We Choose to Go to Space: Agent-driven Human and Multi-Robot Collaboration in Microgravity

TL;DR

This paper introduces SpaceAgents-1, a simulation environment and hierarchical multi-agent system designed to facilitate human and multi-robot collaboration in microgravity, advancing space exploration capabilities.

Contribution

The paper presents a novel microgravity simulation platform and a hierarchical multi-agent architecture guided by foundation models for effective space-based human-robot collaboration.

Findings

Successful simulation of intra-cabin robot configurations

Effective long-horizon collaboration in microgravity scenarios

Hierarchical architecture improves task execution efficiency

Abstract

We present SpaceAgents-1, a system for learning human and multi-robot collaboration (HMRC) strategies under microgravity conditions. Future space exploration requires humans to work together with robots. However, acquiring proficient robot skills and adept collaboration under microgravity conditions poses significant challenges within ground laboratories. To address this issue, we develop a microgravity simulation environment and present three typical configurations of intra-cabin robots. We propose a hierarchical heterogeneous multi-agent collaboration architecture: guided by foundation models, a Decision-Making Agent serves as a task planner for human-robot collaboration, while individual Skill-Expert Agents manage the embodied control of robots. This mechanism empowers the SpaceAgents-1 system to execute a range of intricate long-horizon HMRC tasks.