Gait Modulation for Multi-Directional Burrowing in EMBUR 2.0

The EMerita BUrrowing Robot (EMBUR), inspired by the mole crab, was the first legged robot capable of self-burrowing vertically downward by excavating grains. Ongoing studies demonstrate that downward burrowing performance improves with gait switching in experiments using an updated version of the robot, EMBUR 2.0. In these experiments, gait selection is informed by body orientation relative to gravity during burrowing. We now investigate these concepts in the context of horizontal burrowing to enable EMBUR 2.0 to move multi-directionally. Preliminary data suggests that horizontal motion is achieved when all robotic legs rotate towards the robot's anterior, however the subsequent generation of granular mounds impedes forward locomotion. We propose and test the use of multiple different gaits to overcome these mounds, which we expect to detect using body orientation measures. We anticipate that multiple gait switches may be necessary for EMBUR 2.0 to burrow substantial horizontal displacements along a granular surface – as the robot creates and overcomes mounds or traverses uneven terrains. Results from this work show that legged, vertically self-burrowing robots can also be equipped with the ability to burrow horizontally, or 'walk.'