Isaac Lab

https://isaac-sim.github.io/IsaacLab/main/
Admin
Isaac Lab is a unified and modular framework for robot learning, built on NVIDIA Isaac Sim, designed...
Tokens:2,079,602
Snippets:11,768
Trust Score:8.3
Update:1 day ago
Show doc for...
Tokens:
Raw
### Start container with container.py (after migration)

Source: https://isaac-sim.github.io/IsaacLab/main/source/refs/release_notes.html

The new command to start the container using the Python script container.py. This is shown as the 'Now' example in the release notes.

```bash
./docker/container.py start
```

--------------------------------

### Install Ray on local machine (ray[default]==2.31.0)

Source: https://isaac-sim.github.io/IsaacLab/main/source/features/ray.html

Install Ray version 2.31.0 with default dependencies on the local machine, a shared step for both KubeRay and pure Ray setups.

```bash
python3 -p -m pip install ray[default]==2.31.0
```

--------------------------------

### Clean up existing installations for MicroK8s setup

Source: https://isaac-sim.github.io/IsaacLab/main/source/deployment/cloudxr_teleoperation_cluster.html

Optional cleanup commands to remove MicroK8s, Helm, and Docker to ensure a fresh start before installing MicroK8s.

```bash
# Clean up the system to ensure we start fresh
sudo snap remove microk8s
sudo snap remove helm
sudo apt-get remove docker-ce docker-ce-cli containerd.io
# If you have snap docker installed, remove it as well
sudo snap remove docker
```

--------------------------------

### Main Function Setup

Source: https://isaac-sim.github.io/IsaacLab/main/source/tutorials/05_controllers/run_osc.html

Initializes the simulation context with a time step of 0.01 seconds and the device from command-line arguments, and sets the main camera.

```python
def main():
    """Main function."""
    # Load kit helper
    sim_cfg = sim_utils.SimulationCfg(dt=0.01, device=args_cli.device)
    sim = sim_utils.SimulationContext(sim_cfg)
    # Set main camera
```

--------------------------------

### Get Prim Type Name Example

Source: https://isaac-sim.github.io/IsaacLab/main/source/api/lab/isaaclab.sim.utils.html

Shows how to use get_prim_type_name to retrieve the type name of a prim at a given path. The example imports isaaclab.sim as sim_utils, gets the current stage, and queries the prim at '/World/Cube'. The output is 'Cube'.

```python
>>> import isaaclab.sim as sim_utils
>>>
>>> stage = sim_utils.get_current_stage()
>>> prim = stage.GetPrimAtPath("/World/Cube")
>>> prim.GetTypeName()
"Cube"
```

--------------------------------

### Run launch_app.py with --help

Source: https://isaac-sim.github.io/IsaacLab/main/source/tutorials/00_sim/launch_app.html

Executes the tutorial script with the --help flag to display combined help output from custom arguments and AppLauncher arguments. The output shows usage, options, and app_launcher arguments, with INFO messages indicating which arguments configure the SimulationApp.

```bash
./isaaclab.sh -p scripts/tutorials/00_sim/launch_app.py --help

[INFO] Using python from: /isaac-sim/python.sh
[INFO][AppLauncher]: The argument 'width' will be used to configure the SimulationApp.
[INFO][AppLauncher]: The argument 'height' will be used to configure the SimulationApp.
usage: launch_app.py [-h] [--size SIZE] [--width WIDTH] [--height HEIGHT] [--headless] [--livestream {0,1,2}]
                     [--enable_cameras] [--verbose] [--experience EXPERIENCE]

Tutorial on running IsaacSim via the AppLauncher.

options:
-h, --help            show this help message and exit
--size SIZE           Side-length of cuboid
--width WIDTH         Width of the viewport and generated images. Defaults to 1280
--height HEIGHT       Height of the viewport and generated images. Defaults to 720

app_launcher arguments:
--headless            Force display off at all times.
--livestream {0,1,2}
                      Force enable livestreaming. Mapping corresponds to that for the "LIVESTREAM" environment variable.
--enable_cameras      Enable cameras when running without a GUI.
--verbose             Enable verbose terminal logging from the SimulationApp.
--experience EXPERIENCE
                      The experience file to load when launching the SimulationApp.

                      * If an empty string is provided, the experience file is determined based on the headless flag.
                      * If a relative path is provided, it is resolved relative to the `apps` folder in Isaac Sim and
                        Isaac Lab (in that order).
```

--------------------------------

### Install Robomimic dependencies and module

Source: https://isaac-sim.github.io/IsaacLab/main/source/overview/imitation-learning/augmented_imitation.html

Install system dependencies and the Robomimic Python module via Isaac Lab's setup script. Requires sudo for apt install.

```bash
# install the dependencies
sudo apt install cmake build-essential
# install python module (for robomimic)
./isaaclab.sh -i robomimic
```

--------------------------------

### Launch environment with --enable_cameras

Source: https://isaac-sim.github.io/IsaacLab/main/source/overview/core-concepts/sensors/camera.html

Command-line example showing how to launch a reinforcement learning training script with the --enable_cameras argument, which is required when working with rendering.

```bash
python scripts/reinforcement_learning/rl_games/train.py --task=Isaac-Cartpole-RGB-Camera-Direct-v0 --headless --enable_cameras
```

--------------------------------

### get_subtask_start_signals

Source: https://isaac-sim.github.io/IsaacLab/main/source/api/lab/isaaclab.envs.html

Gets a dictionary of start signal flags for each subtask in a task. The flag is 1 when the subtask has started and 0 otherwise. This method is required for automatic subtask start signal annotation.

```APIDOC
## get_subtask_start_signals

### Description
Gets a dictionary of start signal flags for each subtask in a task. The flag is 1 when the subtask has started and 0 otherwise. This method is required for automatic subtask start signal annotation.

### Method
Function

### Signature
`get_subtask_start_signals(env_ids: Sequence[int] | None = None) -> dict[str, torch.Tensor]`

### Parameters
- **env_ids** (Sequence[int] | None) - Optional - Environment indices to get the start signals for. If None, all envs are considered.

### Returns
- **dict[str, torch.Tensor]** - A dictionary of start signal flags (False or True) for each subtask.
```

--------------------------------

### Manual environment creation (previous tutorial)

Source: https://isaac-sim.github.io/IsaacLab/main/source/tutorials/03_envs/register_rl_env_gym.html

Shows the manual approach to creating an environment by importing the environment class and configuration class, then instantiating with a config. This is the baseline that the tutorial replaces with gymnasium registration.

```python
    # create environment configuration
    env_cfg = CartpoleEnvCfg()
    env_cfg.scene.num_envs = args_cli.num_envs
    env_cfg.sim.device = args_cli.device
    # setup RL environment
    env = ManagerBasedRLEnv(cfg=env_cfg)
```

--------------------------------

### __init__

Source: https://isaac-sim.github.io/IsaacLab/main/source/api/lab_mimic/isaaclab_mimic.envs.html

Initialize the environment.

```APIDOC
## __init__

### Description
Initialize the environment.

### Method
`__init__(cfg: ManagerBasedRLEnvCfg, render_mode: str | None = None, **kwargs)`

### Parameters
- **cfg** (ManagerBasedRLEnvCfg) - Required - The configuration for the environment.
- **render_mode** (str | None) - Optional - The render mode.
- **kwargs** - Optional - Additional keyword arguments.
```

--------------------------------

### Example pip show output for rsl-rl

Source: https://isaac-sim.github.io/IsaacLab/main/source/how-to/add_own_library.html

Example output from the pip show command, showing the library is installed from the cloned directory.

```text
Name: rsl_rl
Version: 3.0.1
Summary: Fast and simple RL algorithms implemented in pytorch
Home-page: https://github.com/leggedrobotics/rsl_rl
Author: ETH Zurich, NVIDIA CORPORATION
Author-email:
License: BSD-3
Location: /home/user/git/rsl_rl
Requires: torch, torchvision, numpy, GitPython, onnx
Required-by:
```

--------------------------------

### Install specific learning framework on Windows

Source: https://isaac-sim.github.io/IsaacLab/main/source/setup/installation/binaries_installation.html

Installs only the specified learning framework for Isaac Lab on Windows. For example, to install only 'rl_games'. Valid options: all, rl_games, rsl_rl, sb3, skrl, robomimic, none.

```batch
isaaclab.bat --install rl_games :: or "isaaclab.bat -i rl_games"
```

--------------------------------

### Run launch_app.py with LIVESTREAM=2

Source: https://isaac-sim.github.io/IsaacLab/main/source/tutorials/00_sim/launch_app.html

Runs the tutorial script with LIVESTREAM=2, which implies headless mode and enables WebRTC livestreaming. Spawns a 0.5m3 cuboid. No GUI appears. Kill with Ctrl+C.

```bash
LIVESTREAM=2 ./isaaclab.sh -p scripts/tutorials/00_sim/launch_app.py --size 0.5
```

--------------------------------

### Install specific learning framework on Linux

Source: https://isaac-sim.github.io/IsaacLab/main/source/setup/installation/binaries_installation.html

Installs only the specified learning framework for Isaac Lab on Linux. For example, to install only 'rl_games'. Valid options: all, rl_games, rsl_rl, sb3, skrl, robomimic, none.

```bash
./isaaclab.sh --install rl_games  # or "./isaaclab.sh -i rl_games"
```

--------------------------------

### DirectMARLEnv.__init__

Source: https://isaac-sim.github.io/IsaacLab/main/source/api/lab/isaaclab.envs.html

Initialize the environment with a configuration object and optional render mode.

```APIDOC
## DirectMARLEnv.__init__

### Description
Initialize the environment.

### Method
Constructor

### Signature
__init__(cfg: DirectMARLEnvCfg, render_mode: str | None = None, **kwargs)

### Parameters
- **cfg** (DirectMARLEnvCfg) - Required - The configuration object for the environment.
- **render_mode** (str | None) - Optional - The render mode for the environment. Defaults to None, which is similar to "human".

### Raises
- **RuntimeError** - If a simulation context already exists. The environment must always create one since it configures the simulation context and controls the simulation.
```

--------------------------------

### Se3Keyboard.__init__(cfg)

Source: https://isaac-sim.github.io/IsaacLab/main/source/api/lab/isaaclab.devices.html

Initializes the keyboard layer with the given configuration.

```APIDOC
## Se3Keyboard.__init__(cfg)

### Description
Initializes the keyboard layer with the given configuration.

### Method
Constructor

### Parameters
- **cfg** (Se3KeyboardCfg) - Required - Configuration object for keyboard settings.

```

--------------------------------

### Initialize and Setup Scene in IsaacLabTutorialEnv

Source: https://isaac-sim.github.io/IsaacLab/main/source/setup/walkthrough/technical_env_design.html

Replaces the __init__ and _setup_scene methods in isaac_lab_tutorial_env.py. The __init__ finds joint indices from the robot, and _setup_scene creates the robot articulation, spawns a ground plane, clones environments, adds the robot to the scene, and adds a dome light.

```python
class IsaacLabTutorialEnv(DirectRLEnv):
    cfg: IsaacLabTutorialEnvCfg

    def __init__(self, cfg: IsaacLabTutorialEnvCfg, render_mode: str | None = None, **kwargs):
        super().__init__(cfg, render_mode, **kwargs)

        self.dof_idx, _ = self.robot.find_joints(self.cfg.dof_names)

    def _setup_scene(self):
        self.robot = Articulation(self.cfg.robot_cfg)
        # add ground plane
        spawn_ground_plane(prim_path="/World/ground", cfg=GroundPlaneCfg())
        # clone and replicate
        self.scene.clone_environments(copy_from_source=False)
        # add articulation to scene
        self.scene.articulations["robot"] = self.robot
        # add lights
        light_cfg = sim_utils.DomeLightCfg(intensity=2000.0, color=(0.75, 0.75, 0.75))
        light_cfg.func("/World/Light", light_cfg)
```

--------------------------------

### Start container with container.sh (before migration)

Source: https://isaac-sim.github.io/IsaacLab/main/source/refs/release_notes.html

The previous command to start the container using the bash script container.sh. This is shown as the 'Before' example in the release notes.

```bash
./docker/container.sh start
```

--------------------------------

### Run pick_and_place.py demo on Linux

Source: https://isaac-sim.github.io/IsaacLab/main/source/overview/showroom.html

Launches the interactive pick-and-place demo with a parallel robot. Use this command on Linux systems. The goal is to pick up the purple cube and drop it on the red sphere. Hold the A key to have the gripper track the cube position.

```bash
./isaaclab.sh -p scripts/demos/pick_and_place.py
```

--------------------------------

### RmpFlowGalbotCubeStackAbsMimicEnv.get_subtask_start_signals

Source: https://isaac-sim.github.io/IsaacLab/main/source/api/lab_mimic/isaaclab_mimic.envs.html

Gets a dictionary of start signal flags for each subtask in a task.

```APIDOC
## get_subtask_start_signals

### Description
Gets a dictionary of start signal flags for each subtask in a task.

### Method
get_subtask_start_signals

### Parameters
- **env_ids** (optional) - Environment IDs.

### Returns
Not specified in source.
```

--------------------------------

### Install KubeRay dependencies (kubernetes, Jinja2)

Source: https://isaac-sim.github.io/IsaacLab/main/source/features/ray.html

Install Python packages required for KubeRay cluster setup on Google GKE with the batteries-included cluster launch file.

```bash
python3 -p -m pip install kubernetes Jinja2
```

--------------------------------

### Isaac Lab _setup_scene() method

Source: https://isaac-sim.github.io/IsaacLab/main/source/migration/migrating_from_isaacgymenvs.html

Shows the Isaac Lab replacement for create_sim(), using _setup_scene() to add actors, ground plane, clone environments, and add lights.

```python
def _setup_scene(self):
  self.cartpole = Articulation(self.cfg.robot_cfg)
  # add ground plane
  spawn_ground_plane(prim_path="/World/ground", cfg=GroundPlaneCfg()
  # clone, filter, and replicate
  self.scene.clone_environments(copy_from_source=False)
  self.scene.filter_collisions(global_prim_paths=[])
  # add articulation to scene
  self.scene.articulations["cartpole"] = self.cartpole
  # add lights
  light_cfg = sim_utils.DomeLightCfg(intensity=2000.0)
  light_cfg.func("/World/Light", light_cfg)

```

--------------------------------

### get_subtask_start_signals() (isaaclab.envs.ManagerBasedRLMimicEnv method)

Source: https://isaac-sim.github.io/IsaacLab/main/genindex.html

Method to get subtask start signals for the environment.

```APIDOC
## get_subtask_start_signals()

### Description
Retrieves the subtask start signals for the environment.

### Method
- **get_subtask_start_signals()**

### Class
- `isaaclab.envs.ManagerBasedRLMimicEnv`

### Module
- `isaaclab.envs`

### Parameters
- No parameters documented.

### Returns
- Not documented in source.
```

--------------------------------

### Collect and replay demonstrations for Isaac-Stack-Cube-Franka-IK-Rel-v0

Source: https://isaac-sim.github.io/IsaacLab/main/source/overview/imitation-learning/teleop_imitation.html

These commands create a dataset folder, collect demonstrations using a teleoperation device (spacemouse, keyboard, or handtracking), and replay the collected dataset. Replace <teleop_device> with your preferred input device. About 10 successful demonstrations are required for subsequent steps.

```bash
# step a: create folder for datasets
mkdir -p datasets
# step b: collect data with a selected teleoperation device. Replace <teleop_device> with your preferred input device.
# Available options: spacemouse, keyboard, handtracking
./isaaclab.sh -p scripts/tools/record_demos.py --task Isaac-Stack-Cube-Franka-IK-Rel-v0 --device cpu --teleop_device <teleop_device> --dataset_file ./datasets/dataset.hdf5 --num_demos 10
# step a: replay the collected dataset
./isaaclab.sh -p scripts/tools/replay_demos.py --task Isaac-Stack-Cube-Franka-IK-Rel-v0 --device cpu --dataset_file ./datasets/dataset.hdf5
```

--------------------------------

### get_subtask_start_signals([env_ids])

Source: https://isaac-sim.github.io/IsaacLab/main/source/api/lab/isaaclab.envs.html

Get a dictionary of start signal flags for each subtask in a task.

```APIDOC
## get_subtask_start_signals([env_ids])

### Description
Gets a dictionary of start signal flags for each subtask in a task.

### Method
Class method

### Parameters
- **env_ids** (Tensor, optional) - Optional - Environment IDs to query. If None, all environments are used.

### Returns
- **start_signals** (dict) - A dictionary of start signal flags for each subtask.
```

--------------------------------

### Create prim example (create_prim)

Source: https://isaac-sim.github.io/IsaacLab/main/source/api/lab/isaaclab.sim.utils.html

Shows how to create a cube and a sphere using sim_utils.create_prim. The cube is placed at a world position, while the sphere uses local translation and scale. The output shows the created Usd.Prim objects.

```python
>>> import isaaclab.sim as sim_utils
>>>
>>> # Create a cube at world position (1.0, 0.5, 0.0)
>>> sim_utils.create_prim(
...     prim_path="/World/Parent/Cube",
...     prim_type="Cube",
...     position=(1.0, 0.5, 0.0),
...     attributes={"size": 2.0},
... )
Usd.Prim(</World/Parent/Cube>)
>>>
>>> # Create a sphere with local translation relative to its parent
>>> sim_utils.create_prim(
...     prim_path="/World/Parent/Sphere",
...     prim_type="Sphere",
...     translation=(0.5, 0.0, 0.0),
...     scale=(2.0, 2.0, 2.0),
... )
Usd.Prim(</World/Parent/Sphere>)
```

--------------------------------

### Get prim path using GetPath().pathString

Source: https://isaac-sim.github.io/IsaacLab/main/source/api/lab/isaaclab.sim.utils.html

Deprecated since 2.3.0. Use USD APIs directly. This example shows how to get the path of a USD prim.

```python
>>> import isaaclab.sim as sim_utils
>>>
>>> stage = sim_utils.get_current_stage()
>>> prim = stage.GetPrimAtPath("/World/Cube")
>>> prim.GetPath().pathString
"/World/Cube"
```

--------------------------------

### Run pick_and_place.py demo on Windows

Source: https://isaac-sim.github.io/IsaacLab/main/source/overview/showroom.html

Launches the interactive pick-and-place demo with a parallel robot. Use this command on Windows systems. The goal is to pick up the purple cube and drop it on the red sphere. Hold the A key to have the gripper track the cube position.

```batch
isaaclab.bat -p scripts\demos\pick_and_place.py
```

--------------------------------

### Get prim at path using GetPrimAtPath

Source: https://isaac-sim.github.io/IsaacLab/main/source/api/lab/isaaclab.sim.utils.html

Deprecated since 2.3.0. Use USD APIs directly. This example shows how to get a USD prim at a specified path.

```python
>>> import isaaclab.sim as sim_utils
>>>
>>> stage = sim_utils.get_current_stage()
>>> stage.GetPrimAtPath("/World/Cube")
Usd.Prim(</World/Cube>)
```

--------------------------------

### Main training script setup and environment creation

Source: https://isaac-sim.github.io/IsaacLab/main/source/tutorials/03_envs/configuring_rl_training.html

Sets the IO descriptors export flag for manager-based RL environments, sets the log directory, creates the Isaac environment, converts to single-agent if needed, and wraps for video recording.

```python
151    # set the IO descriptors export flag if requested
152    if isinstance(env_cfg, ManagerBasedRLEnvCfg):
153        env_cfg.export_io_descriptors = args_cli.export_io_descriptors
154    else:
155        logger.warning(
156            "IO descriptors are only supported for manager based RL environments. No IO descriptors will be exported."
157        )
158
159    # set the log directory for the environment (works for all environment types)
160    env_cfg.log_dir = log_dir
161
162    # create isaac environment
163    env = gym.make(args_cli.task, cfg=env_cfg, render_mode="rgb_array" if args_cli.video else None)
164
165    # convert to single-agent instance if required by the RL algorithm
166    if isinstance(env.unwrapped, DirectMARLEnv):
167        env = multi_agent_to_single_agent(env)
168
169    # wrap for video recording
170    if args_cli.video:
171        video_kwargs = {
172            "video_folder": os.path.join(log_dir, "videos", "train"),
173            "step_trigger": lambda step: step % args_cli.video_interval == 0,
174            "video_length": args_cli.video_length,
175            "disable_logger": True,
176        }
177        print("[INFO] Recording videos during training.")
178        print_dict(video_kwargs, nesting=4)
179        env = gym.wrappers.RecordVideo(env, **video_kwargs)
```

--------------------------------

### Get stage up axis using UsdGeom.GetStageUpAxis

Source: https://isaac-sim.github.io/IsaacLab/main/source/api/lab/isaaclab.sim.utils.html

Deprecated since 2.3.0. Use USD APIs directly. This example shows how to get the up axis of the current stage using UsdGeom.GetStageUpAxis.

```python
>>> import isaaclab.sim as sim_utils
>>> from pxr import UsdGeom
>>>
>>> UsdGeom.GetStageUpAxis(sim_utils.get_current_stage())
'Z'
```

--------------------------------

### __init__(cfg, render_mode=None, **kwargs)

Source: https://isaac-sim.github.io/IsaacLab/main/source/api/lab/isaaclab.envs.html

Initializes the environment with the given configuration and render mode.

```APIDOC
## __init__(cfg, render_mode=None, **kwargs)

### Description
Initialize the environment.

### Method
Python constructor

### Parameters
- **cfg** (ManagerBasedRLEnvCfg) - The configuration for the environment.
- **render_mode** (str | None, optional) - The render mode for the environment. Defaults to None, which is similar to "human".
```

--------------------------------

### Example usage of build_simulation_context

Source: https://isaac-sim.github.io/IsaacLab/main/source/api/lab/isaaclab.sim.html

Shows how to use the context manager to build a simulation context, design the scene, and run the simulation loop. The example uses default settings and includes reset, is_playing, and step calls.

```python
with build_simulation_context() as sim:
    # Design the scene

    # Play the simulation
    sim.reset()
    while sim.is_playing():
        sim.step()
```

--------------------------------

### get_subtask_start_signals() (isaaclab_mimic.envs.pick_place_mimic_env.PickPlaceRelMimicEnv method)

Source: https://isaac-sim.github.io/IsaacLab/main/genindex.html

Method to get subtask start signals for the pick-place relative mimic environment.

```APIDOC
## get_subtask_start_signals()

### Description
Retrieves the subtask start signals for the pick-place relative mimic environment.

### Method
- **get_subtask_start_signals()**

### Class
- `isaaclab_mimic.envs.pick_place_mimic_env.PickPlaceRelMimicEnv`

### Module
- `isaaclab_mimic.envs`

### Parameters
- No parameters documented.

### Returns
- Not documented in source.
```

--------------------------------

### AppLauncher.__init__

Source: https://isaac-sim.github.io/IsaacLab/main/source/api/lab/isaaclab.app.html

Creates a SimulationApp instance based on input settings. Accepts launcher arguments and keyword arguments, merging them with environment variables and CLI arguments.

```APIDOC
## AppLauncher.__init__

### Description
Creates a SimulationApp instance based on the input settings. The launcher arguments and keyword arguments are merged and used to configure the simulation app.

### Method
Constructor

### Signature
__init__(launcher_args: Namespace | dict | None = None, **kwargs)

### Parameters
- **launcher_args** (Namespace | dict | None) - Optional - Input arguments to parse using the AppLauncher and set into the SimulationApp. Defaults to None, which is equivalent to passing an empty dictionary.
- **\*\*kwargs** - Optional - Additional keyword arguments that will be merged into `launcher_args`. Duplicated arguments with `launcher_args` will raise a ValueError.

### Raises
- **ValueError** - If there are common/duplicated arguments between `launcher_args` and `kwargs`.
- **ValueError** - If combination of `launcher_args` and `kwargs` are missing the necessary arguments that are needed by the AppLauncher to resolve the desired app configuration.
- **ValueError** - If incompatible or undefined values are assigned to relevant environment values, such as `LIVESTREAM`.

### Example
```python
from isaaclab.app import AppLauncher

# Launch with default settings
app_launcher = AppLauncher()

# Launch with custom arguments
app_launcher = AppLauncher(launcher_args={'headless': True})
```
```

--------------------------------

### get_subtask_start_signals() (isaaclab_mimic.envs.pick_place_mimic_env.PickPlaceAbsMimicEnv method)

Source: https://isaac-sim.github.io/IsaacLab/main/genindex.html

Method to get subtask start signals for the pick-place absolute mimic environment.

```APIDOC
## get_subtask_start_signals()

### Description
Retrieves the subtask start signals for the pick-place absolute mimic environment.

### Method
- **get_subtask_start_signals()**

### Class
- `isaaclab_mimic.envs.pick_place_mimic_env.PickPlaceAbsMimicEnv`

### Module
- `isaaclab_mimic.envs`

### Parameters
- No parameters documented.

### Returns
- Not documented in source.
```
Tokens:
Raw