DRIMS2 Simulation Environment

First steps

0. Update the docker repo git pull DRIMS2_Docker (⚠️ Even you already cloned this repo, be sure to pull.)
1. Start the docker : ./start.sh
2. Compile and source the drims_ws workspace

cd drims_ws/
catkin_make
source devel/setup.bash

Gazebo Simulation

To run the simulation:

1. Load Gazebo environment and robot simulation:

roslaunch drims_dice_demo arm_gazebo.launch robot:=yumi # or robot:=gofa

2. Launch low-level ROS Services Server for robot planning and control:

roslaunch abb_wrapper_control launchControlServer.launch robot:=yumi # or robot:=gofa

3. Launch high-level ROS Services Server for robot planning and control:

roslaunch abb_wrapper_control launchTaskServer.launch robot:=yumi # or robot:=gofa

4. Spawn the dice in simulation:

roslaunch drims_dice_demo spawn_dice.launch

NOTE: Each command should run in its own terminal

Real robot

⚠️ To use the real robot ask a tutor.

Steps to use the real robot (both Yumi or Gofa):

• You need an Ubuntu machine
• Connect your notebook to the physical robot via Ethernet
• Set your IP as Static IP: (192.168.125.100, Netmask: 255.255.255.0, Empty gateway)
• Test the connection with: ping 192.168.125.1 (from outside Docker)

⚠️ NOTE: If you are using the Gofa robot, before launching the real robot, you must run the following command:

export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/local/lib

Now you are ready to work:

1. Launch the real robot:

roslaunch drims_dice_demo real_robot.launch robot:=yumi # or gofa

2. Open another terminal and connect to the docker ./connect.sh, and launch low-level ROS Services Server for robot planning and control:

roslaunch abb_wrapper_control launchControlServer.launch robot:=yumi # or robot:=gofa

3. Open another terminal and connect to the docker ./connect.sh, and launch high-level ROS Services Server for robot planning and control:

roslaunch abb_wrapper_control launchTaskServer.launch robot:=yumi # or robot:=gofa

Example nodes

You can test the functions with the following commands (⚠️Keep in mind the notes that are given below):

Python:

rosrun drims_template take_the_dice_example.py

C++:

roslaunch drims_template test_example.launch

This launcher runs the script test_example.cpp located in the DRIMS_template ROS package.

⚠️ NOTE: remember to press 'next' in the RvizVisualToolsGui window to execute the motion on the robot. (Check the terminal where you launched the launchControlServer.launch to visualize the previous blue message)

⚠️ NOTE 2: the example assumes the dice localization is active. If you want to test the nodes with the real robots and without the vision, you can fake the dice localization with:

roslaunch drims_template test_spawn_dice.launch

Both the examples present a basic use of the services to move the robot:

• /plan_and_execute_pose: plan a collision-free joint trajectory to a given goal Cartesian pose and execute it.
• /plan_and_execute_slerp: plan a collision-free joint trajectory by using SLERP interpolation to a given a goal Cartesian pose and execute it.
• /plan_and_execute_joint: plan a collision-free joint trajectory to a given a Joint goal and execute it.
• /open_gripper: open the gripper.
• /close_gripper: close the gripper.

The following gives the ROS service definition, located in the srv folder of abb_wrapper_msgs ROS package:

Available ROS services to move the robot

Service structure abb_wrapper_msgs/plan_and_execute_pose

# == Request ==
geometry_msgs/Pose goal_pose
bool is_relative
---
# == Response ==
bool success
string message

Service structure abb_wrapper_msgs/plan_and_execute_joint

# == Request ==
float64[] joint_goal
---
# == Response ==
bool success
string message

Service structure abb_wrapper_msgs/plan_and_execute_slerp

# == Request ==
geometry_msgs/Pose goal_pose
bool is_relative
---
# == Response ==
bool success
string message

Service structure abb_wrapper_msgs/open_gripper and abb_wrapper_msgs/open_gripper

bool in_flag
---
bool out_flag
string message