First Arena Environment

After setting up the docker container and installing isaaclab_arena, learn how to compose your first simple IsaacLab Arena environment by combining assets, scenes, and tasks.

Once within the docker container, run the following command to compile your first IsaacLab Arena environment:

python isaaclab_arena/examples/compile_env_notebook.py

The compiled environment will spawn in an Isaac Sim instance and run for some steps with zero actions. You should see the following scene:

Code Explanation

The following script demonstrates how to create the simple kitchen environment from above with a Franka robot and a cracker box object using the isaaclab_arena API.

Create a Simple Environment

import torch
import tqdm

import pinocchio  # noqa: F401
from isaaclab.app import AppLauncher

# Launch the Isaac Sim application
print("Launching simulation app")
simulation_app = AppLauncher()

from isaaclab_arena.assets.asset_registry import AssetRegistry
from isaaclab_arena.cli.isaaclab_arena_cli import get_isaaclab_arena_cli_parser
from isaaclab_arena.environments.arena_env_builder import ArenaEnvBuilder
from isaaclab_arena.environments.isaaclab_arena_environment import IsaacLabArenaEnvironment
from isaaclab_arena.scene.scene import Scene
from isaaclab_arena.tasks.dummy_task import DummyTask
from isaaclab_arena.utils.pose import Pose

# Step 1: Initialize and get the assets from the registry
asset_registry = AssetRegistry()

background = asset_registry.get_asset_by_name("kitchen")()
embodiment = asset_registry.get_asset_by_name("franka")()
cracker_box = asset_registry.get_asset_by_name("cracker_box")()
cracker_box.set_initial_pose(
    Pose(position_xyz=(0.4, 0.0, 0.1), rotation_wxyz=(1.0, 0.0, 0.0, 0.0))
)

# Step 2: Create a scene with the assets
scene = Scene(assets=[background, cracker_box])

# Step 3: Create a task
task = DummyTask()

# Step 4: Create the IsaacLab Arena environment
isaaclab_arena_environment = IsaacLabArenaEnvironment(
    name="my_first_arena_env",
    embodiment=embodiment,
    scene=scene,
    task=task,
    teleop_device=None,
)

# Step 5: Build and compile the environment
args_cli = get_isaaclab_arena_cli_parser().parse_args([])
env_builder = ArenaEnvBuilder(isaaclab_arena_environment, args_cli)
env = env_builder.make_registered()
env.reset()

# Step 6: Run the simulation with zero actions
NUM_STEPS = 1000
for _ in tqdm.tqdm(range(NUM_STEPS)):
    with torch.inference_mode():
        actions = torch.zeros(env.action_space.shape, device=env.unwrapped.device)
        env.step(actions)

Step-by-Step Breakdown

1. Initialize and interac with the Asset Registry

asset_registry = AssetRegistry()

The AssetRegistry provides access to all available assets including robots, objects, and backgrounds. It automatically discovers registered assets through the registration system.

See Assets Design for details on asset architecture.

background = asset_registry.get_asset_by_name("kitchen")()
embodiment = asset_registry.get_asset_by_name("franka")()
cracker_box = asset_registry.get_asset_by_name("cracker_box")()

2. Compose the Scene

scene = Scene(assets=[background, cracker_box])

See Scene Design for scene composition details.

3. Create a Task

A task defines the objective, success criteria, and behavior logic for the environment. For this example, we use the DummyTask.

task = DummyTask()

See Tasks Design for task creation details.

4. Create the IsaacLab Arena Environment

isaaclab_arena_environment = IsaacLabArenaEnvironment(
    name="my_first_arena_env",
    embodiment=embodiment,
    scene=scene,
    task=DummyTask(),
    teleop_device=None,
)

This puts everything together into an IsaacLabArenaEnvironment object.

See Environment Design for environment composition details.

5. Build the Environment

args_cli = get_isaaclab_arenaena_cli_parser().parse_args([])
env_builder = ArenaEnvBuilder(isaaclab_arena_environment, args_cli)
env = env_builder.make_registered()
env.reset()

The ArenaEnvBuilder compiles the high-level environment description into Isaac Lab configurations. See Environment Compilation Design for compilation details.

6. Run the Simulation

for _ in range(NUM_STEPS):
    with torch.inference_mode():
        actions = torch.zeros(env.action_space.shape, device=env.unwrapped.device)
        env.step(actions)

This is just a standard Isaac Lab simulation loop with zero actions.

Next Steps

Now that you have created your first environment, explore:

• Tasks Design - Create custom tasks with rewards and terminations

• Assets Design - Discover available assets and create custom ones

• Affordances Design - Add interactive behaviors to objects

Explore pre-built example environments in isaaclab_arena/examples/example_environments/ for more complex scenarios.

To move on to data generation and training (Imitation Learning), please refer to the G1 Loco-Manipulation Box Pick and Place Task or GR1 Open Microwave Door Task pages.