Skip to main content

Chapter 2: ROS 2 Communication Model

Learning Objectives​

  • Understand the fundamental ROS 2 communication patterns: nodes, topics, services, and actions
  • Explain the differences between asynchronous (topics) and synchronous (services, actions) communication
  • Describe how execution graphs organize ROS 2 components
  • Apply appropriate communication patterns to humanoid robot sensor and actuator scenarios

Introduction​

In this chapter, we'll explore the core communication mechanisms that make ROS 2 powerful for robotics applications. You'll learn how different components of a robot system communicate with each other through nodes, topics, services, and actions. Understanding these patterns is essential for designing effective robotic systems, particularly for humanoid robots with complex sensorimotor requirements.

Main Content​

Nodes and Execution Graphs​

In ROS 2, a node is a process that performs computation. Nodes are the fundamental building blocks of a ROS 2 system. Each node typically handles a specific task, such as sensor data processing, control algorithm execution, or user interface management.

Nodes organize into an execution graph, which represents the structure and communication flow of your robotic system. This graph shows:

  • Which nodes exist in your system
  • How nodes communicate with each other
  • The flow of data between components
  • Dependencies between different system components

The execution graph is dynamic - nodes can be added, removed, or reconfigured while the system runs, providing flexibility for different robot configurations and operational modes.

Topics: Asynchronous Communication​

Topics provide asynchronous, many-to-many communication in ROS 2. They follow a publish-subscribe pattern where:

  • Publishers send messages to a topic without knowing who will receive them
  • Subscribers receive messages from a topic without knowing who sent them
  • Multiple publishers can send to the same topic
  • Multiple subscribers can receive from the same topic

Topics are ideal for:

  • Sensor data distribution (e.g., camera images, laser scans)
  • Continuous state updates (e.g., robot position, joint angles)
  • Event notifications (e.g., button presses, obstacle detection)

The asynchronous nature means publishers and subscribers don't need to run simultaneously - messages are queued until subscribers are available.

Services: Synchronous Communication​

Services provide synchronous, request-response communication between two nodes. This pattern involves:

  • A client node sends a request to a service
  • A server node processes the request and sends a response back
  • The client waits for the response before continuing

Services are appropriate for:

  • Operations with a clear beginning and end
  • Requesting specific computations (e.g., path planning, image processing)
  • Configuration changes that require confirmation
  • Any operation where the client needs to know the result before proceeding

Actions: Synchronous with Feedback​

Actions are goal-oriented communications that provide feedback during execution and can be canceled. They combine features of both topics and services:

  • A client sends a goal to an action server
  • The server provides feedback about progress during execution
  • The client can cancel the goal before completion
  • The server returns a result when the goal completes

Actions are perfect for:

  • Long-running operations (e.g., navigation to a distant location)
  • Tasks where progress monitoring is important (e.g., arm trajectory execution)
  • Operations that might need interruption (e.g., emergency stops)

Communication Pattern Selection​

Choosing the right communication pattern depends on your specific use case:

  • Use Topics when you need:

    • Continuous data flow
    • Multiple subscribers to the same information
    • Decoupled timing between publishers and subscribers
    • Example: Publishing camera images for multiple vision algorithms

  • Use Services when you need:

    • A specific response to a request
    • Confirmation that an operation completed
    • Synchronous behavior
    • Example: Requesting a map from a mapping service

  • Use Actions when you need:

    • Long-running operations with progress feedback
    • Ability to cancel operations
    • Asynchronous behavior with result guarantees
    • Example: Moving a robot arm to a specific position with progress updates

Key Concepts​

Node​

A node is a process that performs computation in ROS 2. Nodes are organized into packages to form a ROS system. Each node typically handles a specific task or function.

Topic​

A topic is an asynchronous, many-to-many communication mechanism in ROS 2. Multiple nodes can publish to or subscribe to the same topic without direct knowledge of each other.

Service​

A service is a synchronous, request-response communication pattern in ROS 2 between two nodes. The client waits for a response from the service server.

Action​

An action is a goal-oriented communication pattern that provides feedback during execution and can be canceled. It's ideal for long-running operations.

Execution Graph​

The execution graph represents the structure and communication flow of a ROS 2 system, showing how nodes connect and communicate with each other.

Humanoid-Specific Examples​

Consider a humanoid robot performing a reaching task:

Topics for continuous monitoring:

  • Joint position sensors publish to /joint_states for multiple controllers to monitor
  • IMU data publishes to /imu/data for balance and orientation systems
  • Camera images publish to /camera/image_raw for vision processing nodes

Services for discrete operations:

  • A trajectory planner service accepts requests to compute joint trajectories
  • A configuration service allows updating robot parameters with confirmation
  • A calibration service performs sensor calibration and reports results

Actions for goal-oriented tasks:

  • An arm controller action accepts reaching goals and provides progress feedback
  • A walking controller action manages complex locomotion with interrupt capability
  • A grasping action executes manipulation tasks with force feedback

Summary​

This chapter covered the core communication patterns in ROS 2:

  • Nodes as the fundamental computational units
  • Topics for asynchronous, many-to-many communication
  • Services for synchronous request-response interactions
  • Actions for goal-oriented operations with feedback

You now understand when to use each communication pattern and how they enable complex robotic behaviors in humanoid systems.

Next Steps​

The next chapter will explore how Python-based AI agents connect to ROS 2 systems using the rclpy client library, bridging your AI development skills with robotic systems.

Further Reading​

  • ROS 2 Documentation: Nodes and Topics
  • ROS 2 Documentation: Services and Actions
  • "Effective Robotics Programming with ROS" by Anil Mahtani