By mid-21st century, a team of fully autonomous humanoid robot soccer players shall win the soccer game, comply with the official rule of the FIFAThere will be several technological advancements, even if the goal of the robotic soccer team is not reached, starting with Team-Partitioned, Opaque-Transition Reinforcement Learning (TPOT-RL) [#!Sto98!#] which has found application in the domain of packet routing in computer networks. TPOT-RL is a distributed learning method in domains where ``agents have limited information about environmental state transitions'' ([#!Sto98!#],, against the winner of the most recent World Cup [#!robocuphomepage2000:goal!#].
In most RoboCup leagues, the teams consist of either robots or programs that co-operate in order to defeat the opponent team. RoboCup Rescue and the commentator exhibition diverge from the other RoboCup leagues. The goal of defeating an opponent would raise ethical issues in RoboCup Rescue, since we cannot assign comparable utilities to human lives and buildings. Hence, the focus in RoboCup Rescue is on the co-operative efforts between heterogeneous agents. In the commentator exhibition, the goal is to observe and comment.
Besides the commentator exhibition and RoboCup Rescue, the main body of the RoboCup challenge consists of several leagues for soccer playing. However, as this manual is about the simulated league we will only focus on it.