LAUSR

Continuous-nondifferentiable terrains are extremely challenging for the environment adaption of six-legged robots. Previous researches have focused on gait planning methods to improve inherent ability of legged robots to walk over moderate terrains. However, most six-legged robots utilize relatively monotonic gait so that they still cannot well adapt tough terrains. As a result, the current legged robots easily get stuck and fall when encountering continuous-nondifferentiable terrains, such as stairs. Therefore, a method of gait switching is proposed so that six-legged robots can flexibly generate multiple gaits to adapt complex terrains. This study investigated the relationship between six-legged robot gait topologies and physical constraints, such as robot stability and robot-terrain interference. The proposed gait switcher can generate 0–6, 1–5, 2–4 and 3–3 gaits, which is instructed by the stability and interference criteria. Simulations and experiments were performed on a novel six-legged robot Hexa-XIII that succeeded climbing stairs over 45°. The effectiveness of the gait switching method is validated by the experiment results.