ABSTRACT
Based on the design and formulation of a multi-agent reinforcement learning technique provided in the preceding chapter, this chapter will present the results of several experiments that examine the effectiveness of learning group behaviors in autonomous robots and highlight some of the key observations obtained from these experiments as well as their underlying conditions. The aim is to demonstrate the interrelationshipsbetween the local autonomy of individual robots and their emergent spatial properties as a result of the dynamic interactions between robots of the same class and their environment that may involve robots of other classes.
