Scholarlink

To encourage students’ interest in modern technologies-specifically Artificial Intelligent (AI) and robotic/automation-a training program, using a mobile robot platform with visual feedback has been developed for high-school or undergraduate students. This educational mobile robot allows students to gain various skills and concepts, including micro-controller and AI programming, visual feedback from camera, and data communication via WiFi. In this study, we use a two wheeled differential-drive mobile robot controlled by an LN298 motor driver. This robot is also equipped with an ultrasonic sensor for object detection and an ESP32-CAM for visual feedback. Thus, this setup helps students to learn about mechanical construction, electronics, and Arduino programing. Using the web-based Teachable Machine, students can explore concepts of deep learning and learn how to train/create AI models, allowing for sign images to be visual-based navigation for this robot. Lastly, sign images captured by the ESP32-CAM can be classified in real-time by the deep-learning model on a web client, which can be accessed on a smartphone, through two-way communication using a WebSocket program.

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

Visual-Based Navigation of Educational Mobile Robot Using Transfer Learning with Teachable Machine

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Nutchanon Yongsatianchot,
  • Pradya Prempraneerach,
  • Pakorn Uttayopas,
  • Yutana Chongjarearn,
  • Akkharawoot Takhom,
  • Pattiya Thongkruer,
  • Kanchana Silawarawet

摘要

To encourage students’ interest in modern technologies-specifically Artificial Intelligent (AI) and robotic/automation-a training program, using a mobile robot platform with visual feedback has been developed for high-school or undergraduate students. This educational mobile robot allows students to gain various skills and concepts, including micro-controller and AI programming, visual feedback from camera, and data communication via WiFi. In this study, we use a two wheeled differential-drive mobile robot controlled by an LN298 motor driver. This robot is also equipped with an ultrasonic sensor for object detection and an ESP32-CAM for visual feedback. Thus, this setup helps students to learn about mechanical construction, electronics, and Arduino programing. Using the web-based Teachable Machine, students can explore concepts of deep learning and learn how to train/create AI models, allowing for sign images to be visual-based navigation for this robot. Lastly, sign images captured by the ESP32-CAM can be classified in real-time by the deep-learning model on a web client, which can be accessed on a smartphone, through two-way communication using a WebSocket program.