Online 3D workpiece packing is crucial in logistics, but achieving high space utilization without prior knowledge of incoming item sequences remains challenging. Current methods struggle to capture dynamic 3D spatial topologies and heavily rely on rigid heuristics, which limits exploration diversity and causes space fragmentation. A deep reinforcement learning framework integrating a spatially weighted graph attention network and a diffusion model is proposed. The network accurately extracts the global topological dependencies and local geometric constraints of the packing state. Concurrently, the diffusion model synthesizes highly diverse, physically feasible candidate placement poses, fundamentally expanding the agent’s exploration boundary. Experiments demonstrate that our method significantly outperforms baselines in space utilization and efficiency. Its effectiveness and robustness are further validated in a real robotic system, offering a practical hybrid reinforcement learning solution for online 3D workpiece packing.
- Open Access
- Article
A Grasping and Placement Learning Framework for Workpieces Packing in Complex Scenes
- Zhilong Liu,
- Feiyue Wu,
- Qinghui Pan,
- Jie Lian *
Author Information
Received: 01 May 2026 | Revised: 16 Jun 2026 | Accepted: 25 Jul 2026 | Published: 17 Aug 2026
Abstract
Keywords
online 3D workpiece packing | deep reinforcement learning | graph attention network | diffusion model
References
- 1.
Dosa, G.; Sgall, J. Optimal Analysis of Best Fit Bin Packing. In Automata, Languages, and Programming, Proceedings of the 41st International Colloquium, ICALP 2014, Copenhagen, Denmark, 8–11 July 2014; Springer: Berlin/Heidelberg, Germany, 2014; pp. 429–441. https://doi.org/10.1007/978-3-662-43948-7 36.
- 2.
Parreno, F.; Alvarez-Vald´es, R.; Tamarit, J.M.; et al. A Maximal-Space Algorithm for the Container Loading Problem. INFORMS J. Comput. 2008, 20, 412–422. https://doi.org/10.1287/ijoc.1070.0254.
- 3.
Agarwal, M.; Biswas, S.; Sarkar, C.; et al. Jampacker: An Efficient and Reliable Robotic Bin Packing System for Cuboid Objects. IEEE Robot. Autom. Lett. 2021, 6, 319–326. https://doi.org/10.1109/LRA.2020.3043168.
- 4.
Ha, C.T.; Nguyen, T.T.; Bui, L.T.; et al. An Online Packing Heuristic for the Three-Dimensional Container Loading Problem in Dynamic Environments and the Physical Internet. In Applications of Evolutionary Computation, Proceedings of the 20th European Conference, EvoApplications 2017, Amsterdam, The Netherlands, 19–21 April 2017; Springer: Cham, Switzerland, 2017; pp. 140–155. https://doi.org/10.1007/978-3-319-55792-2_10.
- 5.
Hu, R.Z.; Xu, J.Z.; Chen, B.; et al. TAP-Net: Transport-and-Pack Using Reinforcement Learning. ACM Trans. Graph. 2020, 39, 1–15. https://doi.org/10.1145/3414685.3417796.
- 6.
Nguyen, T.H.; Nguyen, X.T. Space Splitting and Merging Technique for Online 3D Bin Packing. Mathematics 2023, 11, 1912. https://doi.org/10.3390/math11081912.
- 7.
Chen, L.; Zhang, S.; Ding, M.M.; et al. Solving a Large-Scale 3D Packing Problem with Robot Constraints. Comput. Oper. Res. 2026, 188, 107373. https://doi.org/10.1016/j.cor.2025.107373.
- 8.
Rao, Z.H.; Wu, Y.C.; Yang, Z.F.; et al. Visual Navigation with Multiple Goals Based on Deep Reinforcement Learning. IEEE Trans. Neural Netw. Learn. Syst. 2021, 32, 5445–5455. https://doi.org/10.1109/TNNLS.2021.3057424.
- 9.
Liu, Y.; Liu, S.; Chen, B.; et al. Fusion-Perception-to-Action Transformer: Enhancing Robotic Manipulation with 3D Visual Fusion Attention and Proprioception. IEEE Trans. Robot. 2025, 41, 1553–1567. https://doi.org/10.1109/TRO.2025.3539193.
- 10.
Que, Q.Q.; Yang, F.; Zhang, D.F. Solving 3D Packing Problem Using Transformer Network and Reinforcement Learning. Expert Syst. Appl. 2023, 214, 119153. https://doi.org/10.1016/j.eswa.2022.119153.
- 11.
Zhao, H.; She, Q.J.; Zhu, C.Y.; et al. Online 3D Bin Packing with Constrained Deep Reinforcement Learning. In Proceedings of the AAAI Conference on Artificial Intelligence, Online, 2–9 February 2021; Volume 35, pp. 741–749. https://doi.org/10.1609/aaai.v35i1.16155.
- 12.
Zhao, H.; Yu, Y.; Xu, K. Learning Efficient Online 3D Bin Packing on Packing Configuration Trees. In Proceedings of the International Conference on Learning Representations, Online, 25–29 April 2022; pp. 1–18.
- 13.
Xiong, H.; Ding, K.; Ding, W.; et al. Towards Reliable Robot Packing System Based on Deep Reinforcement Learning. Adv. Eng. Inform. 2023, 57, 102028. https://doi.org/10.1016/j.aei.2023.102028.
- 14.
Wang, B.Y.; Yu, W.; Zhang, X.D.; et al. Robotic Bin Packing via Hierarchical Reinforcement Learning. IEEE Trans. Autom. Sci. Eng. 2026, 23, 8634–8646. https://doi.org/10.1109/TASE.2026.3678655.
- 15.
Liu, H.T.; Lian, J.; Chen, Y.F.; et al. A Dual-Functional Push-Grasping Strategy for Goal-Agnostic and Goal-Oriented Tasks Based on Deep Reinforcement Learning. In Proceedings of the 2025 Joint International Conference on Automation-Intelligence-Safety (ICAIS) & International Symposium on Autonomous Systems (ISAS), Xi’an, China, 23–25 May 2025; pp. 1–6. https://doi.org/10.1109/ICAISISAS64483.2025.11051742.
- 16.
Hu, H.Y.; Zhang, X.D.; Yan, X.W.; et al. Solving a New 3D Bin Packing Problem with Deep Reinforcement Learning Method. arXiv 2017, arXiv:1708.05930.
This work is licensed under a Creative Commons Attribution 4.0 International License.
Suite 4002 Level 4, 447 Collins Street, Melbourne, Victoria 3000, Australia
General Inquiries: info@sciltp.com