Sample Efficient Learning of Body-Environment Interaction of an Under-Actuated System

2026-01-20

AI Summary

This study addresses the challenge of modeling locomotion in underactuated systems operating in high-friction environments, where motion arises from complex interactions between body deformation and the surroundings—particularly difficult when data are scarce. Within a geometric mechanics framework, the authors systematically evaluate four models of varying complexity (from linear to nonlinear) using motion-tracking data collected from a physical robot. They assess each model’s ability to predict the "mobility map"—the mapping from shape changes to body velocity—across diverse conditions, including different gaits and speeds. The findings reveal a critical trade-off between model complexity and data efficiency: simpler models outperform complex ones with limited data, whereas richer datasets enable nonlinear models to achieve significantly higher predictive accuracy. This work thus provides both theoretical insight and practical guidance for modeling underactuated interaction dynamics under data constraints.

Technology Category

• Intelligent Robots: Motion and Path Planning
• Reasoning under Uncertainty: Relational Probabilistic Models
• Planning, Routing, and Scheduling: Model-Based Reasoning

Application Category

• User Modeling, Personalization and Recommendation: User modeling and simulation for interactive and conversational systems
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs

Abstract

Geometric mechanics provides valuable insights into how biological and robotic systems use changes in shape to move by mechanically interacting with their environment. In high-friction environments it provides that the entire interaction is captured by the ``motility map''. Here we compare methods for learning the motility map from motion tracking data of a physical robot created specifically to test these methods by having under-actuated degrees of freedom and a hard to model interaction with its substrate. We compared four modeling approaches in terms of their ability to predict body velocity from shape change within the same gait, across gaits, and across speeds. Our results show a trade-off between simpler methods which are superior on small training datasets, and more sophisticated methods, which are superior when more training data is available.

Problem

• sample efficient learning
• under-actuated system
• body-environment interaction
• motility map
• geometric mechanics

Innovation

• sample efficient learning
• under-actuated systems
• geometric mechanics
• motility map
• body-environment interaction