Looming Escape (Fly Swatter)

experimentNo. 0070categoryLabs & sensesrecordingpendingdataCC BY 4.0

The fly feeds on spilled sugar on a kitchen counter while four swats approach at constant velocity. In closed loop, a measured DNp01 giant-fiber spike before contact triggers a pre-programmed takeoff.

Setup

• Six stages over 52 s: sugar on the counter, four swat trials of 9 s each, and a tally.
• l/v schedule: a sweep of 80, 40, 20 and 10 ms, or a fixed value from 10 to 120 ms.
• Swatter half-width 50 mm, approached at 60° elevation; without a qualifying spike it lands just in front of the fly as a near miss with a sugar puff.
• Cameras: Counter / feeding fly, Over the shoulder / incoming swatter, Macro / proboscis on sugar, Kitchen / whole swing, Fly POV / approach corridor, Tally card.

Measurement

• Each 200 ms window carries ten 20 ms retinal frames, so the loom expands step by step; measured readouts are DNp01 and LC4 + LPLC2.
• Closed loop: a measured DNp01 spike in a window that ends before counter contact triggers the takeoff. Open loop presents the stimulus only.
• An optional, bounded antennal wind drive scaled by loom size is off by default.
• Escapes and near misses are tallied by l/v from measured giant-fiber windows only.

Limitations

• The takeoff motion itself is pre-programmed; only its trigger comes from a measured spike.
• Neural weights are fixed; the stimuli follow Card & Dickinson (2008) and von Reyn et al. (2014, 2017) at fly scale.

Sources

• Looming-escape reference (Nature Neuroscience)
• DNp01 giant-fiber escape reference