A Study on the Research Trends of Smart Safety Technology in the Construction Industry for Establishing an Analysis-based Safety and Health Management System
,Yong Hwa Kwon,Joon Won Lee,Won Souk Eoh,Hyun Woong Yun,Youn June Hwang
DOI: https://doi.org/10.52902/kjsc.2024.32.85
2024-07-31
Forum of Public Safety and Culture
Abstract:This study aims to propose the research direction for establishing an analysis-based safety management system. As for a study method, it collected academic papers related to smart safety in the construction industry and analyzed their research trends. For the data analysis, 1,326 papers were collected from the Research Information Sharing Service (RISS) after the keywords “smart safety” were searched for and the filters 'full text provided' and 'KCI indexed' were applied. Duplicate papers were checked, and the scope was narrowed down to those papers containing the term “construction.” With the exclusion of the papers unrelated to the research topic, 48 papers were finally selected. Morphological analysis was conducted on the titles, keywords, and abstracts to extract nouns, followed by topic modeling and semantic network analysis. Topic modeling and semantic network analysis are techniques of text mining used for analyzing unstructured big data. Textom was used as the tool topic modeling and semantic network analysis. The study results are as follows: First, according to the word-frequency based keyword analysis, the top 10 keywords in academic papers on smart safety were 'Internet of Things', 'sensor', 'accident', 'bridge', 'prevention', 'environment', 'drone', 'equipment', 'law', and 'ICT'. Second, based on the keywords derived from the word frequency analysis, 10 topics were extracted through topic analysis (LDA topic modeling), and the occurrence ratio of keywords per topic was calculated. Third, the text relationships between topics were clustered by keyword through centrality analysis and CONCOR analysis, and their characteristics were categorized. Implications from this study are as follows: First, all keywords in the field of construction industry smart safety that have been researched so far were analyzed, and it was possible to understand research trends through the extraction of keywords covered in existing studies. Second, the keyword analysis revealed that existing research has been biased towards hardware-centric studies, such as IoT-centric sensor networking and the integration of construction machinery or devices. This indicates a relative lack of research based on software and databases. Third, even though it is crucial to establish a management system for short-term achievements focusing on sensing and monitoring, it is necessary to research and discuss the application of smart safety technologies to build a long-term, analysis-based safety and health management system. To build an analysis-based safety and health management system, the following suggestions are made: First, a safety and health management database should be established. It is necessary to build an integrated database to manage all data generated from systems operated in the field. Second, it should be possible to analyze the data accumulated in the integrated database. For the analysis, analytical tools and methodologies are needed. Third, the analyzed results should be reflected in the safety and health management system. This management system should be established to be reflected throughout the entire lifecycle (PDCA) of the field. It is considered that only when such a management system is built, it is possible to establish an analysis-based safety management system and apply it to accident prediction and prevention.