Jacobian Estimation for Pneumatic Catheter Control

Soft pneumatic catheters are difficult to model analytically because their shape depends on nonlinear material behavior and contact. This project estimated the catheter’s Jacobian online — the local mapping from actuation pressures to tip motion — enabling closed-loop control without an explicit physical model.

The empirical Jacobian was updated continuously from observed tip motion, letting the controller adapt as conditions changed.

Contributions

1. Built the pneumatic actuation and electromagnetic tip-tracking testbed.
2. Implemented online Jacobian estimation from measured tip motion.
3. Demonstrated closed-loop tip positioning under model-free control.