space_validation:

Validation of Space Robotics in Underwater Environments via Disturbance Robustness Equivalency.

This is the code accompanying the paper in which we present an experimental validation framework for space robotics that leverages underwater environments to approximate microgravity dynamics. While neutral buoyancy conditions make underwater robotics an excellent platform for space robotics validation, there are still dynamical and environmental differences that need to be overcome. Given a high-level space mission specification, expressed in terms of a Signal Temporal Logic specification, we overcome these differences via the notion of maximal disturbance robustness of the mission. We formulate the motion planning problem such that the original space mission and the validation mission achieve the same disturbance robustness degree. The validation platform then executes its mission plan using a near-identical control strategy to the space mission where the closed-loop controller considers the spacecraft dynamics. Evaluating our validation framework relies on estimating disturbances during execution and comparing them to the disturbance robustness degree, providing practical evidence of operation in the space environment.

Our evaluation features a dual-experiment setup: an underwater robot operating under near-neutral buoyancy conditions to validate the planning and control strategy of either an experimental planar spacecraft platform or a CubeSat in a high-fidelity space dynamics simulator.

External Requirements

• Planner: create a python environment using the environment.yml file. Solving the planner requires a Gurobi license, which can be freely obtained for academic users.
• Control: the MPC control node relies on ROS 2 Humble and the acados optimization library. The robots (in SITL and hardware) rely on the PX4-Autopilot firmware and the micro-ROS agent for ROS 2 communication.

Workspace setup

Create a new ROS 2 workspace and clone this repo plus px4-offboard inside it.

mkdir -p ~/space_validation_ws/src
cd ~/space_validation_ws/src
git clone <this_repo_url> space_validation
git clone https://github.com/joris997/px4-offboard

Initialize submodules for this repo (this gets the BlueROV dynamics model in the correct place):

cd ~/space_validation_ws/src/space_validation
git submodule update --init --recursive

Install PX4-Autopilot

Follow the official PX4-Autopilot installation instructions and keep all defaults. Also ensure to create a ros2 workspace of the px4_msgs package and build it. Add the install/setup.bash to the .bashrc file and be sure to export PX4_SPACE_SYSTEMS_DIR.

Install acados

Follow the official acados installation instructions and keep all defaults (default options, default build type, default dependencies). Be sure to update the LD_LIBRARY_PATH in the .bashrc file to include the path to the acados library.

BlueROV setup

Ensure the BlueROV is equiped with a PX4 FMU v6x and that the micro-ROS agent is running on the same network as the BlueROV. Flash the BlueROV with the px4_fmu-v6x_uuv firmware.

1. In PX4-Autopilot, run PX4_UXRCE_DDS_NS=snap make px4_sitl_uuv gz_uuv_bluerov2_heavy
2. Start microros: micro-xrce-dds-agent udp4 -p 8888

Flashing BlueROV px4: PX4_UXRCE_DDS_NS=itrl_rov_1 make px4_fmu-v6x_uuv upload given that the BlueROV is connected to the computer via USB.

Then create a second workspace utils_ws for the BlueROV2 utilities. Most importantly, we run the motion capture node locally:

mkdir -p ~/utils_ws/src
cd ~/utils_ws/src
git clone git@github.com:smarc-project/motion_capture_system.git
git clone git@github.com:DISCOWER/discower_launch.git
git clone git@github.com:DISCOWER/bluerov_launch.git
git clone git@github.com:DISCOWER/srl_vehicle_mocap_odom.git
cd ~/utils_ws
colcon build --symlink-install
source install/setup.bash

And add the source command to the .bashrc file.

ATMOS setup

We refer to atmos.discower.io for the latest instructions on how to set up ATMOS. If you're on hardware, everything should be good to go.

For SITL, you can run the following command to start the simulator:

PX4_UXRCE_DDS_NS=snap make px4_sitl_spacecraft gz_atmos
micro-xrce-dds-agent udp4 -p 8888
./startQGC

Running the planner

The planner lives in the Planning package. Run it from the workspace so it can access configuration and solution files.

cd ~/space_validation_ws/src/space_validation
conda activate space_validation
python3 -m space_validation.Planning.main

You can add new scenarios after which all results are dumped into the Planning/solutions folder. In order for the controller to later use these solutions, create a new subfolder in Planning/solutions with the name of the experiment (e.g., exp_3) and move all the generated solution files into that folder!

Running the ROS 2 node

On ATMOS

Make sure the motion-capture PC is running the mo-cap node and that you are on the correct network. Then follow the instructions below.

On the BlueROV

Make sure the motion-capture PC is running downstairs. Connect the BlueROV to the computer via the USB interface and run the following command to start the micro-ROS agent and start the mo-cap node locally:

micro-xrce-dds-agent udp4 -p 8888
ros2 launch bluerov_launch bluerov_obc.launch.py

Always

Build the space_validation workspace and source the setup file:

cd ~/space_validation_ws
colcon build --symlink-install
source install/setup.bash

Now we decide which experiment to run. Please take a look at the launch files in space_validation/space_validation/Launch to see how we enable Feedback Equivalence Control and Offset-Free MPC.

First determine whether you want to run in sitl or hardware

# For SITL
ros2 launch space_validation sitl.launch.py
# For hardware
ros2 launch space_validation hardware.launch.py

Now determine whether you want to run the spacecraft or underwater experiment

# For spacecraft
ros2 launch space_validation space_mpc.launch.py
# For underwater
ros2 launch space_validation subsea_mpc.launch.py