Lab Course: Cognitive Robotics Implementation Challenge

Organizer: Stefan Profanter, Alexander Perzylo, Markus Rickert, Prof. Knoll

Modul: IN2106, IN2257

Registration: Via Matching System or TUMonline, and email to profante@in.tum.de (see below)

NOTE: Due to short term cancellation, there still is one open position. You can register apply for that position by registering directly in TUMonline

Type: Master Lab Course

Semester: WS 2018/19

Time & Location: Weekly meetings during the semester.

These weekly meetings are mandatory for all participating students. Additionally you will need to come to our lab to work with the other team members and with the robots.

Location: fortiss (U6 Nordfriedhof), Room Dijkstra, 2nd floor

  1. Thursday, 18. October 2018, 15:00-16:00
  2. Thursday, 25. October 2018, 15:00-16:00
  3. Tuesday, 30. October 2018, 16:00-17:00
  4. Tuesday, 06. November 2018, 16:00-17:00
  5. Tuesday, 13. November 2018, 16:00-17:00
  6. Tuesday, 20. November 2018, 16:00-17:00
  7. Tuesday, 27. November 2018, 16:00-17:00
  8. Tuesday, 04. December 2018, 16:00-17:00
  9. Tuesday, 11. December 2018, 16:00-17:00
  10. Tuesday, 18. December 2018, 16:00-17:00
  11. Tuesday, 08. January 2019, 16:00-17:00
  12. Tuesday, 15. January 2019, 16:00-17:00

Preparation:

If you already want to prepare yourself for the lab course, you can install Ubuntu 18.04 natively on your notebook (e.g. dual boot, or other comparable linux systems). The performance of a VM will probably not suffice since we are going to use a 3D physics simulator. If you are not familiar with ROS (Robot Operating System), you can also already install that on your system (ROS Melodic) and do some basic tutorials there (http://wiki.ros.org/ROS/Tutorials). Further tutorials will follow after the first appointment.

General Info:

This Lab Course will cover the fundamentals in the area of cognitive robotics by a hands-on implementation challenge on a 6-axis robot like the Comau e.DO (see picture).

This includes understanding forward- and backward kinematics, controlling robots via joint- and cartesian movement, and controlling robot tools like a gripper through middlewares such as OPC UA or ROS.

In combination with cognitive abilities such as creating a world model using knowledge representation and automatic inference and reasoning the student's task is to solve common problems in the area of industrial robotics.

The final task will be a game-like challenge between two student groups within the same course.

Expected previous knowledge:

  • Good knowledge in either C/C++ or Java
  • Knowledge in one of the following fields would be beneficial: Robot Control, Computer Vision, AI

Objective:

The students will obtain knowledge on real-world industrial robot applications and will then know how to build an integrated system for cognitive robot cells. They will also gain experience on working in teams, leading teams, and distributing and integrating tasks between team members.

Registration:

For registration, you have to be identified in TUMonline as a student. The registration will be held over the matching system of the Department of Informatics. However, personal applications will be preferred in the matching system. Please specify your related skill, what you are interested in and send an email with CV to profante@in.tum.de before July 11!