Our goal is to demonstrate the ability of bio-inspired techniques to solve the problem of piloting an autonomous 40-grams aircraft within textured indoor environments. Because of severe weight and energy constraints, inspiration has been taken from the fly and only visual and vestibular-like sensors are employed. This paper describes the models and algorithms that will be used for altitude control and frontal obstacle avoidance, mainly relying on optic flow. For experimental convenience, both mechanisms have first been implemented and tested on a small wheeled robot featuring the same hardware as the targeted aircraft.