We recently spoke with Viktor Seib, PhD student at the University of Koblenz-Landau about the upcoming RoCKIn2015 Challenge. Here’s what they had to say about their mission, vision and preparation for the final event!
Hello Viktor - thank you for joining us today. To start with, could you give us a bit of background information on your team, its aims and mission?
Our team, homer@UniKoblenz, typically consists of students who are supervised by a PhD student (me in this case). We have a computer vision background and focus on enabling the robots to perceive and understand their environment. This also includes active vision, requiring the robot to adjust or re-adjust the sensor positions for a better view.
What appeals to you about RoCKIn and how does it complement your research and personal development goals? What about robot competitions more generally?
My PhD research focuses on object recognition. RoCKIn provides a great opportunity to test the algorithms in practice under reproducible circumstances. The benchmarking idea is what makes the results valuable for further research. Robot competitions in general are great to test the integration of different components into one system. They are also useful to motivate and interest new students for robotics.
Have you been involved in robotic competitions before? If so, what were some of the highlights for you?
Our @home team has participated in the RoboCup competitions since 2008. The highlights are of course to see the robots doing what they are intended to do! However, the greatest moment was to win RoboCup 2015 with our highly motivated team!
What role do you think competitions have in furthering innovation in robotics?
A competition is always a good motivation to put all your energy into a project and attract new people to a research field. However, it is important to create tasks that are difficult to solve, in order to push innovation. This is not trivial, since a competition has to be appealing to the audience as well.
How are your preparations for RoCKIn2015 going? What are you finding to be the most challenging areas?
Preparations are going well. Our hardware is repaired and functional. Additional security measures should prevent it from burning out like during the last year's competition. The most challenging areas for us are manipulation and speech processing.
And finally, what do you think the future holds for domestic service robotics?
There have been some low-cost robotic platforms released recently. I think, having a functional low-cost robotics platform will help many researchers focus on the important algorithms. This should boost the development of new techniques. I am also curious to see how deep convolutional neural networks will affect the performance of service robots in the near future.
Thanks for your time Viktor - we're looking forward to seeing you in Lisbon!