Multi-Agent Collaboration on the Nao Platform
RoboCup is an international research initiative with the stated goal of creating a team of humanoid robots to beat the human soccer champion team by the year 2050. The robot soccer domain poses a multiagent collaboration challenge wherein individual robots in a team must work together towards the common objective of winning soccer matches. Given the hardware specifications and computational constraints of the Aldebaran Nao humanoid robot used in the Standard Platform League of RoboCup, this paper describes the development of adaptive strategies to improve the collaboration capabilities of TT-UT Austin Villa, a joint team comprising Texas Tech University and The University of Texas at Austin. Fundamental aspects of soccer strategy and teamwork such as positioning, passing and scoring are addressed, and the associated algorithms are implemented and evaluated on Nao robots.