Closed-Chain Affordance Planning for the BD Spot using ROS2

This repository contains the cca_spot package, which implements the closed-chain affordance planning framework on the Boston Dynamics Spot robot. The demonstration video of the underlying paper showcasing simulation and real-world tasks is available here. An interactive RViz plugin is also available as one of many ways to interface with the planner. See the plugin demo videos for a glance at the framework's capabilities.

Requirements

• C++20
• ROS Humble

Dependencies

• CCA Libraries
• CCA ROS Interface
• spot_ros
• spot_manipulation

If you're not using a real robot, from spot_ros and spot_manipulation, it's sufficient to build only the spot_description and spot_moveit_config packages, which are used for visualization and collision checking:

colcon build --packages-select spot_description spot_moveit_config

Otherwise, build all packages with:

colcon build

Tip: If you'd like to use your own robot description or MoveIt config for spot, simply update the relevant entries in the cca_spot_settings.py module in the launch/ directory.

Installation

1. Navigate to your ROS2 workspace src folder and clone the repository:

   cd ~/<ros2_ws_name>/src

   git clone git@github.com:UTNuclearRoboticsPublic/closed_chain_affordance_spot.git

2. Build and source the cca_spot package:

   cd ~/<ros2_ws_name>
   colcon build --packages-select cca_spot --cmake-args -DCMAKE_BUILD_TYPE=Release
   source install/setup.bash

Usage

With Physical Robot

To execute CCA-generated joint trajectories on the Spot robot:

1. Run the Spot driver:

   ros2 launch spot_bringup bringup.launch.py hostname:=192.168.50.3 has_arm:=True kinematic_model:="mobile_manipulation"

2. Launch the CCA visualizer for Spot:

   ros2 launch cca_spot cca_spot_val_and_viz.launch.py launch_rviz:=true

   This launches both the visualizer and an interactive RVIZ plugin for code-free planning and execution. To plan and execute trajectories using the RViz plugin, start the following action server:

   ros2 launch cca_spot cca_spot_action_server.launch.py

   At this point, you should be able to interactively plan and execute trajectories using the Rviz plugin.

3. Alternatively, for programmatic task definition, launch the CCA planner which will plan for the tasks defined in src/cca_spot_node.cpp:

   ros2 launch cca_spot cca_spot.launch.py

Without Physical Robot

You can plan and visualize joint trajectories for the BD Spot using the CCA framework without needing a physical robot.

1. Launch the Moveit fake controller to simulate trajectory execution:

   ros2 launch spot_moveit_config fake_controller.launch.py kinematic_model:="mobile_manipulation"

   Alternatively, you may launch the Spot offline state publisher to visualize the robot in Rviz at a named configuration and to use this state for planning:

   ros2 launch spot_description offline_state_publisher.launch.py has_arm:=True kinematic_model:="mobile_manipulation" configuration:="unstowed"

2. Launch the CCA visualizer for Spot:

   ros2 launch cca_spot cca_spot_val_and_viz.launch.py launch_rviz:=true 

   This launches both the visualizer and an interactive RVIZ plugin for code-free planning and execution. To plan and execute trajectories using the RViz plugin, start the following action server:

   ros2 launch cca_spot cca_spot_action_server.launch.py

   At this point, you should be able to interactively plan trajectories using the Rviz plugin.

3. Alternatively, for programmatic task definition, launch the CCA planner which will plan for the tasks defined in src/cca_spot_node.cpp:

   ros2 launch cca_spot cca_spot.launch.py

4. You may also run the demo by launching the CCA planner demo node:

   ros2 launch cca_spot cca_spot_demo.launch.py
   

Task Examples

• Refer to the CCA-ROS README.md Code Tutorial section here for various planning examples.

Other Recommendations

• Explore the interactive RVIZ plugin for code-free planning and execution
• Modify demo node tasks to create custom trajectories

Author

Janak Panthi (aka Crasun Jans)

Support

For issues, feature requests, or contributions, please open an issue in the GitHub repository.