Module API

py_trees_ros

ROS extensions, behaviours and utilities for py_trees.

py_trees_ros.action_clients

class py_trees_ros.action_clients.FromBlackboard(name, action_type, action_name, key, generate_feedback_message=None, wait_for_server_timeout_sec=-3.0)

Bases: Behaviour

An action client interface that draws goals from the blackboard. The lifecycle of this behaviour works as follows:

• initialise(): check blackboard for a goal and send

• update(): if a goal was sent, monitor progress

• terminate(): if interrupted while running, send a cancel request

As a consequence, the status of this behaviour can be interpreted as follows:

• FAILURE: no goal was found to send, it was rejected or it failed while executing

• RUNNING: a goal was sent and is still executing on the server

• SUCCESS: sent goal has completed with success

To block on the arrival of a goal on the blackboard, use with the py_trees.behaviours.WaitForBlackboardVariable behaviour. e.g.

sequence = py_trees.composites.Sequence(name="Sequence", memory=True)
wait_for_goal = py_trees.behaviours.WaitForBlackboardVariable(
    name="WaitForGoal",
    variable_name="/my_goal"
)
action_client = py_trees_ros.aciton_clients.FromBlackboard(
    action_type=py_trees_actions.Dock,
    action_name="dock",
    name="ActionClient"
)
sequence.add_children([wait_for_goal, action_client])

To customise a more interesting feedback message, pass in a method to the constructor, for example:

action_client = py_trees_ros.action_clients.FromBlackboard(
    action_type=py_trees_actions.Dock,
    action_name="dock",
    name="ActionClient",
    generate_message=lambda msg: "{:.2f}%%".format(msg.feedback.percentage_completed)
)

Parameters

• name (str) – name of the behaviour

• action_type (Any) – spec type for the action (e.g. move_base_msgs.action.MoveBase)

• action_name (str) – where you can find the action topics & services (e.g. “bob/move_base”)

• key (str) – name of the key on the blackboard

• generate_feedback_message (Callable[[Any], str]) – formatter for feedback messages, takes action_type.Feedback messages and returns strings (default: None)

• wait_for_server_timeout_sec (float) – use negative values for a blocking but periodic check (default: -3.0)

Note

The default setting for timeouts (a negative value) will suit most use cases. With this setting the behaviour will periodically check and issue a warning if the server can’t be found. Actually aborting the setup can usually be left up to the behaviour tree manager.

cancel_response_callback(future)

Immediate callback for the result of a cancel request. This will set the behaviour’s feedback message accordingly.

Parameters

future (rclpy.task.Future) – incoming cancellation result delivered from the action server

feedback_callback(msg)

Default generator for feedback messages from the action server. This will update the behaviour’s feedback message with a stringified version of the incoming feedback message.

Parameters

msg (Any) – incoming feedback message (e.g. move_base_msgs.action.MoveBaseFeedback)

get_result_callback(future)

Immediate callback for the result, saves data into local variables so that the update method can react accordingly.

Parameters

future (rclpy.task.Future) – incoming goal result delivered from the action server

goal_response_callback(future)

Handle goal response, proceed to listen for the result if accepted.

Parameters

future (rclpy.task.Future) – incoming goal request result delivered from the action server

initialise()

Reset the internal variables and kick off a new goal request.

send_cancel_request()

Send a cancel request to the server. This is triggered when the behaviour’s status switches from RUNNING to INVALID (typically a result of a priority interrupt).

send_goal_request(goal)

Send the goal, get a future back and start lining up the chain of callbacks that will lead to a result.

Parameters

goal (Any) –

setup(**kwargs)

Setup the action client services and subscribers.

Parameters

**kwargs (dict) – distribute arguments to this behaviour and in turn, all of it’s children

Raises

• KeyError – if a ros2 node isn’t passed under the key ‘node’ in kwargs

• TimedOutError – if the action server could not be found

shutdown()

Clean up the action client when shutting down.

terminate(new_status)

If running and the current goal has not already succeeded, cancel it.

Parameters

new_status (Status) – the behaviour is transitioning to this new status

update()

Check only to see whether the underlying action server has succeeded, is running, or has cancelled/aborted for some reason and map these to the usual behaviour return states.

Returns

py_trees.common.Status

class py_trees_ros.action_clients.FromConstant(name, action_type, action_name, action_goal, generate_feedback_message=None, wait_for_server_timeout_sec=-3.0)

Bases: FromBlackboard

Convenience version of the action client that only ever sends the same goal.

Parameters

• name (str) – name of the behaviour

• action_type (Any) – spec type for the action (e.g. move_base_msgs.action.MoveBase)

• action_name (str) – where you can find the action topics & services (e.g. “bob/move_base”)

• action_goal (Any) – the goal to send

• generate_feedback_message (Callable[[Any], str]) – formatter for feedback messages, takes action_type.Feedback messages and returns strings (default: None)

• wait_for_server_timeout_sec (float) – use negative values for a blocking but periodic check (default: -3.0)

Note

The default setting for timeouts (a negative value) will suit most use cases. With this setting the behaviour will periodically check and issue a warning if the server can’t be found. Actually aborting the setup can usually be left up to the behaviour tree manager.

py_trees_ros.battery

Getting the most out of your battery.

class py_trees_ros.battery.ToBlackboard(name, topic_name, qos_profile, threshold=30.0)

Bases: ToBlackboard

Subscribes to the battery message and writes battery data to the blackboard. Also adds a warning flag to the blackboard if the battery is low - note that it does some buffering against ping-pong problems so the warning doesn’t trigger on/off rapidly when close to the threshold.

When ticking, updates with RUNNING if it got no data, SUCCESS otherwise.

Blackboard Variables:

• battery (sensor_msgs.msg.BatteryState)[w]: the raw battery message

• battery_low_warning (bool)[w]: False if battery is ok, True if critically low

Parameters

• name (str) – name of the behaviour

• topic_name (str) – name of the battery state topic

• qos_profile (rclpy.qos.QoSProfile) – qos profile for the subscriber

• threshold (float) – percentage level threshold for flagging as low (0-100)

update()

Call the parent to write the raw data to the blackboard and then check against the threshold to determine if the low warning flag should also be updated.

Returns

SUCCESS if a message was written, RUNNING otherwise.

Return type

Status

py_trees_ros.blackboard

This module provides the py_trees_ros.blackboard.Exchange class, a ROS wrapper around a Blackboard that permits introspection of either the entire board or a window onto a smaller part of the board over a ROS API via the py-trees-blackboard-watcher command line utility.

class py_trees_ros.blackboard.BlackboardView(node, topic_name, variable_names, filter_on_visited_path, with_activity_stream)

Bases: object

Utility class that enables tracking and publishing of relevant parts of the blackboard for a user. This is used by the Exchange operator.

Parameters

• node (rclpy.node.Node) – an rclpy node for communications handling

• topic_name (str) – name of the topic for the publisher

• variable_names (Set[str]) – requested variables to view

• filter_on_visited_path (bool) – constrain dynamically to the visited region of the blackboard

• with_activity_stream (bool) –

is_changed(visited_clients)

Dynamically adjusts the tracking parameters for the sub-blackboard against the clients that were visited (if desired) and proceeds to determine if there was a change since the last check.

Warning

Since this caches the current blackboard, it can’t be used multiple times in succession.

Returns

bool

Parameters

visited_clients (Set[UUID]) –

Return type

bool

shutdown()

Shutdown the temporarily created publisher.

class py_trees_ros.blackboard.BlackboardWatcher(namespace_hint=None)

Bases: object

The blackboard watcher sits on the other side of the exchange and is a useful mechanism from which to pull all or part of the blackboard contents. This is useful for live introspection, or logging of the blackboard contents.

Parameters

namespace_hint (str) – (optionally) used to locate the blackboard if there exists more than one

See also

py-trees-blackboard-watcher

create_service_client(key)

Convenience api for opening a service client and waiting for the service to appear.

Parameters

key (str) – one of ‘open’, ‘close’.

Raises

• NotReadyError – if setup() wasn’t called to identify the relevant services to connect to.

• TimedOutError – if it times out waiting for the server

echo_blackboard_contents(msg)

Very simple formatter of incoming messages.

Parameters

msg (std_msgs.String) – incoming blackboard message as a string.

setup(timeout_sec)

Creates the node and checks that all of the server-side services are available for calling on to open a connection.

Parameters

timeout_sec (float) – time (s) to wait (use common.Duration.INFINITE to block indefinitely)

Raises

• NotFoundError – if no services were found

• MultipleFoundError – if multiple services were found

shutdown()

Perform any ros-specific shutdown.

class py_trees_ros.blackboard.Exchange

Bases: object

Establishes ros communications around a Blackboard that enable users to introspect or watch relevant parts of the blackboard.

ROS Publishers:

• ~/_watcher_<N> (std_msgs.msg.String)

  • streams the (sub)blackboard over a blackboard watcher connection

ROS Services:

• ~/get_variables (py_trees_msgs.srv.GetBlackboardVariables)

  • list all the blackboard variable names (not values)

• ~/open (py_trees_msgs.srv.OpenBlackboardStream)

  • request a publisher to stream a part of the blackboard contents

• ~/close (py_trees_msgs.srv.CloseBlackboardStream)

  • close a previously opened watcher

Watching could be more simply enabled by just providing a get style service on a particular variable(s), however that would introduce an asynchronous interrupt on the library’s usage and subsequently, locking. Instead, a watcher service requests for a stream, i.e. a publisher to be created for private use that streams only a subsection of the blackboard to the user. Additional services are provided for closing the stream and listing the variables on the blackboard.

See also

py_trees_ros.trees.BehaviourTree (in which it is used) and py-trees-blackboard-watcher (working with the watchers).

post_tick_handler(visited_client_ids=None)

Update blackboard watcher views, publish changes and clear the activity stream. Publishing is lazy, depending on blackboard watcher connections.

Typically you would call this from a tree custodian (e.g. py_trees_ros.trees.BehaviourTree) after each and every tick.

Parameters

visited_client_ids (Optional[List[UUID]]) – blackboard client unique identifiers

setup(node)

This is where the ros initialisation of publishers and services happens. It is kept outside of the constructor for the same reasons that the familiar py_trees setup() method has - to enable construction of behaviours and trees offline (away from their execution environment) so that dot graphs and other visualisations of the tree can be created.

Parameters

node (Node) – node to hook ros communications on

Examples

It is expected that users will use this in their own customised tree custodian:

class MyTreeManager(py_trees.trees.BehaviourTree):

    def __init__(self):
        pass

    def setup(self, timeout):
        super(MyTreeManager, self).setup(timeout)
        self.exchange = py_trees_ros.blackboard.Exchange()
        self.exchange.setup(timeout)

See also

This class is used as illustrated above in BehaviourTree.

class py_trees_ros.blackboard.SubBlackboard(node)

Bases: object

Dynamically track the entire blackboard or part thereof and flag when there have been changes. This is a nice to have that only works if the blackboard contains fundamental variables (i.e. objects that can be pickled).

Parameters

node (rclpy.node.Node) – the node handle for ros logging of warnings if needed

update(variable_names)

Check for changes to the blackboard scoped to the provided set of variable names (may be nested, e.g. battery.percentage). Checks the entire blackboard when variable_names is None.

Parameters

variable_names (Set[str]) – constrain the scope to track for changes

py_trees_ros.conversions

Converter methods for transferring information back and forth between py_trees objects and ros messages.

py_trees_ros.conversions.activity_stream_to_msgs()

Convert the blackboard activity stream to a message.

Returns

A list of activity item messages.

Return type

List[<Mock name=’mock.msg.ActivityItem’ id=’140605245237824’>]

py_trees_ros.conversions.additional_detail_to_str(behaviour)

Provide, e.g. policy information about the behaviour (i.e. black magic details under the hood). Useed for debugging, so only strings needed.

Todo

Unlock unicode support across the wire, currently only ascii arrives ungarbled.

Parameters

behaviour (Behaviour) – investigate the policies for this behaviour

Returns

an informative additional detail string

Return type

str

py_trees_ros.conversions.behaviour_to_msg(behaviour)

Convert a behaviour to a message.

Parameters

behaviour (Behaviour) – behaviour to convert

Returns

a ros message representation of a behaviour

Return type

<Mock name=’mock.msg.Behaviour’ id=’140605245237488’>

py_trees_ros.conversions.behaviour_type_to_msg_constant(behaviour)

Convert a behaviour class type to a message constant.

Parameters

behaviour (Behaviour) – investigate the type of this behaviour

Returns

from the type constants in py_trees_ros_interfaces.msg.Behaviour

Return type

uint8

py_trees_ros.conversions.blackbox_enum_to_msg_constant(blackbox_level)

Convert a blackbox level enum to a message constant.

Parameters

blackbox_level (BlackBoxLevel) – blackbox level of a behaviour

Returns

from the type constants in py_trees_ros_interfaces.msg.Behaviour

Return type

uint8

py_trees_ros.conversions.msg_constant_to_behaviour_type(value)

Convert one of the behaviour type constants in a py_trees_ros_interfaces.msg.Behaviour message to a type.

Parameters

value (int) – see the message definition for details

Returns

a behaviour class type (e.g. py_trees.composites.Sequence)

Raises

TypeError – if the message type is unrecognised

Return type

Any

py_trees_ros.conversions.msg_constant_to_blackbox_level_enum(value)

Convert one of the blackbox level constants in a Behaviour message to a py_trees status enum.

Parameters

value (int) – see the message definition for details

Returns

a py_trees blackbox level

Raises

TypeError – if the status type is unrecognised

Return type

BlackBoxLevel

py_trees_ros.conversions.msg_constant_to_status_enum(value)

Convert one of the status constants in a py_trees_ros_interfaces.msg.Behaviour message to a py_trees status enum.

Parameters

value (int) – see the message definition for details

Returns

a py_trees status

Raises

TypeError – if the status type is unrecognised

Return type

Status

py_trees_ros.conversions.msg_to_behaviour(msg)

Convert behaviour message to a py_trees behaviour. This doesn’t completely recreate the original behaviour (doesn’t have any of the custom state), but allows the user to compose a tree that can be utilised for debugging or visualisation applications.

Parameters

msg (<Mock name='mock.msg.Behaviour' id='140605245237488'>) – a ros message representation of a behaviour

Returns

converted, skeleton of the original behaviour

Return type

Behaviour

py_trees_ros.conversions.msg_to_uuid4(msg)

Convert a uuid4 python object to a ros unique identifier, UUID type.

Parameters

msg (<Mock name='mock.msg.UUID' id='140605245237776'>) – the ros message type

Returns

the behaviour’s uuid, python style

Return type

UUID

py_trees_ros.conversions.rclpy_duration_to_float(duration)

Convert a ros2 duration (seconds/nanoseconds) to a float.

Parameters

• time – time to convert

• duration (rclpy.time.Duration) –

Returns

time (seconds) as a float

Return type

float

py_trees_ros.conversions.rclpy_time_to_float(time)

Convert a ros2 time (seconds/nanoseconds) to a float.

Parameters

time (rclpy.time.Time) – time to convert

Returns

time (seconds) as a float

Return type

float

py_trees_ros.conversions.status_enum_to_msg_constant(status)

Convert a status to a message constant.

Parameters

status (Status) – status enum of a behaviour

Returns

from the status constants in py_trees_ros_interfaces.msg.Behaviour

Return type

uint8

py_trees_ros.conversions.uuid4_to_msg(uuid4=UUID('4274e85a-7a5f-46d5-98ac-58cea39e68d5'))

Convert a uuid4 python object to a ros unique identifier, UUID type.

Parameters

uuid4 (UUID) – unique identifier to convert, defaults to auto-generated uuid4

Returns

the ros message type

Return type

<Mock name=’mock.msg.UUID’ id=’140605245237776’>

py_trees_ros.exceptions

Custom exception types for py_trees_ros.

exception py_trees_ros.exceptions.MultipleFoundError

Bases: Exception

Middleware connection found when but one was expected.

exception py_trees_ros.exceptions.NotFoundError

Bases: Exception

Middleware connection not found.

exception py_trees_ros.exceptions.NotReadyError

Bases: Exception

Typically used when methods have been called that expect, but have not pre-engaged in the ROS2 specific setup typical of py_trees_ros classes and behaviours.

exception py_trees_ros.exceptions.ServiceError

Bases: Exception

Failure in a service request-response process (usually no response).

exception py_trees_ros.exceptions.TimedOutError

Bases: Exception

Timed out waiting (typically) for middleware connections to be established.

py_trees_ros.publishers

Convenience behaviours for publishing ROS messages.

class py_trees_ros.publishers.FromBlackboard(name, topic_name, topic_type, qos_profile, blackboard_variable)

Bases: Behaviour

This behaviour looks up the blackboard for content to publish … and publishes it.

This is a non-blocking behaviour - if there is no data yet on the blackboard it will tick with FAILURE, otherwise SUCCESS.

To convert it to a blocking behaviour, simply use with the py_trees.behaviours.WaitForBlackboardVariable. e.g.

sequence = py_trees.composites.Sequence(name="Sequence", memory=True)
wait_for_data = py_trees.behaviours.WaitForBlackboardVariable(
    name="WaitForData",
    variable_name="/my_message"
)
publisher = py_trees_ros.publishers.FromBlackboard(
    topic_name="/foo",
    topic_type=std_msgs.msg.Empty
    qos_profile=my_qos_profile,
    blackboard_variable="/my_message"
)
sequence.add_children([wait_for_data, publisher])

The various set/unset blackboard variable behaviours can also be useful for setting and unsetting the message to be published (typically elsewhere in the tree).

Parameters

• name (str) – name of the behaviour

• topic_name (str) – name of the topic to connect to

• topic_type (Any) – class of the message type (e.g. std_msgs.msg.String)

• qos_profile (rclpy.qos.QoSProfile) – qos profile for the subscriber

• blackboard_variable (str) – name of the variable on the blackboard (can be nested)

setup(**kwargs)

Initialises the publisher.

Parameters

**kwargs (dict) – distribute arguments to this behaviour and in turn, all of it’s children

Raises

KeyError – if a ros2 node isn’t passed under the key ‘node’ in kwargs

update()

Publish the specified variable from the blackboard.

Raises

TypeError if the blackboard variable is not of the required type –

Returns

FAILURE (variable does not exist on the blackboard) or SUCCESS (published)

py_trees_ros.subscribers

ROS subscribers are asynchronous communication handles whilst py_trees are by their nature synchronous. They tick, pause, then tick again and provide an assumption that only one behaviour or function is running at any single moment. Firing off a subscriber callback in the middle of that synchronicity to write to a blackboard would break this assumption.

To get around that, subscriber behaviours run the ros callbacks in a background thread and constrain locking and a local cache inside the behaviour. Only in the update function is a cached variable unlocked and then permitted to be used or written to the blackboard.

class py_trees_ros.subscribers.CheckData(name, topic_name, topic_type, qos_profile, variable_name, expected_value, comparison_operator=<built-in function eq>, fail_if_no_data=False, fail_if_bad_comparison=False, clearing_policy=ClearingPolicy.ON_INITIALISE)

Bases: Handler

Check a subscriber to see if it has received data.

It optionally checks whether that data, or part of it has a specific value.

Usage Patterns

Sequence Guard: RUNNING until there is a successful comparison

• fail_if_no_data=False

• fail_if_bad_comparison=False

Selector Priority Chooser: FAILURE until there is a successful comparison

• fail_if_no_data=True

• fail_if_bad_comparison=True

Parameters

• name (str) – name of the behaviour

• topic_name (str) – name of the topic to connect to

• topic_type (Any) – class of the message type (e.g. std_msgs.msg.String)

• qos_profile (rclpy.qos.QoSProfile) – qos profile for the subscriber

• variable_name (str) – name of the variable to check

• expected_value (Any) – expected value of the variable

• comparison_operator – one from the python operator module

• fail_if_no_data – FAILURE instead of RUNNING if there is no data yet

• fail_if_bad_comparison – FAILURE instead of RUNNING if comparison failed

• clearing_policy – when to clear the data

Tip

Prefer ToBlackboard and the various blackboard checking behaviours instead. It will provide you with better introspection capability and less complex behaviour construction to manage.

update()

Handles all the logic for determining what result should go back.

Returns

depending on the checking results

Return type

Status

class py_trees_ros.subscribers.EventToBlackboard(name, topic_name, qos_profile, variable_name)

Bases: Handler

Listen for events (std_msgs.msg.Empty) on a subscriber and writes the result to the blackboard.

This will write True to the variable on the blackboard if a message was received since the last tick, False otherwise. The behaviour itself always returns SUCCESS.

Tip

Ideally you need this at the very highest part of the tree so that it gets triggered every time - once this happens, then the rest of the behaviour tree can utilise the variables.

Parameters

• name (str) – name of the behaviour

• topic_name (str) – name of the topic to connect to

• qos_profile (rclpy.qos.QoSProfile) – qos profile for the subscriber

• variable_name (str) – name to write the boolean result on the blackboard

update()

Check for data and write to the board.

Returns

Always returns SUCCESS

class py_trees_ros.subscribers.Handler(name, topic_name, topic_type, qos_profile, clearing_policy=ClearingPolicy.ON_INITIALISE)

Bases: Behaviour

Warning

Do not use - it will always return INVALID. Subclass it to create a functional behaviour.

Not intended for direct use, as this just absorbs the mechanics of setting up a subscriber for inheritance by user-defined behaviour classes. There are several options for the mechanism of clearing the data so that a new result can be processed.

Always Look for New Data

This will clear any currently stored data upon entry into the behaviour (i.e. when initialise() is called). This is useful for a behaviour in a sequence that is looking to start fresh as it is about to tick and be in a RUNNING state until new data arrives.

Look for New Data only after SUCCESS

This will clear any currently stored data as soon as the behaviour returns SUCCESS. This is useful for catching new triggers/events from things like buttons where you don’t want to miss things even though you may not actually be ticking.

Use the Latest Data

Even if this data was received before entry into the behaviour. In this case initialise() does not do anything with the currently stored data. Useful as a blocking behaviour to wait on some some topic having been initialised with some data (e.g. CameraInfo).

Parameters

• name (str) – name of the behaviour

• topic_name (str) – name of the topic to connect to

• topic_type (Any) – class of the message type (e.g. std_msgs.msg.String)

• qos_profile (rclpy.qos.QoSProfile) – qos profile for the subscriber

• clearing_policy (ClearingPolicy) – when to clear the data

initialise()

If the clearing policy is set to ON_INITIALISE it will clear the internally saved message, otherwise it does nothing.

setup(**kwargs)

Initialises the subscriber.

Parameters

**kwargs (dict) – distribute arguments to this behaviour and in turn, all of it’s children

Raises

KeyError – if a ros2 node isn’t passed under the key ‘node’ in kwargs

class py_trees_ros.subscribers.ToBlackboard(name, topic_name, topic_type, qos_profile, blackboard_variables={}, initialise_variables={}, clearing_policy=ClearingPolicy.ON_INITIALISE)

Bases: Handler

Saves the latest message to the blackboard and immediately returns success. If no data has yet been received, this behaviour blocks (i.e. returns RUNNING).

Typically this will save the entire message, however sub fields can be designated, in which case they will write to the specified keys.

Parameters

• topic_name (str) – name of the topic to connect to

• topic_type (Any) – class of the message type (e.g. std_msgs.msg.String)

• qos_profile (rclpy.qos.QoSProfile) – qos profile for the subscriber

• blackboard_variables (Dict[str, Any]) – blackboard variable string or dict {names (keys) - message subfields (values)}, use a value of None to indicate the entire message

• initialise_variables (Dict[str, Any]) – initialise the blackboard variables to some defaults

• name (str) – name of the behaviour

• clearing_policy – when to clear the data

Examples

To capture an entire message:

chatter_to_blackboard = ToBlackboard(topic_name="chatter",
                                     topic_type=std_msgs.msg.String,
                                     blackboard_variables = {'chatter': None}
                                     )

or to get rid of the annoying sub-data field in std_msgs/String:

chatter_to_blackboard = ToBlackboard(topic_name="chatter",
                                     topic_type=std_msgs.msg.String,
                                     blackboard_variables = {'chatter': 'data'}
                                     )

combinations of multiple entities inside the message can also be saved:

blackboard_variables={"pose_with_covariance_stamped": None, "pose": "pose.pose"}

update()

Writes the data (if available) to the blackboard.

Returns

RUNNING (no data) or SUCCESS

Return type

Status

class py_trees_ros.subscribers.WaitForData(name, topic_name, topic_type, qos_profile, clearing_policy=ClearingPolicy.ON_INITIALISE)

Bases: Handler

Waits for a subscriber’s callback to be triggered. This doesn’t care about the actual data, just whether it arrived or not, so is useful for catching triggers (std_msgs/Empty messages) or blocking (in the behaviour sense) until some data has arrived (e.g. camera configuration). There are two use cases:

Usage Patterns

Got Something, Sometime

• clearing_policy == NEVER

Don’t care when data arrived, just that it arrived. This could be for something like a map topic, or a configuration that you need to block. Once it returns SUCCESS, it will always return SUCCESS.

Wating for the Next Thing, Starting From Now

• clearing_policy == ON_INTIALISE

Useful as a gaurd at the start of a sequence, that is waiting for an event to trigger after the sequence has started (i.e. been initialised).

Wating for the Next Thing, Starting from the Last

• clearing_policy == ON_SUCCESS

Useful as a guard watching for an event that could have come in anytime, but for which we do with to reset (and subsequently look for the next event). e.g. button events.

Parameters

• topic_name (str) – name of the topic to connect to

• topic_type (Any) – class of the message type (e.g. std_msgs.msg.String)

• qos_profile (rclpy.qos.QoSProfile) – qos profile for the subscriber

• name (str) – name of the behaviour

• clearing_policy – when to clear the data

update()

Returns

RUNNING (no data) or SUCCESS

Return type

Status

py_trees_ros.transforms

Various behaviours that enable common interactions with ROS transforms.

class py_trees_ros.transforms.FromBlackboard(name, variable_name, target_frame, source_frame, static, qos_profile, static_qos_profile=None)

Bases: Behaviour

Broadcast a transform from the blackboard using the transform broadcaster mechanisms.

If no it fails to find a geometry_msgs.Transform object on the blackboard, or the blackboard variable is None, this behaviour will update with status FAILURE.

Parameters

• variable_name (str) – name of the transform variable on the blackboard

• target_frame (str) – name of the frame to transform into

• source_frame (str) – name of the input frame

• static (bool) – designate whether it is a static transform or otherwise

• qos_profile (rclpy.qos.QoSProfile) – qos profile for the non-static broadcaster

• static_qos_profile (rclpy.qos.QoSProfile) – qos profile for the static broadcaster (default: use tf2_ros’ defaults)

• name (str) – name of the behaviour

setup(**kwargs)

Initialises the transform broadcaster.

Parameters

**kwargs (dict) – distribute arguments to this behaviour and in turn, all of it’s children

Raises

KeyError – if a ros2 node isn’t passed under the key ‘node’ in kwargs

update()

Retrieves the transform from the blackboard, stamps it and subsequently broadcasts it.

Raises

TypeError – if the blackboard variable is of the wrong type.

class py_trees_ros.transforms.ToBlackboard(name, variable_name, target_frame, source_frame, qos_profile, static_qos_profile=None, clearing_policy=ClearingPolicy.ON_INITIALISE)

Bases: Behaviour

Blocking behaviour that looks for a transform and writes it to a variable on the blackboard.

If it fails to find a transform immediately, it will update with status RUNNING and write ‘None’ to the blackboard.

Tip

To ensure consistent decision making, use this behaviour up-front in your tree’s tick to record a transform that can be locked in for the remainder of the tree tick.

Usage Patterns

• clearing_policy == ON_INTIALISE

Use if you have subsequent behaviours that need to make decisions on whether the transform was received or not.

• clearing_policy == NEVER

Never clear the result. Useful for static transforms or if you are doing your own lookup on the timestamps for any relevant decision making.

Parameters

• variable_name – name of the key to write to on the blackboard

• target_frame (str) – name of the frame to transform into

• source_frame (str) – name of the input frame

• qos_profile (rclpy.qos.QoSProfile) – qos profile for the non-static subscriber

• static_qos_profile (rclpy.qos.QoSProfile) – qos profile for the static subscriber (default: use tf2_ros’ defaults)

• name (str) – name of the behaviour

• clearing_policy (ClearingPolicy) –

Raises

TypeError – if the clearing policy is neither ON_INITIALISE or NEVER

initialise()

Clear the blackboard variable (set to ‘None’) if using the ON_INTIALISE policy.

setup(**kwargs)

Initialises the transform listener.

Parameters

**kwargs (dict) – distribute arguments to this behaviour and in turn, all of it’s children

Raises

KeyError – if a ros2 node isn’t passed under the key ‘node’ in kwargs

update()

Checks for the latest transform and posts it to the blackboard if available.

py_trees_ros.trees

The py_trees_ros.trees.BehaviourTree class extends the core py_trees.trees.BehaviourTree class with both blackboard and (tree) snapshot streaming services. Interact with these services via the py-trees-blackboard-watcher and py-trees-tree-watcher command line utilities.

class py_trees_ros.trees.BehaviourTree(root, unicode_tree_debug=False)

Bases: BehaviourTree

Extend the py_trees.trees.BehaviourTree class with a few bells and whistles for ROS.

ROS Parameters:

• default_snapshot_stream: enable/disable the default snapshots stream in ~/snapshots` (default: False)

• default_snapshot_period: periodically publish (default: math.inf)

• default_snapshot_blackboard_data: include tracking and visited variables (default: True)

• default_snapshot_blackboard_activity: include the blackboard activity (default: False)

• setup_timeout: time (s) to wait (default: math.inf)

  • if math.inf, it will block indefinitely

ROS Publishers:

• ~/snapshots (py_trees_interfaces.msg.BehaviourTree): the default snapshot stream, if enabled

ROS Services:

• ~/blackboard_streams/close (py_trees_ros_interfaces.srv.CloselackboardWatcher)

• ~/blackboard_streams/get_variables (py_trees_ros_interfaces.srv.GetBlackboardVariables)

• ~/blackboard_streams/open (py_trees_ros_interfaces.srv.OpenBlackboardWatcher)

• ~/snapshot_streams/close (py_trees_ros_interfaces.srv.CloseSnapshotsStream)

• ~/snapshot_streams/open (py_trees_ros_interfaces.srv.OpenSnapshotsStream)

• ~/snapshot_streams/reconfigure (py_trees_ros_interfaces.srv.ReconfigureSnapshotsStream)

Topics and services are not intended for direct use, but facilitate the operation of the utilities py-trees-tree-watcher and py-trees-blackboard-watcher.

Parameters

• root (Behaviour) – root node of the tree

• unicode_tree_debug (bool) – print to console the visited ascii tree after every tick

Raises

AssertionError – if incoming root variable is not the correct type

setup(node=None, node_name='tree', timeout=Duration.INFINITE, visitor=None, **kwargs)

Setup the publishers, exchange and add ROS relevant pre/post tick handlers to the tree. Ultimately relays this call down to all the behaviours in the tree.

Parameters

• node (Optional[rclpy.node.Node]) – Optional ROS Node object. If None (default), creates its own node.

• node_name (str) – Name of ROS node created. Only takes effect if node is None.

• timeout (float) – time (s) to wait (use common.Duration.INFINITE to block indefinitely)

• visitor (Optional[VisitorBase]) – runnable entities on each node after it’s setup

• **kwargs (int) – distribute args to this behaviour and in turn, to it’s children

Raises

• rclpy.exceptions.NotInitializedException – rclpy not yet initialised

• TypeError – node argument is not a valid rclpy.node.Node object

• Exception – be ready to catch if any of the behaviours raise an exception

Parameters

• node (Optional[rclpy.node.Node]) –

• node_name (str) –

• timeout (float) –

• visitor (Optional[VisitorBase]) –

• kwargs (int) –

shutdown()

Cleanly shut down rclpy timers and nodes.

tick_tock(period_ms, number_of_iterations=-1, pre_tick_handler=None, post_tick_handler=None)

Tick continuously at the period specified.

This is a re-implementation of the tick_tock() tick_tock that takes advantage of the rclpy timers so callbacks are interleaved inbetween rclpy callbacks (keeps everything synchronous so no need for locks).

Parameters

• period_ms (float) – sleep this much between ticks (milliseconds)

• number_of_iterations (int) – number of iterations to tick-tock

• pre_tick_handler (func) – function to execute before ticking

• post_tick_handler (func) – function to execute after ticking

class py_trees_ros.trees.SnapshotStream(node, topic_name=None, parameters=None)

Bases: object

SnapshotStream instances are responsible for creating / destroying a snapshot stream as well as the configurable curation of snapshots that are published on it.

Parameters

• node (rclpy.node.Node) –

• topic_name (str) –

• parameters (SnapshotStream.Parameters) –

class Parameters(blackboard_data=False, blackboard_activity=False, snapshot_period=Duration.INFINITE)

Bases: object

Reconfigurable parameters for the snapshot stream.

Parameters

• blackboard_data (bool) – publish blackboard variables on the visited path

• blackboard_activity (bool) – enable and publish blackboard activity in the last tick

• snapshot_period (float) – period between snapshots (use /inf to only publish on tree status changes)

static expand_topic_name(node, topic_name)

Custom name expansion depending on the topic name provided. This is part of the stream configuration on request which either provides no hint (automatic name generation), a simple hint (relative name expanded under the snapshot streams namespace) or a complete hint (absolute name that does not need expansion).

Parameters

• node (rclpy.node.Node) – node used to reference the absolute namespace if the hint is not absolute

• topic_name (str) – hint for the topic name

Returns

the expanded topic name

Return type

str

publish(root, changed, statistics, visited_behaviour_ids, visited_blackboard_client_ids)

” Publish a snapshot, including only what has been parameterised.

Parameters

• root (Behaviour) – the tree

• changed (bool) – whether the tree status / graph changed or not

• statistics (<Mock name='mock.msg.Statistics' id='140605245586304'>) – add tree statistics to the snapshot data

• visited_behaviour_ids (Set[UUID]) – behaviours on the visited path

• visited_blackboard_client_ids (Set[UUID]) – blackboard clients belonging to behaviours on the visited path

shutdown()

Shutdown the (temporarily) created publisher.

class py_trees_ros.trees.Watcher(topic_name=None, namespace_hint=None, parameters=None, statistics=False, mode=WatcherMode.SNAPSHOTS)

Bases: object

The tree watcher sits on the other side of a running BehaviourTree and is a useful mechanism for quick introspection of it’s current state.

Parameters

• topic_name (str) – location of the snapshot stream (optional)

• namespace_hint (str) – refine search for snapshot stream services if there is more than one tree (optional)

• parameters (Parameters) – snapshot stream configuration controlling both on-the-fly stream creation and display

• statistics (bool) – display statistics

• mode (WatcherMode) –

See also

py_trees_ros.programs.tree_watcher

callback_snapshot(msg)

Formats the string message coming in.

Args

msg (py_trees_ros_interfaces.msg.BehaviourTree):serialised snapshot

create_service_client(key)

Convenience api for opening a service client and waiting for the service to appear.

Parameters

key (str) – one of ‘open’, ‘close’.

Raises

• NotReadyError – if setup() wasn’t called to identify the relevant services to connect to.

• TimedOutError – if it times out waiting for the server

setup(timeout_sec)

Creates the node and checks that all of the server-side services are available for calling on to open a connection.

Parameters

timeout_sec (float) – time (s) to wait (use common.Duration.INFINITE to block indefinitely)

Raises

• NotFoundError – if services/topics were not found

• MultipleFoundError – if multiple services were found

• TimedOutError – if service clients time out waiting for the server

class py_trees_ros.trees.WatcherMode(value)

Bases: Enum

An enumerator specifying the view mode for the watcher

DOT_GRAPH = 'DOT_GRAPH'

Render with the dot graph representation of the static tree (using an application or text to console).

SNAPSHOTS = 'SNAPSHOTS'

Render ascii/unicode snapshots from a snapshot stream

py_trees_ros.utilities

Assorted utility functions.

class py_trees_ros.utilities.Publishers(node, publisher_details, introspection_topic_name='publishers')

Bases: object

Utility class that groups the publishers together in one convenient structure.

Parameters

(obj (publisher_details) – tuple): list of (str, str, msgType, bool, int) tuples representing (unique_name, topic_name, publisher_type, latched) specifications for creating publishers

Examples

Convert the incoming list of specifications into proper variables of this class.

publishers = py_trees.utilities.Publishers(
    [
        ('foo', '~/foo', std_msgs.String, True, 5),
        ('bar', '/foo/bar', std_msgs.String, False, 5),
        ('foobar', '/foo/bar', std_msgs.String, False, 5),
    ]
)

class py_trees_ros.utilities.Services(node, service_details, introspection_topic_name='services')

Bases: object

Utility class that groups services together in one convenient structure.

Parameters

(obj (service_details) – tuple): list of (str, str, srvType, func) tuples representing (unique_name, topic_name, service_type, callback) specifications for creating services

Examples

Convert the incoming list of specifications into proper variables of this class.

services = py_trees.utilities.Services(
    [
        ('open_foo', '~/get_foo', foo_interfaces.srv.OpenFoo, open_foo_callback),
        ('open_foo', '/foo/open', foo_interfaces.srv.OpenFoo, self.open_foo_callback),
        ('get_foo_bar', '/foo/bar', foo_interfaces.srv.GetBar, self.foo.get_bar_callback),
    ]
)

class py_trees_ros.utilities.Subscribers(node, subscriber_details, introspection_topic_name='subscribers')

Bases: object

Utility class that groups subscribers together in one convenient structure.

Parameters

(obj (subscriber_details) – tuple): list of (str, str, msgType, bool, func) tuples representing (unique_name, topic_name, subscriber_type, latched, callback) specifications for creating subscribers

Examples

Convert the incoming list of specifications into proper variables of this class.

subscribers = py_trees.utilities.Subscribers(
    [
        ('foo', '~/foo', std_msgs.String, True, foo),
        ('bar', '/foo/bar', std_msgs.String, False, self.foo),
        ('foobar', '/foo/bar', std_msgs.String, False, foo.bar),
    ]
)

py_trees_ros.utilities.basename(name)

Generate the basename from a ros name.

Parameters

name (str) – ros name

Returns

name stripped up until the last slash or tilde character.

Return type

str

Examples

basename("~dude")
# 'dude'
basename("/gang/dude")
# 'dude'

py_trees_ros.utilities.create_anonymous_node_name(node_name='node')

Creates an anonoymous node name by adding a suffix created from a monotonic timestamp, sans the decimal.

Returns

the unique, anonymous node name

Return type

str

py_trees_ros.utilities.find_service(node, service_type, namespace=None, timeout=0.5)

Discover a service of the specified type and if necessary, under the specified namespace.

Parameters

• node (rclpy.node.Node) – nodes have the discovery methods

• service_type (str) – primary lookup hint

• namespace (str) – secondary lookup hint

• timeout (float) – immediately post node creation, can take time to discover the graph (sec)

Returns

fully expanded service name

Return type

str

Raises

• NotFoundError – if no services were found

• MultipleFoundError – if multiple services were found

py_trees_ros.utilities.find_topics(node, topic_type, namespace=None, timeout=0.5)

Discover a topic of the specified type and if necessary, under the specified namespace.

Parameters

• node (rclpy.node.Node) – nodes have the discovery methods

• topic_type (str) – primary lookup hint

• namespace (Optional[str]) – secondary lookup hint

• timeout (float) – check every 0.1s until this timeout is reached (can be None -> checks once)

Return type

List[str]

Returns

list of fully expanded topic names (can be empty)

Parameters

• node (rclpy.node.Node) –

• topic_type (str) –

• namespace (Optional[str]) –

• timeout (float) –

Return type

List[str]

py_trees_ros.utilities.get_py_trees_home()

Find the default home directory used for logging, bagging and other esoterica.

py_trees_ros.utilities.qos_profile_latched()

Convenience retrieval for a latched topic (publisher / subscriber)

py_trees_ros.utilities.qos_profile_unlatched()

Default profile for an unlatched topic (in py_trees_ros).

py_trees_ros.utilities.resolve_name(node, name)

Convenience function for getting the resolved name (similar to ‘publisher.resolved_name’ in ROS1).

Parameters

• node (rclpy.node.Node) – the node, namespace it should be relevant to

• (obj (name) – str): topic or service name

Note

This entirely depends on the user providing the relevant node, name pair.

py_trees_ros.visitors

ROS Visitors are entities that can be passed to a ROS tree implementation (e.g. BehaviourTree) and used to either visit each and every behaviour in the tree, or visit behaviours as the tree is traversed in an executing tick. At each behaviour, the visitor runs its own method on the behaviour to do as it wishes - logging, introspecting).

Warning

Visitors should not modify the behaviours they visit.

See also

The base interface and core visitors in py_trees.visitors

class py_trees_ros.visitors.SetupLogger(node)

Bases: VisitorBase

Use as a visitor to py_trees_ros.trees.TreeManager.setup() to log the name and timings of each behaviours’ setup to the ROS debug channel.

Parameters

node (rclpy.node.Node) – an rclpy node that will provide debug logger

finalise()

Override if any work needs to be performed after ticks (i.e. showing data).

initialise()

Initialise the timestamping chain.

run(behaviour)

Converse with the behaviour.

This method gets run as each behaviour is ticked. Override it to perform some activity - e.g. introspect the behaviour to store/process logging data for visualisations.

Parameters

behaviour (Behaviour) – behaviour that is ticking

class py_trees_ros.visitors.TreeToMsgVisitor

Bases: VisitorBase

Visits the entire tree and gathers all behaviours as messages for the tree logging publishers.

Variables

tree (py_trees_msgs.msg.BehaviourTree) – tree representation in message form

initialise()

Initialise and stamp a py_trees_msgs.msg.BehaviourTree instance.

run(behaviour)

Convert the behaviour into a message and append to the tree.

Parameters

behaviour (Behaviour) – behaviour to convert

py_trees_ros.programs

Entry points to the programs (tools and utilities).

py_trees_ros.programs.blackboard_watcher

Open up a window onto the blackboard!

Introspect on the entire blackboard or a part thereof and receive a stream of updates whenever values change.

Examples:

# list all keys on the blackboard
$ py-trees-blackboard-watcher --list
# stream all variables
$ py-trees-blackboard-watcher 
# stream only visited variables and access details
$ py-trees-blackboard-watcher --visited --activity
# stream a single variable
$ py-trees-blackboard-watcher odometry
# stream only a single field within a variable
$ py-trees-blackboard-watcher odometry.pose.pose.position

usage: py-trees-blackboard-watcher [-h] [-l] [-a] [-v] [-n [NAMESPACE]] ...

Positional Arguments

variables

space separated list of blackboard variable names (may be nested) to watch

Default: []

Named Arguments

-l, --list

list the blackboard variable names

-a, --activity

include the logged activity stream for recent changes

Default: False

-v, --visited

filter selected keys from those associated with behaviours on the most recent tick’s visited path

Default: False

-n, --namespace

namespace of blackboard services (if there should be more than one blackboard)

Example interaction with the services of an py_trees_ros.blackboard.Exchange:

py_trees_ros.programs.blackboard_watcher.main(command_line_args=['-T', '-E', '-W', '--keep-going', '-b', 'html', '-d', '_build/doctrees', '-D', 'language=en', '.', '/home/docs/checkouts/readthedocs.org/user_builds/py-trees-ros/checkouts/devel/_readthedocs//html'])

Entry point for the blackboard watcher script.

py_trees_ros.programs.tree_watcher

Open up a window onto the behaviour tree!

Render a oneshot snapshot of the tree as a dot graph, or stream it and it’s state continuously as unicode art on your console. This utility automatically discovers the running tree and opens interfaces to that, but if there is more than one tree executing use the namespace argument to differentiate between trees.

Examples:

# render the tree as a dot graph (does not include runtime information)
$ py-trees-tree-watcher --dot-graph
# connect to an existing snapshot stream (e.g. the default, if it is enabled)
$ py-trees-tree-watcher /tree/snapshots
# open and connect to a snapshot stream, visualise the tree graph and it's changes only
$ py-trees-tree-watcher 
# open a snapshot stream and include visited blackboard variables
$ py-trees-tree-watcher -b
# open a snapshot stream and include blackboard access details (activity)
$ py-trees-tree-watcher -a
# open a snapshot stream and include timing statistics
$ py-trees-tree-watcher -s

usage: py-trees-tree-watcher [-h] [-n [NAMESPACE_HINT]] [-a] [-b] [-s]
                             [--snapshots | --dot-graph]
                             [topic_name]

Positional Arguments

topic_name

snapshot stream to connect to, will create a temporary stream if none specified

Named Arguments

-n, --namespace-hint

namespace hint snapshot stream services (if there should be more than one tree)

-a, --blackboard-activity

show logged activity stream (streaming mode only)

Default: False

-b, --blackboard-data

show visited path variables (streaming mode only)

Default: False

-s, --statistics

show tick timing statistics (streaming mode only)

Default: False

--snapshots

render ascii/unicode snapshots from a snapshot stream

--dot-graph

render the tree as a dot graph

Command line utility that introspects on a running BehaviourTree instance over it’s snapshot stream interfaces. Use to visualise the tree as a dot graph or track tree changes, timing statistics and blackboard variables visited by the tree on each tick.

py_trees_ros.programs.tree_watcher.echo_blackboard_contents(contents)

Parameters

contents (str) – blackboard contents

Note

The string comes pre-formatted with bash color identifiers and newlines. This is currently not especially good for anything other than debugging.

py_trees_ros.programs.tree_watcher.main()

Entry point for the tree watcher script.