Learning Contact Representation for Leg Odometry

2026-06-03

AI Summary

This work addresses the challenge of foot contact state detection in legged robots, which commonly relies on force sensors that are costly, scarcely available, and insensitive to disturbances such as slippage, thereby limiting odometry accuracy. The authors propose a self-supervised representation learning framework that leverages only standard joint encoder measurements—without requiring force sensors—to probabilistically model contact states and effectively distinguish between stance and swing phases. To the best of the authors’ knowledge, this is the first method to achieve highly robust contact detection without external force sensing, eliminating dependence on sensor calibration or additional hardware. Experimental results demonstrate that the proposed approach outperforms existing supervised methods and conventional probabilistic baselines in both accuracy and robustness.

Abstract

The estimation of odometry in legged robots depends on the assumption that the velocity of the foot with respect to the world remains zero during the stance phase. Feedback for the main body velocity is derived from the kinematic serial chain of the feet making accurate leg phase detection is a critical subproblem. A considerable number of studies employ ground reaction force sensors mounted at the tip of the foot to classify, yet these sensors may not be universally available for all legged robots. Additionally, these sensors are often unresponsive to unaccounted disturbances, such as slippage, while the foot remains in contact with the ground. In this study, we propose a self-supervised representation learning framework for contact detection that utilizes the standard sensor set of joint encoders without reliance on force sensor augmentations. We employ learned representations to model the stance and swing phases probabilistically. The experimental results obtained confirm the efficacy of the proposed self-supervised contact detector. Our framework exhibited superior performance in comparison to supervised methods which necessitate sensor set augmentation and labeling, as well as baseline probabilistic approaches. Additionally, we make our code available to the public.

Problem

• legged odometry
• contact detection
• phase classification
• sensor limitation
• slippage

Innovation

• self-supervised learning
• contact detection
• legged odometry
• representation learning
• sensor-free