Visualization-Based Safety Training Framework for Foreign Workers Using F×S Risk Analysis and VR/BIM

Dayeon Song, Sungjin Ahn

Foreign construction workers in Korea experience disproportionately high accident rates due to language barriers and ineffective text-based safety training. This study proposes a visualization-based safety education model tailored to their specific vulnerabilities and trade-specific accident risks. By analyzing accident statistics in high-risk trades such as scaffolding and lifting, the research develops an integrated framework that utilizes the Frequency-Severity (F×S) index to prioritize educational content. The model recommends a tiered application of visualization media, assigning high-immersion tools like Virtual Reality (VR) for critical risk scenarios and Building Information Modeling (BIM) animations for procedural training. This approach minimizes reliance on text and maximizes intuitive understanding, providing a practical, learner-centered framework to effectively reduce accident vulnerability in multicultural construction environments.