me597

ME59700AS Final Project Repo

Project Archives: https://drive.google.com/drive/folders/12VroA-d8rddguSaayqEc3N36t1FRXnAG?usp=drive_link

Final Project: TurtleBot Autonomous Mapping, Navigation, and Object Localization

This repository presents the Fall 2025 ME59700AS final project: a ROS 2 TurtleBot3 Gazebo autonomy stack for autonomous mapping, map-based navigation, static-obstacle avoidance, local replanning, and colored-object search/localization.

The implementation is centered in final_project/sim_ws_Fall2025. It uses occupancy grids, TF, AMCL, LaserScan data, A* planning, RRT* local replanning, PID control, OpenCV-based color detection, and RViz visualization.

Visual Overview

The map preview above is generated from the saved final-project occupancy grid used by the localization and navigation tasks.

flowchart LR
  Sim["Gazebo TurtleBot3 Simulation"]
  Map["Occupancy Grid Map"]
  Scan["LaserScan"]
  Camera["Camera Frames"]
  Pose["TF / AMCL Pose"]
  Planner["A* Global Planner"]
  Replanner["RRT* Local Replanner"]
  Safety["Obstacle Safety Checks"]
  Perception["HSV Object Detection"]
  Localization["World-Frame Object Localization"]
  Control["PID / Path Following"]
  Outputs["RViz Markers, Paths, and Object Topics"]

  Sim --> Scan
  Sim --> Camera
  Sim --> Pose
  Map --> Planner
  Scan --> Safety
  Scan --> Localization
  Camera --> Perception
  Pose --> Planner
  Pose --> Localization
  Planner --> Replanner
  Safety --> Replanner
  Replanner --> Control
  Perception --> Localization
  Control --> Sim
  Localization --> Outputs
  Planner --> Outputs

Loading

flowchart TD
  T1["Task 1: Explore unknown world and build map"]
  MapOut["Saved occupancy map"]
  T2["Task 2: Navigate to goals with static obstacle handling"]
  Bonus["Task 2 Bonus: RRT* local detours"]
  T3["Task 3: Search, detect, and localize colored objects"]
  Demo["Project archive: presentation and demo videos"]

  T1 --> MapOut
  MapOut --> T2
  T2 --> Bonus
  MapOut --> T3
  Bonus --> T3
  T3 --> Demo

Loading

Project Tasks

Task 1: Autonomous Mapping

task1.py explores an unknown Gazebo environment and builds a map using frontier exploration.

Key ideas:

• Detects and clusters frontier cells from /map and /map_updates.
• Uses TF to locate the robot in the map frame.
• Selects safe exploration goals with obstacle clearance.
• Plans with grid-based A* and smooths the path where line-of-sight is safe.
• Follows paths with PID control and LaserScan-based safety checks.
• Stops cleanly when no useful frontiers remain.

Task 2: Navigation With Static Obstacles

task2.py navigates to RViz goals using the map from Task 1 while avoiding static and sensed obstacles.

Key ideas:

• Uses AMCL pose, map data, LaserScan data, and RViz goal poses.
• Inflates static and dynamic obstacles to preserve robot clearance.
• Plans global routes with A*.
• Detects blocked path segments during execution.
• Uses pure-pursuit-style path following with speed reduction near turns and goals.
• Falls back to reactive avoidance when obstacles are too close.

Task 2 Bonus: RRT* Local Replanning

task2_bonus.py extends the Task 2 navigation stack with local RRT* replanning.

When the active A* path becomes blocked, the node searches for a safe reconnect point, generates a local RRT* detour, publishes it as /local_plan, and splices it into the active route.

Task 3: Object Search and Localization

task3.py autonomously searches for red, green, and blue objects and estimates their world-frame locations.

Key ideas:

• Generates coverage waypoints from free-space structure.
• Navigates between waypoints using A* with RRT* detours.
• Performs 360-degree scans at search locations.
• Detects colored balls with HSV segmentation, shape filtering, and multi-frame stability checks.
• Combines camera bearing, LiDAR range, and AMCL pose to estimate object positions.
• Publishes RViz markers, PoseArray, and /red_pos, /green_pos, /blue_pos.

Important Project Files

Path	Purpose
final_project/sim_ws_Fall2025	Main ROS 2 final project workspace.
final_project/sim_ws_Fall2025/src/turtlebot3_gazebo	Gazebo worlds, TurtleBot3 models, launch files, maps, params, and task scripts.
final_project/sim_ws_Fall2025/src/turtlebot3_gazebo/src/lab4	Main task implementations.
final_project/sim_ws_Fall2025/src/turtlebot3_gazebo/launch	Launch files for simulation, mapping, localization, navigation, and object spawning.
final_project/sim_ws_Fall2025/src/turtlebot3_gazebo/maps	Saved maps used for localization and navigation.
final_project/sim_ws_Fall2025/src/turtlebot3_gazebo/params	AMCL, map server, and SLAM Toolbox configuration.
final_project/sim_ws_Fall2025/src/sim_utils	Supplemental simulation utilities.
final_project/helper	Helper maps, ROS topic notes, and project support files.

Setup

Build and source the final project workspace:

cd final_project/sim_ws_Fall2025
colcon build --symlink-install
source install/local_setup.bash
export TURTLEBOT3_MODEL=waffle

Useful supporting packages include:

sudo apt install ros-humble-turtlebot3-teleop
sudo apt install ros-humble-slam-toolbox
sudo apt install ros-humble-navigation2
pip install pynput

Useful launch entry points:

ros2 launch turtlebot3_gazebo turtlebot3_house_norviz.launch.py
ros2 launch turtlebot3_gazebo turtlebot3_house.launch.py
ros2 launch turtlebot3_gazebo task_6.launch.py

Project Artifacts

The archive link at the top of this README contains larger final-project artifacts such as presentation files, task demo videos, maps, and submitted deliverables that are better kept outside the Git repository.