Learning the Geometric Mechanics of Robot Motion Using Gaussian Mixtures

2025/02/07 by Ruizhen Hu, Hu, Ruizhen, Shai Revzen +1 · 1 citation
Engineering · Computer Science · #Robotic Mechanisms and Dynamics #Robot Manipulation and Learning #Robotic Path Planning Algorithms

paper · pdf · doi:10.48550/arxiv.2502.05309

Abstract

Data-driven models of robot motion constructed using principles from Geometric Mechanics have been shown to produce useful predictions of robot motion for a variety of robots. For robots with a useful number of DoF, these geometric mechanics models can only be constructed in the neighborhood of a gait. Here we show how Gaussian Mixture Models (GMM) can be used as a form of manifold learning that learns the structure of the Geometric Mechanics "motility map" and demonstrate: [i] a sizable improvement in prediction quality when compared to the previously published methods; [ii] a method that can be applied to any motion dataset and not only periodic gait data; [iii] a way to pre-process the data-set to facilitate extrapolation in places where the motility map is known to be linear. Our results can be applied anywhere a data-driven geometric motion model might be useful.

Cited by

• Asymmetric Friction in Geometric Locomotion

Related

• Single-Example Learning in a Mixture of GPDMs with Latent Geometries
• Modeling Kinematic Uncertainty of Tendon-Driven Continuum Robots via Mixture Density Networks
• Generalizing Robot Trajectories from Single-Context Human Demonstrations: A Probabilistic Approach
• Learning from demonstration with model-based Gaussian process
• Improving Out-of-Distribution Generalization of Learned Dynamics by Learning Pseudometrics and Constraint Manifolds