Workshop W2

AI challenges for humanoid soccer-playing robots

Abstract

The workshop will be centered around the RoboCup-soccer benchmark for humanoid robots. The idea for the benchmark dates back to a paper by Alan Mackworth, entitled "On Seeing Robots" presented at VI-92, 1992, and 1993 published in a book "Computer Vision: System, Theory, and Applications". It may be considered as being inspired by Shannon's 1950 chess challenge.

Using the soccer game as a benchmark in robotics research and development is one of the pillars of the RoboCup initiative, the 'dream' of RoboCup: 'By the middle of the 21st century, a team of fully autonomous humanoid robot soccer players shall win a soccer game, complying with the official rules of FIFA, against the winner of the most recent World Cup.' ( www.robocup.org ).

The challenges in the field of soccer-playing robots cover almost the entire range of AI research. Specifically it addresses computer vision, perception, planning and scheduling, heterogeneous multi agent systems and uncertainty in AI. The workshop should foster exchange on AI research in view of cooperating robots. Relevance for a wider AI community stems from the specific scenario of interacting robots that eventually will play against humans.