Multi-Sensor Fusion for Quadruped Robot State Estimation on Challenging Terrain
Year: 2024
Type: conference-paper
Abstract: This paper presents the preliminary results of MUSE, a MUlti-sensor State Estimator for quadruped robots, combining data from IMU, encoders, and cameras to accurately estimate pose and velocity, even in challenging environments such as uneven terrain. Experiments on a Unitree Aliengo robot, tested on stairs, rocks, and slippery surfaces, show MUSE's superior performance compared to using only a T265 tracking camera, providing reliable and high-frequency state estimation.
Authors: Nisticò, Ylenia, Soares, João Carlos Virgolino, Fink, Geoff, Semini, Claudio
Language: English
Subfield: Aerospace Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W6968524257
FWCI:
Cites:
Cited by:
Related to: