In modern industrial environments, automation paradigms that replace hazardous human work areas with robots are accelerating alongside technological advancement. The economic efficiency and task performance capabilities of robotic systems have emerged as key factors for successful industrial implementation. Manufacturing and hazardous industries recognize hardware cost reduction and software advancement as essential prerequisites for robot adoption. Quadruped robots have gained attention in industrial environments due to their exceptional mobility in irregular terrain and complex environments. This study implements rapid and safe mission execution through LiDAR sensor-based real-time obstacle avoidance and gesture recognition systems. Furthermore, we develop systems that perform specific tasks according to operators' intuitive commands by introducing gesture recognition as a human-robot interaction interface. This research designs and implements quadruped robots that integrate obstacle avoidance and gesture-based task command processing modules while recognizing static and dynamic environmental objects in real-time through LiDAR sensors. We aim to significantly enhance robot utilization in industrial sites and present practical robotic solutions capable of replacing humans in hazardous work environments.

• 금동현(Dept of Computer Engineering, Inje University) | Dong-Hyun Geum
• 김건(Dept of Computer Engineering, Inje University) | Geon Kim
• 박현준(Dept of Computer Engineering, Inje University) | Hyun-Jun Park
• 최윤성(Professor, Dept of AI Software, Inje University) | Youn-Sung Choi Corresponding author