Various competitions within RoboCup have a so-called Scientific Challenge, in which teams share the latest scientific developments. This exchange of knowledge is an important goal of RoboCup, to accelerate the development of robotics and AI.
The MSL teams' Scientific Challenge has just taken place.
Our own Hao was allowed to kick off on behalf of Tech United with a wonderful contribution about how knowledge on soccer strategy can be used by the robots. A convincing story with enormous potential for the further development of robot soccer.
This is followed by the French RCT with a story about using Q-learning for learning actions, the combined VDL/Falcons team with a further explanation of all the ins and outs of the story presented yesterday about standardization of interfaces, then team LAR@MSL explaining how they make decisions based on optimisation of soccer-related cost functions, and finally team UM Croatia presenting a number of innovative designs based on low-cost printed and laser-cut plastic parts while maintaining performance.
All in all, an interesting session where the teams showed their best scientific side. The score will follow tomorrow.