OCELOT: Odometry and Contact Estimation for Legged Robots

2026-05-20

AI Summary

This work addresses the challenge of achieving high-precision odometry for legged robots relying solely on proprioceptive sensing, which is highly susceptible to slippage in complex terrains. To this end, the authors propose a legged odometry method based on an Error-State Extended Kalman Filter (ESEKF) that fuses data from IMUs, joint encoders, and force sensors. The approach incorporates state correction via stationary support foot detection and introduces an innovative slip rejection mechanism that explicitly identifies and discards slipping contacts by combining a force-based Gaussian Mixture Model Finite State Machine (GMM-FSM) with a kinematics-based Generalized Likelihood Ratio Test (GLRT), yielding continuous contact reliability scores. Experimental evaluation on a 2.4-kilometer multi-terrain dataset demonstrates that the proposed method significantly outperforms existing proprioceptive and exteroceptive approaches, maintaining high accuracy and robustness even in slippery environments.

Abstract

One of the significant challenges in legged robotics is achieving accurate odometry using only onboard proprioceptive sensors. In this study, we present a complete leg odometry pipeline based on an Error-State EKF (ESEKF) that relies exclusively on proprioceptive data: a body fixed IMU, joint encoders, and force sensors, where filter's state is corrected by feet determined to be in a stationary stance. The core of our contribution is fused contact detection and an uncertainty quantification module designed to explicitly identify and reject slippage. This module runs two detectors in parallel for each foot, 1) a debounced, force-based Gaussian Mixture Model (GMM) guided Finite State Machine (FSM) to confirm physical contact, and 2) a kinematic-based Generalized Likelihood Ratio Test (GLRT) on the estimated velocity of the foot. The continuous quality scores from both estimators are fused to detect if the foot is both physically loaded and kinematically stationary and served as an uncertainty signal for each contact. To validate our approach, we collected a multi-modal dataset of 29 sequences spanning diverse indoor and outdoor terrains (e.g., concrete, grass, pebble, and rock) total of 2.4 km long. We benchmarked our approach against both proprioceptive and exteroceptive methods. The results demonstrate our method's efficacy in providing accurate odometry estimates, robustly handling slippage-prone environments. We also share our code and real-time ROS2 package as open-source.

Problem

• legged robotics
• odometry
• proprioceptive sensors
• slippage
• contact estimation

Innovation

• Legged Odometry
• Contact Detection
• Slippage Rejection
• Error-State EKF
• Proprioceptive Sensing