Within the animal kingdom, flies are the true masters of flight, capable of a
wide variety of very quick aerial maneuvers. Thanks to highly specialized
neuromuscular circuits, flies achieve a high wingbeat frequency combined with
precise control of the wingbeat amplitudes. The underlying control principles,
however, are still poorly understood.

I will report on experimental recordings and computational analysis of the
wingbeat of tethered flying fruit flies (Drosophila melanogaster). To determine
if the observed flight maneuvers arise from the activation of a small number of
independent control modes, we developed a statistical method based on
independent component analysis (ICA). The method identifies contributions to
the wingbeat amplitudes that are mutually statistically independent; such
components may be viewed as the basic control modes of the flight neuromotor
system. In the second part of the talk, I will discuss the role of
mechanosensory feedback in frequency control. Mechanically vibrating the flies,
we stimulated their specialized mechanosensory organ (the halteres), and
measured the stimulus parameters for which the flies phase-locked their
wingbeat with the stimulus. Based on the observed structure of regions of
synchronization, we conclude that the wingbeat is governed by a limit cycle
control system. Implications for robotics (micro air vehicles based on flapping
flight) will be briefly discussed.