Implementation and Design of Agile and Multipurpose Autonomous Robot Using ROS

This paper discusses an implementation of autonomous navigation functionality on a differential drive mobile platform called AMAR (Agile and Multipurpose Autonomous robot) based on the Robot Operating System (ROS). The paper's primary contribution is to describe an unmanned platform capable of mapping, localizing, and navigating in an enclosed environment that is well-suited for industrial operations. The approach used for mapping is through the ROS framework, a technology stack containing multiple software stacks for controlling, visualization, and debugging the robot, wrapped with the help of sensors such as RGBD camera and LiDAR (Light Detection and Ranging). For navigation, Adaptive Monte Carlo Localization is used for localizing the vehicle and thus leading to autonomous operation. The proposed solution is computationally efficient and capable of dynamic obstacle avoidance. The simulations were performed using the 3D physics simulator Gazebo and the results were visualized using RViz.