Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain
the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in
Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles
and JavaScript.
The apparent motion of a flier’s surroundings is shown to stabilize its flight by providing information about its orientation. Lapses in information are overcome through the effects of sensor noise and body oscillations.
These days, a buzzing sound overhead is perhaps just as likely to signal the approach of a camera drone as that of a bee. Either way, the sound results from wings or rotors whose fast cyclical motion serves the joint functions of supporting the flier’s weight, and stabilizing and controlling its flight. Achieving stability through feedback control requires an accurate estimate of the attitude angle that describes a flier’s orientation with respect to the direction of gravity. Drones estimate their attitude using gravity-sensing devices called accelerometers, but it’s not yet clear how insects do so without accelerometers1, because vision offers a direct estimate of attitude only when the horizon is visible. Writing in Nature, de Croon et al.2 show how vision can be used to infer attitude dynamically from the apparent motion of features in a flier’s surroundings — and demonstrate their point by flying unstable drones without accelerometers.