Title:
Simplifying robotic locomotion by escaping traps via an active tail

Thumbnail Image
Author(s)
Soto, Daniel
Authors
Advisor(s)
Goldman, Daniel I.
Hammond, Frank L., III
Advisor(s)
Editor(s)
Associated Organization(s)
Series
Supplementary to
Abstract
Legged systems offer the ability to negotiate and climb heterogeneous terrains, more so than their wheeled counterparts \cite{freedberg_2012}. However, in certain complex environments, these systems are susceptible to failure conditions. These scenarios are caused by the interplay between the locomotor's kinematic state and the local terrain configuration, thus making them challenging to predict and overcome. These failures can cause catastrophic damage to the system and thus, methods to avoid such scenarios have been developed. These strategies typically take the form of environmental sensing or passive mechanical elements that adapt to the terrain. Such methods come at an increased control and mechanical design complexity for the system, often still being susceptible to imperceptible hazards. In this study, we investigated whether a tail could serve to offload this complexity by acting as a mechanism to generate new terradynamic interactions and mitigate failure via substrate contact. To do so, we developed a quadrupedal C-leg robophysical model (length and width = 27 cm, limb radius = 8 cm) capable of walking over rough terrain with an attachable actuated tail (length = 17 cm). We investigated three distinct tail strategies: static pose, periodic tapping, and load-triggered (power) tapping, while varying the angle of the tail relative to the body. We challenged the system to traverse a terrain (length = 160 cm, width = 80 cm) of randomized blocks (length and width = 10 cm, height = 0 to 12 cm) whose dimensions were scaled to the robot. Over this terrain, the robot exhibited trapping failures independent of gait pattern. Using the tail, the robot could free itself from trapping with a probability of 0 to 0.5, with the load-driven behaviors having comparable performance to low frequency periodic tapping across all tested tail angles. Along with increasing this likelihood of freeing, the robot displayed a longer survival distance over the rough terrain with these tail behaviors. In summary, we present the beginning of a framework that leverages mechanics via tail-ground interactions to offload limb control and design complexity to mitigate failure and improve legged system performance in heterogeneous environments.
Sponsor
Date Issued
2022-08-02
Extent
Resource Type
Text
Resource Subtype
Thesis
Rights Statement
Rights URI