NeSyPack: A Neuro-Symbolic Framework for Bimanual Logistics Packing

TL;DR

NeSyPack is a neuro-symbolic framework that enhances bimanual logistics packing by combining hierarchical reasoning, symbolic skill management, and data-driven models, leading to explainability, robustness, and superior performance.

Contribution

It introduces a novel neuro-symbolic hierarchical system for logistics packing that improves generalization, data efficiency, and reusability over traditional end-to-end models.

Findings

Outperformed end-to-end models in logistics packing tasks.

Won First Prize at the 2025 IEEE ICRA WBCD competition.

Demonstrated robustness and adaptability in complex packing scenarios.

Abstract

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.