Visual Upwind Flight Control

Recent experimental results demonstrate that flies posses a robust tendency to orient towards the frontally-centered focus of the visual motion field that typically occurs during upwind flight. We have developed a closed loop flight model (Figure 1), with a control algorithm based on feedback of the location of the visual focus of contraction, which is affected by changes in wind direction. We have demonstrated the feasibility of visually guided upwind orientation with a model derived from current understanding of the biomechanics and sensorimotor computation of insects. The matched filter approach used to model the visual system computations compares extremely well with open-loop experimental data. To test the ability of the closed loop system to orient the fly in the upwind direction, we presented 'step inputs' to the control system, where the fly was given an initial velocity and orientation and the wind was set at a fixed magnitude and
Direction (Figure 2). It is clear from the step responses that the tracking works, in the sense that the steady state error is driven to zero, resulting in upwind orientation. Closed loop simulations show stable upwind orientation behavior over the range of behaviorally-relevant wind speeds and sensitivity only to very low frequency disturbances. In future work we expect to extend this simulation to three dimensions and six degrees of freedom, and investigate vision algorithms that take advantage of global optic flow cues. Also of immediate interest is the velocity control problem associated with transition from backwards to forward flight in the upwind direction.

1 Closed loop model of upwind flight with wind disturbances.

2 Closed loop simulation results show the robustness of the tracking behavior. (A) Simulated 20 second flight trajectories, with fly positions plotted every 1.5 seconds. (B) Step responses in wind direction, showing zero steady-state error.

return to research index