Optimization-Driven Design of Monolithic Soft-Rigid Grippers

TL;DR

Addresses the sim-to-real gap in soft robotics by designing WaveJoint-based soft-rigid grippers through a global optimization framework that maps joint geometry to target stiffness. Uses a surrogate-based optimization with Radial Basis Functions (RBF) via the RBFOpt library and compares with an incremental neural network approach, leveraging FEM simulations and real-world tests. Demonstrates that the method can identify manufacturable joint geometries that achieve the desired stiffness with few measurements, and validates with noisy real-world experiments. Shows that the feasible stiffness space can be characterized and that the approach reduces prototyping iterations, enabling faster deployment of soft robotic grippers.

Abstract

Sim-to-real transfer remains a significant challenge in soft robotics due to the unpredictability introduced by common manufacturing processes such as 3D printing and molding. These processes often result in deviations from simulated designs, requiring multiple prototypes before achieving a functional system. In this study, we propose a novel methodology to address these limitations by combining advanced rapid prototyping techniques and an efficient optimization strategy. Firstly, we employ rapid prototyping methods typically used for rigid structures, leveraging their precision to fabricate compliant components with reduced manufacturing errors. Secondly, our optimization framework minimizes the need for extensive prototyping, significantly reducing the iterative design process. The methodology enables the identification of stiffness parameters that are more practical and achievable within current manufacturing capabilities. The proposed approach demonstrates a substantial improvement in the efficiency of prototype development while maintaining the desired performance characteristics. This work represents a step forward in bridging the sim-to-real gap in soft robotics, paving the way towards a faster and more reliable deployment of soft robotic systems.

Figures (9)

• Figure 1: Scheme of the WaveJoint with the parameters that can be varied in the design process based on the desired stiffness.
• Figure 2: Flowchart of the optimization-based procedure with RBFOpt as black-box optimization software library.
• Figure 3: Results in terms of displacements of a representative simulation for stiffness evaluation based on FEM structural analysis. For each simulation a total of 1 N force was applied with the following parameters of the WaveJoint: $\textcolor{black}{n_r}=4$, $\textcolor{green}{l_t}= 25 \text{mm}$, $\textcolor{blue}{t_h}=0.5\text{mm}$, $\textcolor{cyan}{\alpha}=10^\circ$, and $\textcolor{orange}{h_t}=5 \text{mm}$. (a) Longitudinal $(k_\xi)$ bending simulation. (b) Lateral $(k_\eta)$ bending simulation. (c) Torsional $(k_\zeta)$ simulation.
• Figure 4: Test on bench for a WaveJoint sample. (a) Longitudinal $(k_\xi)$ bending test. (b) Lateral $(k_\eta)$ bending test. (c) Torsional $(k_\zeta)$ test.
• Figure 5: Progress of residual along RBFOpt iterations for the 14 problem instances related to Table \ref{['tab::preliminary']}. (a) 7 instances -- No data point provided at RBFOpt initialization. (b) 7 instances -- 3043 data points provided at RBFOpt initialization.