SLAM & Autonomous Navigation

Real-time mapping and autonomous navigation for Mini Pupper using ROS 2 Nav2 stack with SLAM Toolbox and SmacPlanner2D.

Overview

Mini Pupper supports two key navigation capabilities:

• SLAM (Mapping): Create maps of your environment using the mini_pupper_slam package
• Autonomous Navigation: This package provides path planning and obstacle avoidance using maps created by SLAM

This navigation package depends on mini_pupper_slam to generate maps for real robot operation.

Architecture

This implementation integrates several ROS 2 navigation components:

• SLAM Toolbox: Real-time graph-based SLAM with loop closure detection
• AMCL: Adaptive Monte Carlo Localization for robot pose estimation
• SmacPlanner2D: Hybrid A* global path planner with obstacle-aware search
• DWB Controller: Dynamic Window Based local trajectory planning and execution
• Nav2 Behavior Trees: Coordinated navigation behaviors and recovery actions

Requirements

• Mini Pupper 2 with lidar sensor and IMU enabled (check mini_pupper_bringup/config/mini_pupper_2.yaml)
• PC for running SLAM and navigation nodes
• Network connection between robot and PC

Quick Start

1. SLAM (Mapping)

Create a map of your environment by driving the robot around while SLAM builds the map in real-time.

Start the robot:

# Terminal 1 (SSH to robot)
source ~/ros2_ws/install/setup.bash
ros2 launch mini_pupper_bringup bringup.launch.py

Launch SLAM on PC:

# Terminal 2 (PC)
source ~/ros2_ws/install/setup.bash
ros2 launch mini_pupper_slam slam_toolbox.launch.py

Control the robot to build the map:

# Terminal 3 (PC) - Keyboard control
source ~/ros2_ws/install/setup.bash
ros2 run teleop_twist_keyboard teleop_twist_keyboard

Save the completed map:

# Terminal 4 (PC)
source ~/ros2_ws/install/setup.bash
ros2 run nav2_map_server map_saver_cli -f ~/map

This creates two files: map.pgm and map.yaml in your home directory.

2. Navigation

Use your saved map for autonomous navigation with obstacle avoidance.

Start the robot:

# Terminal 1 (SSH to robot)
source ~/ros2_ws/install/setup.bash
ros2 launch mini_pupper_bringup bringup.launch.py

Launch navigation with your map:

# Terminal 2 (PC)
source ~/ros2_ws/install/setup.bash
ros2 launch mini_pupper_navigation navigation_smacplanner.launch.py map:=$HOME/map.yaml

Set goals in RViz:

1. Robot will automatically localize if started from the same position as SLAM mapping
2. Use “Nav2 Goal” to set navigation targets
3. Watch the robot autonomously navigate to the goal

Tips for Best Results

For SLAM:

• Drive the robot slowly and smoothly
• Ensure good lidar visibility of walls and obstacles
• Cover the entire area you want to map
• Avoid rapid rotations that can confuse the mapping algorithm

For Navigation:

• Start the robot in the exact same location where mapping began

File truncated at 100 lines see the full file