Hierarchical Performance-Based Design Optimization Framework for Soft Grippers

This paper presents a hierarchical, performance-based framework for the design optimization of multi-fingered soft grippers. To address the need for systematically defined performance indices, the framework employs a multi-objective optimization approach, structuring the optimization process into three integrated layers: Task Space, Motion Space, and Design Space. In the Task Space, performance indices are defined as core objectives, while the Motion Space interprets these into specific movement primitives. Finally, the Design Space applies parametric, topological, and field optimization techniques to refine the geometry and material distribution of the system, achieving a balanced design across key performance metrics. The framework's layered structure enhances SG design, balancing competing design goals without predefined weight assignments, enabling flexible trade-off exploration and scalability for complex tasks.