burlap.ros

Class AsynchronousRosEnvironment

java.lang.Object

burlap.ros.AsynchronousRosEnvironment

All Implemented Interfaces:

burlap.oomdp.singleagent.environment.Environment, ros.RosListenDelegate

Deprecated.

@Deprecated
public class AsynchronousRosEnvironment
extends Object
implements burlap.oomdp.singleagent.environment.Environment, ros.RosListenDelegate

NOTE: This class is now deprecated! You should use the more general RosEnvironment or subclass the AbstractRosEnvironment instead. If you want actions to send their message as a string of their name, set the ActionPublisher for all actions to a ActionStringPublisher.

An environment wrapper for state information provided over ROS with BURLAP actions that are published to ROS. This environment connects to ROS using ROSBridge, which must be running. State information from ROS is expected to use be of type burlap_msgs/burlap_state. The burlap_state message is parsed into an actual BURLAP State object using the object classes defined in a provided BURLAP Domain. This parsed state may be further modified before the environment's current state is set to it by overriding the onStateReceive(burlap.oomdp.core.states.State) method, which receives the parsed state and returns a state object to which the environment's current state will be updated.

When this environment has an action request (via executeAction(burlap.oomdp.singleagent.GroundedAction)), it turns the request into a GroundedAction object and a string rep of the object is retrieved (via the GroundedAction.toString() method, and then published to a ROS topic. The calling thread is then stalled for some delay (giving time for the action to be executed on the ROS robot and the state updated) before the executeAction(burlap.oomdp.singleagent.GroundedAction) method returns. The fact that there is no "action completion" checking is why this is considered an asynchronous environment.

After creating an environment, it may be a good idea to call the blockUntilStateReceived() method before doing anything with it. This method will block the calling thread until the environment receives a state message from ROS which it uses to set its current state.

A BURLAP reward function and terminal function can also be set (setRewardFunction(burlap.oomdp.singleagent.RewardFunction) and setTerminalFunction(burlap.oomdp.core.TerminalFunction)) so that the environment will returns meaningful messages from the getLastReward() and isInTerminalState() methods.

Note that the the environment's current state will be updated as frequently as ROSBridge provides updates, but the states before and after actions returned are those from fixed time intervals set by the client (and described above).

You may want to specify a throttle rate and queue size for how quickly the environment receives state message from ROS Bridge. If States are published too quickly, you may find the environment's state is lagging behind as it processes each received state in order.

Author:

James MacGlashan.

Nested Class Summary

Nested classes/interfaces inherited from interface ros.RosListenDelegate

ros.RosListenDelegate.LegacyFormat

Field Summary

Fields

Modifier and Type	Field and Description
protected ros.Publisher	actionPub Deprecated.
protected int	actionSleepMS Deprecated.
protected boolean	addHeader Deprecated.
protected burlap.oomdp.core.states.State	curState Deprecated.
protected int	debugCode Deprecated.
protected burlap.oomdp.core.Domain	domain Deprecated.
protected double	lastReward Deprecated.
protected boolean	printStateAsReceived Deprecated.
protected Boolean	receivedFirstState Deprecated.
protected burlap.oomdp.singleagent.RewardFunction	rf Deprecated.
protected ros.RosBridge	rosBridge Deprecated.
protected burlap.oomdp.core.TerminalFunction	tf Deprecated.

Constructor Summary

Constructors

Constructor and Description
AsynchronousRosEnvironment(burlap.oomdp.core.Domain domain, String rosBridgeURI, String rosStateTopic, String rosActionTopic, int actionSleepMS) Deprecated. Creates an environment wrapper for state information provided over ROS with BURLAP actions needing to be published to ROS.
AsynchronousRosEnvironment(burlap.oomdp.core.Domain domain, String rosBridgeURI, String rosStateTopic, String rosActionTopic, int actionSleepMS, int rosBridgeThrottleRate, int rosBridgeQueueLength) Deprecated. Creates an environment wrapper for state information provided over ROS with BURLAP actions needing to be published to ROS.
AsynchronousRosEnvironment(burlap.oomdp.core.Domain domain, String rosBridgeURI, String rosStateTopic, String rosActionTopic, String rosStateTopicMessageType, String rosActionTopicMessageType, int actionSleepMS, int rosBridgeThrottleRate, int rosBridgeQueueLength) Deprecated. Creates an environment wrapper for state information provided over ROS with BURLAP actions needing to be published to ROS.

Method Summary

Modifier and Type	Method and Description
void	blockUntilStateReceived() Deprecated. A method you can call that forces the calling thread to wait until the first state from ROS has been received.
burlap.oomdp.singleagent.environment.EnvironmentOutcome	executeAction(burlap.oomdp.singleagent.GroundedAction ga) Deprecated.
burlap.oomdp.core.states.State	getCurrentObservation() Deprecated.
Map<String,Object>	getHeader() Deprecated. Specifies the header data that will be added to an action message if this environment has been set to add a header (by default headers are not added).
double	getLastReward() Deprecated.
ros.RosBridge	getRosBridge() Deprecated. Returns the RosBridge object to which this environment is connected.
boolean	isInTerminalState() Deprecated.
protected burlap.oomdp.core.states.State	JSONPreparedToState(List<Map<String,Object>> objects) Deprecated.
protected burlap.oomdp.core.states.State	JSONToState(com.fasterxml.jackson.databind.JsonNode objects) Deprecated. Takes a JsonNode that represents the BURLAP state and turns it into a BURLAP State object.
protected burlap.oomdp.core.states.State	onStateReceive(burlap.oomdp.core.states.State s) Deprecated. This method is called whenever a state message from ROS is received.
void	receive(com.fasterxml.jackson.databind.JsonNode data, String stringRep) Deprecated.
void	receive(Map<String,Object> data, String stringRep) Deprecated.
void	resetEnvironment() Deprecated.
void	setAddHeader(boolean addHeader) Deprecated. Specify whether a header object should be added to the action message.
void	setPrintStateAsReceived(boolean printStateAsReceived) Deprecated. Sets whether the default implementation of onStateReceive(burlap.oomdp.core.states.State) will print the state information to the terminal.
void	setRewardFunction(burlap.oomdp.singleagent.RewardFunction rf) Deprecated. Sets a BURLAP RewardFunction to use to provide reward signals from the getLastReward() method.
void	setTerminalFunction(burlap.oomdp.core.TerminalFunction tf) Deprecated. Sets the BURLAP TerminalFunction to use to provide terminal state checks from the isInTerminalState().

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Field Detail

domain

protected burlap.oomdp.core.Domain domain

Deprecated.

rosBridge

protected ros.RosBridge rosBridge

Deprecated.

actionPub

protected ros.Publisher actionPub

Deprecated.

actionSleepMS

protected int actionSleepMS

Deprecated.

curState

protected burlap.oomdp.core.states.State curState

Deprecated.

addHeader

protected boolean addHeader

Deprecated.

rf

protected burlap.oomdp.singleagent.RewardFunction rf

Deprecated.

tf

protected burlap.oomdp.core.TerminalFunction tf

Deprecated.

lastReward

protected double lastReward

Deprecated.

receivedFirstState

protected Boolean receivedFirstState

Deprecated.

printStateAsReceived

protected boolean printStateAsReceived

Deprecated.

debugCode

protected int debugCode

Deprecated.

Constructor Detail

AsynchronousRosEnvironment

public AsynchronousRosEnvironment(burlap.oomdp.core.Domain domain,
                                  String rosBridgeURI,
                                  String rosStateTopic,
                                  String rosActionTopic,
                                  int actionSleepMS)

Deprecated.

Creates an environment wrapper for state information provided over ROS with BURLAP actions needing to be published to ROS. State information from ROS is expected to use be of type burlap_msgs/burlap_state. The burlap_state message is parsed into an actual BURLAP State object using the object classes defined in a provided BURLAP Domain.

When this environment has an action request (via executeAction(burlap.oomdp.singleagent.GroundedAction)), it turns the request into a GroundedAction object and a string rep of the object is retrieved (via the GroundedAction.toString() method, and then published to a ROS topic. The calling thread is then stalled for some delay (giving time for the action to be executed on the ROS robot and the state updated) before the executeAction(burlap.oomdp.singleagent.GroundedAction) method returns.

Parameters:

domain - the domain into which ROS burlap_state messages are parsed

rosBridgeURI - the URI of the ros bridge server. Note that by default, ros bridge uses port 9090. An example URI is ws://localhost:9090

rosStateTopic - the name of the ROS topic that publishes the burlap_msgs/burlap_state messages.

rosActionTopic - the name of the ROS topic to which BURLAP actions are published (as strings)

actionSleepMS - the amount of time that the executeAction(burlap.oomdp.singleagent.GroundedAction) method stalls after publishing an action.

AsynchronousRosEnvironment

public AsynchronousRosEnvironment(burlap.oomdp.core.Domain domain,
                                  String rosBridgeURI,
                                  String rosStateTopic,
                                  String rosActionTopic,
                                  int actionSleepMS,
                                  int rosBridgeThrottleRate,
                                  int rosBridgeQueueLength)

Deprecated.

Creates an environment wrapper for state information provided over ROS with BURLAP actions needing to be published to ROS. State information from ROS is expected to use be of type burlap_msgs/burlap_state. The burlap_state message is parsed into an actual BURLAP State object using the object classes defined in a provided BURLAP Domain.

When this environment has an action request (via executeAction(burlap.oomdp.singleagent.GroundedAction)), it turns the request into a GroundedAction object and a string rep of the object is retrieved (via the GroundedAction.toString() method, and then published to a ROS topic. The calling thread is then stalled for some delay (giving time for the action to be executed on the ROS robot and the state updated) before the executeAction(burlap.oomdp.singleagent.GroundedAction) )} method returns.

Parameters:

domain - the domain into which ROS burlap_state messages are parsed

rosBridgeURI - the URI of the ros bridge server. Note that by default, ros bridge uses port 9090. An example URI is ws://localhost:9090

rosStateTopic - the name of the ROS topic that publishes the burlap_msgs/burlap_state messages.

rosActionTopic - the name of the ROS topic to which BURLAP actions are published (as strings)

actionSleepMS - the amount of time that the executeAction(burlap.oomdp.singleagent.GroundedAction) method stalls after publishing an action.

rosBridgeThrottleRate - the ROS Bridge server throttle rate: how frequently the server will send state messages

rosBridgeQueueLength - the ROS Bridge queue length: how many messages are queued on the server; queueing is a consequence of the throttle rate

AsynchronousRosEnvironment

public AsynchronousRosEnvironment(burlap.oomdp.core.Domain domain,
                                  String rosBridgeURI,
                                  String rosStateTopic,
                                  String rosActionTopic,
                                  String rosStateTopicMessageType,
                                  String rosActionTopicMessageType,
                                  int actionSleepMS,
                                  int rosBridgeThrottleRate,
                                  int rosBridgeQueueLength)

Deprecated.

Creates an environment wrapper for state information provided over ROS with BURLAP actions needing to be published to ROS. Note that although you can specify a different message type for the state and action using this constructor, the State message should adhere to the same type in as the burlap_msgs/burlap_state (State information from ROS is expected to use be of type burlap_msgs/burlap_state (https://github.com/h2r/burlap_msgs) and the action topic message type should either be a std/String, or a message adhering to the time stamped string burlap_msgs/timed_action (https://github.com/h2r/burlap_msgs). The burlap_state message is parsed into an actual BURLAP State object using the object classes defined in a provided BURLAP Domain.

When this environment has an action request (via executeAction(burlap.oomdp.singleagent.GroundedAction)), it turns the request into a GroundedAction object and a string rep of the object is retrieved (via the GroundedAction.toString() method, and then published to a ROS topic. The calling thread is then stalled for some delay (giving time for the action to be executed on the ROS robot and the state updated) before the executeAction(burlap.oomdp.singleagent.GroundedAction) method returns.

Parameters:

domain - the domain into which ROS burlap_state messages are parsed

rosBridgeURI - the URI of the ros bridge server. Note that by default, ros bridge uses port 9090. An example URI is ws://localhost:9090

rosStateTopic - the name of the ROS topic that publishes the burlap_msgs/burlap_state messages.

rosActionTopic - the name of the ROS topic to which BURLAP actions are published (as strings)

rosStateTopicMessageType - the ROS message type used for states.

rosActionTopicMessageType - the ROS message type used for actions.

actionSleepMS - the amount of time that the executeAction(burlap.oomdp.singleagent.GroundedAction) method stalls after publishing an action.

rosBridgeThrottleRate - the ROS Bridge server throttle rate: how frequently the server will send state messages

rosBridgeQueueLength - the ROS Bridge queue length: how many messages are queued on the server; queueing is a consequence of the throttle rate

Method Detail

getRosBridge

public ros.RosBridge getRosBridge()

Deprecated.

Returns the RosBridge object to which this environment is connected.

Returns:

the RosBridge object to which this environment is connected.

setRewardFunction

public void setRewardFunction(burlap.oomdp.singleagent.RewardFunction rf)

Deprecated.

Sets a BURLAP RewardFunction to use to provide reward signals from the getLastReward() method.

Parameters:

rf - RewardFunction to use to provide reward signals.

setTerminalFunction

public void setTerminalFunction(burlap.oomdp.core.TerminalFunction tf)

Deprecated.

Sets the BURLAP TerminalFunction to use to provide terminal state checks from the isInTerminalState().

Parameters:

tf - the BURLAP TerminalFunction to use to provide terminal state checks

getCurrentObservation

public burlap.oomdp.core.states.State getCurrentObservation()

Deprecated.

Specified by:

getCurrentObservation in interface burlap.oomdp.singleagent.environment.Environment

isInTerminalState

public boolean isInTerminalState()

Deprecated.

Specified by:

isInTerminalState in interface burlap.oomdp.singleagent.environment.Environment

resetEnvironment

public void resetEnvironment()

Deprecated.

Specified by:

resetEnvironment in interface burlap.oomdp.singleagent.environment.Environment

setPrintStateAsReceived

public void setPrintStateAsReceived(boolean printStateAsReceived)

Deprecated.

Sets whether the default implementation of onStateReceive(burlap.oomdp.core.states.State) will print the state information to the terminal.

Parameters:

printStateAsReceived - if true, then the default implementation of onStateReceive(burlap.oomdp.core.states.State) will print states to the screen; if false, the state receiving is silent by default.

setAddHeader

public void setAddHeader(boolean addHeader)

Deprecated.

Specify whether a header object should be added to the action message. If yes, then the header will be retrieved from the method getHeader() (which you can override). By default, the returned header is unpopulated, which will cause RosBridge to auto populate the sequence and time stamp (but leave the frame as an empty string).

Parameters:

addHeader - if true, then a header will be added to the action message; if false, no header is added.

blockUntilStateReceived

public void blockUntilStateReceived()

Deprecated.

A method you can call that forces the calling thread to wait until the first state from ROS has been received.

receive

public void receive(com.fasterxml.jackson.databind.JsonNode data,
                    String stringRep)

Deprecated.

Specified by:

receive in interface ros.RosListenDelegate

receive

@Deprecated
public void receive(Map<String,Object> data,
                                String stringRep)

Deprecated.

onStateReceive

protected burlap.oomdp.core.states.State onStateReceive(burlap.oomdp.core.states.State s)

Deprecated.

This method is called whenever a state message from ROS is received. The new current state of the environment will be set to whatever this method returns. By default, this method simply returns the same reference and if this environment's printStateAsReceived data member is set to true, then it will print to the terminal the string representation of the state. Override this method to provide special handling of used state (e.g., adding objects to the state that ROS does not perceive).

Parameters:

s - the parsed state from the ROS message received.

executeAction

public burlap.oomdp.singleagent.environment.EnvironmentOutcome executeAction(burlap.oomdp.singleagent.GroundedAction ga)

Deprecated.

Specified by:

executeAction in interface burlap.oomdp.singleagent.environment.Environment

getLastReward

public double getLastReward()

Deprecated.

Specified by:

getLastReward in interface burlap.oomdp.singleagent.environment.Environment

getHeader

public Map<String,Object> getHeader()

Deprecated.

Specifies the header data that will be added to an action message if this environment has been set to add a header (by default headers are not added). Like all JSON messages, the header is represented as a map from the JSON field name to its values. A populated ROS header message map should have entries for the field names seq, time, and frame_id. By default, this method will return an empty Map, which will cause ROS Bridge to auto populate the seq and time fields and set the frame_id to the string "". If you want to set some of those values, override this method.

Returns:

a map specifying header field values.

JSONToState

protected burlap.oomdp.core.states.State JSONToState(com.fasterxml.jackson.databind.JsonNode objects)

Deprecated.

Takes a JsonNode that represents the BURLAP state and turns it into a BURLAP State object. The JSonNode at the top level is a list of objects. Each object has a map with the fields "name", "object_class", and "values". The former are just string definitions. The latter is another list of values. Each value has the fields "attribute" and "value". Attribute specified the name of the BURLAP attribute; "value" specifies the string value of the value. If the attribute is a MULTITARGETRELATIONAL type, then it is assumed the different object names to which it is pointing are separated by a single ',' without spaces.

Parameters:

objects - a JsonNode specifying the BURLAP state objects.

Returns:

A BURLAP State representation of the input JSON state.

JSONPreparedToState

@Deprecated
protected burlap.oomdp.core.states.State JSONPreparedToState(List<Map<String,Object>> objects)

Deprecated.

Takes a JSON prepared data structure from a ROS message representation of a state and turns it into a BURLAP state object. The JSON prepared version is a list of maps. Each map represents an object instance which stores the objects name ('name'), name of the object's class ('object_class'), a list of values ('values'). Each value is a map specifying the attribute name ('attribute') and its value in a string rep form ('value'). If the attribute is a MULTITARGETRELATIONAL type, then it is assumed the different object names to which it is pointing are separated by a single ',' without spaces.

Parameters:

objects - the list of OO-MDP object instances

Returns:

and OO-MDP State object.