Nav2

ROS 2 navigation stack with path planning and obstacle avoidance for mobile robots

License

Apache-2.0

Overview

Nav2 is the navigation stack for mobile robots in ROS 2 and the successor to the ROS 1 navigation stack. It provides path planning, AMCL localization, obstacle avoidance and recovery behaviors, with navigation logic built on behavior trees. Planner plugins, costmap layers and waypoint following let it be configured for different robots and sensors.

Nav2 (Navigation2) is the next-generation navigation stack for ROS 2, providing a complete solution for autonomous mobile robot navigation. It's the successor to the ROS 1 navigation stack with significant architectural improvements.

Key features include: - Multiple planner plugins (NavFn, Smac, Theta*) - Behavior tree-based navigation logic - Dynamic obstacle avoidance - Costmap layers for sensor fusion - AMCL localization - Recovery behaviors - Waypoint following and task execution - Lifecycle management for safety

Nav2 is used by hundreds of companies for warehouse robots, service robots, and autonomous guided vehicles. It's highly configurable and supports various robot platforms and sensor configurations.

Why it matters

Nav2 is the navigation stack that makes ROS 2 robots autonomous. It provides everything from path planning to obstacle avoidance in a production-ready package, and is deployed on thousands of real robots in warehouses, hospitals, and research labs worldwide.

Problem it solves

Making a mobile robot navigate autonomously requires integrating localization, mapping, planning, and recovery behaviors. Nav2 provides this complete stack for ROS 2, tested and maintained by the community.

Key takeaways

• Complete autonomous navigation stack for ROS 2 mobile robots
• Production-ready path planning, localization, and obstacle avoidance
• Behavior tree-based architecture for flexible navigation behaviors
• Supports multiple planning algorithms and costmap configurations
• Deployed in commercial robots across industries

Who it is for

• ROS 2 developers building mobile robots
• Companies deploying autonomous mobile robots
• Researchers studying robot navigation algorithms
• Teams building warehouse, delivery, or service robots

Limitations

• Configuration complexity for non-standard robot platforms
• 2D navigation focus — 3D terrain navigation requires extensions
• Performance tuning requires understanding of navigation theory
• Behavior tree architecture has a learning curve