Intelligent Bin Picking with Universal Robots UR5e Cobot for a semi-random object distribution using Simulink®
An Intelligent Bin Picking application consists of a perception algorithm to detect objects and a motion planning algorithm that creates the trajectory path of the robot to pick up the object. First the positions and orientations of the objects lying in the bin are identified from the Camera images using Deep Learning based Image processing capability from Computer Vision Toolbox. The calculated pose is then passed to motion-planning algorithm as an input which plans the path and the trajectory is generated for the Cobot to pick the objects from the bin and place it. Using the Robotics System Toolbox™ Support Package for Universal Robots UR Series Manipulators, you can establish ROS communication with Universal Robots manipulators and send the generated trajectory for the Cobot to pick up and place the objects at a destination location.This example is a Simulink based solution for the featured examples Semi-Structured Intelligent Bin Picking for UR5e Using YOLO and PCA-Based Object Detection and Gazebo Simulation of Semi-Structured Intelligent Bin Picking for UR5e Using YOLO and PCA-Based Object Detection.
MathWorks Products (https://www.mathworks.com)
Requires MATLAB® release R2023a or higher
- MATLAB®
- Simulink®
- Stateflow®
- Robotics System Toolbox™
- ROS Toolbox
- Computer Vision Toolbox™
- Image Processing Toolbox™
- Deep Learning Toolbox™
- Optimization Toolbox™
- Statistics and Machine Learning Toolbox™
- Robotics System Toolbox Support Package for Universal Robots UR Series Manipulators
Optional Products
Installation instructions
- MATLAB installation: Visit installation instructions webpage to get started with the MATLAB installation process.
- Support package installation: Before proceeding, ensure that the products mentioned under MathWorks Products above are installed. To install the Robotics System Toolbox Support Package for Universal Robots UR Series Manipulators, follow the steps mentioned here.
- Download ROS Virtual Machine For Pre-Installed Gazebo Environment: You can download a virtual machine image that already has ROS and Gazebo installed. This virtual machine is based on Ubuntu® Linux® and is pre-configured to support the examples in ROS Toolbox™. Download and install the ROS virtual machine from this link.
- YOLO Object Detector MAT files and Dataset: The example uses a pre-trained YOLOV4 object detector for identifying the Cuboid objects. Download the detector MAT file from the following links.
Open the the project to get started with bin picking example.The MLX file associated has the necessary background and steps to complete the example..
open('IntelligentBinPickingExampleWithSimulink.prj')
To learn how to communicate with UR Series cobots over ROS, see Getting Started with Connecting and Controlling a UR5e Cobot from Universal Robots.
The license is available in the License file within this repository.
You can post your queries on the MATLAB Central page for the support package. You can also add your questions at MATLAB Answers.
Copyright 2023 The MathWorks, Inc.