intrinsic.solutions.execution.Executive

View source on GitHub

Wrapper for the Executive gRPC service.

intrinsic.solutions.execution.Executive(
stub: executive_service_pb2_grpc.ExecutiveServiceStub,
blackboard_stub: blackboard_service_pb2_grpc.ExecutiveBlackboardStub,
error_loader: error_processing.ErrorsLoader,
simulation: Optional[intrinsic.solutions.simulation.Simulation] = None,
polling_interval_in_seconds: float = _DEFAULT_POLLING_INTERVAL_IN_SECONDS
)

Args
`stub`	The gRPC stub to be used for communication with the executive service.
`blackboard_stub`	The gRPC stub to be used for blackboard related calls.
`error_loader`	Can load ErrorReports about executions
`simulation`	The workcell simulation module (optional).
`polling_interval_in_seconds`	Number of seconds to wait while polling for the operation state in blocking calls such as Executive.suspend().

Attributes
`execution_mode`	Currently set mode of execution (normal or step-wise).
`operation`
`simulation_mode`	Currently set mode of simulation (physics or kinematics).

Child Classes

class ExecutionMode

class ResumeMode

class SimulationMode

Methods

`await_value`

View source

await_value(
value: blackboard_value.BlackboardValue
) -> None

Blocks until a key is available on the blackboard.

While the key is available await_value will always return immediately until the key is removed. Changes in value are not reflected.

Args
`value`	wait for this value to be available on the blackboard

`block_until_completed`

View source

block_until_completed(
*, silence_outputs: bool = False
) -> None

Waits until plan execution has begun and then stops.

Polls executive state every self._polling_interval_in_seconds.

Args
`silence_outputs`	If true, do not show success or error outputs of the execution in Jupyter.

Raises
`ExecutionFailedError`	On unexpected state of the executive during plan execution.
`solutions_errors.UnavailableError`	On executive service not reachable.
`grpc.RpcError`	On any other gRPC error.

`cancel`

View source

cancel() -> None

Cancels plan execution and blocks until execution finishes.

Raises
`solutions_errors.UnavailableError`	On executive service not reachable.
`solutions_errors.InvalidArgumentError`	On operation not in RUNNING state.
`grpc.RpcError`	On any other gRPC error.

`cancel_async`

View source

cancel_async() -> None

Asynchronously cancels plan execution; returns immediately.

Raises
`solutions_errors.UnavailableError`	On executive service not reachable.
`solutions_errors.InvalidArgumentError`	On operation not in RUNNING state.
`grpc.RpcError`	On any other gRPC error.

`connect`

View source

@classmethodconnect(
grpc_channel: grpc.Channel,
error_loader: error_processing.ErrorsLoader,
simulation: Optional[intrinsic.solutions.simulation.Simulation] = None,
polling_interval_in_seconds: float = _DEFAULT_POLLING_INTERVAL_IN_SECONDS
) -> 'Executive'

Connect to a running executive.

Args
`grpc_channel`	Channel to the executive gRPC service.
`error_loader`	Loads error data for executive runs.
`simulation`	The workcell simulation module (optional).
`polling_interval_in_seconds`	Number of seconds to wait while polling for the operation state in blocking calls such as Executive.suspend().

Returns
A newly created instance of the Executive wrapper class.

`get_errors`

View source

get_errors(
print_level: error_processing.PrintLevel = error_processing.PrintLevel.OFF
) -> error_processing.ErrorGroup

Loads and optionally prints errors from current, failed operation.

Args
`print_level`	If not set to 'OFF', error reports are printed.

Returns
Error summaries.

`get_value`

View source

get_value(
value: blackboard_value.BlackboardValue
) -> Any

Gets the actual data written for the specified value on the blackboard.

Args
`value`	the value to get actual data for

Returns
Value as read from the blackboard

Raises
ValueNotAvailable error in case the value has not yet been resolved ValueError if the received value is not of the expected type based on the return value description of the skill

`is_value_available`

View source

is_value_available(
value: blackboard_value.BlackboardValue
) -> bool

Checks whether a value is available on the blackboard.

Args
`value`	check availability for this value

Returns
True if a value has been set False otherwise

`load`

View source

load(
behavior_tree_or_action: Optional[BehaviorTreeOrActionType]
) -> None

Loads an action or behavior tree into the executive.

Note that an action can be a general action or any skill obtained through skills.

Args
`behavior_tree_or_action`	A behavior tree, a list of actions (can be nested one level) or a single action.

Raises
`solutions_errors.UnavailableError`	On executive service not reachable.
`grpc.RpcError`	On any other gRPC error.

`reset`

View source

reset() -> None

Resets the current operation to the state from when it was loaded.

Raises
`solutions_errors.UnavailableError`	On executive service not reachable.
`grpc.RpcError`	On any other gRPC error.

`resume`

View source

resume(
mode: Optional['Executive.ResumeMode'] = None
) -> None

Resumes plan execution (SUSPENDED --> RUNNING).

Args
`mode`	The resume mode, STEP and NEXT work when in step-wise execution mode, CONTINUE will switch to NORMAL execution mode before resume.

Raises
`solutions_errors.UnavailableError`	On executive service not reachable.
`solutions_errors.InvalidArgumentError`	On executive not in SUSPENDED state.
`grpc.RpcError`	On any other gRPC error.

`run`

View source

run(
plan_or_action: Optional[BehaviorTreeOrActionType],
*,
silence_outputs: bool = False,
step_wise: bool = False,
start_node: Optional[intrinsic.solutions.behavior_tree.NodeIdentifierType] = None,
simulation_mode: Optional['Executive.SimulationMode'] = None,
embed_skill_traces: bool = False
)

Executes an action or plan and blocks until completion.

This corresponds to running load and start after one another.

Note that an action can be a general action or any skill obtained through Skills.

If plan_or_action is None, runs the initial plan as specified in the Kubernetes template.

Implicitly calls simulation.reset() if this is the first action or plan being executed (after an executive.reset()). This makes sure that any world edits that might have occurred are reflected in the simulation.

Args
`plan_or_action`	A behavior tree, a list of actions (can be nested one level), or a single action.
`silence_outputs`	If true, do not show success or error outputs of the execution in Jupyter.
`step_wise`	Execute step-wise, i.e., suspend after each node of the tree.
`start_node`	Run the specified node as if it were the root node of a tree instead of the complete tree.
`simulation_mode`	Set the simulation mode on the start request. If None will execute in whatever mode is currently set in the executive.
`embed_skill_traces`	If true, execution traces in Google Cloud will incorporate all information from skill traces, otherwise execution traces contain links to individual skill traces.

Raises
`ExecutionFailedError`	On unexpected state of the executive during plan execution.
`solutions_errors.UnavailableError`	On executive service not reachable.
`grpc.RpcError`	On any other gRPC error.

Returns
The start and end timestamps which can be used to query the logs.

`run_async`

View source

run_async(
plan_or_action: Optional[BehaviorTreeOrActionType] = None,
*,
silence_outputs: bool = False,
step_wise: bool = False,
start_node: Optional[intrinsic.solutions.behavior_tree.NodeIdentifierType] = None,
simulation_mode: Optional['Executive.SimulationMode'] = None,
embed_skill_traces: bool = False
)

Requests execution of an action or plan and returns immediately.

Note that an action can be a general action or any skill obtained through Skills.

If plan_or_action is None, runs the initial plan as specified in the Kubernetes template.

Args
`plan_or_action`	A behavior tree, list of actions, or a single action or skill.
`silence_outputs`	If true, do not show success or error outputs of the execution in Jupyter.
`step_wise`	Execute step-wise, i.e., suspend after each node of the tree.
`start_node`	Run the specified node as if it were the root node of a tree instead of the complete tree.
`simulation_mode`	Set the simulation mode on the start request. If None will execute in whatever mode is currently set in the executive.
`embed_skill_traces`	If true, execution traces in Google Cloud will incorporate all information from skill traces, otherwise execution traces contain links to individual skill traces.

Raises
`solutions_errors.UnavailableError`	On executive service not reachable.
`grpc.RpcError`	On any other gRPC error.

`start`

View source

start(
blocking: bool = True,
silence_outputs: bool = False,
*,
step_wise: bool = False,
start_node: Optional[intrinsic.solutions.behavior_tree.NodeIdentifierType] = None,
simulation_mode: Optional['Executive.SimulationMode'] = None,
embed_skill_traces: bool = False
) -> None

Starts the currently loaded plan.

Args
`blocking`	If True, waits until execution finishes. Otherwise, returns immediately after starting.
`silence_outputs`	If true, do not show success or error outputs of the execution in Jupyter.
`step_wise`	Execute step-wise, i.e., suspend after each node of the tree.
`start_node`	Start only the specified node instead of the complete tree.
`simulation_mode`	Set the simulation mode on the start request. If None will execute in whatever mode is currently set in the executive.
`embed_skill_traces`	If true, execution traces in Google Cloud will incorporate all information from skill traces, otherwise execution traces contain links to individual skill traces.

Raises
`solutions_errors.UnavailableError`	On executive service not reachable.
`grpc.RpcError`	On any other gRPC error.

`suspend`

View source

suspend() -> None

Requests to suspend plan execution and blocks until SUSPENDED.

Since the executive currently cannot preempt running skills, this function waits until all running skills have terminated.

Raises
`solutions_errors.UnavailableError`	On executive service not reachable.
`solutions_errors.InvalidArgumentError`	On executive not in RUNNING state.
`grpc.RpcError`	On any other gRPC error.

`suspend_async`

View source

suspend_async() -> None

Requests to suspend plan execution and returns immediately.

Raises
`solutions_errors.UnavailableError`	On executive service not reachable.
`solutions_errors.InvalidArgumentError`	On executive not in RUNNING state.
`grpc.RpcError`	On any other gRPC error.

Args

Attributes

Child Classes​

Methods​

await_value

block_until_completed

cancel

cancel_async

connect

get_errors

get_value

is_value_available

load

reset

resume

run

run_async

start

suspend

suspend_async

Child Classes

Methods

`await_value`

`block_until_completed`

`cancel`

`cancel_async`

`connect`

`get_errors`

`get_value`

`is_value_available`

`load`

`reset`

`resume`

`run`

`run_async`

`start`

`suspend`

`suspend_async`