Skip to content

Isaac Sim 4.5 on Ubuntu 22.04 (binary)

Jump to bottom
Minho edited this page Sep 21, 2025 · 2 revisions

Isaac Sim

mkdir ~/isaacsim
cd ~/Downloads
unzip "isaac-sim-standalone@4.5.0-rc.36+release.19112.f59b3005.gl.linux-x86_64.release.zip" -d ~/isaacsim
cd ~/isaacsim
./post_install.sh
./isaac-sim.selector.sh
  • If error occurs due to IOMMU, it is because of Intel VT-d technology. You can turn off the VT-d via BIOS setting and modify the following line in the GRUB setting.
sudo nano /etc/default/grub
GRUB_CMDLINE_LINUX_DEFULAT="quiet splash intel_iommu=off"
  • RTX 50 series have the known issue during real-time rendering in Isaac 4.x, Ubuntu 22.04. To detour this, you need to set "Path tracing" and "Samples per pixel per frame" options at the beginning. Otherwise, the renderings will be noisy and blurred. This issue is resolved in Isaac 5.x.
simulation_app = SimulationApp({
    "renderer": "PathTracing",
    "samples_per_pixel_per_frame": 1,
    "headless": False,
})

Verify the Isaac Sim installation

  • Environmental variables: Input the following lines in the bash file:
# Isaac Sim root directory
export ISAACSIM_PATH="${HOME}/isaacsim"
# Isaac Sim python executable
export ISAACSIM_PYTHON_EXE="${ISAACSIM_PATH}/python.sh"
  • Check that the simulator runs as expected:
source ~/.bashrc
# note: you can pass the argument "--help" to see all arguments possible.
${ISAACSIM_PATH}/isaac-sim.sh
  • Check that the simulator runs from a standalone python script:
# checks that python path is set correctly
${ISAACSIM_PYTHON_EXE} -c "print('Isaac Sim configuration is now complete.')"
# checks that Isaac Sim can be launched from python
${ISAACSIM_PYTHON_EXE} ${ISAACSIM_PATH}/standalone_examples/api/isaacsim.core.api/add_cubes.py
${ISAACSIM_PYTHON_EXE} ${ISAACSIM_PATH}/standalone_examples/api/isaacsim.robot.manipulators/franka/follow_target_with_rmpflow.py
  • For 50 series GPUs, please use the latest PyTorch nightly build instead of PyTorch 2.5.1, which comes with Isaac Sim:
cd ~/isaacsim
./python.sh -m pip install --upgrade --pre torch torchvision --index-url https://download.pytorch.org/whl/nightly/cu128
  • Remove old CUDA package 11 to avoid collision:
./python.sh -m pip list | grep -i cuda
./python.sh -m pip uninstall nvidia-cuda-cupti-cu11  
./python.sh -m pip uninstall nvidia-cuda-nvrtc-cu11  
./python.sh -m pip uninstall nvidia-cuda-runtime-cu11
  • Verify PyTorch
import torch
print(torch.cuda.is_available())
print(torch.__version__)

VS Code Setting (Option 1)

"env": {
    "RESOURCE_NAME": "IsaacSim",
    "PYTHONPATH": "${workspaceFolder}/vsprojects:${workspaceFolder}/lecture"
},
  • Terminal--Run Task--"setup_python_env"--"Continue without scanning the task output.
  • Check ".standalone_examples.env" file is created in the ".vscode" folder.
  • Open ${ISAACSIM_PATH}/standalone_examples/api/omni.isaac.franka/follow_target_with_rmpflow.py and start debugging (Ctrk+F5 or F5)

VS Code Setting (Option 2)

  • It works for only Isaac Sim >= 4.1.0

  • It works for any python script outside the project folder.

  • Reference: https://docs.isaacsim.omniverse.nvidia.com/4.5.0/development_tools/vscode.html#id1

  • Launch Isaac Sim 4.1.0, Menu--Window--Extensions

  • Search "vscode" and enable omni.isaac.vscode (AUTOLOAD enabled)

  • Menu--Window--VS Code

  • In the VS Code, add extension "Isaac Sim VS Code Edition"

  • Open it on your left side bar, and run

Isaac Lab 2.0

sudo apt install git cmake build-essential
cd ~ && git clone https://github.com/isaac-sim/IsaacLab.git
  • Create a symbolic link
cd IsaacLab
ln -s ${HOME}/isaacsim _isaac_sim
  • Create conda env
conda remove --name env_isaaclab --all
./isaaclab.sh --conda env_isaaclab
  • Installation
conda activate env_isaaclab
./isaaclab.sh --install
  • For 50 series GPUs, please use the latest PyTorch nightly build instead of PyTorch 2.5.1, which comes with Isaac Sim:
pip install --upgrade --pre torch torchvision --index-url https://download.pytorch.org/whl/nightly/cu128
  • Remove old CUDA package 11 to avoid collision:
pip list | grep -i cuda
pip uninstall nvidia-cuda-cupti-cu11  
pip uninstall nvidia-cuda-nvrtc-cu11  
pip uninstall nvidia-cuda-runtime-cu11

Verify the Isaac Lab installation

# Option 1: Using the isaaclab.sh executable
# note: this works for both the bundled python and the virtual environment
./isaaclab.sh -p scripts/tutorials/00_sim/create_empty.py

# Option 2: Using python in your virtual environment
python scripts/tutorials/00_sim/create_empty.py
  • Test training
./isaaclab.sh -p scripts/reinforcement_learning/rsl_rl/train.py --task=Isaac-Ant-v0 --headless
./isaaclab.sh -p scripts/reinforcement_learning/rsl_rl/train.py --task=Isaac-Velocity-Rough-Anymal-C-v0 --headless

Orbit Surgical

# Set the IsaacLab_PATH environment variable to point to your Isaac Lab installation directory
export IsaacLab_PATH="${HOME}/IsaacLab"
  • Clone ORBIT-Surgical repository to a directory outside the Isaac Lab installation directory:
cd ~
git clone https://github.com/orbit-surgical/orbit-surgical
  • Before install, Isaac Sim 4.5 requires to modify the file structure. In source/extensions/orbit.surgica.ext folder, create pyproject.toml that contains the following:
[build-system]
requires = ["setuptools", "wheel", "toml"]
build-backend = "setuptools.build_meta"

Same procedure to orbit.surgical.assets and orbit.surgical.tasks.

  • Install ORBIT-Surgical
conda activate env_isaaclab
cd orbit-surgical/
./orbitsurgical.sh

Set up IDE

To setup the IDE, please follow these instructions:

  1. Open the ORBIT-Surgical directory on Visual Studio Code IDE.
  2. Run VSCode Tasks, by pressing Ctrl+Shift+P, selecting Tasks: Run Task and running the setup_python_env in the drop down menu. When running this task, you will be prompted to add the absolute path to your Isaac Sim installation.

Set your conda environment as the default interpreter in VSCode by opening the command palette (Ctrl+Shift+P), choosing Python: Select Interpreter and selecting your conda environment.

If everything executes correctly, it should create a file .python.env in the .vscode directory. The file contains the python paths to all the extensions provided by Isaac Sim and Omniverse. This helps in indexing all the python modules for intelligent suggestions while writing code.

Verification using Teleoperation

We provide interfaces for providing commands in SE(2) and SE(3) space for robot control. In case of SE(2) teleoperation, the returned command is the linear x-y velocity and yaw rate, while in SE(3), the returned command is a 6-D vector representing the change in pose.

To play inverse kinematics (IK) control with a keyboard device:

${IsaacLab_PATH}/isaaclab.sh -p source/standalone/environments/teleoperation/teleop_se3_agent.py --task Isaac-Lift-Needle-PSM-IK-Rel-v0 --num_envs 1 --teleop_device keyboard

The script prints the teleoperation events configured. For keyboard, these are as follows:

Keyboard Controller for SE(3): Se3Keyboard
    Reset all commands: L
    Toggle gripper (open/close): K
    Move arm along x-axis: W/S
    Move arm along y-axis: A/D
    Move arm along z-axis: Q/E
    Rotate arm along x-axis: Z/X
    Rotate arm along y-axis: T/G
    Rotate arm along z-axis: C/V

Trouble Shooting

No module named pip

cd ${ISAACSIM_PATH}
./python.sh -m ensurepip (reinstall pip)
./python.sh -m pip install numpy==1.22.2 (compatible version with Isaac Sim)

Diverge when collide with ground

  • ground에 충돌시 반작용으로 로봇이 튕겨나오면서 발산함 (ground는 'fixed' 특성을 가짐.)
  • ground를 obstacle로 인식하도록 하거나 z-position 값이 음수가 되지 않도록 설정해야 함.
  • 유사하게, box를 생성하여 ground 아래로 끌어내리면 튕겨나감. 시뮬레이터 자체 physics가 그렇게 설정되어있는 듯.

UserWarning: FigureCanvasAgg is non-interactive, and thus cannot be shown

  • import matplotlib.pyplot as plt
  • plt.show() 함수 호출시 발생
  • 최신 버젼의 PyQt5를 설치해주면 됨. (Isaac Sim 환경에서는 ./python.sh -m pip install PyQt5==5.14.1)

imshow() during launching simulation_app

  • simulation_app과 cv2.imshow()를 함께 사용할 수 없음. 아마도 Qt를 사용하면서 충돌이 있는듯.
  • 따라서 캡쳐한 이미지를 확인하고 싶으면, plt.show()를 활용해야 함.
import matplotlib.pyplot as plt
...
plt.imshow(seg_rgb)
plt.show()

Clone this wiki locally