Skip to content

Latest commit

 

History

History
198 lines (145 loc) · 9.65 KB

usage_example.md

File metadata and controls

198 lines (145 loc) · 9.65 KB

Using the ur_robot_driver

This tutorial should help you getting started to use the ur_robot_driver in your application. It explains how to start the robot driver and how to control the robot using either MoveIt! or a simple python script.

Prerequisites

In order to use the driver you will need a working robot to test with. This can be either a real robot, connected with an ethernet cable to your ROS machine, or it can be a simulated robot using URSim.

In this tutorial we assume the following IP settings:

Machine IP address
Robot / VM 192.168.56.101
ROS machine 192.168.56.1

If your setup is different, modify the IP addresses in the following sections accordingly.

Furthermore, we will assume that you already built the driver successfully and sourced your workspace containing the built driver.

For this tutorial we will also skip using a calibrated robot description. See the main setup guide

on how to do that.

Real robot

If you have a real robot and want to use it for testing, make sure it is connected to the ROS machine on which you'd like to run the driver. We recommend a direct connection between the robot and the control PC so that other network participants will not disturb the network traffic. Make sure that the network is configured, such that both, the robot and the ROS machine are in the same subnet. See the robot's user manual how to setup a static IP address on the robot.

Using a simulated robot

For simulation we recommend running URSim in a virtual machine. You can directly download a virtual machine image containing a preconfigured URSim from Universal Robot's download page. Please note, that you will have to select "Non-Linux Offline simulator" in order to get the virtual machine.

The following will assume that you have VirtualBox installed on your machine. If not, you can install it by calling sudo apt install virtualbox in a terminal.

Once downloaded, extract the archive and open the file URSim_VIRTUAL-<version>.vbox. It should right open in VirtualBox.

Network setup

Next, setup a host network that will be used for the machine.

For this, click File > Host Network Manager. A dialog will open in which you can click the create button which will add an entry to the list. It should automatically get the static IP address 192.168.56.1. If not, set it up manually in the dialog. Once done, you can close the dialog again to go back to your VirtualBox main window. There, select the URSim_VIRTUAL... machine and click on Settings. Select the Network point in the list on the left and activate at least one network adapter, setting the Attached to dropdown menu to Host-only Adapter. If there are multiple host-only networks available, select the network you just created. Click on the ok button to close the dialog and go back to your main window.

As a side note: A host-only network is not the only option to get this running, but it is the easiest. If you consider using another networking mode, make sure that the virtual machine is reachable via its own IP address from the host machine and that it can reach the host machine as well.

Starting a robot

To start a simulated robot, select your URSim virtual machine and click on the Start button. The machine should boot and finally present a desktop with a shortcut for every robot model. Simply double click on the robot model that you want to use for testing. No further network setup for the simulated robot is necessary.

In this tutorial we will assume that you start a UR5 robot from a URSim version > 5.8.

Starting the driver and visualizing the robot in RViz

In order to start the driver it is not strictly necessary to install the External Control URCap. If you only want to connect to the robot and visualize the robot in RViz, nothing has to be changed on the robot. You should, however, make sure that is is running, as otherwise no joint data will be sent.

To start the driver run the following command in a terminal:

roslaunch ur_robot_driver ur5e_bringup.launch robot_ip:=192.168.56.101

In another terminal, run

roslaunch ur_robot_driver example_rviz.launch

This should open up an RViz window showing the robot. Try moving around the robot's joints using the teach panel and watch how it also moves inside ROS.

RViz window

Control the robot

For controlling the robot you'll need a special program running on the robot that will interpret and execute the commands coming from ROS. This can be either done by installing the External Control URCap as explained here or by using the headless mode. We assume the use of the URCap in the following.

Before any control of the robot can happen, a program containing the External Control program node must be running on the robot. You should see the output Robot ready to receive control commands. in the terminal where you launched the driver. See this troubleshooting entry if you experience problems in this step.

Control the robot using the test_move script

Once you've setup the robot for getting controlled by ROS, you can move it around using the test_move script.

Start the driver and an RViz window as before. Then, in a third terminal run

rosrun ur_robot_driver test_move

This script will guide you through the different trajectory controllers available. For informations about the different controllers, please see TODO: Change once merged #423.

It will first ask about the controller to be used:

Available trajectory controllers:
0 (joint-based): scaled_pos_joint_traj_controller
1 (joint-based): scaled_vel_joint_traj_controller
2 (joint-based): pos_joint_traj_controller
3 (joint-based): vel_joint_traj_controller
4 (joint-based): forward_joint_traj_controller
5 (Cartesian): pose_based_cartesian_traj_controller
6 (Cartesian): joint_based_cartesian_traj_controller
7 (Cartesian): forward_cartesian_traj_controller
Please choose a controller by entering its number (Enter '0' if you are unsure / don't care): 0

After that, you'll have to confirm that the motions the robot is about to make, can safely be executed the way your robot is mounted. If you're running a simulated robot, you can confirm this without thinking twice. However, if a real robot is connected, please make sure that the robot is actually free of any obstacles inside the space where the motion is taking place. See the images below for a visualization of the setpoints:

Joint-based motions: (The cable leaving the base to the back on an e-Series robot) joint_based motion 1

Cartesian motions: (The cable leaving the base to the back on an e-Series robot) Please note: Depending on your starting configuration there might also be a valid elbow-down configuration for each of the poses!

cartesian motion 1

You can play around by changing the trajectories inside the python code, if you like. You can also setup velocities in each waypoint to get a smooth motion through all waypoints.

Control the robot using MoveIt!

For controlling the robot using MoveIt! start the following three launch files (Each in a separate terminal window):

roslaunch ur_robot_driver ur5e_bringup.launch robot_ip:=192.168.56.101
roslaunch ur5e_moveit_config moveit_planning_execution.launch
roslaunch ur5e_moveit_config moveit_rviz.launch rviz_config:=$(rospack find ur5e_moveit_config)/launch/moveit.rviz

You should get an RViz window showing the robot overlayed with an orange version of the robot:

MoveIt RViz window

You can change the tcp target by dragging around the blue ball. The orange robot will show the configuration used to reach the target pose.

MoveIt RViz window planning

By clicking on the Plan button in the left panel a path from the current pose to the target pose is calculated. On success, it gets animated using a semi-transparent version of the robot:

MoveIt RViz window path

By clicking on the Execute button in the left panel the robot executes the planned motion:

MoveIt RViz window execution

In a real-world application you would not use the ur5e_moveit_config package but create your own moveit configuration matching your actual workcell with all obstacles and sensors. See the setup_assistant tutorial for more information on that.