Co-Design of Soft Gripper with Neural Physics

2025-05-26

AI Summary

Soft robotic grasping faces challenges in jointly optimizing stiffness distribution and grasp pose, non-differentiable simulation, and decoupled structural-design-and-control paradigms. Method: This paper proposes an end-to-end differentiable co-optimization framework that integrates a neural physics surrogate model into the optimization pipeline, enabling concurrent design of segmented stiffness topology and grasp pose for soft grippers. A differentiable dynamic model of soft fingers is built upon a uniform-pressure tendon-driven actuation model; training data are generated via stochastic-parameter simulation. The framework supports monolithic multi-stiffness 3D printing. Results: In both simulation and real-world hardware experiments, the method achieves over 40% improvement in grasp success rate compared to baseline approaches, demonstrating superior generalizability, robustness, and manufacturability.

Technology Category

• Intelligent Robots: Manipulation
• Search and Optimization: Sampling/Simulation-based Search
• Reasoning under Uncertainty: Stochastic Optimization

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Experiences and lessons learnt from Web-based algorithms and system deployments
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• User Modeling, Personalization and Recommendation: User modeling and simulation for interactive and conversational systems

Abstract

For robot manipulation, both the controller and end-effector design are crucial. Soft grippers are 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 derived a uniform-pressure tendon model for a flexure-based soft finger, then generated 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 optimize the ideal stiffness configuration and best grasp pose. Finally, we 3D-print the optimized grippers of various stiffness by changing the structural parameters. We demonstrate that our co-designed grippers significantly outperform baseline designs in both simulation and hardware experiments.

Problem

• Optimize soft gripper stiffness and grasp pose
• Develop neural physics model for co-design
• Enhance gripper performance via simulation and 3D printing

Innovation

• Co-design framework optimizes gripper stiffness and pose
• Neural physics model enables fast differentiable simulation
• 3D-printed grippers with varied stiffness outperform baselines