arXiv 2506.065672025-06-06cs.RO

NeSyPack：一种用于双臂物流包装的神经符号框架

NeSyPack: A Neuro-Symbolic Framework for Bimanual Logistics Packing

Carnegie Mellon University（卡内基梅隆大学）

Bowei Li, Peiqi Yu, Zhenran Tang, Han Zhou, Yifan Sun, Ruixuan Liu, Changliu Liu

备注：: 10 pages, 5 figures. Accepted to the RSS 2025 Workshop on Benchmarking Robot Manipulation: Improving Interoperability and Modularity. First Prize in the WBCD competition at ICRA 2025. Equal contribution by Bowei Li and Peiqi Yu

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

本文提出神经符号框架 NeSyPack，通过分层推理与符号技能图组合数据驱动模型，实现可解释、可复用且可靠的双臂物流包装，并在 WBCD 竞赛中获一等奖。

摘要：

本文提出了 NeSyPack，这是一种用于双臂物流包装的神经符号框架。NeSyPack 将数据驱动模型与符号推理相结合，构建了一个可泛化、数据高效且可靠的可解释分层系统。它通过分层推理将任务分解为若干子任务，并进一步分解为由符号技能图管理的原子技能。该图为执行选择技能参数、机器人配置以及特定任务的控制策略。这种模块化设计实现了鲁棒性、适应性和高效复用，优于需要大规模重新训练的端到端模型。借助 NeSyPack，我们团队在 2025 年 IEEE 国际机器人与自动化会议的“What Bimanuals Can Do”（WBCD）竞赛中获得一等奖。

英文摘要：

This paper presents NeSyPack, a neuro-symbolic framework for bimanual logistics packing. NeSyPack combines data-driven models and symbolic reasoning to build an explainable hierarchical system that is generalizable, data-efficient, and reliable. It decomposes a task into subtasks via hierarchical reasoning, and further into atomic skills managed by a symbolic skill graph. The graph selects skill parameters, robot configurations, and task-specific control strategies for execution. This modular design enables robustness, adaptability, and efficient reuse - outperforming end-to-end models that require large-scale retraining. Using NeSyPack, our team won the First Prize in the What Bimanuals Can Do (WBCD) competition at the 2025 IEEE International Conference on Robotics and Automation.

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。