We aim at developing vision-based controllers for microyers capable of ying autonomously in indoor environments. Although signicant advances have been made in this domain over the past decade, only subproblems have been successfully addressed (for example wall avoidance) and no ying robot is yet fully capable of achieving this goal. Since ying insects show remarkable indoor ight capabilities, they serve as a major source of inspiration at the level of both sensor suite and control. Traditional approaches that rely on inertial measurement units, GPS or active distance sensors are indeed impossible indoors due to the weight and consumption of these sensors. Ultimately, our research goal is a deeper understanding of the minimal set of me chanisms required by robots and insects to y in textured environments.