B-Human 2019 - Complex Team Play Under Natural Lighting Conditions

Thomas Röfer; Tim Laue; Gerrit Felsch; Arne Hasselbring; Tim Haß; Jan Oppermann; Philip Reichenberg; Nicole Schrader

In: Stephan Chalup; Tim Niemueller; Jackrit Suthakorn; Mary-Anne Williams (Hrsg.). RoboCup 2019: Robot World Cup XXIII. RoboCup International Symposium (RoboCup-2019), Pages 646-657, Lecture Notes in Artificial Intelligence (LNAI), Vol. 11531, Springer, 2019.


In the RoboCup Standard Platform League 2019, the team B-Human won the main competition and, together with Berlin United - Nao-Team Humboldt, the Mixed Team Competition. For being successful in such a competitive environment, many sophisticated solutions for a variety of robotics tasks need to be found and integrated in a reliable and efficient manner. In this paper, we focus on three aspects that we consider as crucial for this year's success and that have not been published before: a system of neural networks for ball classification and position estimation, a new framework for behavior specification along with its application to passes and set plays, and a set of approaches for maintaining the reliability of our robot team throughout a game.

Weitere Links

Deutsches Forschungszentrum für Künstliche Intelligenz
German Research Center for Artificial Intelligence