TAMS KUKA LWR

Kuka LWR	Perception

Introduction

• Our setup uses the FRI (fast-research-interface) for real-time communication with the arm controller at 100Hz+.
• The ROS node is an improved version of the FRILibrary code from Torsten Kröger and the Reflexxes-TypeII real-time motion generation, providing joint-position goals, velocity goals, trajectory execution, and both joint and Cartesian compliance modes. See the ros_fri package for details.
• Startup of the robot is a bit fiddly, make sure to select the correct tool and coordinate base system at startup, and see below for detailed instructions and tips doc here.
• Once the KUKA controller detects a motion anomaly it basically locks up. Restart the robot and the ROS nodes.

ROS workspace setup

• In your workspace

git clone https://github.com/lianghongzhuo/reflexxes_type2.git
git clone https://github.com/lianghongzhuo/frilibrary.git
git clone https://github.com/TAMS-Group/tams_apriltags.git
git clone git@github.com:TAMS-Group/tams_lwr.git

or use rosinstall file:

cd src
wget https://raw.githubusercontent.com/TAMS-Group/rosinstalls/master/noetic-tams-lwr.rosinstall
mv noetic-tams-lwr.rosinstall .rosinstall
wstool update

• Build your workspace with catkin build

Robot setup

• see doc here for a step to step instruction.

Run robot with ROS

• see doc here for a step to step instruction.

PDF documentations

• All PDF documentations are at: https://git.crossmodal-learning.org/TAMS/tams_lwr_doc