SLAM Navigation Stack — UCR Robotics Project

TL;DR: Full SLAM + navigation stack on TurtleBot3 in Gazebo & real-world testbed. Includes mapping, localization, global/local planners, and Nav2 integration.

Highlights

• Mapping: Gmapping + Cartographer; occupancy grid maps
• Localization: AMCL (particle filter) with adaptive resampling
• Planning: Global (A*) + Local (DWB, TEB)
• Control: Tuned velocity constraints for safe indoor navigation
• Results (UCR lab): ~25% map accuracy improvement (Cartographer vs. Gmapping), stable localization over 10m trajectories

Repo Layout

• src/ — ROS2 packages/nodes
• launch/ — bringup & Nav2 launch files
• config/ — costmap, planner, controller params
• maps/ — saved maps (YAML + PGM)
• scripts/ — helper scripts (bag-to-map, teleop logging)
• media/ — GIF/MP4 demos
• docs/ — project report, diagrams

Quick Start

# Launch sim
ros2 launch slam_nav_tb3 slam_sim.launch.py

# Run Nav2
ros2 launch slam_nav_tb3 nav2.launch.py map:=maps/ucr_lab.yaml