Scholarlink

<p>Safety training is essential for preventing industrial accidents. The integration of Virtual Reality (VR) and Large Language Models (LLMs) holds transformative potential for safety training, as each technology has critical limitations when used alone. While VR provides realistic, risk-free simulations, it lacks adaptability to individual learning needs. Conversely, LLMs offer personalized feedback but lack contextual immersion. Guided by adaptive learning theory, this study proposes a framework which integrating VR and LLMs for safety training, namely AIVR. AIVR consists three core components: an AI tutoring system, that leveraging LLMs for personalized content generation; an adaptive learning system that continuously assesses learner performance and adjusts training content to address individual knowledge gaps; a VR system that contributes adaptive navigation support by visually emphasizing weak points, highlighting relevant areas in the immersive environment, and restructuring interaction paths in response to learner progress.This study then compared AIVR with traditional VR and slide-based instruction in an experiment assessing knowledge acquisition, retention, and user experience. Results demonstrated that AIVR significantly improved knowledge acquisition and retention, and outperformed the other methods in self-efficacy and technology acceptance. The findings also indicated that AIVR's personalization enhanced engagement and reduced anxiety. These results highlight the effectiveness of applying adaptive learning theory to combine immersive and intelligent technologies for enhancing safety training.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Revolutionizing safety training: enhancing safety knowledge acquisition, retention, and user experience with adaptive VR-LLM integration

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年12月3日

查看全文
  • Zuzhen Ji,
  • Yawen Tang,
  • Chaojie Hu,
  • Yuchen Wang,
  • Mingbo Guo,
  • Zhanni Luo,
  • Chenchen Zhou,
  • Dirk Pons,
  • Yong Chen,
  • Shuanghua Yang

摘要

Safety training is essential for preventing industrial accidents. The integration of Virtual Reality (VR) and Large Language Models (LLMs) holds transformative potential for safety training, as each technology has critical limitations when used alone. While VR provides realistic, risk-free simulations, it lacks adaptability to individual learning needs. Conversely, LLMs offer personalized feedback but lack contextual immersion. Guided by adaptive learning theory, this study proposes a framework which integrating VR and LLMs for safety training, namely AIVR. AIVR consists three core components: an AI tutoring system, that leveraging LLMs for personalized content generation; an adaptive learning system that continuously assesses learner performance and adjusts training content to address individual knowledge gaps; a VR system that contributes adaptive navigation support by visually emphasizing weak points, highlighting relevant areas in the immersive environment, and restructuring interaction paths in response to learner progress.This study then compared AIVR with traditional VR and slide-based instruction in an experiment assessing knowledge acquisition, retention, and user experience. Results demonstrated that AIVR significantly improved knowledge acquisition and retention, and outperformed the other methods in self-efficacy and technology acceptance. The findings also indicated that AIVR's personalization enhanced engagement and reduced anxiety. These results highlight the effectiveness of applying adaptive learning theory to combine immersive and intelligent technologies for enhancing safety training.