Skip to content

Latest commit

 

History

History
175 lines (141 loc) · 6.33 KB

setup.md

File metadata and controls

175 lines (141 loc) · 6.33 KB
Raw

Cameras

Azure Kinect (Kinect 4)

RealSense SR300/SR305

First follow installation instructions for librealsense2 here

pip install pyrealsense2
python robot_io/cams/realsense/realsenseSR300_librs2.py  # to test

Framos D435e

groups | grep video  # user must be in video group, otherwise ask Michael K.
file /usr/src/librealsense2  # (see below a)
diff misc/framos_setup_files/setup.py /usr/src/librealsense2/setup.py
cp -r /usr/src/librealsense2 .
cd librealsense2
pip uninstall pyrealsense2
pip install -e .

cd ../robot_io/cams/realsense
python realsense.py  # test script
  • a) If /usr/src/librealsense2 does not exist, download FRAMOS software package from https://www.framos.com/en/industrial-depth-cameras#downloads. Follow installation instructions, make sure to use local admin user (e.g. xam2) to install (file system may NOT be network mounted).
  • b) Use Ethernet sockets on the ceiling for PoE.

Robots

KUKA iiwa

Clone Kuka Java Repository on aisgit git clone https://aisgit.informatik.uni-freiburg.de/hermannl/kuka_java_interface_no_ros

robot_io/robot_interface/iiwa_interface.py is a ROS free python controller for KUKA iiwa. It sends UDP messages to the iiwa_java_controller.

Supported control modes:

  • Joint position control
    • degrees / radians
  • Cartesian position control
    • PTP / LIN motions
    • with / without impedance
    • absolute / relative coordinates

For more information please read README of kuka_java_interface_no_ros

Franka Emika Panda

IK fast

IK fast is an analytic IK solver. In order to use IK fast, first install ikfast-pybind:

git clone --recursive https://github.com/yijiangh/ikfast_pybind
cd ikfast_pybind
# copy panda IK solution .cpp and .h to ikfast_pybind
cp ../robot_io/misc/ik_fast_files/ikfast.h ./src/franka_panda/
cp ../robot_io/misc/ik_fast_files/ikfast0x10000049.Transform6D.0_1_2_3_4_5_f6.cpp ./src/franka_panda/
pip install .

For creating different IK solutions (e.g. in case of a different gripper) please refer to: http://docs.ros.org/en/kinetic/api/framefab_irb6600_support/html/doc/ikfast_tutorial.html

Frankx

git clone [email protected]:lukashermann/frankx
cd frankx
git clone [email protected]:pantor/affx
git clone [email protected]:pantor/ruckig
cd affx; git checkout -b frankx_version dabe0ba; cd ..
cd ruckig; git checkout -b frankx_version 31f50f0; cd ..

conda install pybind11
vim setupy.py  # add "-DFranka_DIR=/opt/ros/noetic/share/franka/cmake/"
pip install -e .
firefox https://192.168.180.87/desk/  # unlock joints
export LD_LIBRARY_PATH=/opt/ros/noetic/lib/:$LD_LIBRARY_PATH

cd robot_io/robot_interface
python panda_frankx_interface.py  # test robot

Input Devices

SpaceMouse

sudo apt install libspnav-dev spacenavd
conda activate robot
pip install spnav

Next test if it works, some common pitfalls are:

  1. Turn on SpaceMouse in the back
  2. May not work while charging.
  3. Wireless range is quite limited.
  4. Comment the following two lines in site-packages/spnav/__init__.py
#pythonapi.PyCObject_AsVoidPtr.restype = c_void_p
#pythonapi.PyCObject_AsVoidPtr.argtypes = [py_object]

To test execute the following program. When moving the mouse you should see numbers scrolling by.

python robot_io/input_devices/space_mouse.py

VR Teleoperation

Install Steam and SteamVR

  • In terminal run $ steam, it will start downloading an update and create a .steam folder in your home directory.
  • If you get an error, try deleting the steam folders on your home directory with rm -rf .local/share/Steam/ and rm -rf .steam
  • In Steam, create user account or use existing account.
  • Install SteamVR
    • If on pickup click Steam -> Settings -> Downloads -> Steam Library Folders -> Add Library Folder -> /media/hdd/SteamLibrary to add the existing installation of SteamVR to your Steam account
    • Otherwise download SteamVR
  • Restart Steam
  • Connect and turn on HTC VIVE
  • Launch Library -> SteamVR (if not shown, check [] Tools box)
  • If SteamVR throws an Error: setcap of vrcompositor-launcher failed, run $ sudo setcap CAP_SYS_NICE+ep /media/hdd/SteamLibrary/steamapps/common/SteamVR/bin/linux64/vrcompositor-launcher
  • Make sure Headset and controller are correctly detected
  • Go through VR setup procedure (standing is sufficient)

Install Bullet

$ git clone https://github.com/bulletphysics/bullet3.git
$ cd bullet3

# Optional: patch bullet for selecting correct rendering device
# (only relevant when using EGL and multi-gpu training)
$ wget https://raw.githubusercontent.com/BlGene/bullet3/egl_remove_works/examples/OpenGLWindow/EGLOpenGLWindow.cpp -O examples/OpenGLWindow/EGLOpenGLWindow 

# For building Bullet for VR  add -DUSE_OPENVR=ON to line 8 of build_cmake_pybullet_double.sh
# Run
$ ./build_cmake_pybullet_double.sh

$ pip install numpy  # important to have numpy installed before installing bullet
$ pip install -e .  # effectively this is building bullet a second time, but importing is easier when installing with pip

# add alias to your bashrc
alias bullet_vr="~/.steam/steam/ubuntu12_32/steam-runtime/run.sh </PATH/TO/BULLET/>bullet3/build_cmake/examples/SharedMemory/App_PhysicsServer_SharedMemory_VR"

# to test VR control
# make sure SteamVR is started
$ bullet_vr
$ cd <PATH/TO/ROBOTIO>/robot_io/robot_io/control
$ python teleop_robot.py

Robot Teleop instructions:

  1. Push dead-man switch (riffled grip right)
  2. Move controller in direction robot is pointing (twoards window)
  3. Push top middle button (with three lines)
  4. Robot should reset to home position
  5. Robot only moves with dead-man-switch activated

Marker Detector

Make sure OpenCV and Eigen are installed:

$ sudo apt install libopencv-dev python3-opencv
$ sudo apt install libeigen3-dev

$ cd robot_io/marker_detection/apriltag_detection
$ python setupBatch.py build_ext --inplace