Skip to main content

Building ROS 2 Packages with Python (rclpy)

Introduction​

Python is a popular language for robotics due to its simplicity and extensive libraries. ROS 2 provides rclpy, a client library for Python, enabling seamless integration of Python scripts with the ROS 2 ecosystem. This section will guide you through building ROS 2 packages using Python and rclpy to interface with ROS controllers.

ROS 2 Packages and Workspace​

In ROS 2, functionality is organized into packages. A package is a directory containing source code, build scripts, configuration files, and other resources. Multiple packages are typically organized within a workspace.

Creating a Workspace​

First, create a ROS 2 workspace:

mkdir -p ~/ros2_ws/src
cd ~/ros2_ws/
colcon build

Creating a New Python Package​

To create a new Python package within your workspace:

cd ~/ros2_ws/src
ros2 pkg create --build-type ament_python my_python_pkg

This command creates a directory named my_python_pkg with a basic structure.

Writing Python Nodes with rclpy​

A Python node using rclpy typically involves:

  1. Importing rclpy: import rclpy
  2. Initializing ROS 2: rclpy.init(args=args)
  3. Creating a node: node = rclpy.create_node('my_node_name')
  4. Implementing logic: Publishers, subscribers, service servers, service clients.
  5. Spinning the node: rclpy.spin(node) (keeps the node alive and processing events)
  6. Shutting down ROS 2: node.destroy_node() and rclpy.shutdown()

Example: A Simple Publisher Node​

import rclpy
from rclpy.node import Node
from std_msgs.msg import String

class MinimalPublisher(Node):

    def __init__(self):
        super().__init__('minimal_publisher')
        self.publisher_ = self.create_publisher(String, 'topic', 10)
        timer_period = 0.5  # seconds
        self.timer = self.create_timer(timer_period, self.timer_callback)
        self.i = 0

    def timer_callback(self):
        msg = String()
        msg.data = 'Hello ROS 2: %d' % self.i
        self.publisher_.publish(msg)
        self.get_logger().info('Publishing: "%s"' % msg.data)
        self.i += 1

def main(args=None):
    rclpy.init(args=args)
    minimal_publisher = MinimalPublisher()
    rclpy.spin(minimal_publisher)
    minimal_publisher.destroy_node()
    rclpy.shutdown()

if __name__ == '__main__':
    main()

Bridging Python Agents to ROS Controllers​

rclpy allows Python-based AI agents (e.g., planning algorithms, reinforcement learning agents) to publish commands to ROS 2 topics (e.g., motor commands) or call ROS 2 services (e.g., for complex motion planning). Conversely, agents can subscribe to sensor data topics to perceive the robot's environment.

This integration is key for developing intelligent robotic behaviors using advanced AI techniques in Python.

Build and Run​

After writing your Python node, you need to build your workspace:

cd ~/ros2_ws
colcon build --packages-select my_python_pkg

And then source your workspace and run your node:

source install/setup.bash
ros2 run my_python_pkg minimal_publisher