Navigation2 is the ROS 2 navigation framework that provides a complete autonomous navigation system for mobile robots. Written primarily in C++, it serves as the successor to the original ROS navigation stack and offers a modern, modular architecture designed for both research and production robotics applications. The framework is actively maintained by Open Navigation LLC, which provides professional support and development services to ensure long-term sustainability of the project.
The repository encompasses a comprehensive set of navigation capabilities including path planning, localization, mapping, sensor integration, and behavior tree-based task execution. Navigation2 implements multiple planning algorithms such as the Smac Planner with Hybrid A*, State Lattice, and 2D planning options, along with the Regulated Pure Pursuit controller for path tracking. The system supports dynamic environments and can handle real-time obstacle avoidance while maintaining kinematically feasible trajectories for various robot morphologies. The framework integrates VSLAM capabilities for visual simultaneous localization and mapping, enabling robots to operate in previously unmapped environments.
The project maintains extensive documentation covering concepts, setup guides, tutorials, configuration, plugin development, and API references. Navigation2 provides Docker containers for simplified deployment and supports multiple ROS distributions. The framework is extensible through a plugin architecture that allows developers to implement custom planners, controllers, and behavior modules. The repository includes comprehensive testing infrastructure with GitHub Actions workflows and code coverage tracking via codecov.
Navigation2 has attracted significant industry backing from sponsors including Nvidia, AMD, Dexory, Polymath Robotics, Stereolabs, 3Laws Robotics, and Staer, reflecting its importance in the robotics ecosystem. The project has generated multiple peer-reviewed publications documenting its algorithms and design, with citations available for the Marathon 2 navigation system, the Smac Planner, the Regulated Pure Pursuit controller, and comparative analyses of VSLAM approaches. The repository's classification spans motion control, robot pathfinding, autonomous navigation, 2D planning, mapping, and sensor integration, positioning it as a foundational component for autonomous mobile robot systems across industrial, research, and service robotics domains.
