A Topological Approach to Gait Generation for Biped Robots

This video supplements the results of the proposed bipedal gait generation framework in [1]. The video shows walking and brachiating gaits of a representative set of 2D and 3D biped robots studied in the literature.
Every gait shown in the video is part of a continuous family of gaits in a space of parameterized hybrid trajectories for the corresponding model being animated. These gait families connect to the equilibria of their models in their respective trajectory spaces. This motivates the use of equilibria as templates, numerical continuation methods to trace these gait families, and an overall topological approach to gait generation as proposed in the paper. In the end, this video demonstrates that equilibria are reliable seeds for the construction of gait families, eliminating the need for random, intuited, or bio-inspired initial guesses at feasible trajectories in an optimization framework. Further details can be found in our paper.

[1] Nelson Rosa Jr. and Kevin M. Lynch: “A Topological Approach to Gait Generation for Biped Robots”, June 2020; http://arxiv.org/abs/2006.03785.

An implementation of the framework can be found here: https://github.com/nr-codes/BipedalGa....