DogBot is a quadruped robot developed by React AI as a platform for researching robotics, AI and more.

The Dogbot repository contains CAD designs, along with software for higher-level systems for control and operation of the React AI DogBot robot via ROS. It utilises the lower-level motor control firmware in the corresponding Firmware repository for interaction with the physical DogBot.

This code has been tested for installation and operation on Ubuntu Linux.

ROS must be installed; the code has been tested against the Kinetic release. The repository contains a catkin workspace under Software/ROS, which should be built with catkin build

You may need to install the packages ros-kinetic-rosparam-shortcuts ros-kinetic-ros-control ros-kinetic-ros-controllers

The motors must be ‘homed’, i.e. know their absolute position, before they can be controlled via the ROS driver, hence they must first be controlled via the UI as described in the Firmware repo.

Note: complete the Firmware UI steps before starting the ROS components. The hardware interface node in ROS will send conflicting signals if it is operational while the UI tries to alter the motor positions.

To get started with ROS control, build and source the code then run dogbot_hardware.launch

cd Software/ROS
catkin build
source devel/setup.bash
roslaunch dogbot_control dogbot_hardware.launch

This launches a hardware interface node of type dogbot_control/dogbot_hw_main, which uses the ros_control package to spawn and interact with ROS controllers. This provides a framework for different controller types, as well as standardising interactions with Gazebo and RViz.

If you are controlling other equipment you may require a different launch file, such as testrig_hardware.launch

The motors can now be controlled by interacting directly with ROS topics or via ROS nodes. Further details are in the ROS directory readme file

If you have any problems with the contents of either repository, or have a question about DogBot, please feel free to contact us

If you wish to contribute please fork the project on GitHub. More details on contributing to follow.

details to follow