platforms_mujoco

MuJoCo Platforms Module.

This module provides platform implementations for the MuJoCo simulator, mirroring the platform interface in robotblockset.platforms.

class robotblockset.mujoco.platforms_mujoco.platform_mujoco(platform_name: str, scene: mjInterface | None = None, host: str = 'localhost', port: int = 50000, **kwargs: Any)[source]

Bases: platform

MuJoCo-backed mobile platform interface using the socket-based server API.

Initialize a MuJoCo platform interface.

Parameters:

  • platform_name (str) – Base name of the platform model in MuJoCo.

  • scene (mjInterface, optional) – Existing MuJoCo interface instance. If None, a new connection is created.

  • host (str, optional) – Hostname of the MuJoCo simulator.

  • port (int, optional) – Port of the MuJoCo simulator.

  • **kwargs (Any) – Additional keyword arguments.

Notes

When constructing objects of this class, the following keyword arguments can explicitly configure model element names. If omitted, names are derived from the platform name:

JointNameslist[str], optional

Explicit list of platform joint names.

ActuatorNameslist[str], optional

Explicit list of actuator names used for velocity commands.

RobotBaseNamestr, optional

Name of the robot base body used as the platform reference frame.

SensorJointPosNameslist[str], optional

Sensor names used to read joint positions.

SensorJointVelNameslist[str], optional

Sensor names used to read joint velocities.

SensorPosNamestr, optional

Sensor name used to read the platform position.

SensorOriNamestr, optional

Sensor name used to read the platform orientation.

SensorLinVelNamestr, optional

Sensor name used to read the platform linear velocity.

SensorRotVelNamestr, optional

Sensor name used to read the platform angular velocity.

SensorForceNamestr, optional

Sensor name used to read force measurements.

SensorTorqueNamestr, optional

Sensor name used to read torque measurements.

Raises:

Exception – If joint or actuator names do not resolve uniquely in the MJCF model.

__init__(platform_name: str, scene: mjInterface | None = None, host: str = 'localhost', port: int = 50000, **kwargs: Any) → None[source]

Initialize a MuJoCo platform interface.

Parameters:

  • platform_name (str) – Base name of the platform model in MuJoCo.

  • scene (mjInterface, optional) – Existing MuJoCo interface instance. If None, a new connection is created.

  • host (str, optional) – Hostname of the MuJoCo simulator.

  • port (int, optional) – Port of the MuJoCo simulator.

  • **kwargs (Any) – Additional keyword arguments.

Notes

When constructing objects of this class, the following keyword arguments can explicitly configure model element names. If omitted, names are derived from the platform name:

JointNameslist[str], optional

Explicit list of platform joint names.

ActuatorNameslist[str], optional

Explicit list of actuator names used for velocity commands.

RobotBaseNamestr, optional

Name of the robot base body used as the platform reference frame.

SensorJointPosNameslist[str], optional

Sensor names used to read joint positions.

SensorJointVelNameslist[str], optional

Sensor names used to read joint velocities.

SensorPosNamestr, optional

Sensor name used to read the platform position.

SensorOriNamestr, optional

Sensor name used to read the platform orientation.

SensorLinVelNamestr, optional

Sensor name used to read the platform linear velocity.

SensorRotVelNamestr, optional

Sensor name used to read the platform angular velocity.

SensorForceNamestr, optional

Sensor name used to read force measurements.

SensorTorqueNamestr, optional

Sensor name used to read torque measurements.

Raises:

Exception – If joint or actuator names do not resolve uniquely in the MJCF model.

Close() → None[source]

Close the MuJoCo connection.

Return type:

None

Init() → None[source]

Initialize MuJoCo handles and cached sensor indices.

Return type:

None

GetState() → None[source]

Update platform state from MuJoCo data buffers.

Return type:

None

isReady() → bool[source]

Check if the platform is connected.

Returns:

True if the simulator connection is active.

Return type:

bool

Restart(qpos: ndarray | List[float] | List[int] | Tuple[float, ...] | Tuple[int, ...] | None = None, u: ndarray | None = None, reset: bool = True, keyframe: int | None = None) → None[source]

Reset the simulation and optionally set joint positions/inputs.

Parameters:

  • qpos (ArrayLike, optional) – Joint positions for the platform.

  • u (JointVelocityType, optional) – Joint velocity command applied after the reset.

  • reset (bool, optional) – If True, reset the simulation state.

  • keyframe (int, optional) – Optional keyframe index for reset.

Stop() → None[source]

Stop platform motion and clear velocity commands.

Return type:

None

Set_vel(v: ndarray | List[float] | List[int] | Tuple[float, ...] | Tuple[int, ...], wait: float | None = None) → int[source]

Update platform velocities (forward, turn) and wait.

Parameters:

  • v (ArrayLike) – Desired planar velocity [vx, wz].

  • wait (float, optional) – Time to wait (in seconds) to synchronize the command to move. If omitted, self.tsamp is used.

Returns:

Status of the move (0 for success, non-zero for error).

Return type:

int

SendRobot_u(u: ndarray) → None[source]

Send joint commands to platform actuators.

Parameters:

u (JointVelocityType) – Actuator command vector.

Raises:

ValueError – If u does not contain exactly self.nj commands.

SendCtrl(u: ndarray | List[float] | List[int] | Tuple[float, ...] | Tuple[int, ...]) → None[source]

Send a full control vector to the MuJoCo platform.

Parameters:

u (ArrayLike) – Full actuator control vector for the MuJoCo scene.

SendAuxCtrl(idx: Sequence[int], val: ndarray | List[float] | List[int] | Tuple[float, ...] | Tuple[int, ...]) → None[source]

Update selected actuator controls by index.

Parameters:

  • idx (Sequence[int]) – Actuator indices to update.

  • val (ArrayLike) – Control values written to the selected actuators.

Return type:

None

Raises:

  • TypeError – If an actuator index is not an integer.

  • ValueError – If an actuator index is out of range or the number of values does not match the number of indices.

GetAuxJointPos(idx: Sequence[int]) → ndarray[source]

Return joint positions for auxiliary joints by index.

Parameters:

idx (Sequence[int]) – Joint indices in the simulator state vector.

Returns:

Joint positions for the requested indices.

Return type:

JointConfigurationType

GetSensor(ide: str | int | Sequence[int] | None = None) → ndarray | None[source]

Read sensor data by name or ID, or return the full sensor array.

Parameters:

ide (str, int, or sequence of int, optional) – Sensor name, scalar sensor ID, or sequence of sensor IDs. Data for multiple IDs is concatenated in the requested order. If None, all sensor samples are returned.

Returns:

Selected sensor data, the full sensor array, or None if a sensor cannot be resolved.

Return type:

np.ndarray or None

Raises:

TypeError – If ide is not a sensor name, integer ID, sequence of integer IDs, or None.

GetContacts(ide: str | int | Sequence[int] | None = None) → ndarray | None[source]

Return contact forces for a geom or for all contacts.

Parameters:

ide (str, int, or sequence of int, optional) – Geom name, scalar geom ID, or sequence of geom IDs. If None, all contacts are reported.

Returns:

Contact force array of shape (N, 3), or None if there are no contacts or a geom cannot be resolved.

Return type:

np.ndarray or None

Raises:

TypeError – If ide is not a geom name, integer ID, sequence of integer IDs, or None.

SetMocapPose(ide: str | int, x: ndarray | List[float] | List[int] | Tuple[float, ...] | Tuple[int, ...]) → None[source]

Set the pose of a mocap body.

Parameters:

  • ide (ObjectIdType) – Mocap body name or ID.

  • x (PoseInputType) – Mocap pose.

Raises:

ValueError – If the pose shape is unsupported.

GetMocapPose(ide: str | int, out: str = 'x') → ndarray | None[source]

Return a mocap-body pose in the requested output format.

Parameters:
Returns:

Mocap-body pose, or None if it cannot be resolved.

Return type:

Pose3DType or HomogeneousMatrixType or Vector3DType or RotationMatrixType or None

GetObjectPose(typ: str, ide: str | int, out: str = 'x') → ndarray | None[source]

Return the pose of a body, site, or geom in the requested format.

Parameters:

  • typ (str) – Object type (“body”, “site”, or “geom”).

  • ide (ObjectIdType) – Object name or id.

  • out (str, optional) – Output pose format understood by robotblockset.transformations.map_pose().

Returns:

Pose of the requested object, or None when the object type or identifier is unsupported.

Return type:

PoseOutputType or None

GetObjectData(ide: str | int) → Any | None[source]

Return raw MuJoCo body data for a body name or ID.

Parameters:

ide (str or int) – Body name or ID.

Returns:

Body data, or None if the body cannot be resolved.

Return type:

Any or None

SetObjectPose(ide: str | int, x: ndarray | List[float] | List[int] | Tuple[float, ...] | Tuple[int, ...]) → None[source]

Set a body pose from a spatial representation.

Parameters:

  • ide (ObjectIdType) – Body name or id.

  • x (PoseInputType) – Pose represented as position, quaternion, pose vector, rotation matrix, or homogeneous transform.

Return type:

None

Raises:

ValueError – If x has an unsupported shape.

SetEquality(ide: str | int, val: int | bool) → None[source]

Set an equality constraint activation flag.

Parameters:

  • ide (ObjectIdType) – Equality constraint name or id.

  • val (Union[int, bool]) – Activation flag.

Return type:

None

SimulatorMessage(msg: str) → None[source]

Send a message to the simulator UI.

Parameters:

msg (str) – Message to display.

Return type:

None

sim(dt: float) → None[source]

Advance the simulator for a fixed duration.

Parameters:

dt (float) – Duration, in seconds, to advance the simulator.

Return type:

None

Raises:

  • ValueError – If dt is not a finite, nonnegative real scalar.

  • RuntimeError – If MuJoCo returns invalid simulation time or simulation time does not advance for one second.

class robotblockset.mujoco.platforms_mujoco.tiagobase(platform_name: str = 'tiagobase', **kwargs: Any)[source]

Bases: platform_mujoco, tiagobase_spec

MuJoCo platform wrapper for the PAL Robotics Tiago Base.

Create a TiagoBase platform backed by MuJoCo.

Parameters:

  • platform_name (str, optional) – Base name of the platform in the MuJoCo model.

  • **kwargs (Any) – Additional keyword arguments passed to platform_mujoco, including optional joint, actuator, base-body, and sensor names.

__init__(platform_name: str = 'tiagobase', **kwargs: Any) → None[source]

Create a TiagoBase platform backed by MuJoCo.

Parameters:

  • platform_name (str, optional) – Base name of the platform in the MuJoCo model.

  • **kwargs (Any) – Additional keyword arguments passed to platform_mujoco, including optional joint, actuator, base-body, and sensor names.

class robotblockset.mujoco.platforms_mujoco.mir100(platform_name: str = 'mir', **kwargs: Any)[source]

Bases: platform_mujoco, mir100_spec

MuJoCo platform wrapper for the MiR100 mobile base.

Create a MiR100 platform backed by MuJoCo.

Parameters:

  • platform_name (str, optional) – Base name of the platform in the MuJoCo model.

  • **kwargs (Any) – Additional keyword arguments passed to platform_mujoco, including optional joint, actuator, base-body, and sensor names.

__init__(platform_name: str = 'mir', **kwargs: Any) → None[source]

Create a MiR100 platform backed by MuJoCo.

Parameters:

  • platform_name (str, optional) – Base name of the platform in the MuJoCo model.

  • **kwargs (Any) – Additional keyword arguments passed to platform_mujoco, including optional joint, actuator, base-body, and sensor names.

Classes

mir100([platform_name])

MuJoCo platform wrapper for the MiR100 mobile base.

platform_mujoco(platform_name[, scene, ...])

MuJoCo-backed mobile platform interface using the socket-based server API.

tiagobase([platform_name])

MuJoCo platform wrapper for the PAL Robotics Tiago Base.