Action Functors

Action Functors#

This page lists all available action terms that can be used with the Action Manager. Action terms are configured using ActionTermCfg and are responsible for processing raw actions from the policy and converting them to the format expected by the robot (e.g., qpos, qvel, qf).

Joint Position Control#

Joint Position Action Terms#

Action Term

Description

DeltaQposTerm

Delta joint position action: current_qpos + scale * action -> qpos. The policy outputs position deltas relative to the current joint positions.

{"func": "DeltaQposTerm", "params": {"scale": 0.1}}

QposTerm

Absolute joint position action: scale * action -> qpos. The policy outputs direct target joint positions.

{"func": "QposTerm", "params": {"scale": 1.0}}

QposDenormalizedTerm

Normalized action in [-1, 1] -> denormalize to joint limits -> qpos. The policy outputs normalized values that are mapped to joint limits. With scale=1.0 (default), action in [-1, 1] maps to [low, high].

{"func": "QposDenormalizedTerm", "params": {"scale": 1.0}}

QposNormalizedTerm

Normalize action from qpos limits -> [range[0], range[1]]. Maps joint positions to a normalized range based on joint limits. Typically used for post-processing action outputs.

{"func": "QposNormalizedTerm", "params": {"range": [0.0, 1.0]}}

End-Effector Control#

End-Effector Action Terms#

Action Term

Description

EefPoseTerm

End-effector pose (6D or 7D) -> IK -> qpos. The policy outputs target end-effector poses which are converted to joint positions via inverse kinematics. Returns ik_success in the output so reward/observation can penalize or condition on IK failures. Supports both 6D (euler angles) and 7D (quaternion) pose representations.

{"func": "EefPoseTerm", "params": {"scale": 0.1, "pose_dim": 7}}

Velocity and Force Control#

Velocity and Force Action Terms#

Action Term

Description

QvelTerm

Joint velocity action: scale * action -> qvel. The policy outputs target joint velocities.

{"func": "QvelTerm", "params": {"scale": 1.0}}

QfTerm

Joint force/torque action: scale * action -> qf. The policy outputs target joint torques/forces.

{"func": "QfTerm", "params": {"scale": 1.0}}

Usage Example#

from embodichain.lab.gym.envs.managers.cfg import ActionTermCfg

# Example: Delta joint position control
actions = {
    "joint_position": ActionTermCfg(
        func="DeltaQposTerm",
        params={
            "scale": 0.1,  # Scale factor for action deltas
        },
    ),
}

# Example: Normalized joint position control
actions = {
    "normalized_joint_position": ActionTermCfg(
        func="QposDenormalizedTerm",
        params={
            "scale": 1.0,  # Full joint range utilization
        },
    ),
}

# Example: Normalize qpos to [0, 1] range (for post-processing)
actions = {
    "normalize_qpos": ActionTermCfg(
        func="QposNormalizedTerm",
        params={
            "range": [0.0, 1.0],  # Normalize to [0, 1] range
        },
    ),
}

# Example: End-effector pose control
actions = {
    "eef_pose": ActionTermCfg(
        func="EefPoseTerm",
        params={
            "scale": 0.1,
            "pose_dim": 7,  # 7D (position + quaternion)
        },
    ),
}

Action Term Properties#

All action terms provide the following properties:

  • action_dim: The dimension of the action space (number of values the policy should output)

  • process_action(action): Method to convert raw policy output to robot control format

Contents