Sensory Integration with Articulated Motion on a Humanoid Robot

Abstract

This paper describes the integration of articulated motion with auditory and visual sensory information that enables a humanoid robot to achieve certain reflex actions that mimic those of people. Reflexes such as reach‐and‐grasp behavior enables the robot to learn, through experience, its own state and that of the world. A humanoid robot with binaural audio input, stereo vision, and pneumatic arms and hands exhibited tightly coupled sensory‐motor behaviors in four different demonstrations. The complexity of successive demonstrations was increased to show that the reflexive sensory‐motor behaviors combine to perform increasingly complex tasks. The humanoid robot executed these tasks effectively and established the groundwork for the further development of hardware and software systems, sensory‐motor vector‐space representations, and coupling with higher‐level cognition.