Abstract
Robotic data generation is a promising paradigm for scaling robot learning without collecting large-scale real-world data. However, generating geometrically diverse yet physically valid data for contact-rich tasks remains challenging, especially when success depends on precise geometric interfaces. Standard shape augmentation methods often distort task-critical interfaces, resulting in invalid contact relationships, e.g., fit mismatches or interpenetration, rendering downstream interactions infeasible. To address these limitations, we propose a function-preserving Real-to-Sim-to-Real framework that generates synthetic demonstrations from reconstructed assets without teleoperated source trajectories. Our method augments task-relevant object geometries through constraint-guided mesh deformation, together with physically consistent transfer of task poses and collision proxies. Visual domain randomization is further applied during simulation rollouts, enabling robust zero-shot policy deployment without real-world fine-tuning. Extensive experiments in both real-world and simulation settings demonstrate that our method enables robust generalization across unseen object geometries and diverse visual conditions in contact-rich and long-horizon tasks. Our method provides a practical path toward scalable robot learning for contact-rich tasks via shape deformation.
Keywords
Subject
Publication details
- Journal
- Not available
- Open access
- Green open access
Cite this article
APA 7
Xiang, T., Xie, X., Cao, J., Luo, A. F., Li, H., & Ma, J. (2026). Function-Preserving Data Generation for Zero-Shot Real-to-Sim-to-Real Manipulation. https://omanscience.com/en/articles/function-preserving-data-generation-for-zero-shot-real-to-sim-to-real-manipulation
MLA 9
Xiang, Tianyi, et al. "Function-Preserving Data Generation for Zero-Shot Real-to-Sim-to-Real Manipulation." https://omanscience.com/en/articles/function-preserving-data-generation-for-zero-shot-real-to-sim-to-real-manipulation.
Chicago (author–date)
Xiang, Tianyi, Xupeng Xie, Jiahang Cao, Andrew F. Luo, Haoang Li, and Jun Ma. 2026. "Function-Preserving Data Generation for Zero-Shot Real-to-Sim-to-Real Manipulation." https://omanscience.com/en/articles/function-preserving-data-generation-for-zero-shot-real-to-sim-to-real-manipulation.
Harvard
Xiang, T., Xie, X., Cao, J., Luo, A. F., Li, H. and Ma, J. (2026) 'Function-Preserving Data Generation for Zero-Shot Real-to-Sim-to-Real Manipulation', Available at: https://omanscience.com/en/articles/function-preserving-data-generation-for-zero-shot-real-to-sim-to-real-manipulation.
Vancouver
Xiang T, Xie X, Cao J, Luo AF, Li H, Ma J. Function-Preserving Data Generation for Zero-Shot Real-to-Sim-to-Real Manipulation. https://omanscience.com/en/articles/function-preserving-data-generation-for-zero-shot-real-to-sim-to-real-manipulation
IEEE
T. Xiang, X. Xie, J. Cao, A. F. Luo, H. Li, and J. Ma, "Function-Preserving Data Generation for Zero-Shot Real-to-Sim-to-Real Manipulation," https://omanscience.com/en/articles/function-preserving-data-generation-for-zero-shot-real-to-sim-to-real-manipulation.