Soft IK with Stiffness Control

Building robots using soft materials comes with numerous advantages, including inherent safety and natural compliance with the environment. However, it is quite challenging to match the strength and precision of rigid robots. In this work we ask how to get the best of both worlds, and make soft robots capable of acting stiff on demand. This would expand the capabilities of soft robots, and work towards the vision of using them to accomplish complex and variable tasks. We extend the open loop control method called Soft IK (soft robot inverse kinematics) to simultaneously control a cable-driven soft robot's stiffness along with its deformed shape. Our method is an interactive multi-objective optimization, which reasons about stiffness by considering an ensemble of load cases. We show that by harnessing redundant actuation, our method is able to hit IK targets with controllable stiffness.