Skip to main content

ROS 2 Core Concepts

This chapter expands on the core concepts introduced in Module 1, with diagrams, patterns and guidelines.

Nodes​

  • Single responsibility principle: each node handles one logical function.
  • Lifecycle management: plan startup/shutdown and error recovery.

Topics​

  • Use topics for streaming sensor data and state information.
  • Define topic naming conventions and message rate expectations.

Services​

  • Use services for calibration calls, configuration changes, or short one-off queries.

Actions​

  • Use actions for missions or tasks that take time and need feedback (navigation, manipulation).

QoS Guidelines​

  • Sensors (high-frequency) may use BEST_EFFORT for performance-sensitive streams.
  • Critical control channels should use RELIABLE and appropriate deadlines.
  • Document QoS choices in design docs for each topic.

Executors & Scheduling​

  • Choose executor strategy (single-threaded, multi-threaded) based on concurrency needs.
  • Plan for real-time requirements: isolate critical threads and prioritize them.

Tools & Workflow​

  • Use topic/service/action naming conventions and a shared data dictionary for teams.
  • Maintain testing procedures for message validation and replay.

Diagrams & Patterns​

  • Include system-level diagrams to show flow between perception -> planner -> controller.
  • Use sequence diagrams for action lifecycles and failure modes.