Autonomous Navigation of a Teo-Jansen Mechanism-Based Robot

Roboin, Robotics club

Yonsei University, Seoul, South Korea

Achievement:

(1) Computational Design Competition 1st Place Record, Korea Institute of Science and Technology Information, 2020

Project: Optimal Design and Autonomous Navigation of a Teo-Jansen Mechanism-Based Robot

Abstract:

This work presents the design and development of an autonomous robot utilizing the Teo-Jansen mechanism, specifically optimized using software tools. The robot is designed to navigate predefined paths using only a Raspberry Pi camera, ensuring it stays on course while recognizing STOP signs and AR markers. The research is divided into three main optimization areas: the robot’s structural design with the Teo-Jansen mechanism, vision sensor integration for path recognition, and directional decision-making for autonomous navigation. The optimized Teo-Jansen mechanism, selected for its stability and speed, uses eight legs, each with carefully calculated dimensions to balance structural integrity and movement efficiency. A vision-based algorithm processes real-time binary images to determine the robot’s path, enabling it to make left or right turns based on the detected path’s features. Additionally, the robot identifies STOP signs using color masking techniques and detects AR markers to execute specific commands. The project combines physical prototyping, algorithm development, and hardware-software integration, highlighting the potential for Teo-Jansen mechanism-based robots in autonomous navigation applications.

What I did:

(1) Hardware design & fabrication

(2) Developed vision algorithms (Stop sign detection learning, path recognition, and AR marker recognition)

etc. WIP robot with MDF legs