. Conclusions: Despite using the same exposure factors to calculate dermal and inhalation exposures, the results varied substantially based on the model's exposure algorithm. Therefore, selecting an exposure model for assessing consumer products should be done with careful consideration.