ROMAMar 23

Can a Robot Walk the Robotic Dog: Triple-Zero Collaborative Navigation for Heterogeneous Multi-Agent Systems

arXiv:2603.2172381.51 citationsh-index: 8Has Code
AI Analysis

This addresses the challenge of real-world deployment for heterogeneous robot cooperation, offering a practical solution by eliminating reliance on training and simulation, though it appears incremental as it builds on existing multimodal language models and robot architectures.

The paper tackles the problem of collaborative navigation for heterogeneous multi-robot systems by introducing Triple Zero Path Planning (TZPP), a framework that requires zero training, zero prior knowledge, and zero simulation, achieving robust, human-comparable efficiency and strong adaptability in diverse environments.

We present Triple Zero Path Planning (TZPP), a collaborative framework for heterogeneous multi-robot systems that requires zero training, zero prior knowledge, and zero simulation. TZPP employs a coordinator--explorer architecture: a humanoid robot handles task coordination, while a quadruped robot explores and identifies feasible paths using guidance from a multimodal large language model. We implement TZPP on Unitree G1 and Go2 robots and evaluate it across diverse indoor and outdoor environments, including obstacle-rich and landmark-sparse settings. Experiments show that TZPP achieves robust, human-comparable efficiency and strong adaptability to unseen scenarios. By eliminating reliance on training and simulation, TZPP offers a practical path toward real-world deployment of heterogeneous robot cooperation. Our code and video are provided at: https://github.com/triple-zeropp/Triple-zero-robot-agent

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes