CoTeSys-ROS Fall School on Cognition-enabled Mobile Manipulation

• Location: Hess Str 134 and Barer Str 21, Munich, Germany (MAP)

• Date: Monday November 1 2010 - Saturday November 6 2010

• Time: 9am -7pm

• Flyer: flyer.pdf

Sponsored by:

Description and Objectives

Autonomous mobile manipulation is expected to make a quantum leap over the next couple of years. The reasons are manifold and include the following ones. Sensor-equipped mobile manipulation platforms are becoming more affordable and more widely available. Leading-edge computer vision and 3D perception get finally applied to real scenes in autonomous robotics applications. Cognition- and AI-based control techniques are being successfully grounded into high-end robot control systems and are becoming mature enough to substantially increase the generality, flexibility, and reliability of robots' activities. Also, comprehensive open-source software libraries that contain leading-edge software tools for perception, learning and planning and all the way to hand manipulations are being developed and are supported by strong communities.

The COTESYS - ROS Fall School on Cognition-enabled Mobile Manipulation will introduce the participants into the exciting research area of autonomous mobile manipulation. Introductory lectures will be given by world renowned experts in the fields of 2D/3D perception, learning, reasoning and planning. The lectures will be complemented with hands-on practical exercises using various modules of the open-source ROS framework. Many of principle developers will be present and available for support and interaction. The application domain for the fall school will be everyday manipulation in human living environments. The goal to be pursued throughout the whole week will be to build a practical application on the high-end robots (e.g. TUM-Rosie, PR2) that will give robots the power to perceive and interpret typical scenes in households, reason about and infer the subsequent course of actions and to finally execute them in the form of mobile manipulation.

Prerequisites

This is a hands-on workshop, with programming exercises for using ROS and the PR2.

Attendees will need prior experience with ROS:

• Make sure you can read the syntax of C++ (www.cplusplus.com) and Python (http://docs.python.org/tutorial/)

• http://www.ros.org/wiki/ROS/Tutorials

• http://www.ros.org/wiki/tf/Tutorials

• http://www.ros.org/wiki/actionlib/Tutorials

• Follow the setup instructions for your local machine on IT Infrastructure

• Verify that Gazebo and Rviz are running properly: /Troubleshooting#Are_Gazebo_and_Rviz_running_properly.3F

• Install Protege and make the tutorial

• Skim through the first five tutorials on Prolog

• Skim through chapters 1-7 of Practical Common Lisp book. Most important are chapters 4, 5 and 6.

Schedule

Monday (before lunch: Hess Str. 134, after lunch: Barer Str. 21)

8:00 AM Registration

8:30 AM Technical Remarks

8:45 AM Welcome

9:00 AM Cognition Enabled Control (Michael Beetz) [ pdf ] [ video ]

10.00 AM ROS Introduction (Radu Bogdan Rusu) [ ROS Overview ] [ video ]

11.00 AM BREAK

11.15 AM ROS Introduction (Radu Bogdan Rusu) [ ROS tutorial ] [ video ]

12.00 AM Navigation tutorial (Brian Gerkey) [ Navigation tutorial ] [ video ]

1:00 PM Lunch

2.00 PM Bus to CCRLI

2.15 PM Lab Session I: Getting to know ROS :: Navigation (Brian Gerkey, Radu Bogdan Rusu)

READ THE SYLLABUS

7:00 PM Gone For the Day or Free Hacking

Tuesday (before lunch: Hess Str. 134, after lunch: Barer Str. 21)

9:00 AM OpenCV (Victor Eruhimov) [ vision ] [ OpenCV ] [ video1 ] [ video2 ]

11.00 AM BREAK

11.15 Point Cloud Library (Radu Bogdan Rusu) [ PCL ] [ Viewpoint Feature Histogram ] [ video ]

1:00 PM Lunch

2.00 PM Bus to CCRLI

2.15 PM Lab Session II: OpenCV / PCL object recognition (Radu Bogdan Rusu, Victor Eruhimov)

READ THE SYLLABUS

7:00 PM Gone For the Day or Free Hacking

Wednesday (all day: Hess Str. 134)

9:00 AM Presentation Willow Garage (Brian Gerkey) [ odp ]

9:20 AM Presentation Kuka (Johannes Kurth)

9:40 AM Presentation Schunk (Roko Tschakarow) [ pptx ]

10:00 AM Presentation Tecnalia (Gorka Azkune) [ odp ]

10:20 AM Presentation REC (Christian Verbeek) [ pdf ]

10:40 AM Presentation Metralabs (Christian Martin) [ pdf ]

11.00 AM BREAK

11.15 AM Company Exhibitions

1:00 PM Lunch

2:00 PM 2-min Presentation of Participants

4:00 PM Demo at CCRLI Lab

7:00 PM Organized Leisure Event at Hofbräuhaus

Thursday (before lunch: Hess Str. 134, after lunch: Barer Str. 21)

9:00 AM 3D Simulaton (Mihai Dolha)[ pdf ] [ video ]

9.30 AM Object-Action Complexes (Tamim Asfour) [ pdf ] [ video ]

10.45 AM BREAK

11.00 AM Manipulation and Grasping in ROS (Gil E. Jones) http://download.ros.org/data/Events/CoTeSys-ROS-School/cotesys_ros_manipulation_and_grasping.pdf, [ video ]

12.00 AM Plan-based Control (Lorenz Moesenlechner) [ video ]

1:00 PM Lunch

2.00 PM Bus to CCRLI

2.15 PM Lab Session III: Manipulation and Grasping in ROS (Gil E. Jones), Plan-based Control (Lorenz Moesenlechner)

READ THE SYLLABUS

7:00 PM Gone For the Day or Free Hacking

Friday (before lunch: Hess Str. 134, after lunch: Barer Str. 21)

9:00 AM Learning Kinematic Models of Articulated Objects [ pdf ] [ articulation stack ] [ webcam demo ] [ video ](Juergen Sturm)

10.00 AM Task Guided Attention Control and Visual Verification (Kei Okada) [ pdf ] [ video ]

11.20 AM BREAK

11.30 Inverse Reinforcement Learning [ pdf ] [ video ] (Pieter Abbeel)

1:00 PM Lunch

2.00 PM Bus to CCRLI

2.15 PM Lab Session IV: Knowledge Processing (Moritz Tenorth et al.) [ KnowRob slides ] [ Exercise ] [ Cheat sheet ]

READ THE SYLLABUS

7:00 PM Gone For the Day or Free Hacking

Saturday (before lunch: Hess Str. 134, after lunch: Barer Str. 21)

9:00 AM Supervised/Reinforcement Learning (Martin Riedmiller) [ pdf ] [ video ]

10.30 AM BREAK

10.40 AM Search-based Planning for Manipulation with Motion Primitives (Maxim Likhachev) [ pdf ] [ video ]

12.10 AM To ROS or not to ROS (Herman Bruyninckx) [ pdf ] [ video ]

1:00 PM Lunch

2.00 PM Bus to CCRLI

2.15 PM Lab Session V-Part 1: SBPL Planning (Ben Cohen)

3.30 PM Lab Session V-Part 2: Inferring Missing Objects from Tabletops and Integration [ pdf ](Dejan Pangercic et al.)

READ THE SYLLABUS

6:00 PM Program Close

IT Infrastructure

IT Infrastructure

Teams

Teams

All Talks

* Cognition Enabled Control (Michael Beetz) [ pdf ] [ video ]

* ROS Introduction (Radu Bogdan Rusu) [ ROS Overview ] [ video ]

* ROS Introduction (Radu Bogdan Rusu) [ ROS tutorial ] [ video ]

* Navigation tutorial (Brian Gerkey) [ Navigation tutorial ] [ video ]

* OpenCV (Victor Eruhimov) [ vision ] [ OpenCV ] [ video1 ] [ video2 ]

* Point Cloud Library (Radu Bogdan Rusu) [ PCL ] [ Viewpoint Feature Histogram ] [ video ]

* 3D Simulaton (Mihai Dolha)[ pdf ] [ video ]

* Object-Action Complexes (Tamim Asfour) [ pdf ] [ video ]

* Manipulation and Grasping in ROS (Gil E. Jones) cotesys_ros_manipulation_and_grasping.pdf, [ video ]

* Plan-based Control (Lorenz Moesenlechner) [ video ]

* Learning Kinematic Models of Articulated Objects [ pdf ] [ articulation stack ] [ webcam demo ] [ video ](Juergen Sturm)

* Task Guided Attention Control and Visual Verification (Kei Okada) [ pdf ] [ video ]

* Inverse Reinforcement Learning [ pdf ] [ video ] (Pieter Abbeel)

* Supervised/Reinforcement Learning (Martin Riedmiller) [ pdf ] [ video ]

* Search-based Planning for Manipulation with Motion Primitives (Maxim Likhachev) [ pdf ] [ video ]

* To ROS or not to ROS (Herman Bruyninckx) [ pdf ] [ video ]