🤖 CargoBot

Autonomous Ground Cargo Robot
Graduation Project — 2026

📋 Table of Contents

Overview
Key Features
Hardware
System Architecture
Software Stack
ROS Packages
Getting Started
- Prerequisites
- Installation
Usage
TF Tree
Project Structure
Configuration Reference
License

Overview

CargoBot is an autonomous ground cargo delivery robot built as a graduation project. It is based on a modified Waveshare JetBot 2GB chassis, upgraded with a 2D LiDAR and an IMU for real-time sensor fusion. The robot uses Google Cartographer for SLAM (Simultaneous Localization and Mapping) and the ROS Navigation Stack for fully autonomous point-to-point navigation.

A custom 3D-printed cargo trailer is attached to the rear of the robot, enabling it to carry and deliver small payloads autonomously within mapped indoor environments.

Key Features

Feature	Details
Autonomous Navigation	Goal-based path planning with obstacle avoidance via `move_base` + DWA local planner
SLAM	Real-time 2D map building with Google Cartographer (LiDAR + IMU fusion)
Localization	AMCL particle-filter localization on pre-built maps with `laser_scan_matcher` odometry
Sensor Fusion	RPLiDAR A1 laser scans fused with MPU-6050 IMU data (Madgwick filter)
Differential Drive	Adafruit Motor HAT controlled via `cmd_vel` with dead-zone and clamping
Cargo Trailer	3D-printed detachable trailer for payload delivery
Edge Computing	All processing runs on-board the NVIDIA Jetson Nano 2GB

Hardware

Component	Specification
Compute	NVIDIA Jetson Nano 2GB Developer Kit
Chassis	Waveshare JetBot 2GB AI Kit (modified)
LiDAR	RPLiDAR A1 — 360° 2D laser scanner (12 m range)
IMU	MPU-6050 6-axis (3-axis gyro + 3-axis accelerometer)
Motor Driver	Adafruit DC Motor HAT (I²C, PCA9685-based)
Motors	2× DC geared motors (differential drive)
Camera	CSI camera module (on-board, for future vision tasks)
Trailer	Custom 3D-printed cargo tray with passive caster wheels
Power	Portable battery pack

System Architecture

Software Stack

Layer	Technology
OS	Ubuntu 18.04 (JetPack 4.x)
Middleware	ROS Melodic
SLAM	Google Cartographer (`cartographer_ros`)
Localization	AMCL (Adaptive Monte Carlo Localization)
Odometry	`laser_scan_matcher` + custom `pose_to_odom` bridge
IMU Filtering	`imu_filter_madgwick` (orientation estimation)
Path Planning	`move_base` — NavfnROS (global) + DWA (local)
Motor Control	Adafruit MotorHAT Python driver (I²C / PCA9685)
Robot Model	URDF (base_link, wheels, camera, laser, imu_link)

ROS Packages

catkin_ws/src/
├── cartographer_config   # Cartographer SLAM configuration & launch
├── imu_bringup           # MPU-6050 driver + Madgwick filter launch
├── imu_tf                # IMU → base_link TF broadcaster
├── jetbot_bringup        # Top-level bringup (URDF + motors)
├── jetbot_description    # URDF model & robot_state_publisher launch
├── jetbot_navigation     # move_base, AMCL, laser_scan_matcher, configs
├── jetbot_ros            # Motor driver, OLED display, teleop scripts
├── mpu_6050_driver       # Raw MPU-6050 I²C reader → /imu/data
├── ros_deep_learning     # NVIDIA deep learning inference nodes (future)
└── rplidar_ros           # RPLiDAR A1 ROS driver → /scan

Package Details

Package	Purpose	Key Files
`cartographer_config`	Google Cartographer 2D SLAM tuning	`config/jetbot_2d.lua`, `launch/cartographer.launch`
`imu_bringup`	Launches MPU-6050 driver + Madgwick filter pipeline	`launch/imu.launch`
`imu_tf`	Broadcasts `base_link → imu_link` transform from filtered IMU	`scripts/imu_tf_broadcaster.py`
`jetbot_bringup`	One-shot bringup of URDF + motors	`launch/bringup_essentials.launch`
`jetbot_description`	URDF model definition (links, joints, sensors)	`urdf/jetbot.urdf`
`jetbot_navigation`	Full navigation stack (AMCL, move_base, DWA, costmaps)	`launch/`, `config/`, `scripts/`
`jetbot_ros`	Low-level motor control + teleop	`scripts/jetbot_motors.py`, `scripts/teleop_key.py`
`mpu_6050_driver`	Raw I²C communication with MPU-6050 sensor	`scripts/imu_node.py`
`rplidar_ros`	RPLiDAR A1 ROS wrapper (publishes `/scan`)	`launch/`, `src/`

Getting Started

Prerequisites

NVIDIA Jetson Nano 2GB with JetPack 4.x (Ubuntu 18.04)
ROS Melodic fully installed (installation guide)
Python 2.7 (ROS Melodic default)
Hardware connected:
- RPLiDAR A1 via USB (/dev/ttyUSB0)
- MPU-6050 via I²C (bus 1)
- Adafruit Motor HAT via I²C (bus 1)

Installation

# 1. Clone this repository
git clone https://github.com/ahmetsalihkaya/GradProject-CargoBot.git
cd CargoBot/workspace/catkin_ws

# 2. Install ROS dependencies
sudo apt-get update
sudo apt-get install -y \
    ros-melodic-cartographer-ros \
    ros-melodic-move-base \
    ros-melodic-amcl \
    ros-melodic-map-server \
    ros-melodic-dwa-local-planner \
    ros-melodic-laser-scan-matcher \
    ros-melodic-imu-filter-madgwick \
    ros-melodic-robot-state-publisher \
    ros-melodic-tf \
    ros-melodic-tf2-ros

# 3. Install Python dependencies
pip install Adafruit-MotorHAT smbus2

# 4. Build the workspace
catkin_make

# 5. Source the workspace
echo "source $(pwd)/devel/setup.bash" >> ~/.bashrc
source devel/setup.bash

# 6. Set USB permissions for RPLiDAR
sudo chmod 666 /dev/ttyUSB0

Usage

1. Bringup

Start the core robot systems — URDF, motors, LiDAR, and IMU:

# Terminal 1: Robot base (URDF + motors)
roslaunch jetbot_bringup bringup_essentials.launch

# Terminal 2: LiDAR
roslaunch rplidar_ros rplidar.launch

# Terminal 3: IMU pipeline (MPU-6050 + Madgwick filter)
roslaunch imu_bringup imu.launch

2. Mapping (SLAM)

Build an occupancy grid map of your environment using Google Cartographer:

# Terminal 4: Start Cartographer SLAM
roslaunch cartographer_config cartographer.launch

# Terminal 5: Visualize in RViz
roslaunch jetbot_description rviz.launch

# Drive the robot around using teleop
rosrun jetbot_ros teleop_key.py

Once the map is complete, save it:

# Save the map for navigation
rosrun map_server map_saver -f $(rospack find jetbot_navigation)/maps/my_map

3. Autonomous Navigation

Load a pre-built map and navigate autonomously:

# Terminal 4: Odometry (laser scan matcher + pose-to-odom bridge)
roslaunch jetbot_navigation laser_scan_matcher.launch

# Terminal 5: Localization (AMCL + map server)
roslaunch jetbot_navigation amcl.launch

# Terminal 6: Path planning & obstacle avoidance
roslaunch jetbot_navigation move_base.launch

4. Send a Goal

Send a navigation goal from the command line:

# Usage: goal_publisher.py <x> <y> <yaw_radians>
rosrun jetbot_navigation goal_publisher.py 1.0 2.5 0.0

Or use RViz → click "2D Nav Goal" to set goals interactively on the map.

TF Tree

map
 └── odom                    (published by AMCL: map → odom)
      └── base_link          (published by pose_to_odom: odom → base_link)
           ├── left_wheel    (robot_state_publisher, from URDF)
           ├── right_wheel   (robot_state_publisher, from URDF)
           ├── camera_link   (robot_state_publisher, from URDF)
           ├── laser         (robot_state_publisher, from URDF)
           └── imu_link      (robot_state_publisher, from URDF)

Project Structure

CargoBot/
├── README.md
├── docs/
│   ├── cargobot_side.jpeg       # Robot photo — side view
│   └── cargobot_rear.jpeg       # Robot photo — rear view with trailer
└── workspace/
    └── catkin_ws/
        ├── .catkin_workspace
        ├── build/
        ├── devel/
        └── src/
            ├── cartographer_config/
            │   ├── config/jetbot_2d.lua
            │   └── launch/cartographer.launch
            ├── imu_bringup/
            │   └── launch/imu.launch
            ├── imu_tf/
            │   └── scripts/imu_tf_broadcaster.py
            ├── jetbot_bringup/
            │   └── launch/bringup_essentials.launch
            ├── jetbot_description/
            │   ├── launch/
            │   │   ├── jetbot_rsp.launch
            │   │   └── rviz.launch
            │   └── urdf/jetbot.urdf
            ├── jetbot_navigation/
            │   ├── config/
            │   │   ├── amcl.yaml
            │   │   ├── laser_scan_matcher.yaml
            │   │   └── move_base/
            │   │       ├── costmap_common_params.yaml
            │   │       ├── dwa_local_planner_params.yaml
            │   │       ├── global_costmap_params.yaml
            │   │       └── local_costmap_params.yaml
            │   ├── launch/
            │   │   ├── amcl.launch
            │   │   ├── laser_scan_matcher.launch
            │   │   └── move_base.launch
            │   ├── maps/
            │   │   ├── my_map.pbstream
            │   │   ├── my_map.pgm
            │   │   └── my_map.yaml
            │   └── scripts/
            │       ├── goal_publisher.py
            │       └── pose_to_odom.py
            ├── jetbot_ros/
            │   └── scripts/
            │       ├── jetbot_motors.py
            │       ├── jetbot_oled.py
            │       ├── teleop_joy.py
            │       └── teleop_key.py
            ├── mpu_6050_driver/
            │   └── scripts/imu_node.py
            ├── ros_deep_learning/
            └── rplidar_ros/

Configuration Reference

Cartographer SLAM (`jetbot_2d.lua`)

Parameter	Value	Notes
`tracking_frame`	`imu_link`	IMU-referenced tracking for better scan alignment
`use_imu_data`	`true`	Fuses IMU orientation with LiDAR scans
`min_range` / `max_range`	0.1 m / 8 m	RPLiDAR A1 effective range
`optimize_every_n_nodes`	`0`	Disabled in localization mode (set to 35 for mapping)

DWA Local Planner

Parameter	Value	Notes
`max_vel_x`	0.4 m/s	Maximum forward speed
`max_vel_theta`	2.0 rad/s	Maximum rotation speed
`xy_goal_tolerance`	0.10 m	10 cm goal reach threshold
`yaw_goal_tolerance`	6.28 rad	Full-circle tolerance (heading not constrained)

AMCL Localization

Parameter	Value	Notes
`min_particles` / `max_particles`	250 / 1000	Particle filter bounds
`laser_model_type`	`likelihood_field`	Probabilistic laser model
`odom_model_type`	`diff-corrected`	Differential drive odometry model
`update_min_d`	0.01 m	Re-localize after 1 cm movement

Robot Physical Parameters

Parameter	Value
Wheel separation	0.125 m
Wheel radius	0.03 m
Robot radius (costmap)	0.10 m
Inflation radius	0.20 m
Max PWM	115

License

This project is open-source and available under the MIT License.

Built with ❤️ as a graduation project — 2026

Name		Name	Last commit message	Last commit date
assets		assets
workspace/catkin_ws/src		workspace/catkin_ws/src
README.md		README.md

Navigation Menu

Repository files navigation

🤖 CargoBot

📋 Table of Contents

Overview

Key Features

Hardware

System Architecture

Software Stack

ROS Packages

Package Details

Getting Started

Prerequisites

Installation

Usage

1. Bringup

2. Mapping (SLAM)

3. Autonomous Navigation

4. Send a Goal

TF Tree

Project Structure

Configuration Reference

Cartographer SLAM (`jetbot_2d.lua`)

DWA Local Planner

AMCL Localization

Robot Physical Parameters

License

About

Releases

Packages

Contributors

Languages

Latest commit

History

Folders and files

Repository files navigation

🤖 CargoBot

📋 Table of Contents

Overview

Key Features

Hardware

System Architecture

Software Stack

ROS Packages

Package Details

Getting Started

Prerequisites

Installation

Usage

1. Bringup

2. Mapping (SLAM)

3. Autonomous Navigation

4. Send a Goal

TF Tree

Project Structure

Configuration Reference

Cartographer SLAM (jetbot_2d.lua)

DWA Local Planner

AMCL Localization

Robot Physical Parameters

License

About

Resources

Stars

Watchers

Forks

Releases

Packages

Contributors

Languages

Cartographer SLAM (`jetbot_2d.lua`)