Creating a Manager-Based RL Environment

Creating a Manager-Based RL Environment#

Having learnt how to create a base environment in Creating a Manager-Based Base Environment, we will now look at how to create a manager-based task environment for reinforcement learning.

The base environment is designed as an sense-act environment where the agent can send commands to the environment and receive observations from the environment. This minimal interface is sufficient for many applications such as traditional motion planning and controls. However, many applications require a task-specification which often serves as the learning objective for the agent. For instance, in a navigation task, the agent may be required to reach a goal location. To this end, we use the envs.ManagerBasedRLEnv class which extends the base environment to include a task specification.

Similar to other components in Isaac Lab, instead of directly modifying the base class envs.ManagerBasedRLEnv, we encourage users to simply implement a configuration envs.ManagerBasedRLEnvCfg for their task environment. This practice allows us to separate the task specification from the environment implementation, making it easier to reuse components of the same environment for different tasks.

In this tutorial, we will configure the cartpole environment using the envs.ManagerBasedRLEnvCfg to create a manager-based task for balancing the pole upright. We will learn how to specify the task using reward terms, termination criteria, curriculum and commands.

The Code#

For this tutorial, we use the cartpole environment defined in isaaclab_tasks.manager_based.classic.cartpole module.

Code for cartpole_env_cfg.py
  1# Copyright (c) 2022-2025, The Isaac Lab Project Developers (https://github.com/isaac-sim/IsaacLab/blob/main/CONTRIBUTORS.md).
  2# All rights reserved.
  3#
  4# SPDX-License-Identifier: BSD-3-Clause
  5
  6# Copyright (c) 2022-2025, The Isaac Lab Project Developers.
  7# All rights reserved.
  8#
  9# SPDX-License-Identifier: BSD-3-Clause
 10
 11import math
 12
 13import isaaclab.sim as sim_utils
 14from isaaclab.assets import ArticulationCfg, AssetBaseCfg
 15from isaaclab.envs import ManagerBasedRLEnvCfg
 16from isaaclab.managers import EventTermCfg as EventTerm
 17from isaaclab.managers import ObservationGroupCfg as ObsGroup
 18from isaaclab.managers import ObservationTermCfg as ObsTerm
 19from isaaclab.managers import RewardTermCfg as RewTerm
 20from isaaclab.managers import SceneEntityCfg
 21from isaaclab.managers import TerminationTermCfg as DoneTerm
 22from isaaclab.scene import InteractiveSceneCfg
 23from isaaclab.utils import configclass
 24
 25import isaaclab_tasks.manager_based.classic.cartpole.mdp as mdp
 26
 27##
 28# Pre-defined configs
 29##
 30from isaaclab_assets.robots.cartpole import CARTPOLE_CFG  # isort:skip
 31
 32
 33##
 34# Scene definition
 35##
 36
 37
 38@configclass
 39class CartpoleSceneCfg(InteractiveSceneCfg):
 40    """Configuration for a cart-pole scene."""
 41
 42    # ground plane
 43    ground = AssetBaseCfg(
 44        prim_path="/World/ground",
 45        spawn=sim_utils.GroundPlaneCfg(size=(100.0, 100.0)),
 46    )
 47
 48    # cartpole
 49    robot: ArticulationCfg = CARTPOLE_CFG.replace(prim_path="{ENV_REGEX_NS}/Robot")
 50
 51    # lights
 52    dome_light = AssetBaseCfg(
 53        prim_path="/World/DomeLight",
 54        spawn=sim_utils.DomeLightCfg(color=(0.9, 0.9, 0.9), intensity=500.0),
 55    )
 56
 57
 58##
 59# MDP settings
 60##
 61
 62
 63@configclass
 64class ActionsCfg:
 65    """Action specifications for the MDP."""
 66
 67    joint_effort = mdp.JointEffortActionCfg(asset_name="robot", joint_names=["slider_to_cart"], scale=100.0)
 68
 69
 70@configclass
 71class ObservationsCfg:
 72    """Observation specifications for the MDP."""
 73
 74    @configclass
 75    class PolicyCfg(ObsGroup):
 76        """Observations for policy group."""
 77
 78        # observation terms (order preserved)
 79        joint_pos_rel = ObsTerm(func=mdp.joint_pos_rel)
 80        joint_vel_rel = ObsTerm(func=mdp.joint_vel_rel)
 81
 82        def __post_init__(self) -> None:
 83            self.enable_corruption = False
 84            self.concatenate_terms = True
 85
 86    # observation groups
 87    policy: PolicyCfg = PolicyCfg()
 88
 89
 90@configclass
 91class EventCfg:
 92    """Configuration for events."""
 93
 94    # reset
 95    reset_cart_position = EventTerm(
 96        func=mdp.reset_joints_by_offset,
 97        mode="reset",
 98        params={
 99            "asset_cfg": SceneEntityCfg("robot", joint_names=["slider_to_cart"]),
100            "position_range": (-1.0, 1.0),
101            "velocity_range": (-0.5, 0.5),
102        },
103    )
104
105    reset_pole_position = EventTerm(
106        func=mdp.reset_joints_by_offset,
107        mode="reset",
108        params={
109            "asset_cfg": SceneEntityCfg("robot", joint_names=["cart_to_pole"]),
110            "position_range": (-0.25 * math.pi, 0.25 * math.pi),
111            "velocity_range": (-0.25 * math.pi, 0.25 * math.pi),
112        },
113    )
114
115
116@configclass
117class RewardsCfg:
118    """Reward terms for the MDP."""
119
120    # (1) Constant running reward
121    alive = RewTerm(func=mdp.is_alive, weight=1.0)
122    # (2) Failure penalty
123    terminating = RewTerm(func=mdp.is_terminated, weight=-2.0)
124    # (3) Primary task: keep pole upright
125    pole_pos = RewTerm(
126        func=mdp.joint_pos_target_l2,
127        weight=-1.0,
128        params={"asset_cfg": SceneEntityCfg("robot", joint_names=["cart_to_pole"]), "target": 0.0},
129    )
130    # (4) Shaping tasks: lower cart velocity
131    cart_vel = RewTerm(
132        func=mdp.joint_vel_l1,
133        weight=-0.01,
134        params={"asset_cfg": SceneEntityCfg("robot", joint_names=["slider_to_cart"])},
135    )
136    # (5) Shaping tasks: lower pole angular velocity
137    pole_vel = RewTerm(
138        func=mdp.joint_vel_l1,
139        weight=-0.005,
140        params={"asset_cfg": SceneEntityCfg("robot", joint_names=["cart_to_pole"])},
141    )
142
143
144@configclass
145class TerminationsCfg:
146    """Termination terms for the MDP."""
147
148    # (1) Time out
149    time_out = DoneTerm(func=mdp.time_out, time_out=True)
150    # (2) Cart out of bounds
151    cart_out_of_bounds = DoneTerm(
152        func=mdp.joint_pos_out_of_manual_limit,
153        params={"asset_cfg": SceneEntityCfg("robot", joint_names=["slider_to_cart"]), "bounds": (-3.0, 3.0)},
154    )
155
156
157##
158# Environment configuration
159##
160
161
162@configclass
163class CartpoleEnvCfg(ManagerBasedRLEnvCfg):
164    """Configuration for the cartpole environment."""
165
166    # Scene settings
167    scene: CartpoleSceneCfg = CartpoleSceneCfg(num_envs=4096, env_spacing=4.0)
168    # Basic settings
169    observations: ObservationsCfg = ObservationsCfg()
170    actions: ActionsCfg = ActionsCfg()
171    events: EventCfg = EventCfg()
172    # MDP settings
173    rewards: RewardsCfg = RewardsCfg()
174    terminations: TerminationsCfg = TerminationsCfg()
175
176    # Post initialization
177    def __post_init__(self) -> None:
178        """Post initialization."""
179        # general settings
180        self.decimation = 2
181        self.episode_length_s = 5
182        # viewer settings
183        self.viewer.eye = (8.0, 0.0, 5.0)
184        # simulation settings
185        self.sim.dt = 1 / 120
186        self.sim.render_interval = self.decimation

The script for running the environment run_cartpole_rl_env.py is present in the isaaclab/scripts/tutorials/03_envs directory. The script is similar to the cartpole_base_env.py script in the previous tutorial, except that it uses the envs.ManagerBasedRLEnv instead of the envs.ManagerBasedEnv.

Code for run_cartpole_rl_env.py
 1# Copyright (c) 2022-2025, The Isaac Lab Project Developers (https://github.com/isaac-sim/IsaacLab/blob/main/CONTRIBUTORS.md).
 2# All rights reserved.
 3#
 4# SPDX-License-Identifier: BSD-3-Clause
 5
 6# Copyright (c) 2022-2025, The Isaac Lab Project Developers.
 7# All rights reserved.
 8#
 9# SPDX-License-Identifier: BSD-3-Clause
10
11"""
12This script demonstrates how to run the RL environment for the cartpole balancing task.
13
14.. code-block:: bash
15
16    ./isaaclab.sh -p scripts/tutorials/03_envs/run_cartpole_rl_env.py --num_envs 32
17
18"""
19
20"""Launch Isaac Sim Simulator first."""
21
22import argparse
23
24from isaaclab.app import AppLauncher
25
26# add argparse arguments
27parser = argparse.ArgumentParser(description="Tutorial on running the cartpole RL environment.")
28parser.add_argument("--num_envs", type=int, default=16, help="Number of environments to spawn.")
29
30# append AppLauncher cli args
31AppLauncher.add_app_launcher_args(parser)
32# parse the arguments
33args_cli = parser.parse_args()
34
35# launch omniverse app
36app_launcher = AppLauncher(args_cli)
37simulation_app = app_launcher.app
38
39"""Rest everything follows."""
40
41import torch
42
43from isaaclab.envs import ManagerBasedRLEnv
44
45from isaaclab_tasks.manager_based.classic.cartpole.cartpole_env_cfg import CartpoleEnvCfg
46
47
48def main():
49    """Main function."""
50    # create environment configuration
51    env_cfg = CartpoleEnvCfg()
52    env_cfg.scene.num_envs = args_cli.num_envs
53    env_cfg.sim.device = args_cli.device
54    # setup RL environment
55    env = ManagerBasedRLEnv(cfg=env_cfg)
56
57    # simulate physics
58    count = 0
59    while simulation_app.is_running():
60        with torch.inference_mode():
61            # reset
62            if count % 300 == 0:
63                count = 0
64                env.reset()
65                print("-" * 80)
66                print("[INFO]: Resetting environment...")
67            # sample random actions
68            joint_efforts = torch.randn_like(env.action_manager.action)
69            # step the environment
70            obs, rew, terminated, truncated, info = env.step(joint_efforts)
71            # print current orientation of pole
72            print("[Env 0]: Pole joint: ", obs["policy"][0][1].item())
73            # update counter
74            count += 1
75
76    # close the environment
77    env.close()
78
79
80if __name__ == "__main__":
81    # run the main function
82    main()
83    # close sim app
84    simulation_app.close()

The Code Explained#

We already went through parts of the above in the Creating a Manager-Based Base Environment tutorial to learn about how to specify the scene, observations, actions and events. Thus, in this tutorial, we will focus only on the RL components of the environment.

In Isaac Lab, we provide various implementations of different terms in the envs.mdp module. We will use some of these terms in this tutorial, but users are free to define their own terms as well. These are usually placed in their task-specific sub-package (for instance, in isaaclab_tasks.manager_based.classic.cartpole.mdp).

Defining rewards#

The managers.RewardManager is used to compute the reward terms for the agent. Similar to the other managers, its terms are configured using the managers.RewardTermCfg class. The managers.RewardTermCfg class specifies the function or callable class that computes the reward as well as the weighting associated with it. It also takes in dictionary of arguments, "params" that are passed to the reward function when it is called.

For the cartpole task, we will use the following reward terms:

  • Alive Reward: Encourage the agent to stay alive for as long as possible.

  • Terminating Reward: Similarly penalize the agent for terminating.

  • Pole Angle Reward: Encourage the agent to keep the pole at the desired upright position.

  • Cart Velocity Reward: Encourage the agent to keep the cart velocity as small as possible.

  • Pole Velocity Reward: Encourage the agent to keep the pole velocity as small as possible.

@configclass
class RewardsCfg:
    """Reward terms for the MDP."""

    # (1) Constant running reward
    alive = RewTerm(func=mdp.is_alive, weight=1.0)
    # (2) Failure penalty
    terminating = RewTerm(func=mdp.is_terminated, weight=-2.0)
    # (3) Primary task: keep pole upright
    pole_pos = RewTerm(
        func=mdp.joint_pos_target_l2,
        weight=-1.0,
        params={"asset_cfg": SceneEntityCfg("robot", joint_names=["cart_to_pole"]), "target": 0.0},
    )
    # (4) Shaping tasks: lower cart velocity
    cart_vel = RewTerm(
        func=mdp.joint_vel_l1,
        weight=-0.01,
        params={"asset_cfg": SceneEntityCfg("robot", joint_names=["slider_to_cart"])},
    )
    # (5) Shaping tasks: lower pole angular velocity
    pole_vel = RewTerm(
        func=mdp.joint_vel_l1,
        weight=-0.005,
        params={"asset_cfg": SceneEntityCfg("robot", joint_names=["cart_to_pole"])},
    )

Defining termination criteria#

Most learning tasks happen over a finite number of steps that we call an episode. For instance, in the cartpole task, we want the agent to balance the pole for as long as possible. However, if the agent reaches an unstable or unsafe state, we want to terminate the episode. On the other hand, if the agent is able to balance the pole for a long time, we want to terminate the episode and start a new one so that the agent can learn to balance the pole from a different starting configuration.

The managers.TerminationsCfg configures what constitutes for an episode to terminate. In this example, we want the task to terminate when either of the following conditions is met:

  • Episode Length The episode length is greater than the defined max_episode_length

  • Cart out of bounds The cart goes outside of the bounds [-3, 3]

The flag managers.TerminationsCfg.time_out specifies whether the term is a time-out (truncation) term or terminated term. These are used to indicate the two types of terminations as described in Gymnasium’s documentation.

@configclass
class TerminationsCfg:
    """Termination terms for the MDP."""

    # (1) Time out
    time_out = DoneTerm(func=mdp.time_out, time_out=True)
    # (2) Cart out of bounds
    cart_out_of_bounds = DoneTerm(
        func=mdp.joint_pos_out_of_manual_limit,
        params={"asset_cfg": SceneEntityCfg("robot", joint_names=["slider_to_cart"]), "bounds": (-3.0, 3.0)},
    )

Defining commands#

For various goal-conditioned tasks, it is useful to specify the goals or commands for the agent. These are handled through the managers.CommandManager. The command manager handles resampling and updating the commands at each step. It can also be used to provide the commands as an observation to the agent.

For this simple task, we do not use any commands. Hence, we leave this attribute as its default value, which is None. You can see an example of how to define a command manager in the other locomotion or manipulation tasks.

Defining curriculum#

Often times when training a learning agent, it helps to start with a simple task and gradually increase the tasks’s difficulty as the agent training progresses. This is the idea behind curriculum learning. In Isaac Lab, we provide a managers.CurriculumManager class that can be used to define a curriculum for your environment.

In this tutorial we don’t implement a curriculum for simplicity, but you can see an example of a curriculum definition in the other locomotion or manipulation tasks.

Tying it all together#

With all the above components defined, we can now create the ManagerBasedRLEnvCfg configuration for the cartpole environment. This is similar to the ManagerBasedEnvCfg defined in Creating a Manager-Based Base Environment, only with the added RL components explained in the above sections.

@configclass
class CartpoleEnvCfg(ManagerBasedRLEnvCfg):
    """Configuration for the cartpole environment."""

    # Scene settings
    scene: CartpoleSceneCfg = CartpoleSceneCfg(num_envs=4096, env_spacing=4.0)
    # Basic settings
    observations: ObservationsCfg = ObservationsCfg()
    actions: ActionsCfg = ActionsCfg()
    events: EventCfg = EventCfg()
    # MDP settings
    rewards: RewardsCfg = RewardsCfg()
    terminations: TerminationsCfg = TerminationsCfg()

    # Post initialization
    def __post_init__(self) -> None:
        """Post initialization."""
        # general settings
        self.decimation = 2
        self.episode_length_s = 5
        # viewer settings
        self.viewer.eye = (8.0, 0.0, 5.0)
        # simulation settings
        self.sim.dt = 1 / 120
        self.sim.render_interval = self.decimation

Running the simulation loop#

Coming back to the run_cartpole_rl_env.py script, the simulation loop is similar to the previous tutorial. The only difference is that we create an instance of envs.ManagerBasedRLEnv instead of the envs.ManagerBasedEnv. Consequently, now the envs.ManagerBasedRLEnv.step() method returns additional signals such as the reward and termination status. The information dictionary also maintains logging of quantities such as the reward contribution from individual terms, the termination status of each term, the episode length etc.

def main():
    """Main function."""
    # 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)

    # simulate physics
    count = 0
    while simulation_app.is_running():
        with torch.inference_mode():
            # reset
            if count % 300 == 0:
                count = 0
                env.reset()
                print("-" * 80)
                print("[INFO]: Resetting environment...")
            # sample random actions
            joint_efforts = torch.randn_like(env.action_manager.action)
            # step the environment
            obs, rew, terminated, truncated, info = env.step(joint_efforts)
            # print current orientation of pole
            print("[Env 0]: Pole joint: ", obs["policy"][0][1].item())
            # update counter
            count += 1

    # close the environment
    env.close()

The Code Execution#

Similar to the previous tutorial, we can run the environment by executing the run_cartpole_rl_env.py script.

./isaaclab.sh -p scripts/tutorials/03_envs/run_cartpole_rl_env.py --num_envs 32

This should open a similar simulation as in the previous tutorial. However, this time, the environment returns more signals that specify the reward and termination status. Additionally, the individual environments reset themselves when they terminate based on the termination criteria specified in the configuration.

result of run_cartpole_rl_env.py

To stop the simulation, you can either close the window, or press Ctrl+C in the terminal where you started the simulation.

In this tutorial, we learnt how to create a task environment for reinforcement learning. We do this by extending the base environment to include the rewards, terminations, commands and curriculum terms. We also learnt how to use the envs.ManagerBasedRLEnv class to run the environment and receive various signals from it.

While it is possible to manually create an instance of envs.ManagerBasedRLEnv class for a desired task, this is not scalable as it requires specialized scripts for each task. Thus, we exploit the gymnasium.make() function to create the environment with the gym interface. We will learn how to do this in the next tutorial.