Heim, S.

Abstract

In this thesis we explore the use of a tail for terrestrial locomotion of animals and robots. We focus on steady-state locomotion in the sagittal plane over smooth terrain, with an emphasis on the exploitation of natural dynamics. We approach the problem in three ways: first we analyze the equations of motion to find the effect on the dynamics of the control input associated with the tail. From this we hypothesize that the tail’s primary function is to stabilize body pitch, and more importantly that the system’s morphology should be optimized in such a way that the control task of body-pitch stabilization is decoupled from the task of energy-input. We also explore the implications of scaling and compare this with a few cases in biology. Second we run optimizations on the derived models to find open-loop periodic solutions. We optimize over both the control inputs as well as morphological parameters. We thus observe if the optimal solutions match with the predictions from our previous hypothesis, as well as the effect of different cost functions.  Third, we have built a simple 1 degree of freedom tail and tested it’s use in finding a bounding gait with the Cheetah-Cub robot developed at the BioRob lab at EPFL. With this robot we have empirically tested the effect of different tail-parameters such as size and weight, as well as determined the importance of the patterns for a Central Pattern Generator (CPG) to generate open-loop inputs to synthesize stable bounding.

By matching the results in all three approaches we shed some light on the opportunities in which incorporating a tail in a robot makes sense, as well as suggest some reasons why most large land animals do not use tails for locomotion.