Co-Design of Soft Gripper with Neural Physics

Sha Yi, Xueqian Bai, Adabhav Singh, Jianglong Ye, Michael T. Tolley, Xiaolong Wang
Proceedings of The 9th Conference on Robot Learning, PMLR 305:4313-4327, 2025.

Abstract

For robot manipulation, both the controller and end-effector design are crucial. Compared with rigid grippers, soft grippers are more generalizable by deforming to different geometries, but designing such a gripper and finding its grasp pose remains challenging. In this paper, we propose a co-design framework that generates an optimized soft gripper’s block-wise stiffness distribution and its grasping pose, using a neural physics model trained in simulation. We adopt a uniform-pressure tendon model, then generate a diverse dataset by randomizing both gripper pose and design parameters. A neural network is trained to approximate this forward simulation, yielding a fast, differentiable surrogate. We embed that surrogate in an end-to-end optimization loop to recover the ideal stiffness configuration and best grasp pose. Finally, we 3D-print the optimized grippers of various stiffness by changing the printing infills and parameters. We demonstrate that our co-designed grippers significantly outperform baseline designs in terms of force closure and success rate.

Cite this Paper

BibTeX
@InProceedings{pmlr-v305-yi25a, title = {Co-Design of Soft Gripper with Neural Physics}, author = {Yi, Sha and Bai, Xueqian and Singh, Adabhav and Ye, Jianglong and Tolley, Michael T. and Wang, Xiaolong}, booktitle = {Proceedings of The 9th Conference on Robot Learning}, pages = {4313--4327}, year = {2025}, editor = {Lim, Joseph and Song, Shuran and Park, Hae-Won}, volume = {305}, series = {Proceedings of Machine Learning Research}, month = {27--30 Sep}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v305/main/assets/yi25a/yi25a.pdf}, url = {https://proceedings.mlr.press/v305/yi25a.html}, abstract = {For robot manipulation, both the controller and end-effector design are crucial. Compared with rigid grippers, soft grippers are more generalizable by deforming to different geometries, but designing such a gripper and finding its grasp pose remains challenging. In this paper, we propose a co-design framework that generates an optimized soft gripper’s block-wise stiffness distribution and its grasping pose, using a neural physics model trained in simulation. We adopt a uniform-pressure tendon model, then generate a diverse dataset by randomizing both gripper pose and design parameters. A neural network is trained to approximate this forward simulation, yielding a fast, differentiable surrogate. We embed that surrogate in an end-to-end optimization loop to recover the ideal stiffness configuration and best grasp pose. Finally, we 3D-print the optimized grippers of various stiffness by changing the printing infills and parameters. We demonstrate that our co-designed grippers significantly outperform baseline designs in terms of force closure and success rate.} }
Endnote
%0 Conference Paper %T Co-Design of Soft Gripper with Neural Physics %A Sha Yi %A Xueqian Bai %A Adabhav Singh %A Jianglong Ye %A Michael T. Tolley %A Xiaolong Wang %B Proceedings of The 9th Conference on Robot Learning %C Proceedings of Machine Learning Research %D 2025 %E Joseph Lim %E Shuran Song %E Hae-Won Park %F pmlr-v305-yi25a %I PMLR %P 4313--4327 %U https://proceedings.mlr.press/v305/yi25a.html %V 305 %X For robot manipulation, both the controller and end-effector design are crucial. Compared with rigid grippers, soft grippers are more generalizable by deforming to different geometries, but designing such a gripper and finding its grasp pose remains challenging. In this paper, we propose a co-design framework that generates an optimized soft gripper’s block-wise stiffness distribution and its grasping pose, using a neural physics model trained in simulation. We adopt a uniform-pressure tendon model, then generate a diverse dataset by randomizing both gripper pose and design parameters. A neural network is trained to approximate this forward simulation, yielding a fast, differentiable surrogate. We embed that surrogate in an end-to-end optimization loop to recover the ideal stiffness configuration and best grasp pose. Finally, we 3D-print the optimized grippers of various stiffness by changing the printing infills and parameters. We demonstrate that our co-designed grippers significantly outperform baseline designs in terms of force closure and success rate.
APA
Yi, S., Bai, X., Singh, A., Ye, J., Tolley, M.T. & Wang, X.. (2025). Co-Design of Soft Gripper with Neural Physics. Proceedings of The 9th Conference on Robot Learning, in Proceedings of Machine Learning Research 305:4313-4327 Available from https://proceedings.mlr.press/v305/yi25a.html.

Related Material