Spot ROS

The Spot robot comes with its own Spot SDK from Boston Dynamics, available in Python3 and C++. However, there is a ROS package that allows you to trigger most functionality of the robot using ROS messages, services and actions.

Installing the ROS package follows the same process as any other ROS package; git clone the repository, install dependencies, then build the package. Instructions from the Github Pages is here. It is recommended to use ROS Noetic for Python3 compatibility. Use a Docker container to get the correct ROS distro setup with all the distro-specific packages.

First, setup your catkin workspace, if you haven't already. Then, download the source code.

mkdir -p ~/catkin_ws/src
source /opt/ros/noetic/setup.bash
git clone https://github.com/heuristicus/spot_ros.git ~/catkin/src/spot_ros

Next, install the required dependencies with `rosdep`.

sudo apt install -y python3-rosdep
rosdep install --from-paths ~/catkin_ws/src --ignore-src -y

Build the package with `catkin build`.

cd ~/catkin_ws
catkin build

Source the built package.

source ~/catkin_ws/devel/setup.bash

Alternatively, you can make use of pre-built Docker images generated by the CI pipeline.

1. Download the image from Docker Hub with `sudo docker pull jeremysee/spot_ros:master`
2. Save the image into a .tar archive with `sudo docker save -o spotros.tar jeremysee/spot_ros:master`
3. Upload the .tar archive to Spot CORE using this guide

The ROS package has certain parameters that can be configured in `spot_driver/config/spot_ros.yaml`.

There are also some flags that can be set to enable certain startup behaviour.

Lastly, some environmental variables you can set if you want to configure other nodes (Velodyne driver or teleop).

Before running the driver, you need to set the following parameters to authenticate with the robot.

To view the robot in RViz, use:

roslaunch spot_viz view_robot.launch

You should see something like this. The colour points represent the depth cloud produced by the 5 stereo cameras on the robot. The white dots will only be available after adding the PointCloud2 LIDAR data (/spot/lidar/points) to the RViz viewport.

Basic control of the robot can be done using the control panel on the right. This allows you to claim the lease, power on the robot and make it stand up. You can also command it to move to a specific location using the `2D Nav Goal` tool in RViz, or set a specific body pose to go to. When issuing any movement commands to the robot, you will need to claim the lease → power on → stand, before the commands will execute.

The documentation of available ROS services, topics and actions are listed below, explaining the capabilities of the spot_ros driver.

The available ROS services and thier descriptions are in this section.

Question 1: How do I upload a Docker container to Spot?

Answer: Follow this guide on how to use Portainer on the Spot CORE.

Question 2: Does Spot have access to the Internet?

Answer: No, not unless you specifically set it up.

This section includes links to additional documentation, tutorials, and other resources.

spot_ros Github

spot_wrapper Github

spot_ros2 Github

A snapshot of these repos are also available on the internal gitlab:

https://git.cs.lth.se/robotlab/spot