Ecosyste.ms: Awesome

An open API service indexing awesome lists of open source software.

Awesome Lists | Featured Topics | Projects

https://github.com/rwth-irt/lbrmed-ros

ROS1 support for the KUKA LBR Med: FRI-based hardware interface, MoveIt, kinematic simulation
https://github.com/rwth-irt/lbrmed-ros

control hardware iiwa kuka lbr med robot ros simulation

Last synced: 16 days ago
JSON representation

ROS1 support for the KUKA LBR Med: FRI-based hardware interface, MoveIt, kinematic simulation

Awesome Lists containing this project

README 

        
[![CI](https://github.com/rwth-irt/lbrmed-ros/actions/workflows/ci.yml/badge.svg)](https://github.com/rwth-irt/lbrmed-ros/actions/workflows/ci.yml)

# KUKA LBR MED STACK

ROS package for the Kuka LBR MED 14, also compatible with the iiwa 14.

We support

- commanding the robot hardware via the KUKA FRI interface and `ros_control`.

- kinematic simulations, Gazebo was never stable enough for contact rich tasks.

- MoveIt support for the hardware and simulation.



- A kinematic simulation environment and the robot hardware.

However, contact forces tend to be unstable.



Please look at the READMEs in the specific hardware packages.

**When using the real robot, always keep a safe distance and set up your safety configuration as required**.



## Why Another KUKA LBR Med / iiwa 14 ROS Package?

* The widely used [iiwa_stack](https://github.com/IFL-CAMP/iiwa_stack) from TUM did not allow us to **control the robot in real-time and with high precision**.

  * Our robot stack **allows implementing low-level controllers** to set joint positions via the KUKA FRI interface, see [lbrmed_control](./lbrmed_control) for an example.

  * A Laptop with a standard Ubuntu installation is sufficient.

  * No need to patch a kernel as we do not use kernel interrupts only the standard `ros_control` loop.

  * You might observe connectivity errors if you launch additional loads, e.g. RViz, or use a network switch.

* We provide a **development container**, so you can start without manually installing any dependencies.

* Otherwise, this package is simple to set up without manual steps using `wstool` / `vcs` and `rosdep`.

* We built collision and visualization meshes for the LBR Med with internal electric connectors.



# Install

## Devcontainer

This is the easiest way to get started

When opening this repository in Visual Studio Code, you should be prompted to reopen it in a container.

On launch, the container will setup all required dependencies via rosinstall and rosdep.



This repository also ships **VS Code tasks** for:

- Building the workspace `catkin build workspace`

- Building the package of the current file only `catkin build package`

- Clean the catkin build files `catkin clean`



After starting the container, the lbrmed-ros package will be opened as **workspace**.

The folder structure is `/workspace/src/lbrmed-ros`, so you will want to `cd` to a top-level folder when using low-level catkin commands



## Manual steps

### Create a catkin workspace

* Create a catkin workspace folder (e.g. ~/catkin_workspace)

* Create a src folder in that workspace (e.g. ~/catkin_workspace/src)

* Initialize the catkin workspace (e.g. `cd ~/catkin_workspace && catkin init`)



### Cloning the repository

We use [`git-lfs`](https://packagecloud.io/github/git-lfs/install) which should be installed before cloning.

Otherwise, you might get errors when the STL files of the robot or the jar files are being loaded.



For the Sunrise Project you have two ways to proceed:

* (a) Clone the repository into the src folder of the catkin workspace

* (b) (preferable) Clone the repository into a separate folder (e.g. ~/git-repos) and symlink it (e.g. `ln -s ~/git-repos/kuka_MED_Stack ~/catkin_workspace/src`, the target of the symlink must be an absolute path) 



### Installing Dependencies

In the catkin workspace, run the following commands to install all workspace and system dependencies.

```bash

rosdep install --from-paths src --ignore-src -r -y

``` 



If you want to use the [FZI Cartesian controllers](https://github.com/fzi-forschungszentrum-informatik/cartesian_controllers), you can find the installation steps in the `lbrmed_control` [README](./lbrmed_control/README.md).



### Building the workspace

To build the packages in parallel, we use the [catkin_tools](https://catkin-tools.readthedocs.io/en/latest/installing.html). After installing them, run:

```bash

cd ~/catkin_workspace

catkin build

```



# LBR MED Stack - Package Organization

*Reference: https://roboticsbackend.com/package-organization-for-a-ros-stack-best-practices/#Simulation_with_Gazebo*



* lbrmed: Metapackage (optional)

* [lbrmed_bringup](./lbrmed_bringup): Main entry point launch files for this robot application, including Parameters and Configuration files. Good starting point to setup your own robot.

* [lbrmed_control](./lbrmed_control): Custom controllers for the LBR Med, e.g., a kind of admittance controller and integration of the FZI cartesian controllers.

* [lbrmed_description](./lbrmed_description): All files that are necessary for visualization (URDFs, Meshes, test launch files, ...)

* [lbrmed_hw_fri](./lbrmed_hw_fri/) (recommended): Implements the `hardware_interface::RobotHW` which can be used by ROS control using the FRI real-time interface.

* [lbrmed_moveit_config](./lbrmed_moveit_config): Default MoveIt configuration for without an end-effector, created with the MoveIt wizard.

* [lbrmed_msgs](./lbrmed_msgs): All custom messages, services and actions for this robot

* [lbrmed_ros_java](./lbrmed_ros_java): Transfer this code to the robot via the Sunrise Workbench. Add your own station setup and safety configuration!