Controller

Motion control and dynamic obstacle avoidance.

The Controller is the real-time “pilot” of your robot. While the Planner looks ahead to find a global route, the Controller deals with the immediate reality — calculating velocity commands to follow the global path (path following) or a global target point (object following) while reacting to dynamic obstacles and adhering to kinematic constraints.

It supports several control algorithms, switched by configuration or at run time: Pure Pursuit, Stanley, DWA, DVZ and the vision followers.

Available Run Types

The Controller typically runs at a high frequency (10Hz-50Hz) to ensure smooth motion.

Timed	Periodic Control Loop. Computes a new velocity command periodically if all necessary inputs are available.
Action Server	Goal Tracking. Offers a ControlPath ROS2 Action. Continuously computes control commands until the goal is reached or the action is preempted.

Inputs

Key Name	Allowed Types	Number	Default
plan	nav_msgs.msg.Path	1	Topic(name="/plan", msg_type="Path")
location	nav_msgs.msg.Odometry, geometry_msgs.msg.PoseStamped, geometry_msgs.msg.Pose	1	Topic(name="/odom", msg_type="Odometry")
sensor_data	sensor_msgs.msg.LaserScan, sensor_msgs.msg.PointCloud2	1	Topic(name="/scan", msg_type="LaserScan")
local_map	nav_msgs.msg.OccupancyGrid	1	Topic(name="/local_map/occupancy_layer", msg_type="OccupancyGrid")
vision_detections	automatika_embodied_agents.msg.Detections, automatika_embodied_agents.msg.Trackings	1	None. Provide it to use vision target tracking.
depth_camera_info	sensor_msgs.msg.CameraInfo	1	None. Required for the RGB-D follower.
vision_depth	sensor_msgs.msg.Image, sensor_msgs.msg.PointCloud2	1	None. A separate depth source, an aligned depth image or a point cloud, paired with detections by time stamp within vision_depth_max_age.

Tip

Provide a vision_detections input topic to the controller to activate the creation of a vision-based target following action server, track_vision_target. See the Vision Tracking tutorial for more details.

Outputs

Key Name	Allowed Types	Number	Default
command	geometry_msgs.msg.Twist, geometry_msgs.msg.TwistStamped	1	Topic(name="/control", msg_type="Twist")
multi_command	kompass_interfaces.msg.TwistArray	1	Topic(name="/control_list", msg_type="TwistArray")
interpolation	nav_msgs.msg.Path	1	Topic(name="/interpolated_path", msg_type="Path")
local_plan	nav_msgs.msg.Path	1	Topic(name="/local_path", msg_type="Path")
tracked_point	nav_msgs.msg.Odometry, geometry_msgs.msg.PoseStamped, geometry_msgs.msg.Pose	1	Topic(name="/tracked_point", msg_type="PoseStamped")
path_samples	nav_msgs.msg.Path	1	Topic(name="/debug_samples", msg_type="Path"). Published with debug=True.

Algorithms

EMOS includes several production-ready control plugins suited for different environments:

• Pure Pursuit — Geometric path tracking towards a look-ahead point on the path. Simple, smooth and predictable.

• Stanley — Geometric path tracking using the front axle as reference. Best for Ackermann steering.

• DVZ — Deformable Virtual Zone. Reactive collision avoidance based on risk zones. Extremely fast for crowded dynamic environments.

• DWA — Dynamic Window Approach. Sample-based collision avoidance with GPU support. Considers kinematics to find optimal velocity.

• Vision followers — VisionRGBFollower and VisionRGBDFollower. Steer the robot to keep a visual target centered, from RGB alone or with depth.

See the Algorithms Reference for detailed descriptions of each algorithm.

Configuration

Field	Default	Effect
algorithm	DWA	The control algorithm, a ControllersID or its name.
control_time_step	0.1 s	Time step of the computed control sequence.
use_direct_sensor	false	Read obstacles from the sensor directly rather than from the local map.
ctrl_publish_type	Array	How commands go out: Sequence, Parallel or Array.
vision_depth_max_age	0.2 s	How far apart a detection and its separate depth may be in time.
debug	false	Publishes the sampled trajectories on path_samples.

Built-in Actions

set_algorithm(algorithm_value) switches the algorithm while the recipe runs, which is what an event uses to move from Pure Pursuit outdoors to DWA indoors. stop_path_tracking() ends the current tracking as if the path had been reached, clears the plan and stops the robot; a new plan resumes tracking. Both return (success, message). With vision tracking enabled, the StopVisionTracking service stops a running target follow.

Usage Example

from kompass.components import Controller, ControllerConfig
from kompass.control import ControllersID
from kompass.ros import Topic

# Setup custom configuration
my_config = ControllerConfig(loop_rate=10.0)

# Init a controller object
my_controller = Controller(component_name="controller", config=my_config)

# Change an input
my_controller.inputs(plan=Topic(name='/global_path', msg_type='Path'))

# Change run type (default "Timed")
my_controller.run_type = "ActionServer"

# Choose the algorithm
my_controller.algorithm = ControllersID.DWA