Space ROS Isaac Sim: Curiosity's Sulfur Stone Discovery in the Gediz Vallis Channel Demo

This repository provides a Docker image for our NASA's Curiosity rover demo based on NVidia Isaac Sim that interfaces with Space ROS. We also built a digital twin of the Gediz Vallis Channel Environment, focusing on the area surrounding the recent event of sulfur stone discovery. The map was developed using the HIRISE Mars DTM (real Mars terrain data) and with custom-made rock assets that seamlessly blend into the environment. It provides a highly detailed replication of the environment for both current and future missions. As of September 2024, Curiosity is still exploring within this region.

See Map Documentation

Highlight of the features

• Docker image of NVIDIA Isaac Sim with Space ROS interface
• Digital twin environment of the Gediz Vallis channel sulfur stone discovery site with enhenced terrain features
• Curiosity rover asset integration into the simulation
• ROS2 interfaces for controlling the rover and sensor feedback

Advantages of NVidia Isaac Sim

• Realistic graphic with Ray-Tracing technology
• PhysX engine for advanced physics simulation
• Built-in DLSS Frame Generation for improved performence
• Direct VR/AR support (with extension), useful in planetory exploration simulations

System Requirements

To run the Curiosity rover simulation in Isaac Sim, ensure your system meets the following requirements:

Requirement	Description
OS	Ubuntu 22.04
GPU	Nvidia RTX 30XX series or higher
RAM	16 GB or higher
CPU	Intel i7 or higher
NVIDIA Driver	ver. 560.35.03 (exact ver. recommended)

For more detailed hardware requirements, refer to the Isaac Sim Hardware Documentation.

Prerequisites

The following software dependencies must be installed on your system:

Dependency	Description
Docker Engine	Docker Installation Guide
NVIDIA Toolkit	NVIDIA Container Toolkit Installation

Required Docker Images

Ensure the following Docker images are built and available on your system before running the simulation:

Image	Description
space-robots	The base Docker image for space robotics research.
moveit2	Provides MoveIt2 for motion planning.
space ros	Core ROS2 components for space robotics.

Building and Running the Container

Start the Docker with

sudo systemctl start docker

To build and launch the container you should cd to docker folder:

To build image use this script:

./build.sh

To run docker container use this script:

./run.sh

That's all you need to start the simulation environment.

Cloning Repositories into Docker

The following repositories are cloned into the Docker container during the build process:

1. isaac_ros2_utils
2. curiosity_rover_description
3. mars_rover_control

Recommended Development Setup

For an optimized development workflow, it is recommended to use Visual Studio Code (VSCode) along with its Docker-related extensions. These tools greatly simplify the process of interacting with Docker containers, especially when debugging and managing development environments.

Recommended VSCode Extensions:

• Dev Containers: Enables you to develop inside a Docker container.
• Docker Extension for VSCode: Simplifies the management of Docker containers directly from VSCode.

By using these tools, you can streamline your workflow and improve the experience of working within a Dockerized development environment.

Usage

IMPORTANT: Before launching scripts, execute this command to attach to space_ros_isaac_sim container or use VSCode remote connection.

sudo docker exec -it space_ros_isaac_sim /bin/bash

In the first terminal, enter the Docker container(if not already) and launch Isaac Sim with one of the following two environments

• For spawning Curiosity in the sulfur stone area in the Gediz Vallis channel environment

  python ~/curiosity_sim/scripts/gale_crater_scene.py 
  

• For spawning Curiosity in the test ground environment

  python ~/curiosity_sim/scripts/test_scene.py
  

Note: DO NOT source demos_ws/install in this terminal (that launches Isaac Sim).

In the second terminal, enter the Docker container and launch the control system:

source ~/demos_ws/install/setup.bash
ros2 launch mars_rover_control mars_rover_control.launch.py

In the third terminal, for manual control using the keyboard:

source ~/demos_ws/install/setup.bash
ros2 run teleop_twist_keyboard teleop_twist_keyboard

Note: this terminal needs to be focused during keyboard control

ROS Service Commands

In addition to ROS2 topics, the control graph supports various ROS2 service commands, allowing for precise control of different parts of the rover:

• Drive the rover forward:

  ros2 service call /move_forward std_srvs/srv/Empty

• Stop the rover:

  ros2 service call /move_stop std_srvs/srv/Empty

• Turn left:

  ros2 service call /turn_left std_srvs/srv/Empty

• Turn right:

  ros2 service call /turn_right std_srvs/srv/Empty

• Open the tool arm:

  ros2 service call /open_arm std_srvs/srv/Empty

• Close the tool arm:

  ros2 service call /close_arm std_srvs/srv/Empty

• Open the mast (camera arm):

  ros2 service call /mast_open std_srvs/srv/Empty

• Close the mast (camera arm):

  ros2 service call /mast_close std_srvs/srv/Empty

Manuevering Curiosity on a Slope

There are many hills in the Gediz Vallis channel environment. Curiosity may drift downhill on a slope and it will require increased speed to counteract the incline.

Recommended Render Settings for Isaac Sim

To optimize performance and maintain high FPS in the simulation, it is recommended to adjust the render settings in Isaac Sim as follows:

• Enable DLSS: Deep Learning Super Sampling (DLSS) is a technology that uses AI to upscale images, which can significantly boost performance without compromising visual quality. Enabling DLSS is especially useful for maintaining high FPS in complex scenes.

  • For more information, refer to the NVIDIA DLSS documentation.

• Adjust Viewport Resolution: If your system struggles to maintain a stable FPS, reducing the resolution of the Viewport window can help. Lowering the viewport resolution reduces the computational load, improving performance during simulation.

  • You can find more details on adjusting the viewport resolution in the Viewport Settings documentation.

By using these settings, you can balance visual quality with performance, ensuring smoother operation of the simulation even on lower-end systems.

Detailed Documentation

The following table provides links to detailed documentation on various components of the Curiosity Mars Rover simulation, covering control, sensors, and more.

Document	Description
Maps	Information on maps and environments used in the simulation, providing context for different scenarios.
Curiosity Description	Description and overview of the Curiosity rover model used in the simulation.
Curiosity Control	Detailed documentation on the control system of the Curiosity rover using ROS2 and action graphs.
Curiosity Sensors	Explanation of the sensors used in the simulation, including Lidar, camera, and odometry configurations.