Robot football, artificial life, and complexity

Abstract

Autonomous football-playing robots provide a stimulating research challenge in the sciences of complexity and artificial life. Currently, the game is dominated by problems of making the robots move sufficiently accurately. Even so, the dynamics of robot football are clearly chaotic, requiring some higher level control strategy. A mathematics of therelations between the robots, the ball, and the pitch is introduced. This mathematics supports a theory of structural time necessary for higher level dynamics and cognitive functions. In comparison with computer chess, robot football is more complex and may supplant it as a bench-mark test. Many systems considered to be complex have behaviour which emerges from interacting autonomous agents.Simulation is a new paradigm on which a science of such systems is being built. However, simulation currently suffers from the “can you trust it” syndrome: for many systems it is impractical to do experiments to test the simulation. However, robot football is a system which can be both simulated and built. It is suggested that this makes it an important scientific laboratory subject for understanding the relationship between simulation and real complex system behaviour.