UR5 Pick and Place Simulation in Ros/Gazebo

Table of contents

• Description
• Requirement
• Folder
• Setup
• Usage
• Contributors

Description

This repository demonstrates UR5 pick-and-place in ROS and Gazebo. The UR5 uses a Xbox Kinect cam to detect eleven types of Lego Bricks, and publish its position and angolation.

The goals of this project are:

• simulate the iteration of a UR5 robot with Lego bricks
• The robotic arm must be able to move a block from position A to B and construct a castle by assembling different bricks

Folder

UR5-Pick-and-Place-Simulation/catkin_ws/
├── levelManager
├── vision
├── motion_planning
├── gazebo_ros_link_attacher
├── robot

• levelManager: the task of this package is to launch the world and spawn the different bricks
• vision: the task of this package is to recognize the object type and orientation of the bricks
• motion_planning: the task is to move the robot and pick and place the lego
• gazebo_ros_link_attacher: A gazebo plugin definable from URDF to inform a client of a collision with an object
• robot: the task is to define the robot model with appropriate PID settings

Requirements

For running each sample code:

• Ros Noetic: http://wiki.ros.org/noetic/Installation
• Gazebo: https://classic.gazebosim.org/tutorials?tut=ros_installing&cat=connect_ros
• Yolov5 https://github.com/ultralytics/yolov5
• Catkin https://catkin-tools.readthedocs.io/en/latest/

Setup

After installing the libraries needed to run the project. Clone this repo:

git clone https://github.com/pietrolechthaler/UR5-Pick-and-Place-Simulation/

Setup the project:

cd UR5-Pick-and-Place-Simulation/catkin_ws
source /opt/ros/noetic/setup.bash
catkin build
source devel/setup.bash
echo "source $PWD/devel/setup.bash" >> $HOME/.bashrc

Clone and install YoloV5:

cd ~
git clone https://github.com/ultralytics/yolov5
cd yolov5
pip3 install -r requirements.txt

Usage

Launch the world

roslaunch levelManager lego_world.launch

Choose the level (from 1 to 4):

rosrun levelManager levelManager.py -l [level]

Start the kinematics process

rosrun motion_planning motion_planning.py

Start the localization process

rosrun vision vision.py -show

• -show : show the results of the recognition and localization process with an image

Contributors

Name	Github
Davice Cerpelloni	https://github.com/davidecerpelloni
Leonardo Collizzolli	https://github.com/leocolliz
Pietro Lechthaler	https://github.com/pietrolechthaler
Stefano Rizzi	https://github.com/StefanoRizzi