Baxter Pick and Place

This repository contains the implementation of the control module from the paper:

Antipodal Robotic Grasping using Generative Residual Convolutional Neural Network

Sulabh Kumra, Shirin Joshi, Ferat Sahin

arxiv | video

If you use this project in your research or wish to refer to the baseline results published in the paper, please use the following BibTeX entry:

@article{kumra2019antipodal,
  title={Antipodal Robotic Grasping using Generative Residual Convolutional Neural Network},
  author={Kumra, Sulabh and Joshi, Shirin and Sahin, Ferat},
  journal={arXiv preprint arXiv:1909.04810},
  year={2019}
}

Prerequisites

• A developer workstation meeting the minimum system requirements is available for use.
• The developer workstation has been set up successfully with Ubuntu and ROS - instructions here.
• You have configured your Development Workstation to access Github. Instructions here.
• A grasp generator. Our implementation of the GR-ConvNet based grasp generator is available at: https://github.com/skumra/robotic-grasping

Installation

• Download and Install the SDK dependencies

$ sudo apt-get install python-wstool python-rosdep

• Create a new catkin workspace.

$ mkdir -p ~/ros_ws/src
$ cd ~/ros_ws/src

• Download and copy the rosinstall file to the root of your Catkin workspace using wstool.

$ wstool init .
$ wstool merge https://raw.github.com/RethinkRobotics/baxter/master/baxter_sdk.rosinstall
$ wstool update
$ cd ..

This pulls down the development branches from all of our repositories into your source directory.

For Baxter Gazebo simulation, checkout the baxter_simulator repository:

$ cd ~/ros_ws/src
$ git clone https://github.com/RethinkRobotics/baxter_simulator.git

• Check out the Baxter PNP package into your ROS workspace

$ cd ~/ros_ws/src
$ git clone https://github.com/skumra/baxter-pnp.git

• Source ros setup if you haven't yet

$ source /opt/ros/kinetic/setup.bash

• Build and Install

$ cd ~/ros_ws
$ catkin_make
$ catkin_make install

• Use the baxter.sh script for proper environment setup

$ cp src/baxter/baxter.sh .

Run PNP in Simulation

• Run the baxter.sh script with sim specified

$ cd ~/ros_ws
$ ./baxter.sh sim

• Start simulation with controllers

$ roslaunch baxter_gazebo baxter_world.launch

• Start pick and place

$ cd ~/ros_ws
$ source devel/setup.bash
$ rosrun baxter-pnp run_pnp.py

Run PNP on Robot

• Run the baxter.sh

$ cd ~/ros_ws
$ ./baxter.sh

• Start pick and place

$ cd ~/ros_ws
$ source devel/setup.bash
$ rosrun baxter-pnp run_pnp.py