navigation2

ROS 2 Navigation Framework and System

2,461

1,244

2,461

View on GitHub

Top Related Projects

grid_map

2,580

Universal grid map library for mobile robotic mapping

navigation

2,298

ROS Navigation stack. Code for finding where the robot is and how it can get somewhere else.

moveit

1,638

:robot: The MoveIt motion planning framework

Quick Overview

Navigation2 (Nav2) is the next generation of the ROS Navigation Stack, designed for ROS 2. It provides a set of modules for robot navigation, including path planning, obstacle avoidance, and localization. Nav2 is highly configurable and extensible, making it suitable for a wide range of robotic applications.

Pros

Fully compatible with ROS 2, taking advantage of its improved performance and security features
Modular architecture allows for easy customization and extension of navigation capabilities
Supports various types of robots, including differential drive, omnidirectional, and legged robots
Includes advanced features like behavior trees for complex navigation tasks and 3D navigation support

Cons

Steeper learning curve compared to the original ROS Navigation Stack
Documentation, while improving, may still be lacking in some areas for advanced use cases
Requires more computational resources compared to the ROS 1 navigation stack
Some legacy ROS 1 navigation plugins may not be directly compatible and require porting

Code Examples

Creating a basic navigation node:

from nav2_simple_commander.robot_navigator import BasicNavigator
import rclpy

rclpy.init()
navigator = BasicNavigator()

# Set initial pose
initial_pose = navigator.get_initial_pose()
navigator.setInitialPose(initial_pose)

# Wait for navigation to fully activate
navigator.waitUntilNav2Active()

Sending a goal to the navigation stack:

from geometry_msgs.msg import PoseStamped

goal_pose = PoseStamped()
goal_pose.header.frame_id = 'map'
goal_pose.pose.position.x = 2.0
goal_pose.pose.position.y = 1.5
goal_pose.pose.orientation.w = 1.0

navigator.goToPose(goal_pose)

Using behavior trees for complex navigation tasks:

from nav2_msgs.action import NavigateToPose
from nav2_simple_commander.robot_navigator import BasicNavigator

navigator = BasicNavigator()

# Load and use a behavior tree
navigator.loadBehaviorTree('navigate_w_replanning_and_recovery.xml')
goal = NavigateToPose.Goal()
goal.pose.header.frame_id = 'map'
goal.pose.pose.position.x = 1.0
goal.pose.pose.position.y = 1.0
goal.pose.pose.orientation.w = 1.0

navigator.goToPose(goal)

Getting Started

To get started with Navigation2:

Install ROS 2 and the Nav2 stack:

sudo apt install ros-<ros2-distro>-navigation2 ros-<ros2-distro>-nav2-bringup

Create a new ROS 2 package for your robot:

ros2 pkg create --build-type ament_cmake my_robot_nav

Configure your robot's URDF, map, and Nav2 parameters.

Launch Nav2 with your robot:

ros2 launch nav2_bringup bringup_launch.py use_sim_time:=True map:=/path/to/map.yaml

Use the BasicNavigator API or RViz2 to send navigation goals to your robot.

Competitor Comparisons

grid_map

2,580

Universal grid map library for mobile robotic mapping

Pros of grid_map

Specialized for 2D grid map operations, offering efficient data structures and algorithms for map manipulation
Provides a wide range of map-specific functions, including interpolation, filtering, and geometric operations
Lightweight and focused, making it easier to integrate into existing robotics projects

Cons of grid_map

Limited scope compared to navigation2, focusing primarily on grid map operations rather than full navigation stack
May require additional packages or custom development for complete navigation solutions
Less active community and fewer contributors compared to the larger navigation2 project

Code Comparison

navigation2 (C++):

nav2_msgs::msg::Costmap costmap_msg;
nav2_costmap_2d::Costmap2D* costmap = costmap_ros_->getCostmap();
nav2_costmap_2d::Costmap2DPublisher::toMsg(*costmap, costmap_msg);

grid_map (C++):

grid_map::GridMap map;
map.setGeometry(grid_map::Length(1.0, 1.0), 0.01);
map.add("elevation");
map["elevation"].setConstant(0.0);

The code snippets demonstrate the different focus areas of the two projects. navigation2 deals with costmaps and navigation-specific data structures, while grid_map provides a more general-purpose grid map manipulation interface.

navigation

2,298

ROS Navigation stack. Code for finding where the robot is and how it can get somewhere else.

Pros of navigation

More mature and stable, with a longer history of development and use
Wider community support and extensive documentation
Compatible with older ROS 1 systems, beneficial for legacy projects

Cons of navigation

Limited to ROS 1, not compatible with ROS 2 without significant modifications
Less modular architecture, making it harder to extend or customize
Lacks some modern features and optimizations present in navigation2

Code Comparison

navigation (C++):

void DWAPlanner::calculateVelocities(...) {
  // Velocity calculation logic
}

navigation2 (C++):

void DWBLocalPlanner::computeVelocityCommands(...) {
  // More modular velocity computation
}

The navigation2 codebase generally demonstrates a more modular and flexible approach, with improved separation of concerns and easier extensibility compared to the original navigation stack.

moveit

1,638

:robot: The MoveIt motion planning framework

Pros of MoveIt

Specialized for robotic arm manipulation and motion planning
Extensive inverse kinematics solvers and collision checking capabilities
Rich ecosystem of plugins and tools for various robotic applications

Cons of MoveIt

Steeper learning curve for beginners compared to Navigation2
Less focused on mobile robot navigation and SLAM
Can be resource-intensive for simpler navigation tasks

Code Comparison

MoveIt (C++):

#include <moveit/move_group_interface/move_group_interface.h>

moveit::planning_interface::MoveGroupInterface move_group("arm");
geometry_msgs::Pose target_pose;
move_group.setPoseTarget(target_pose);
move_group.move();

Navigation2 (Python):

from nav2_simple_commander.robot_navigator import BasicNavigator

navigator = BasicNavigator()
goal_pose = PoseStamped()
navigator.goToPose(goal_pose)

MoveIt focuses on arm manipulation with detailed control over joint movements, while Navigation2 emphasizes mobile robot navigation with a simpler interface for goal-oriented movement. MoveIt's code involves more setup for specific robot configurations, whereas Navigation2 provides a more straightforward approach to navigation tasks.

Convert designs to code with AI

Introducing Visual Copilot: A new AI model to turn Figma designs to high quality code using your components.

Try Visual Copilot

README

Nav2

For detailed instructions on how to:

Please visit our documentation site. Please visit our community Slack here (if this link does not work, please contact maintainers to reactivate).

â ï¸ If you need professional services related to Nav2, please contact Open Navigation at info@opennav.org.

Our Sponsors

Please thank our amazing sponsors for their generous support of Nav2 on behalf of the community to allow the project to continue to be professionally maintained, developed, and supported for the long-haul! Open Navigation LLC provides project leadership, maintenance, development, and support services to the Nav2 & ROS community.

Dexory develops robotics and AI logistics solutions to drive better business decisions using a digital twin of warehouses to provide inventory insights.

Nvidia develops GPU and AI technologies that power modern robotics, autonomous driving, data centers, gaming, and more.

Polymath Robotics creates safety-critical navigation systems for industrial vehicles that are radically simple to enable and deploy.

Stereolabs produces the high-quality ZED stereo cameras with a complete vision pipeline from neural depth to SLAM, 3D object tracking, AI and more.

Confidential is just happy to support Nav2's mission!

Citation

If you use the navigation framework, an algorithm from this repository, or ideas from it please cite this work in your papers!

S. Macenski, F. MartÃn, R. White, J. Clavero. The Marathon 2: A Navigation System. IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2020.

@InProceedings{macenski2020marathon2,
  title = {The Marathon 2: A Navigation System},
  author = {Macenski, Steve and MartÃn, Francisco and White, Ruffin and GinÃ©s Clavero, Jonatan},
  year = {2020},
  booktitle = {2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)},
  url = {https://github.com/ros-planning/navigation2},
  pdf = {https://arxiv.org/abs/2003.00368}
}

If you use any of the algorithms in Nav2 or the analysis of the algorithms in your work, please cite this work in your papers!

S. Macenski, T. Moore, DV Lu, A. Merzlyakov, M. Ferguson, From the desks of ROS maintainers: A survey of modern & capable mobile robotics algorithms in the robot operating system 2, Robotics and Autonomous Systems, 2023.

  @article{macenski2023survey,
        title={From the desks of ROS maintainers: A survey of modern & capable mobile robotics algorithms in the robot operating system 2}, 
        author={S. Macenski, T. Moore, DV Lu, A. Merzlyakov, M. Ferguson},
        year={2023},
        journal = {Robotics and Autonomous Systems}
  }

If you use the Smac Planner (Hybrid A*, State Lattice, 2D), please cite this work in your papers!

S. Macenski, M. Booker, J. Wallace, Open-Source, Cost-Aware Kinematically Feasible Planning for Mobile and Surface Robotics. 2024.

@article{macenski2024smac,
      title={Open-Source, Cost-Aware Kinematically Feasible Planning for Mobile and Surface Robotics}, 
      author={Steve Macenski and Matthew Booker and Josh Wallace},
      year={2024},
      journal = {Arxiv}
}

If you use the Regulated Pure Pursuit Controller algorithm or software from this repository, please cite this work in your papers!

S. Macenski, S. Singh, F. Martin, J. Gines, Regulated Pure Pursuit for Robot Path Tracking. Autonomous Robots, 2023.

@article{macenski2023regulated,
      title={Regulated Pure Pursuit for Robot Path Tracking}, 
      author={Steve Macenski and Shrijit Singh and Francisco Martin and Jonatan Gines},
      year={2023},
      journal = {Autonomous Robots}
}

If you use our work on VSLAM and formal comparisons for service robot needs, please cite the paper:

A. Merzlyakov, S. Macenski. A Comparison of Modern General-Purpose Visual SLAM Approaches. IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2021.

@InProceedings{vslamComparison2021,
  title = {A Comparison of Modern General-Purpose Visual SLAM Approaches},
  author = {Merzlyakov, Alexey and Macenski, Steven},
  year = {2021},
  booktitle = {2021 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)},
  pdf = {https://arxiv.org/abs/2107.07589}
}

Build Status

Package	iron Source	iron Debian
navigation2
nav2_amcl
nav2_behavior_tree
nav2_behaviors
nav2_bringup
nav2_bt_navigator
nav2_collision_monitor
nav2_common
nav2_constrained_smoother
nav2_controller
nav2_core
nav2_costmap_2d
nav2_docking	N/A	N/A
nav2_dwb_controller
nav2_graceful_controller	N/A	N/A
nav2_lifecycle_manager
nav2_map_server
nav2_mppi_controller
nav2_msgs
nav2_navfn_planner
nav2_planner
nav2_regulated_pure_pursuit
nav2_rotation_shim_controller
nav2_rviz_plugins
nav2_simple_commander
nav2_smac_planner
nav2_smoother
nav2_system_tests
nav2_theta_star_planner
nav2_util
nav2_velocity_smoother
nav2_voxel_grid
nav2_waypoint_follower

Top Related Projects

Convert designs to code with AI

Introducing Visual Copilot: A new AI model to turn Figma designs to high quality code using your components.

Try Visual Copilot