Evolving a vision-driven robot controller for real-world indoor navigation

Abstract

In this paper, we use genetic programming (GP) to evolve a vision-driven robot controller capable of navigating in a real-world environment. To this aim, we extract visual primitives from the video stream provided by a camera mounted on the robot and let them to be interpreted by a GP individual. The response of GP expressions is then used to control robot's servos. Thanks to the primitive-based approach, evolutionary process is less constrained in the process of synthesizing image features. Experiments concerning navigation in indoor environment indicate that the evolved controller performs quite well despite very limited human intervention in the design phase. © 2008 Springer-Verlag Berlin Heidelberg.