Inverse Kinematics Modeling of a Cable-Actuated Soft Robotic Neck in SOFA

Abstract

This work presents a comparative study of two approaches to solving the inverse kinematics (IK) of a cable-driven soft robotic
neck. The first approach is based on a constant curvature analytical model, which assumes an internal and uniform distribution
of cables along the structure; this hypothesis proves inadequate when the cables are fixed to an external platform at the top of
the robot, leading to growing deviations due to cable sagging. The second method obtains the inverse kinematics through a finite
element simulation coupled with a mechanical constraints solver within the Simulation Open Framework Architecture (SOFA)
environment, allowing for a more accurate modeling of the physical interactions and actual deformations of the system. The
results demonstrate that the simulation-based solution offers significantly greater robustness and accuracy than the analytical
model when the geometric assumptions are no longer valid, validating the effectiveness of advanced physical simulation tools
for controlling soft robots in complex environments and configurations.