agents.components.map_encoding

agents.components.map_encoding#

Module Contents#

Classes#

MapEncoding

Map encoding component that encodes text information as a semantic map based on the robots localization. It takes in map layers, position topic, map occupancy grid topic, and a vector database client. Map layers can be arbitrary text based outputs from other components such as MLLMs or Vision.

API#

class agents.components.map_encoding.MapEncoding(*, layers: List[agents.ros.MapLayer], position: agents.ros.Topic, map_topic: agents.ros.Topic, config: agents.config.MapConfig, db_client: agents.clients.db_base.DBClient, trigger: Union[agents.ros.Topic, List[agents.ros.Topic], float] = 10.0, callback_group=None, component_name='map_encoder_component', **kwargs)#

Bases: agents.components.component_base.Component

Map encoding component that encodes text information as a semantic map based on the robots localization. It takes in map layers, position topic, map occupancy grid topic, and a vector database client. Map layers can be arbitrary text based outputs from other components such as MLLMs or Vision.

Parameters:

  • layers (list[MapLayer]) – A list of map layer objects to be encoded.

  • position (Topic) – The topic for the current robot position.

  • map_topic (Topic) – The OccupancyGrid topic for storing and retrieving map data.

  • config (MapConfig) – The configuration for the map encoding component.

  • db_client (DBClient) – A database client to store and retrieve map data.

  • trigger (Union[Topic, list[Topic], float]) – An optional trigger value or topic that triggers the map encoding process.

  • callback_group – An optional callback group for the map encoding component.

  • kwargs – Additional keyword arguments.

Example usage:

position_topic = Topic(name="position", msg_type="Odometry")
map_topic = Topic(name="map", msg_type="OccupancyGrid")
config = MapConfig(map_name="map")
db_client = DBClient(db=ChromaDB("database_name"))
layers = [MapLayer(subscribes_to=text1, resolution_multiple=3),
          MapLayer(subscribes_to=detections1, temporal_change=True)]  # text1 and detections1 are String topics that are being published on by other components
map_encoding_component = MapEncoding(
    layers=layers,
    position=position_topic,
    map_topic=map_topic,
    config=config,
    db_client=db_client,
)
activate()#

activate.

deactivate()#

deactivate.

add_point(layer: agents.ros.MapLayer, point: tuple[numpy.ndarray, str]) → None#

Component action to add a user defined point to the map collection. This action can be executed on an event.

Parameters:

  • layer (MapLayer) – Layer to which the point should be added

  • point (tuple[np.ndarray, str]) – A tuple of position (numpy array) and text data (str)

Return type:

None