Autonomous-Ball-Pick-up-Car

Introduction

This project involves designing an autonomous ball pick-up car equipped with mecanum wheels. The car performs active Laser SLAM to map the room and then autonomously navigates to collect ping-pong balls from every corner.

Structure

All structural files are stored in the structure directory. The naming conventions for the parts and assembly files are as follows:

• Individual parts: <name>-<manufacture_method>_<date>.SLDPRT
• Assembly file: picker_assembly_20241130.SLDASM

Laser SLAM

SLAM is based on RPLIDAR A1. When performing SLAM, ensure all related nodes are activated.

• Hector SLAM: Follow these steps to configure and run Laser SLAM:

cd ~/Autonomous-Ball-Pick-up-Car/workspace
catkin_make
source /opt/ros/melodic/setup.bash
source devel/setup.bash
sudo chmod 666 /dev/laser
roslaunch rplidar_ros rplidar_slam.launch

• Cartographer: For better performance, Cartographer is used. Three terminal windows are required:

# Terminal 1
cd ~/Autonomous-Ball-Pick-up-Car/workspace
catkin_make
source /opt/ros/melodic/setup.bash
source devel/setup.bash
sudo chmod 666 /dev/laser
roslaunch rplidar_ros rplidar.launch

# Terminal 2
cd ~/Autonomous-Ball-Pick-up-Car/cartographer_workspace
./carto_slam.sh

# Terminal 3
cd ~/Autonomous-Ball-Pick-up-Car/cartographer_workspace
./map_save.sh

# Once the map is successfully saved, quit all terminals.

Active SLAM

To improve mapping efficiency, a simple exploration algorithm is implemented. Follow these steps:

sudo chmod 666 /dev/arduino
roslaunch explore explore.launch
roslaunch map map_save.launch

Odometry

The package rf2o_laser_odometry is primarily used for odometry. Execute the following command to run it:

roslaunch rf2o_laser_odometry rf2o_laser_odometry.launch

Localization

AMCL Localization

• The AMCL algorithm is applied for robot localization. Use the following command:

roslaunch point_to_point_nav amcl_test.launch

Lidar-Based Localization

• Alternatively, a laser lidar-based localization method is provided. After mapping the target area, run the following command:

roslaunch point_to_point_nav point_to_point_nav.launch

Keyboard Control

The chassis is controlled by an ESP32-S3 board. To control movement using a keyboard, follow these steps:

1. Burn controlled_move.ino into the ESP32-S3 board.
2. Open the serial monitor and set the baud rate to 115200.
3. Use the following keys to control movement:
   • w: Move forward
   • a: Turn left
   • s: Move backward
   • d: Turn right
   • x: Stop

Navigation

A customized local planner is applied for the navigation stack. Follow these steps:

roslaunch point_to_point_nav point_to_point_nav.launch

# Use 2D pose estimate to calibrate the initial position.

# Use 2D Nav Goal to set the destination.