ALife in Humanoids: Developing a Framework to Employ Artificial Life Techniques for High-Level Perception and Cognition Tasks on Humanoid Robots

We describe our recent research and advances in building a framework enabling artifical life ( ALife ) systems on real robotic hardware. Our framework allows our iCub humanoid to build better visual perception, improve its motion capabilities and even provide a sense of proprioception. This paper presents how we can use various techniques, such as, e.g., genetic programming, to build subsystems for these specific areas. Our framework runs in parallel with the hardware system and is updated with new information from the real robot. We plan to use this framework in the future for developing higher cognitive tasks, such as, scene understanding, prediction of action outcomes, and reasoning on our robot.