ROS
- Also called: Robot Operating System, ROS 2, ROS2
- Category: Software & tooling › Tools & formats
- Tags: ros, ros2, middleware, robot-frameworks, software
The open-source middleware and ecosystem most robotics software is built on — and the source of much of the field's shared vocabulary.
Definition
ROS (Robot Operating System) is not an operating system. It is middleware: a set of libraries and conventions that let independent programs on one or many machines discover each other and exchange data, plus the tooling and package ecosystem built around that.
The core idea is decomposition. A robot is built as many small programs — nodes — each doing one job: reading a camera, estimating pose, planning a path, driving motors. Nodes communicate by publishing and subscribing to named topics, or by calling services and long-running actions. Because the transport is anonymous, a node does not know or care whether the camera on the other end is real hardware, a simulator, or a recording being replayed. That single property is why the same control code runs in simulation and on the robot, and it shapes how almost all modern robotics software is structured.
ROS 2 is the current generation and a substantial rewrite: it replaced the original custom transport with DDS, removed the single point of failure that the old master node represented, and added the real-time and multi-robot support the first version lacked. New work should assume ROS 2.
Its wider significance is vocabulary. URDF, tf frames, point cloud message types and the node-and-topic model came out of this ecosystem and are now used far beyond it, including by projects that never import a single ROS library. Understanding ROS is therefore less about one framework than about reading the field's shared language.
Why it matters
ROS is the closest thing robotics has to a common platform, and its concepts leak everywhere. Many other terms in this glossary — Transform, SLAM, Odometry, URDF, Sensor Fusion among them — describe themselves partly in ROS's language. Knowing it is what lets you read most robotics repositories at all, and knowing that it is middleware rather than an operating system is what stops the first hour of confusion.
Key applications
- Structuring a robot as independent nodes rather than one monolithic program
- Swapping simulated and real hardware beneath unchanged control code
- Recording sensor data on the robot and replaying it for offline debugging
- Reusing community packages for navigation, manipulation and perception
- Coordinating multiple robots over a shared network
- Describing robot geometry and coordinate frames in a portable way
In practice
A working system is a graph of nodes launched together, wired by topic names, described by a URDF and a tree of coordinate frames. Development leans on a handful of tools: RViz for visualising what the robot believes, rosbag for recording and replaying real sensor data, and launch files for bringing the graph up reproducibly. In practice, the ability to record a real run and replay it into unchanged code is what makes debugging a robot tractable at all.
Common misconceptions
- ROS is an operating system — it is middleware that runs on top of one, usually Linux
- ROS is real-time by nature — the original was not; ROS 2 makes real-time possible with the right executor and DDS settings, but it is not automatic
- ROS 2 is ROS 1 with a version bump — it is a rewrite with a different transport and no master node, and the two do not interoperate without a bridge
- Using ROS means using it for everything — many robots run ROS for perception and coordination while the real-time control loop lives on a microcontroller
Who should know this
- Anyone reading or contributing to open-source robotics code
- Builders deciding how to structure software across a robot's computers
- Anyone who wants simulation and hardware to run identical control code
- Engineers debugging a robot from recorded runs rather than live retries
- Students meeting terms like node, topic, tf or URDF for the first time
Related terms
Related projects
- oomwoo: Open-source DIY robot vacuum: 3D-printed, ROS2 + 2D LiDAR, fully local.
- PX4 Autopilot: Open-source flight control software for drones and other unmanned vehicles
- Cartographer: Google's real-time 2D and 3D SLAM system for multi-platform robotics
- OpenMower: Open-source project converting commercial mowers into autonomous RTK-GPS guided robots
- ROS 2: Next-gen Robot Operating System for real-time production robotics
- PlotJuggler: Fast time-series visualization tool for ROS topics, rosbags, and robot debugging
Videos
- Introduction to ROS Part 1: What is the Robot Operating System? | DigiKey: DigiKey
- What is ROS and Why use ROS to Learn Robotics?: Kevin Wood | Robotics & AI