건설현장의 스마트 안전장비 성능과 도입 효과 분석

초록

This study aims to empirically analyze the risk detection accuracy and accident reduction effect of smart safety equipment at small and medium sized construction sites, and to suggest policy directions for government support. Using data from 13 sites equipped with AI-based CCTV in 2023, we evaluated the detection accuracy of safety helmet violations, falls, and fire occurrences based on the YoLo v4 algorithm. Results showed that fire and helmet-related risks exceeded the acceptable threshold of 90.0%, while fall detection was below it. Although the sample size was limited, sites with smart equipment recorded a 0.0% accident rate, compared to 0.461% at general small sites, demonstrating a clear preventive effect. These findings confirm the effectiveness of smart safety technologies in field conditions. To enhance accuracy and applicability, adoption of advanced algorithms and improved equipment is necessary. Additionally, the government should provide financial incentives and institutional support to promote broader deployment of smart safety systems.

키워드