Gait Regulation and Feedback on a Robotic Climbing Hexapod

This paper proposes a novel method of applying feedback control for legged robots, by directly modifying parameters of a robot’s gait pattern. Gaits are a popular means of producing stable locomotion for legged robots, through the use of cyclic feedforward motion patterns, while requiring little to no sensory information. We are interested in incorporating feedback with these systems, and make use of salient parameters, found in gait patterns, to produce behaviors that span the space of possible gaits. These concepts are applied to a robotic hexapod, which, through the use of compliant microspines on its feet, is capable of climbing hard vertical textured surfaces, such as stucco. Experimental results are obtained comparing the use of a purely feedforward gait pattern to a behavior that actively modifies gait parameters while climbing, based upon sensory data.