OpenRAVE-IKFAST-Generator

//////////////////////////////////////////////////////////////////////////
///  \author  Gu Dingyi
///
///  \file    README.md
///
///  \brief   OpenRAVE-IKFAST-Generator
///
///  \version 1.0
///
///  \date    Apr/03/2024
//////////////////////////////////////////////////////////////////////////

• Getting Started with OpenRAVE environment. Guidance could be followed at Stéphane Caron's blog.

• Error may be frequently encountered during compile. Alternatively, docker image is recommended.

  Docker setup

• Check personalrobotics/ros-openrave

  sudo docker pull personalrobotics/ros-openrave:latest

• Since the image is pre-built, no need to docker build with Dockerfile. Just run it. Mirror the /root/data to your current workspace.

  sudo docker run -it --rm -v /home/dingyi/workspace/dingyi-code/side-project/openrave-test:/root/data --network host ada6b60e98b7 bash

• Where image id ada6b60e98b7 could be obtained using sudo docker images.

  Closed-form solver generation

• Now go to /root/data in the docker image will direct to the same workspace.

• xmlfile like ur5.robot.xml represents the robot model. Use ikfastpy to generate the ikfast solver.

  python `openrave-config --python-dir`/openravepy/_openravepy_/ikfast.py --robot=ur5.robot.xml --iktype=transform6d --baselink=0 --eelink=6 --savefile=ikfast61.cpp --maxcasedepth 1

• It takes several minutes to generate the closed-form ikfast solver, which is a cpp file named by you. Then ikfast61.cpp is shown in your workspace.

• Now you can exit the docker image.

  Test the solver

• Compile the ikfast solver using gcc(need -llapack).

  gcc -fPIC -lstdc++ -DIKFAST_NO_MAIN -DIKFAST_CLIBRARY -shared -Wl,-soname,libik.so -o libik.so ikfast61.cpp

• libik.so comes out in your workspace.

• Compile the executable demo for testing purpose.

  g++ ikfastdemo.cpp -lstdc++ -llapack -o compute -lrt

• Run compute for easy tests for fk and ik.

./compute fk -3.0, -1.6, 1.6, -1.6, -1.6, 0.0
./compute ik 0.457016 0.172972 0.434512 -0.001461 0.755053 -0.655338 0.020595

• Verify your results with the real robot.

  Reference

• For more information and to explore other interesting projects, please refer to the following research papers:

  @phdthesis{diankov_thesis,
    author = "Rosen Diankov",
    title = "Automated Construction of Robotic Manipulation Programs",
    school = "Carnegie Mellon University, Robotics Institute",
    month = "August",
    year = "2010",
    number= "CMU-RI-TR-10-29",
    url={http://www.programmingvision.com/rosen_diankov_thesis.pdf},
  }

  @inproceedings{zeng2018learning,
  title={Learning synergies between pushing and grasping with self-supervised deep reinforcement learning},
  author={Zeng, Andy and Song, Shuran and Welker, Stefan and Lee, Johnny and Rodriguez, Alberto and Funkhouser, Thomas},
  booktitle={2018 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)},
  pages={4238--4245},
  year={2018},
  organization={IEEE}
  }