arXivDaily arXiv每日学术速递 周一至周五更新
​通知:联邦节假日调休:哥伦布日 / 原住民日(Columbus Day / Indigenous Peoples' Day),今日(13日)无更新。
arXiv 2402.14299cs.ROcs.AI

我们选择前往太空:微重力下智能体驱动的人机与多机器人协作

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

Beijing University of Posts and Telecommunications(北京邮电大学);TongJi University(同济大学);Beihang University(北京航空航天大学);Beijing Normal University(北京师范大学);Nanjing University of Aeronautics and Astronautics(南京航空航天大学);Carnegie Mellon University(卡内基梅隆大学)

Miao Xin, Zhongrui You, Zihan Zhang, Taoran Jiang, Tingjia Xu, Haotian Liang, Guojing Ge, Yuchen Ji, Shentong Mo, Jian Cheng

AI总结:

本文提出 SpaceAgents-1 系统,通过微重力仿真环境和分层异构多智能体架构,实现由基础模型引导的人机与多机器人长期协作任务执行。

摘要:

我们提出 SpaceAgents-1,一个用于在微重力条件下学习人机与多机器人协作(HMRC)策略的系统。未来的太空探索需要人类与机器人协同工作。然而,在地面实验室中,要在微重力条件下习得熟练的机器人技能和灵活的协作能力面临重大挑战。为解决这一问题,我们开发了一个微重力仿真环境,并给出了三种典型的舱内机器人构型。我们提出一种分层异构多智能体协作架构:在基础模型的引导下,决策智能体(Decision-Making Agent)作为人机协作的任务规划器,而各个技能专家智能体(Skill-Expert Agents)负责机器人的具身控制。该机制使 SpaceAgents-1 系统能够执行一系列复杂的长期人机与多机器人协作任务。

英文摘要:

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.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑