Skip to main content
Autonomous navigation enables the TurtleBot3 to navigate to goal positions without manual control. Using Navigation2 (Nav2) with a pre-built map, the robot plans paths, avoids obstacles, and reaches destinations autonomously.

​
Prerequisites

Before using autonomous navigation, you must have:
  1. A saved map: Create a map using SLAM first
  2. Map files: Both .pgm and .yaml files in ~/maps/
  3. Matching simulation world: Use the same world where you created the map
You cannot navigate autonomously without a map. Complete the SLAM mapping process first before attempting navigation.

​
Starting Navigation2

​
Using the alias

The pre-configured alias starts Navigation2 with your saved map: 
tb3_nav
This alias expects your map at ~/maps/my_map.yaml.

​
Using the full command

For more control or different map locations: 
ros2 launch turtlebot3_navigation2 navigation2.launch.py use_sim_time:=True map:=$HOME/maps/my_map.yaml
Replace my_map.yaml with your actual map filename:
ros2 launch turtlebot3_navigation2 navigation2.launch.py use_sim_time:=True map:=$HOME/maps/turtlebot3_world.yaml

​
Complete navigation workflow

1

Launch simulation

Start the same world you used for mapping:
tb3_world
The simulation environment must match your saved map.
2

Start Navigation2

In a second terminal, launch Nav2 with your map:
ros2 launch turtlebot3_navigation2 navigation2.launch.py use_sim_time:=True map:=$HOME/maps/turtlebot3_world.yaml
RViz2 will open automatically showing your map.
3

Access RViz2

Open the VNC desktop in your browser:
http://localhost:6080
Password: ros
4

Set initial pose

In RViz2:
  • Click the “2D Pose Estimate” button in the toolbar
  • Click on the map where the robot is currently located
  • Drag to set the robot’s orientation
  • Release the mouse button
5

Set navigation goal

In RViz2:
  • Click the “Navigation2 Goal” button in the toolbar
  • Click on your desired destination on the map
  • Drag to set the desired final orientation
  • Release the mouse button
6

Watch autonomous navigation

The robot will:
  • Plan a path to the goal (shown as a green line)
  • Start moving autonomously
  • Avoid obstacles dynamically
  • Navigate to the destination

​
Using RViz2 for navigation

​
Initial pose estimation

Telling Nav2 where the robot is located on the map. Steps:
  1. Click “2D Pose Estimate” button (top toolbar)
  2. Click on the map at the robot’s current location
  3. Drag the arrow to match the robot’s orientation
  4. Release to set the pose
Visual feedback:
  • Green particle cloud appears around the robot
  • Particles should converge to a tight cluster
  • If particles are spread out, your pose estimate is inaccurate
Compare the robot’s position in Gazebo with the map in RViz2 to accurately set the initial pose.

​
Setting navigation goals

Telling the robot where to go. Steps:
  1. Click “Navigation2 Goal” button (top toolbar)
  2. Click on your desired destination on the map
  3. Drag to set the final orientation you want
  4. Release to start navigation
What happens:
  • Nav2 computes a global path (green line)
  • Robot begins moving toward the goal
  • Local planner adjusts for dynamic obstacles
  • Robot stops when goal is reached

​
Monitoring navigation

In RViz2, watch:
  • Global path (green line): Planned route to goal
  • Local costmap (colored grid around robot): Obstacle awareness
  • Robot footprint (small square): Robot’s physical size
  • Laser scan (red dots): Real-time sensor data
Navigation status:
  • Robot moving: Navigation is active
  • Robot stopped at goal: Success!
  • Robot oscillating: Path planning difficulty
  • Robot stopped mid-path: Obstacle blocking or navigation failure

​
Workflow examples

Terminal 1: Launch world 
tb3_world
Terminal 2: Start navigation 
ros2 launch turtlebot3_navigation2 navigation2.launch.py use_sim_time:=True map:=$HOME/maps/turtlebot3_world.yaml
In RViz2 (VNC browser at http://localhost:6080):
  1. Set initial pose where robot starts
  2. Click Navigation2 Goal
  3. Choose a destination across the map
  4. Watch the robot navigate autonomously
Terminal 1: Launch house 
tb3_house
Terminal 2: Start navigation 
ros2 launch turtlebot3_navigation2 navigation2.launch.py use_sim_time:=True map:=$HOME/maps/house_complete.yaml
In RViz2:
  1. Set initial pose in the starting room
  2. Set navigation goal in a different room
  3. Watch the robot navigate through doorways autonomously

​
Best practices

​
Accurate initial pose

Critical for success:
  • Look at robot position in Gazebo
  • Match that position on the RViz2 map
  • Align the arrow with robot’s facing direction
  • If navigation fails, reset the initial pose
Verification:
  • Particle cloud should be tight and small
  • Laser scan should align with map walls
  • Robot’s position should match Gazebo

​
Choosing good goals

Select destinations that:
  • Are in open, free space (white areas on map)
  • Have clear paths from current location
  • Avoid narrow passages for first attempts
  • Are reachable without complex maneuvering
Avoid goals that:
  • Are in or near obstacles (black areas)
  • Require passing through very narrow spaces
  • Are in unknown areas (gray on map)

​
Dealing with navigation failures

If the robot stops moving:
1

Check the goal

Ensure the goal is in free space, not inside an obstacle or too close to walls.
2

Reset initial pose

The robot may have lost localization. Set a new initial pose estimate.
3

Choose closer goal

Try a shorter navigation distance first, then navigate to the original goal in steps.
4

Check for obstacles

Verify nothing is blocking the path in Gazebo.

​
Understanding Nav2 behavior

​
Path planning

Nav2 uses two planners: Global planner:
  • Plans overall route from start to goal
  • Uses the static map
  • Shown as green line in RViz2
  • Computed once per goal
Local planner:
  • Adjusts path in real-time
  • Avoids dynamic obstacles
  • Follows global path while avoiding collisions
  • Continuously updated

​
Costmaps

Nav2 uses costmaps to represent obstacle information: Global costmap:
  • Based on your saved map
  • Covers entire environment
  • Static obstacles
Local costmap:
  • Small area around robot
  • Updated with live sensor data
  • Detects dynamic obstacles
  • Shown as colored grid in RViz2

​
Recovery behaviors

If navigation gets stuck, Nav2 automatically tries:
  1. Rotation: Spin in place to find a path
  2. Backing up: Move backward to escape tight spots
  3. Clear costmap: Reset obstacle information and retry

​
Troubleshooting

​
Robot won’t move

1

Check initial pose

Set a new initial pose estimate. Incorrect pose prevents navigation.
2

Verify goal location

Ensure goal is in free space (white area), not in obstacles (black).
3

Check simulation

Verify Gazebo simulation is running and robot is loaded:
ros2 node list | grep turtlebot3
4

Restart navigation

Stop Nav2 (Ctrl+C) and relaunch:
ros2 launch turtlebot3_navigation2 navigation2.launch.py use_sim_time:=True map:=$HOME/maps/my_map.yaml

​
Map doesn’t match environment

Cause: Using a map from a different world. Solution:
  • Always use the map created for the current simulation world
  • If you’re in tb3_world, use the map created when you explored tb3_world
  • Re-create the map if you’ve modified the environment

​
Robot gets stuck

1

Cancel current goal

In RViz2, set a new goal in an easier location.
2

Use teleoperation

Temporarily control the robot manually:
tb3_teleop
Move to open space, then restart navigation.
3

Adjust goal

Set a goal in more open space, away from obstacles.

​
Poor navigation performance

Symptoms:
  • Robot takes inefficient paths
  • Excessive oscillation
  • Frequent recovery behaviors
Solutions:
  • Create a higher quality map with SLAM (drive slower during mapping)
  • Ensure initial pose is very accurate
  • Choose goals in open areas
  • Adjust Nav2 parameters for your environment (advanced)

​
RViz2 not showing map

1

Verify map file exists

ls -lh ~/maps/my_map.*
Both .pgm and .yaml files must be present.
2

Check map path

Ensure the path in your launch command is correct:
ros2 launch turtlebot3_navigation2 navigation2.launch.py use_sim_time:=True map:=$HOME/maps/my_map.yaml
3

Verify map topic

ros2 topic list | grep map
ros2 topic echo /map --once

​
Advanced navigation

​
Multi-goal navigation

Navigate to multiple points sequentially:
  1. Set first goal and wait for completion
  2. Set second goal and wait
  3. Continue for each destination

​
Waypoint following

Nav2 supports waypoint following through the Nav2 Waypoint Follower for complex routes.

​
Parameter tuning

Advanced users can tune Nav2 parameters:
  • Planner settings
  • Controller gains
  • Costmap configurations
  • Recovery behavior thresholds
Parameters are configured in Navigation2 launch files and configuration YAML files.

Build docs developers (and LLMs) love

Get started for freeTalk to us