The Robot Operating System, or ROS, is an open-source framework for robot software. It is not an operating system in the traditional sense. It is a collection of libraries, tools, and conventions that help developers build robot applications. It handles communication between processes, hardware abstraction, and common algorithms.

ROS started at Stanford in 2007 and was later maintained by Willow Garage and the Open Source Robotics Foundation. ROS 2, the current version, improved real-time performance, security, and multi-robot support. Thousands of packages are available: drivers for sensors, implementations of SLAM, motion planning libraries, and simulation interfaces.

Core ROS concepts

• Nodes: processes that perform computation.
• Topics: publish-subscribe message channels.
• Services: request-response communication.
• Actions: long-running tasks with feedback.
• Packages: organized collections of code.

ROS is not perfect. It has a steep learning curve. Real-time performance requires careful configuration. ROS 2 is better but still not a hard real-time system. Some commercial robots use ROS for prototyping and then rewrite for production. Others use it directly. The strength of ROS is the community. A developer can find a package for almost any sensor or algorithm. That saves months of work. ROS has become the lingua franca of robotics research. Most papers include ROS code. Most labs use it. It is not the only framework, but it is the most widely used. For anyone entering robotics, learning ROS is almost mandatory.