It's also possible to connect both the Rpi and the computer to the same WiFi network. I haven't included in my GitHub the RPi network configuration but basically the RPi creates an Ad-Hoc network I connect to with my computer in order to send SSH commands. In order to navigate Nox you will need to start two launch files using SSH and ROS nodes over network. You can install and build the package by copying the 'nox' and 'nox_description' folders in 'your_catkin_workspace/src' on both your computer and the Raspberry Pi and running: ROSSerial package (for connecting to the Arduino Mega).TF, Joint State and Robot State Publishers.The freenect package (for connecting to the Kinect).To use Nox you will need the following packages (most of them should already be installed by default or requested when building the nox packages): For more information you can check the ROS tutorials. This readme doesn't explain how to install ROS and packages. More recent versions should work as well but might require some tweaking. The software for the Nox project was developped with ROS Kinetic and Ubuntu 16.04. It also sends back the encoder data needed for the odometry to the controller node through ROSSerial. The arduino node receives the command order from the base controller node and commands the motor speeds. The package also includes several launch files used to start the navigation (see 'How to use Nox' below). It also computes the odometry (with the encoder data and according to the robot kinematic model) from data received from the Arduino Mega. This node will convert velocity command data into proper command data for the robot (according to the robot kinematic model) and send it to the Arduino Mega (through ROSSerial). This package includes the base controller node 'nox_controller.cpp'. The launch file 'nox_rviz.launch' launches the joint and robot state publisher as well as RViz to display the model.

This package includes the URDF description of the robot and the associated CAD files for display in RViz. More information (such as components used) can be found in the Hackster page of the project. On the rear part of the robot plates can be stacked to put the electronic boards on. The rest of the structure is made mainly of wood and metal brackets, easy to find in any DIY retail store.

The base is made of wood with two caster wheels for support. Nox is a differential drive robot with the motors placed on the same axis. It can plan a path to a given goal and drive to it avoiding obstacles. In its current state the robot can use SLAM (gmapping) to create a map of its surroundings (using the Kinect depth perception to detect wall and obstacles) and localize itself within the map. It is powered by ROS running on a Raspberry Pi 3B and an Arduino that controls two motors with encoders. Nox is a nice (and time-consuming for me) robot which uses SLAM (ROS) with a Kinect to navigate in its environment.

Solidworks with kinect