Autonomous Quadruped Robot System: Lidar Sensor Navigation And Task Execution With A Robotic Arm

Abstract

The aim of this thesis project was to develop an integrated motion planning and autonomous navigation system for the Unitree B1 quadruped robot, which is equipped with a Unitree Z1 robotic arm and a Robosense Helios LIDAR sensor. The objective was to navigate the Robot from one point to another within a specific envi-ronment autonomously while avoiding obstacles and perform required tasks using the integrated robotic arm.

The thesis project was mainly implemented by using the Robot Operating System (ROS) framework. ROS melodic was used for this project. Simultaneous Localization and Mapping (SLAM) algorithm was used for mapping the environment, while Adaptive Monte Carlo Localization (AMCL) was used for the position and orientation of robot within a generated map. Move Base from ROS Navigation stack was used for path planning and obstacle avoidance. The motion planning of robotic arm was done using its SDK in C++ to perform manipulation tasks. A 3D visualization tool named Robot Visualization (Rviz) was used for visualization and assigning the navigation goals during the development and testing phases.

Experimental results demonstrated that the robot successfully navigated to goals while avoiding obstacles. The planned motion was performed by robotic arm as required. Thus, the results indicate that the proposed integrated system provides reliable autonomous quadruped robot system with motion planning and autonomous navigation capabilities.