Large-scale automated tracking of ant running kinematics and foot contact dynamics


Meeting Abstract

P2-201  Friday, Jan. 5 15:30 – 17:30  Large-scale automated tracking of ant running kinematics and foot contact dynamics CLIFTON, GT*; GRAVISH, N; Univ. of California, San Diego; UCSD glenna.clifton@gmail.com

Ants have evolved to inhabit and therefore move through diverse habitats, requiring walking, running, climbing and maneuvering. How do ants coordinate limb movements and modulate foot contact under varying locomotion behaviors? Previous research on ant locomotion and foot mechanics focuses on in situ experiments or straight running during one or a few isolated strides. Here, we present a new approach to automatically record, track and analyze ant kinematics and foot contact dynamics across many strides and during non-stereotyped behaviors. Argentine ants, Linepithema humile (body length ~2-3 mm, ~3.5 mg), are placed within a darkened enclosure resting on a smooth glass surface illuminated with LEDs from the side. Ants within the field of view are automatically detected and recorded from above and below using two high-speed cameras at 200 fps. The body of each ant is automatically tracked while foot contact timing and area is visualized using frustrated Total Internal Reflection. The above camera has a field of view of ~28×38 mm and reliably identifies contact areas above 0.0003 mm^2. The lower camera provides a detailed view of foot contact with a view of ~4×6 mm and a contact area resolution of 0.0001 mm^2. We record ants reaching running speeds of 60 mm/s (corresponding to 26 body lengths/s) and an instantaneous acceleration of over 450 body lengths/s^2 within a single stride. Most ants employ an alternating tripod gait with a maximum stride frequency of 14 Hz. The hindlimbs generally contact the surface less strongly than the fore- and midlimbs, except while braking or pivoting within a stride. To run along a curve, ants shift the touchdown of the outside forelimb towards the body midline. Findings using this technique can inform legged robot design and control to accomplish diverse tasks.

the Society for
Integrative &
Comparative
Biology