Introduction

  This project was carried out in 2021, which was a required capstone project for the Robotics Software Engineer Nanodegree course of Udacity. Here a two-wheeled autonomous mobile robot (AMR) was developed, which was acting like a home service robot in a virtual building and was executing pick-and-place tasks on virtual objects. These robots are usually used in warehouses like Amazon, where they transport products from one place to another and organize them in shelf’s. Nevertheless, there are plenty of other application fields where such robots get applied nowadays.

 

Purpose

  The main goal was to simulate a pick-and-place task, which can be used to creare robotics applications for real robots. To achieve thia different robotics algorithms like Localization, Mapping, SLAM and Navigation were implemented. These were also carried out in previous projects in the course (links are available bellow). The steps applied were the following ones:

• ROS1 programming with C++ and Python in ROS Noetic
• Simulation in Gazebo and RViz
• Design of robots using URDF and Xacro
• Simulation of a Lidar sensor and RGB-D camera
• Localization with AMCL (shown in the Where Am I project)
• SLAM using the RTAB-Map ROS package (shown in the Map My World project)
• Navigation using the ROS Navigation Stack (shown in Home Service Robot project)

 

Development

Development system	Ubuntu 20.04 LTS, Windows 10, Interl i7, 32GB RAM, Nvidia GTX1080, 1TB HDD
Development tools	VS Code, Gazebo, RViz, RQT Graph
Programming languages	C++, Python
ROS Packages	ROS Noetic, RTAB-Map, map_server, AMCL, Map Creator, Navigation Stack, ROS Teleop
Repository

 

Images