Design and Develop Automated Guided Vehicle (AGV) Base for Mapping and Navigation Task

Abstract

The environment scanning has been important in mobile robot studies and applications that may help people to investigate, monitor and study a variety of environments such as unexplored or hazardous surroundings. The main issue regarding this research is the efficiency of the mobile robot system and the accuracy of the laser sensor. This project presents designing and developing of an Autonomous Guided Vehicle (AGV) with indoor mapping capability only using a LIDAR as primary sensor and without odometry information of the AGV base. Hector mapping SLAM algorithm is adopted for building the map around the robot which is equipped with single 1D laser range finder in ROS platform and viewed through ROS visualization Rviz. To evaluate the performance of the sensor and the AGV base, usability testing was carried out in multiple experiments. The Lidar results for object distance measurement show over 97% accuracy for targets within 9 meter radius of the AGV base and the mapping accuracy is 83% for areas larger than 34sq.m and over 94% for areas smaller. Arduino Uno is used to communicate and control the motors. The AGV is remotely controlled via wireless communication. It is also designed to transport load in closed areas such as labs or libraries therefore, the design base is capable of carrying upto 20 kilograms at a speed of 0.6 m/s. Discussion of errors and troubleshooting methods are presented after the experiment results.