Korean Journal of Construction Engineering and Management

ISO Journal Title : Korean J. Constr. Eng. Manag.
Open Access Journal Bimonthly
  • ISSN (Print) : 2005-6095
  • ISSN (Online) : 2465-9703

An Examination of Reporting and Recording Practices in Construction Safety Management: Analyzing Improvements under the Construction Technology Promotion Act

Jaehoon Lee ; Moonseo Park ; Changbum Ryan Ahn

https://dx.doi.org/10.6106/KJCEM.2025.26.5.003

With the increasing frequency and severity of accidents in the construction industry, the importance of effective safety management is being emphasized more than ever. A major issue in the current safety management practice of construction industry is the excessive documentation required at construction sites. This excessive documentation leads safety managers to neglect their primary duties of overseeing on-site safety. Previous studies have not adequately addressed the real reasons behind safety practitioners’ neglect of their duties, nor have they provided tangible solutions. To address these gaps, this study focuses on problems of safety management documentation practices at construction sites, identify the main causes of inefficiency. Field surveys targeting safety practitioners were conducted to gather opinions on the redundancy, necessity, and costs associated with the current documentation practice. Based on these insights, the study diagnosed the problems of the existing practice and identified documents in need of improvement. This study provides a detailed diagnosis of the current safety management documentation practices by reflecting the opinions of construction site documentation personnel.

A Study of Multi-Artificial Intelligence Agent-Based Risk Assessment in Construction Sites - A Qualitative Approach to Practical Competency -

Byunghee Yoo ; Sungwon Ahn ; Chanbum Ryan Ahn

https://dx.doi.org/10.6106/KJCEM.2025.26.5.013

Risk assessment in construction projects is a critical process for preventing accidents and ensuring worker safety. However, traditional risk assessment methods predominantly rely on the expertise and intuition of safety managers, leading to variability in assessments due to differences in individual experience and judgment. Additionally, newly implemented safety regulations have increased the demand for structured and objective risk evaluation frameworks, further highlighting the limitations of conventional approaches. In response, this study proposes a Multi-Artificial Intelligence (AI) Agent-based System that integrates multimodal data by combining textual and visual inputs to systematically identify hazards, infer potential risk scenarios, assess severity and frequency, and propose mitigation strategies. The system employs multiple specialized AI agents to conduct hazard identification, risk assessment, and develop mitigation measures, thereby reducing dependency on human expertise while enhancing consistency and comprehensiveness. The results show that the AI agents performed comparably to safety managers with over 20 years of experience in risk identification and inference, surpassing them in the number of identified risk factors. However, variability was observed in the proposed mitigation strategies and the overall validity of risk assessments, indicating areas for further refinement. These findings suggest that AI-driven risk assessment systems can serve as valuable decision-support tools, particularly for less experienced safety managers, while complementing expert judgment in ensuring construction site safety.

Comparative Study on the Performance of Multi-view Image Learning-based Penetrability Analysis Model for Classifying Irrelevant Clashes in BIM model

Hyunwoo Lee ; Youngsu Yu ; Wonbok Lee ; Bonsang Koo

https://dx.doi.org/10.6106/KJCEM.2025.26.5.026

BIM models are developed in a fragmented manner by discipline and integrated during the detailed design phase, leading to numerous clashes. Many of these are irrelevant clashes that do not require intervention, yet significant effort is needed for classification. Sleeve installation feasibility assessment is a key task in this process, and various studies have explored automated penetrability analysis. However, existing methods relied on manually defined inference rules or failed to capture detailed clash patterns. This study developed penetrability analysis models using Multi-view CNN (MVCNN) and Multi-view Vision Transformer (MVT), both of which enable multi-view image training. Experimental results showed that MVT achieved an accuracy (ACC) of 0.98, outperforming MVCNN by 0.13 ACC. MVT’s superiority was attributed to its attention mechanism, which focused on clash-prone regions, unlike MVCNN’s emphasis on overall object geometry. These findings demonstrate practical value by enabling the early identification of irrelevant clashes to enhance design efficiency and accuracy, while supporting more precise construction planning through detailed penetration information.

A Study on the Job Analysis for the Development of a Curriculum Linked to the Functional Grade System of Construction Workers

Ae-ri Han ; Ryong-Jae Lee ; Son-hyo Park

https://dx.doi.org/10.6106/KJCEM.2025.26.5.040

This study used the DACUM technique to define jobs and identify necessary competencies for 29 construction occupations under a skill rating system. A DACUM Committee conducted a workshop to draft education and training content and a DACUM chart, which was later validated. Job definitions were aligned with legal standards, and work levels ranged from assistant (beginner) to supervisor (special level). The study identified 459 competencies and 864 educational subjects, averaging 12?26 competencies and 22?41 subjects per occupation. The results led to a new DACUM chart and suggested the need for skill grade system development, foreign worker training, and online education.

The Effect of Housing Location Factors on Lease Renewal Intention among Single-Person Households

Yi-jun Park ; Young-dai Lee

https://dx.doi.org/10.6106/KJCEM.2025.26.5.053

This study examines the impact of housing location factors on lease renewal intentions among single-person households, with a particular focus on the mediating role of residential satisfaction. Hypothesis testing revealed that the educational environment, convenience facilities, medical and welfare services, and undesirable facilities all had significant positive effects on renewal intention, with undesirable facilities having the strongest influence. Similarly, the residential environment, safety, convenience facilities, medical and welfare services, and undesirable facilities significantly influenced residential satisfaction, with convenience facilities having the greatest impact. Residential satisfaction itself had a strong positive effect on renewal intention and was confirmed to mediate the relationship between both the safety environment and the medical and welfare environment with renewal intention. However, this study has several limitations, including the limited representativeness of the sample, as the online survey was conducted solely in the Busan area, and the use of a self-reported, cross-sectional research design. Future research should expand the geographic scope, incorporate qualitative methods, employ a longitudinal design, and explore variations based on housing type.

Optimization of Rapid Deployment Framework for Modular Mobile Hospital (MMH) in Disaster Response

Seyeon Lee ; Sung Hyun Kim

https://dx.doi.org/10.6106/KJCEM.2025.26.5.066

This study proposes a systematic deployment framework for Modular Mobile Hospitals (MMH) designed to provide rapid and effective emergency medical response in disaster and pandemic situations. The research analyzes 117 major disaster cases in Korea from 2014 to 2022 and reviews domestic and international disaster response guidelines and mobile hospital case studies. Disaster response phases are cat egorized into acute (0?2 hours, 2?72 hours), subacute (3?60 days), and chronic (over 60 days) stages, with 30 essential emergency medical programs and 24 corresponding medical units identified for each stage. The study also establishes optimal modular unit specifications (3.3m × 6.6/9.9m × 3.7m) based on transportability, assembly efficiency, and on-site adaptability. Furthermore, a dual utility system strategy?combining centralized (shared) and independent configurations?is proposed to ensure flexible and scalable operation in various disaster contexts. Building on these foundations, a five-stage MMH deployment process is developed, enabling rapid site installation, adaptive medical unit combinations, and efficient resource allocation. This study enhances the feasibility of MMH as a practical solution for emergency medical response and provides a comprehensive framework for disaster healthcare infrastructure planning. The findings offer valuable insights for the development of design guidelines, policy recommendations, and standardization efforts in modular healthcare facilities. Ultimately, this research contributes to strengthening disaster resilience and improving the speed and quality of emergency medical services in large-scale emergencies.

An Evaluation of the Importance of Critical Success Factors in Small and Medium-Sized Construction Project Financing Development Projects

Hyeoncheol Cho ; Seongseok Go

https://dx.doi.org/10.6106/KJCEM.2025.26.5.079

The present study aimed to identify the critical success factors to be considered in the operation of small & medium sized construction project financing (PF) development projects and to assess their relative importance. A comprehensive literature review along with Delphi study leaded to the development of four major categories, including ‘project feasibility,’ ‘financial strategy & financing structure,’ ‘project management,’ and ‘marketing and sales strategy’. In addition, a total of 13 sub-factors corresponding to the four categories were identified. A survey was administered to one-hundred experts to collect pairwise comparison data. The analytic hierarchy process (AHP) was then performed to evaluate the relative importance of both the major categories and their corresponding sub-factors. The AHP results revealed that, ‘project feasibility’ emerged as the most significant among the major categories. At the sub-factor level, ‘accuracy of market research & demand analysis,’ ‘capability in financial terms negotiation,’ ‘project schedule management,’ and ‘appropriate pricing policy & flexible sales planning’ emerged as the most critical within each respective category. Among the 13 sub-factors evaluated, ‘accuracy of market research & demand analysis,’ ‘economic feasibility & profitability analysis,’ and ‘appropriateness of site selection’ were identified as the most influential success factors.

A Study on Construction Material Recognition Using YOLO and Virtual Images

Hyunwoo Kim ; Jeongseop Kim ; Minkoo Kim

https://dx.doi.org/10.6106/KJCEM.2025.26.5.090

This study presents a construction material recognition method using virtual images to enhance safety and productivity at construction sites. Real and virtual images for eight major materials were collected using Pythonbased web crawling and OpenAI’s DALL-E, with data augmentation techniques applied to construct training data. Using the YOLOv8 object recognition algorithm, the study analyzed model performance based on the proportion of virtual images used. The results are as follows (i) The model using only real images had an mAP@0.5 of 0.765, while incorporating approximately 75% virtual images increased the mAP@0.5 to 0.830?an 8.5% improvement. This highlights the importance of appropriately combining virtual and real images to enhance model performance. (ii) Recognition rates for materials like Cement Bag, Wood Plank, Glass, and Stone significantly improved with the inclusion of virtual images, compensating for the lack of diversity in real images. Notably, the mAP@0.5 for Glass increased from 0.368 to 0.740 (101.09% improvement), and for Cement Bag from 0.547 to 0.721 (31.81% improvement). (iii) For materials such as Brick, Rebar, Sand, and Pipe, adding virtual images had little effects or slightly decreased performance, indicating that virtual images are more beneficial for materials with limited data or diversity. The study confirms the potential of using virtual images for construction material recognition, contributing to the advancement of smart construction technology.

A System Architecture for a Collaborative Mixed Reality Workplace for City-Scale Facility Management

Sungjin Choi ; Jaehong Cho ; Sungpyo Kim ; Sanghyeok Kang

https://dx.doi.org/10.6106/KJCEM.2025.26.5.102

As urban facility management becomes increasingly complex, there is a growing need for an effective remote collaboration environment. However, existing studies on Mixed Reality (MR) have been limited to specific tasks or single-user scenarios. This study, therefore, aims to propose and validate an effective system architecture for an MR-based digital workplace that supports real-time collaboration among multiple remote users for managing largescale, heterogeneous urban spatial data. To address this, we designed a client-server architecture composed of a Digital Workplace Server (DWS) for centralized data processing and synchronization, and an MR Operation System (MOS) for user-side visualization and interaction. Performance validation based on a virtual scenario confirmed that the proposed system could render 1.94 GB of large-scale 3D data and synchronize interactions between users, with both tasks being completed in under 3 seconds on average. These results demonstrate that the proposed architecture is a viable solution for building a collaborative MR environment for city-scale facility management. The primary contribution of this research lies in presenting and empirically validating a system architecture that addresses the limitations of previous studies.

A Study on GPT-based Tagging System with Optimized Keywords for Automatic Classification of Construction Site Images

Sungil Son ; Ali Akbar ; Jungtaek Hong ; Soonwook Kwon

https://dx.doi.org/10.6106/KJCEM.2025.26.5.112

The construction industry generates an increasing amount of visual data through sources such as drones, CCTV, and wearable cameras; however, the practical utilization of this data is often limited by a lack of systematic metadata. This study proposes an automatic image tagging system based on a GPT model designed to classify construction site images using multiple keywords. Keywords for image classification were derived through a combination of prior research analysis and expert surveys. The most practical keywords were subsequently selected by optimizing their relevance through performance testing. For the quantitative performance evaluation of the image classification platform, experimental conditions were designed based on three prompt engineering configurations and fine-tuning. Prompt engineering was divided into three types: (1) no prompt application, (2) providing a basic prompt (keyword list), and (3) providing an advanced prompt (keyword list and image derivation case). Fine-tuning was conducted using a multi-keyword image dataset derived from the collected training data. The experimental results demonstrated that both prompt engineering and fine-tuning significantly improved the accuracy, precision, and F1-score. The combination of prompt engineering and fine-tuning produced the most accurate results in complex scenarios. This study highlights the potential of GPT-based models for construction image analysis and demonstrates their value in building a smart site management platform. Future research can expand applicability and scalability, such as extending image-based tagging results to visual tracking and quality evaluation in conjunction with Building Information Modeling (BIM).