본문 바로가기
  • Home

Implementation of a Natural Language-Based Robot Control System: A Case Study on Claw Machine Application

  • Journal of Internet of Things and Convergence
  • Abbr : JKIOTS
  • 2026, 12(1), pp.39~45
  • Publisher : The Korea Internet of Things Society
  • Research Area : Engineering > Computer Science > Internet Information Processing
  • Received : December 22, 2025
  • Accepted : February 3, 2026
  • Published : February 28, 2026

Hee-Jong Yoon 1 Hyeon-Seong Jo 1 Jae-Hyeon Yoon 1 Ji-Hwan Kim 1 KIM, TAEKOOK 1

1국립부경대학교

Accredited

ABSTRACT

This paper presents the design and implementation of a robot control system based on natural language commands. Using a claw machine as a testbed, the proposed system integrates the GPT-4 language model, YOLOv5 object detection, and a Raspberry Pi-based control system to establish an intuitive user interface. When a user describes a target object in natural language, the system performs detection and analysis via YOLOv5, while GPT-4 interprets the results to generate precise control commands for the robotic arm. These commands are then transmitted to the claw mechanism through the Raspberry Pi to execute the task. Furthermore, Server-Sent Events (SSE) technology is implemented to provide real-time updates, including object detection status, execution progress, and exception handling. Experimental results verify that the system accurately identifies and successfully retrieves the user-described targets. This study demonstrates the potential of replacing complex robot control with intuitive natural language interaction, contributing to more efficient robotic operations in future manufacturing and logistics sectors.

Citation status

* References for papers published after 2024 are currently being built.

This paper was written with support from the National Research Foundation of Korea.