Learning plays a vital role in the development of autonomous agents. I
n this paper, we explore the use of reinforcement learning to ''shape'
' a robot to perform a predefined target behavior. We connect both sim
ulated and real robots to Alecsys, a parallel implementation of a lear
ning classifier system with an extended genetic algorithm. After class
ifying different kinds of Animat-like behaviors, we explore the effect
s on learning of different types of agent's architecture and training
strategies. We show that the best results are achieved when both the a
gent's architecture and the training strategy match the structure of t
he behavior pattern to be learned. We report the results of a number o
f experiments carried out both in simulated and in real environments,
and show that the results of simulations carry smoothly to physical ro
bots. While most of our experiments deal with simple reactive behavior
, in one of them we demonstrate the use of a simple and general memory
mechanism. As a whole, our experimental activity demonstrates that cl
assifier systems with genetic algorithms can be practically employed t
o develop autonomous agents.