This book introduces an algorithm to determine a humanoid robot s position in an environment of known characteristics. The algorithm is based on evolutionary method as a machine learning strategy and takes into account the characteristics and constraints to which the humanoid robot is subject. The method raises the possibility to store a set of databases holding the minimal information from the images that describe the navigation area or possible obstacles in the robot's memory. The information in the databases can later be used by the robot at run time to navigate the area.