embodichain.lab.task_program.integrations.simulation

embodichain.lab.task_program.integrations.simulation#

Simulation bindings and factories for Task Program integrations.

These immutable declarations bind canonical Task Program scene and robot contracts to explicitly selected simulation entities and control parts.

AntipodalGraspAffordanceBinding

Build one antipodal grasp affordance from a selected rigid-object mesh.

ContainerAffordanceBinding

Declare one exact object target frame inside a container parent.

ControlPartCommandPreset

Named one-dimensional joint commands for one exact control part.

ControlPartEndpointBinding

Profile endpoint backed by one explicit robot control part.

ControlPartResourceBinding

Joint-backed robot resource containing control-part endpoints.

SimulationArticulationBinding

Explicit binding for one simulation articulation.

SimulationArticulationLinkBinding

Explicit canonical link backed by one native articulation link.

SimulationRigidObjectBinding

Explicit binding for one simulation rigid object.

SimulationRobotSkillProfileBinding

Build a profile from typed resources with strict native validation.

SimulationSceneBinding

Build one authoritative registry from explicit simulation bindings.

SupportSurfaceAffordanceBinding

Declare one exact object target frame on a support parent.

Classes:

AntipodalGraspAffordanceBinding

Build one antipodal grasp affordance from a selected rigid-object mesh.

ContainerAffordanceBinding

Declare one exact object target frame inside a container parent.

ControlPartCommandPreset

Named one-dimensional joint commands for one exact control part.

ControlPartEndpointBinding

Profile endpoint backed by one explicit robot control part.

ControlPartResourceBinding

Joint-backed robot resource containing control-part endpoints.

SimulationArticulationBinding

Explicit binding for one simulation articulation.

SimulationArticulationLinkBinding

Explicit canonical link backed by one native articulation link.

SimulationRigidObjectBinding

Explicit binding for one simulation rigid object.

SimulationRobotSkillProfileBinding

Build a profile from typed resources with strict native validation.

SimulationSceneBinding

Build one authoritative registry from explicit simulation bindings.

SupportSurfaceAffordanceBinding

Declare one exact object target frame on a support parent.

class embodichain.lab.task_program.integrations.simulation.AntipodalGraspAffordanceBinding[source]

Build one antipodal grasp affordance from a selected rigid-object mesh.

internal_axis upgrades the payload to AxisAlignAffordance while preserving the same antipodal grasp capability. The axis is expressed in the rigid object’s local frame and can be reused by held-object skills such as Pour.

Methods:

__init__(entity_id, object_id, native_name, ...)

__init__(entity_id, object_id, native_name, revision, aliases=(), relative_pose=(1.0, 0.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 1.0), mesh_env_id=0, internal_axis=None)
class embodichain.lab.task_program.integrations.simulation.ContainerAffordanceBinding[source]

Declare one exact object target frame inside a container parent.

Parameters:
  • entity_id (str) – Canonical ID of the placement affordance.

  • parent_id (str) – Canonical object, articulation, or link parent ID.

  • native_name (str) – Stable native name of this target frame.

  • aliases (tuple[str, ...]) – Optional non-authoritative lookup aliases.

  • object_target_pose (tuple[float, ...]) – Desired object pose relative to the parent.

  • minimum_confidence (float) – Minimum parent/affordance observation confidence.

  • is_default (bool) – Whether this is the parent’s default Place(inside=...) frame.

Methods:

__init__(entity_id, parent_id, native_name)

__init__(entity_id, parent_id, native_name, aliases=(), object_target_pose=(1.0, 0.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 1.0), minimum_confidence=0.0, is_default=False)
class embodichain.lab.task_program.integrations.simulation.ControlPartCommandPreset[source]

Named one-dimensional joint commands for one exact control part.

Methods:

__init__(preset_id, control_part, commands)

build(*, control_dof)

Build a command profile after validating the native control width.

declare()

Build a provider-free command profile from declared tuple widths.

__init__(preset_id, control_part, commands)
build(*, control_dof)[source]

Build a command profile after validating the native control width.

Return type:

ControlPartCommandProfile

declare()[source]

Build a provider-free command profile from declared tuple widths.

Return type:

ControlPartCommandProfile

class embodichain.lab.task_program.integrations.simulation.ControlPartEndpointBinding[source]

Profile endpoint backed by one explicit robot control part.

Methods:

__init__(endpoint_id, control_part, capabilities)

build(robot)

Build a joint-backed endpoint after native control-part validation.

declare()

Return the endpoint contract without reading a robot.

__init__(endpoint_id, control_part, capabilities, command_preset=None)
build(robot)[source]

Build a joint-backed endpoint after native control-part validation.

Return type:

ResourceEndpoint

declare()[source]

Return the endpoint contract without reading a robot.

Return type:

ResourceEndpoint

class embodichain.lab.task_program.integrations.simulation.ControlPartResourceBinding[source]

Joint-backed robot resource containing control-part endpoints.

Methods:

__init__(resource_id[, endpoints, members])

build(robot)

Build a resource containing strictly validated control-part endpoints.

declare()

Return the resource graph without reading native control parts.

__init__(resource_id, endpoints=(), members=())
build(robot)[source]

Build a resource containing strictly validated control-part endpoints.

Return type:

RobotResource

declare()[source]

Return the resource graph without reading native control parts.

Return type:

RobotResource

class embodichain.lab.task_program.integrations.simulation.SimulationArticulationBinding[source]

Explicit binding for one simulation articulation.

Methods:

__init__(entity_id, simulation_uid[, ...])

__init__(entity_id, simulation_uid, aliases=(), dynamics=SceneDynamics.UNKNOWN, collision_role=SceneCollisionRole.NONE, semantic_type=None, geometry_provider=None)
class embodichain.lab.task_program.integrations.simulation.SimulationArticulationLinkBinding[source]

Explicit canonical link backed by one native articulation link.

Methods:

__init__(entity_id, articulation_id, ...[, ...])

__init__(entity_id, articulation_id, native_link_name, aliases=(), dynamics=SceneDynamics.UNKNOWN, semantic_type=None)
class embodichain.lab.task_program.integrations.simulation.SimulationRigidObjectBinding[source]

Explicit binding for one simulation rigid object.

Methods:

__init__(entity_id, simulation_uid[, ...])

__init__(entity_id, simulation_uid, aliases=(), dynamics=SceneDynamics.UNKNOWN, collision_role=SceneCollisionRole.NONE, semantic_type=None, default_grasp_affordance=None, geometry_provider=None)
class embodichain.lab.task_program.integrations.simulation.SimulationRobotSkillProfileBinding[source]

Build a profile from typed resources with strict native validation.

Methods:

__init__(profile_id, resources[, ...])

build(robot)

Build the existing profile after validating every typed resource.

declare()

Project the complete provider-free robot skill profile.

__init__(profile_id, resources, command_presets=(), defaults=<factory>, presets=(), default_preset=None, skill_presets=<factory>, grounding_providers=<factory>)
build(robot)[source]

Build the existing profile after validating every typed resource.

Parameters:

robot (Robot) – Live robot selected by the simulation factory.

Return type:

RobotSkillProfile

Returns:

Reusable, engine-independent robot skill profile.

declare()[source]

Project the complete provider-free robot skill profile.

Return type:

RobotSkillProfile

class embodichain.lab.task_program.integrations.simulation.SimulationSceneBinding[source]

Build one authoritative registry from explicit simulation bindings.

Methods:

__init__(registry_id[, rigid_objects, ...])

build(simulation)

Build the existing authoritative scene registry.

declare()

Project the complete provider-free scene declaration.

__init__(registry_id, rigid_objects=(), articulations=(), links=(), antipodal_grasps=(), support_surfaces=(), containers=(), collision_world_mode=None)
build(simulation)[source]

Build the existing authoritative scene registry.

Parameters:

simulation (SimulationManager) – Live simulation used only for explicitly named lookups.

Return type:

SceneRegistry

Returns:

Immutable registry with typed roots, links, and affordances.

declare()[source]

Project the complete provider-free scene declaration.

Return type:

SceneManifest

class embodichain.lab.task_program.integrations.simulation.SupportSurfaceAffordanceBinding[source]

Declare one exact object target frame on a support parent.

Parameters:
  • entity_id (str) – Canonical ID of the placement affordance.

  • parent_id (str) – Canonical object, articulation, or link parent ID.

  • native_name (str) – Stable native name of this target frame.

  • aliases (tuple[str, ...]) – Optional non-authoritative lookup aliases.

  • object_target_pose (tuple[float, ...]) – Desired object pose relative to the parent.

  • minimum_confidence (float) – Minimum parent/affordance observation confidence.

  • is_default (bool) – Whether this is the parent’s default Place(on=...) frame.

Methods:

__init__(entity_id, parent_id, native_name)

__init__(entity_id, parent_id, native_name, aliases=(), object_target_pose=(1.0, 0.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 1.0), minimum_confidence=0.0, is_default=False)