Sim-to-Real for Soft Robots using Differentiable FEM: Recipes for Meshing, Damping, and Actuation
This work addresses the sim-to-real gap for soft robots, which is crucial for researchers and engineers in robotics, but it is incremental as it builds on existing FEM and differentiable simulation methods.
The paper tackled the problem of accurately modeling soft robots in simulation for optimal control by comparing differentiable Finite Element Method (FEM) models with physical measurements, achieving improved parameter calibration and proposing a predictive model for pneumatic actuation.
An accurate, physically-based, and differentiable model of soft robots can unlock downstream applications in optimal control. The Finite Element Method (FEM) is an expressive approach for modeling highly deformable structures such as dynamic, elastomeric soft robots. In this paper, we compare virtual robot models simulated using differentiable FEM with measurements from their physical counterparts. In particular, we examine several soft structures with different morphologies: a clamped soft beam under external force, a pneumatically actuated soft robotic arm, and a soft robotic fish tail. We benchmark and analyze different meshing resolutions and elements (tetrahedra and hexahedra), numerical damping, and the efficacy of differentiability for parameter calibration using a simulator based on the fast Differentiable Projective Dynamics (DiffPD). We also advance FEM modeling in application to soft robotics by proposing a predictive model for pneumatic soft robotic actuation. Through our recipes and case studies, we provide strategies and algorithms for matching real-world physics in simulation, making FEM useful for soft robots