Jonathan E. Clark1 and Mark R. Cutkosky2
1University of Pennsylvania, Philadelphia, PA
19104
2Stanford University, Stanford, CA 94305
Estimating the effect of design changes on the stability of robots designed to run over rough terrain is a difficult task. No current predictor or measure is currently universally accepted. The most common techniques, including the ‘stability margins’ commonly used in walking machines and the return-map eigenvalue analysis used on simple models, are not fully applicable to these fast and complex systems. This paper describes a comparison of a number of approaches to measuring stability applied to a redesign of our hexapedal robot.