Original title: Fin-QD: A Computational Design Framework for Soft Grippers: Integrating MAP-Elites and High-fidelity FEM
Authors: Yue Xie, Xing Wang, Fumiya Iida, David Howard
The article explores computational design for soft robotics, particularly in optimizing soft grippers’ designs for diverse objects. While past research has focused on individual finger design, the overall gripper structure remains underexplored. The challenge lies in managing the vast design space and complexities arising from material properties and contact dynamics. This study introduces a novel computational framework employing a quality-diversity approach to optimize soft grippers. It explores a wide design space encompassing finger arrangement and employs Finite Element Modelling (FEM) for accurate grasp analysis. By considering various gripper features, such as volume and workspace, the framework generates diverse gripper designs capable of handling distinct object shapes. This innovative approach fills the gap in exploring the complex design space of soft grippers, enabling effective grasping of diverse objects with simplified control methods.
Original article: https://arxiv.org/abs/2311.12477