Kurzfassung

Since 2005, the Active Vision Group's *Robbie* student projects enable practical insights into the field of domestic robotics for Bachelor and Master students of Computational Visualisitics and Computer Science at Campus Koblenz.

The project's hardware and software architecture is iteratively developed and extended by each year's student team in an agile project development manner. The project is currently supervised by Active Vision Group research associate Daniel Müller as project lead with...
Since 2005, the Active Vision Group's *Robbie* student projects enable practical insights into the field of domestic robotics for Bachelor and Master students of Computational Visualisitics and Computer Science at Campus Koblenz.

The project's hardware and software architecture is iteratively developed and extended by each year's student team in an agile project development manner. The project is currently supervised by Active Vision Group research associate Daniel Müller as project lead with support through student assistant Niklas Yann Wettengel.

Starting in early October, participating students get to know our robots, Lisa and TIAGo, and at first learn how to control such complex systems using top-level commands. After crash courses in the beginning, students quickly learn how to implement more extensive programs and gain in-depth knowledge in common robot functionality such as handling arbitrary sensory data. Furthermore, each student is assigned to a more project specific role besides working as a programmer.

Current *Robbie* projects each conclude with participating in the @Home league of RoboCup GermanOpen in April, where we compete together as team *homer* for University of Koblenz-Landau. If successful enough, there is also the possibility to compete at the world championship of RoboCup @Home league on a voluntary basis. With lots of effort, passion and experience over the past years, we have managed to even win several world championship titles.


Developing and running our robots' functional modules is currently done using Ubuntu 18.04 and ROS Melodic. Our software modules span a wide range of applications, such as
- mapping and navigation,
- object recognition and segmentation,
- person, gesture and speech recognition,
- people following and guiding and
- object manipulation.

In order to being able to achieve autonomous solutions in these fields, knowledge and techniques from computer vision and machine learning are combined and researched further in the scope of our projects.

Please visit our blog for most recent information and updates on our current team!
» weiterlesen» einklappen