Embodied Intelligence Observer

RAEM: Robust Autonomous Exploration for Multi-Floor Environments with a Quadruped Robot

Research

Source: arXiv cs.ROPublish time unverified

arXiv:2608.25366v1 Announce Type: new Abstract: In this paper, we propose RAEM, a robust autonomous exploration framework for quadruped robots operating in multi-floor environments. Most existing ground-robot exploration approaches rely on planar traversability representations, which cannot adequately represent the overlapping structures and cross-floor connectivity of multi-floor buildings.

RAEM: Robust Autonomous Exploration for Multi-Floor Environments with a Quadruped Robot | Embodied Intelligence Observer