A ROS topic is a named channel that carries a continuous stream of messages between nodes — how a robot's sensor data, commands, and state flow through its software, one publish–subscribe pipe at a time.
A ROS topic is a named pipe for data — like '/camera/image' or '/cmd_vel'. One program pumps messages in, others read them out. It's how streaming data moves around a robot's software.
If nodes are a robot's programs, topics are the pipes between them — the named channels that carry the constant flow of sensor data, commands, and state.
What it is
A ROS topic is a named, typed channel. A node publishes messages onto it; other nodes subscribe to receive them — the publish–subscribe pattern in action. The name (like /scan, /camera/image, /cmd_vel) is how everyone finds it, and the message type (e.g. sensor_msgs/LaserScan) defines the data's structure so publisher and subscriber agree on the format.
Data streams along named topics
Each topic is one typed stream. Nodes connect by publishing to and subscribing from topic names, never to each other directly.
Services — quick request/response: "reset the map," "is the gripper open?" One-shot, blocking.
Actions — long goals with feedback: "navigate to the kitchen," which takes time and can report progress or be cancelled.
Sensor and command data almost always flow over topics.
The gotcha: QoS
In ROS 2, every topic has a Quality of Service profile — reliability, durability, history depth — and a publisher and subscriber only connect if their profiles are compatible. The classic beginner bug is a running publisher and running subscriber that never exchange a message because their QoS settings don't match. Sensor streams typically use "best-effort," commands use "reliable."
Everyday tools
The command line makes topics tangible: ros2 topic list shows every channel, ros2 topic echo /scan prints live messages, and ros2 topic hz /scan measures the rate — the first things anyone reaches for when debugging a robot.
Why it matters
Topics are the circulatory system of a ROS robot — the streams that carry data from sensors through perception and planning to the motors. Understanding topics (and their QoS) is fundamental to building or debugging any ROS system.
