Bartender Project 2016/2017

This is the result of a two term master project at the Technical Aspects of Multimodal Systems (TAMS) Group at the University of Hamburg.

Objective

The goal of the project is to let an UR5 robot arm with Robotiq 3-Finger Adaptive Robot Gripper perform a complex grasping and pouring manipulation based on visual data. In other words, the robot arm is supposed prepare a drink following a cocktail recipe and based on user input from a web interface.

Running The Bartender

After getting the host repo and all the dependencies do a catkin_make and off you might go: The demo can be run both with and without hardware. Each method only requires launching a single launch file, which will start all necessary processes, drivers and a simple web interface from which the cocktails can be ordered.

Run On Hardware

roslaunch tams_ur5_bartender bartender.launch

Run In Simulation

The demo can be simulated and visualized in RVIZ completely by running

roslaunch tams_ur5_bartender simulation.launch 

Packages

• tams_ur5_bartender_msgs

• tams_ur5_bartender_coordination

• tams_ur5_bartender_gui

• tams_ur5_bartender_image_processing

• tams_ur5_bartender_manipulation

Dependencies

• tams_ur5_bartender_manipulation
  • MoveIt!
    • we encountered a couple of bugs which we bypassed in the original code
  • tams_ur5_setup with branch 'project2016'
  • tams_ur5
• tams_ur5_bartender_image_processing
  • OpenCV
  • Kinect driver
  • apriltags_ros
  • Camera Positioner
• tams_ur5_bartender_gui
  • rosbridge_suite