Scholarlink

Traditional training methods face challenges such as high costs, operational risks, and limited scalability. Digital twinning enables real-time simulation and optimization, while AI-driven assistants offer personalized instruction. This study explores the development and evaluation of an AI-powered virtual training system, emphasizing its benefits in knowledge acquisition and training performance. It presents the design and implementation of an AI-powered immersive virtual training system to enhance trainee learning in the assembly of machine parts in a virtual environment. The system provides real-time guidance and feedback by integrating Generative AI, Extended Reality (XR), and voice interaction. A virtual assistant, powered by OpenAI’s ChatGPT API, interprets trainee queries and delivers step-by-step instructions. WhisperSpeech-to- Text (STT) and Google Text-to-Speech (TTS) facilitate seamless voice interaction. Built using Unity and tested with an HTC VIVE headset, the system allows trainees to interactively assemble a lathe machine without prior knowledge of its components. The experimental evaluation demonstrated significant improvements in user engagement, task performance, and knowledge retention compared to non-AI-supported training. These findings highlight the potential of AI-integrated digital twins in optimizing industrial training methodologies.

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

Digital Twinning: Generative AI for Virtual Training Systems

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Ahmed Refaat,
  • Nada Nasser,
  • Nada Sharaf,
  • Slim Abdennadher

摘要

Traditional training methods face challenges such as high costs, operational risks, and limited scalability. Digital twinning enables real-time simulation and optimization, while AI-driven assistants offer personalized instruction. This study explores the development and evaluation of an AI-powered virtual training system, emphasizing its benefits in knowledge acquisition and training performance. It presents the design and implementation of an AI-powered immersive virtual training system to enhance trainee learning in the assembly of machine parts in a virtual environment. The system provides real-time guidance and feedback by integrating Generative AI, Extended Reality (XR), and voice interaction. A virtual assistant, powered by OpenAI’s ChatGPT API, interprets trainee queries and delivers step-by-step instructions. WhisperSpeech-to- Text (STT) and Google Text-to-Speech (TTS) facilitate seamless voice interaction. Built using Unity and tested with an HTC VIVE headset, the system allows trainees to interactively assemble a lathe machine without prior knowledge of its components. The experimental evaluation demonstrated significant improvements in user engagement, task performance, and knowledge retention compared to non-AI-supported training. These findings highlight the potential of AI-integrated digital twins in optimizing industrial training methodologies.